Human Health Risk Assessment Onondaga Lake

Lake Bottom Subsite: Sediment Consolidation Area Camillus, NY

June 2010

U.S. Environmental Protection Agency, Region II Emergency and Remedial Response Division

290 Broadway New York, NY 10007

2

TABLE OF CONTENTS Executive Summary....................................................................................................................3 1. Introduction...................................................................................................................10

1.1 Background on Onondaga Lake HHRA..............................................................10 1.2 Overview of the Sediment Consolidation Area...................................................11 1.3 Method Used to Calculate Risks for the SCA .....................................................13

2. Conceptual Site Model and Human Exposure Pathways ................................................13 2.1 Potential Human Exposure Pathways Related to Sediment Remediation ............14 2.2 Potential Human Exposure Pathways Related to Hypothetical Release of Sediment

at the SCA ............................................................................................................16 3. Hazard Identification .....................................................................................................16

3.1 Data Collection and Evaluation ..........................................................................17 3.2 Criteria for Selecting COPCs .............................................................................17 3.3 Calculation of the Exposure Point Concentration................................................19

4. Exposure Assessment ....................................................................................................20 4.1 Exposure Assumptions.......................................................................................20 4.2 Estimating Exposure ..........................................................................................21

5. Toxicity Assessment......................................................................................................22 5.1 Health Effects Criteria for Non-Carcinogens ......................................................22 5.2 Health Effects Criteria for Carcinogens..............................................................23

6. Risk Characterization ....................................................................................................24 6.1 Risk Characterization for Carcinogens ...............................................................25 6.1.1 Methods.............................................................................................................25 6.1.2 Quantification of Carcinogenic Risk...................................................................25 6.2 Quantification of Hazard Indices for Effects other than Cancer ..........................26 6.2.1 Methods.............................................................................................................26 6.2.2 Quantification of Noncarcinogenic Risks ...........................................................27 6.3 Off-Site Workers................................................................................................27 6.4 Uncertainty Assessment .....................................................................................28 6.4.1 Hexavalent Chromium .......................................................................................28 6.2.2 Methyl mercury..................................................................................................28 6.2.3 Cobalt ................................................................................................................29 6.2.4 Toxicity Values..................................................................................................29 6.2.5 Potential for Overestimation within Exposure Scenarios ....................................30 6.2.6 Exposure Assumptions.......................................................................................30 6.2.7 Inhalation of Volatilized Chemicals ...................................................................30 6.2.8 Application of the Highest Receptor Location in Air Estimates ..........................30 6.2.9 TICs...................................................................................................................30

7. Conclusions...................................................................................................................32 8. References.....................................................................................................................33

Appendix A: Figures Appendix B: RAGS Part D Tables Appendix C: Air Dispersion Model Documentation Appendix D: Length Weighted Average Procedure and Sample Locations Appendix E: Air Quality Bench Testing Summary Appendix F: ProUCL Outputs

3

Executive Summary Prior Onondaga Lake Assessment Indicated Need for Sediment Remediation to Protect Ecological Receptors Present In and Near the Lake and Humans who Consume Fish A Record of Decision (ROD) selecting a remedy for the Onondaga Lake Bottom subsite was issued jointly by the New York State Department of Environmental Conservation (NYSDEC) and USEPA in July 2005 (NYSDEC and EPA, 2005). The selected remedy includes hydraulically dredging sediments from the lake, piping the water/sediment mixture up to a sediment consolidation area (SCA) at Wastebed 13 and into geotextile tubes, collecting and treating the water that drains from the geotextile tubes, and encapsulating the geotextile tubes containing sediments in a lined cell on the wastebed. The SCA will then be capped, maintained, and monitored to ensure that it is protective of human health and the environment. The need to remediate lake sediments was based on a Baseline Human Health Risk Assessment (HHRA) (TAMS 2002a) and Baseline Ecological Risk Assessment (BERA) (TAMS 2002b) that was performed for the sediments and other media within the lake. The reports identified risks exceeding acceptable levels for people consuming fish from Onondaga Lake and for ecological receptors present in and near the lake. The human health risk assessment also evaluated risks associated with direct contact with sediments (inadvertently ingesting small amounts of sediment or having sediment contact the skin) and this did not result in unacceptable risks. In response to a recent request from the community and elected officials, EPA has prepared this supplemental HHRA to identify any potential risks posed by sediment management and dewatering activities which will take place at the SCA. This assessment incorporated numerous conservative assumptions, and indicates all potential risks are within levels identified by EPA as acceptable. This Supplemental Assessment is Focused on the Sediment Consolidation Area (SCA) Consistent with EPA risk assessment guidance, a four-step process was utilized for assessing potential human health risks for the SCA: 1 Hazard Identification – Identifies the contaminants of potential concern associated with

site-related contaminants based on several factors such as toxicity, frequency of occurrence and concentration.

2 Exposure Assessment – Estimates the magnitude of actual and/or potential human exposures, the frequency and duration of these exposures and the exposure pathways (e.g.., ingesting contaminated sediment) under future exposure scenarios and under the reasonable maximum exposure anticipated.

3 Toxicity Assessment – Determines the types of adverse health effects associated with chemical exposures, and the relationship between magnitude of exposure (dose) and severity of adverse effects (response).

4 Risk Characterization – Summarizes and combines outputs of the exposure and toxicity assessments to provide a quantitative assessment of site-related risks and hazards, and presents a discussion of the uncertainties of the process.

Each of these steps is discussed below.

4

Hazard Identification Concentrations of chemicals in sediments were compared with health protective screening values to determine which chemicals should be considered in the risk assessment (termed contaminants of potential concern [COPCs]). All COPCs were then further considered in the HHRA. Exposure Assessment The exposure assessment evaluates pathways by which people are or can be exposed to the contaminants of concern in different media (e.g., sediment, soil, air). This exposure assessment considers only hypothetical future exposure scenarios. The quantification of exposure is based on factors including, but not limited to, the concentrations that people are or can be exposed to, the potential frequency of exposure (number of days per year), and the duration of exposure (number of years). The exposure assessment is based on the maximum site-specific parameters that can reasonably be expected at the site, which is termed the reasonable maximum exposure or RME. All potential ways people could be exposed to chemicals in sediments (termed exposure scenarios) were evaluated. Two hypothetical future exposure scenarios were evaluated here: 1. Offsite exposure to chemicals that might volatilize from sediments and water during

sediment management and dewatering in the SCA and migrate beyond the SCA. A five year time frame is evaluated consistent with the anticipated duration of activities at the SCA.

2. Onsite exposure to chemicals in sediments in the SCA if somehow the sediment

containment system was to fail and people were to come into contact with the sediments. Although this scenario is unlikely due to the design and engineering of the SCA, this hypothetical exposure, which evaluates potential risks during the time between a hypothetical failure and when the materials are again secured, was included at the request of the community. This assessment was conducted using the assumptions typically used to estimate residential exposures and is a very health-protective approach because in the unlikely event of a failure of the SCA, any potential exposure that might occur to the sediments would require people coming onto Wastebed 13 and contacting sediments on or near the SCA. The exposure scenario requires that these individuals would need to contact the sediments daily for the 45 day period after the release. During this period, engineering controls such as additional fencing and/or cover material would be implemented to mitigate exposures and corrective actions would be initiated.

Both scenarios are hypothetical and apply conservative health-protective assumptions to evaluate people of all ages in residential settings. For the sediment exposure scenario, a representative site concentration was calculated for all COPCs as a conservative estimate of the mean (i.e, the 95% upper confidence limit on the mean concentration). For the inhalation scenario, in order to conservatively estimate the air concentrations of chemicals at receptors, health-based air concentrations at the work zone perimeter were used as a starting point. Control measures will be implemented within the SCA to ensure that these criteria are met. An air dispersion model was then used to estimate the chemical concentrations in offsite air at community receptors assuming all chemicals were

5

present at the boundary at the full extent of the allowable concentration. This approach likely overestimates risk because it assumes that site boundary concentrations are at the maximum level for all chemicals, when in reality the worst case is that one or a select few chemicals may be approaching this level while the majority of chemical concentrations would be significantly below the criteria. All exposure assumptions were selected based on EPA guidance on deriving estimates that represent the reasonable maximum exposure (RME) and as such, these estimates are intended not to underestimate risks, and they likely overestimate risks for most individuals. Toxicity Assessment The toxicity assessment conservatively estimates the types of adverse health effects potentially associated with exposures to contaminants at the site and the relationship between the magnitude of exposure (dose) and severity of adverse effects (response). EPA has identified toxicity values for both carcinogenic effects (termed slope factors or unit risks) and non-carcinogenic effects associated with the COPCs. The toxicity values for effects other than cancer include oral reference dose (RfD), the absorbed RfD for dermal exposure, and the inhalation reference concentrations (RfC). The non-cancer health endpoints (e.g., the target organ) are also assembled and considered in the risk assessment. These toxicity values applied in the risk assessment are selected to represent effects on the most sensitive endpoints and life stages and as such provide a health protective means to evaluate risks. Risk Characterization In risk characterization, quantitative exposure estimates and toxicity factors are combined to calculate numerical estimates of potential health risk. In this section, potential cancer and noncancer health risks are estimated assuming long-term exposure to chemicals detected in site media. A 1×10–6 cancer risk represents a one-in-one-million additional probability that an individual may develop cancer over a 70-year lifetime as a result of exposure under the conditions and scenarios evaluated. The findings presented here are compared with the range of acceptable risk levels cited in the National Oil and Hazardous Substances Pollution Contingency Plan (NCP), of 1×10–6 to 1×10–4. No cancer risk estimates were greater than the 1×10-4 risk level that is the upper end of the EPA target risk range. The following cancer risk estimates were derived:

• Inhalation – Adults and Children: The hypothetical future inhalation cancer risk estimates

for offsite adults and children were 4×10-6 with primary contributors to risk being ethylbenzene and naphthalene.

• Contact with Sediments – Adults: The hypothetical future cancer risk estimates associated with exposure to sediments (oral and dermal exposure routes) was 1×10-6 for all carcinogenic chemicals combined.

• Contact with Sediments – Children: The hypothetical future cancer risk estimate associated with exposure to sediments (oral and dermal routes) was 3×10-5 for children ages zero to 2 and 1×10-5 for children ages 2-6, with the primary contributors to risk being hexavalent chromium and the carcinogenic PAHs.

• Contact with Sediments – Adolescents: The hypothetical future cancer risk estimate associated with exposure to sediments (oral and dermal routes) was 3×10-6 for adolescents

6

ages 6 to 16 (Appendix B Table 8.1) also associated primarily with hexavalent chromium and the carcinogenic PAHs.

The following noncancer risk estimates were derived: All Scenarios: All noncancer risk estimates had hazard indices less than or equal to one indicating low potential for adverse effects and that non-cancer risks were within acceptable levels. Detailed risk estimates are provided in Appendix B Tables 7-1 through 7-3 and Appendix B, Table 8-2 (inhalation exposure) and Table 8-3 (sediment exposure) show the hazard indices (HI) for COPCs summed based on the primary target organs (or systems) that they effect. Tables 8-2 and 8-3 also show the primary chemicals contributing to the total hazard index for each pathway-endpoint combination. For adult exposure to sediments (see Table 7-2), none of the scenarios had hazard quotients or combined hazard indices (0.2) greater than 1.0. Although the total HI (approximately 2) for child exposure to sediments (ingestion and dermal contact) slightly exceeds 1 (as shown in Table 7-1), the HI values for each endpoint do not exceed 1 (as shown in Table 8-3). In addition, although the total HI (approximately 4) for the inhalation exposure exceeds 1 (as shown in Table 7-3), the HI values for all receptors for each endpoint do not exceed 1 (as shown in Table 8-2). Hypothetical risk estimates were derived using conservative assumptions as a means to evaluate plans for the SCA and to ensure the SCA activities are safe and protective of human health. The findings from this assessment can be used by risk managers to ensure that the SCA is designed, constructed, monitored, and managed within acceptable risk levels. Many of the assumptions applied here tend to overestimate potential site risks. These include the following:

1. Hexavalent Chromium. Chromium can exist is several forms, and two forms are most commonly considered in an HHRA, the trivalent form and the hexavalent form. Results for samples collected from 4 locations near the Crucible Lake Pump Station site that were analyzed for both total and hexavalent chromium did not identify any hexavalent chromium; these results were therefore considered to be in the trivalent form, which is significantly less harmful, and were ultimately screened out and not considered further in this HHRA. All other chromium results were analyzed only for total chromium, and were conservatively considered to be in the hexavalent form, which is more toxic. Under this conservative assumption, hexavalent chromium risk estimates were responsible for approximately 90% of the cancer risk estimates for ingestion of sediments. This is likely an overestimate of the true risk, since hexavalent chromium is reasonably assumed only to be a small percentage of the total chromium in the sediments, and the limited data which are available indicate that the chromium is not present as hexavalent chromium.

2. Methyl mercury. Mercury is another chemical that can exist in several forms.

However, these forms were not analyzed in most sediment samples, and mercury was reported as total mercury. For the hypothetical sediment exposure pathway in this HHRA, all mercury risk estimates were calculated assuming mercury was present as methyl mercury in sediments, which is the most conservative approach when assessing oral and dermal exposures. However, existing sediment data indicate the maximum percentage of methyl mercury is 1.4%. Nevertheless, all hazard indices for

7

mercury for both the sediment and the inhalation exposure pathways were within acceptable levels.

3. Cobalt. One exposure scenario estimates potential risk from direct contact from

sediments in the hypothetical event of a failure of the SCA and individuals come into contact with the sediments. Under this scenario, cobalt contributes approximately 19% to the overall hazard index to the child from exposure to sediments (the total noncancer hazard index to the adult are less than 1, so no further evaluation is necessary). The sediment data set for cobalt consists of 133 samples collected from 37 locations within the area to be dredged. A length weighted average (LWA) was calculated for each location and these 37 LWAs were used to calculate the exposure point concentration for cobalt that was used in the risk assessment. Of these 37 locations, two locations near the Crucible Lake Pump Station site have significantly higher LWA cobalt concentrations than all other locations in the area to be dredged and strongly influence the overall exposure point concentration. These two locations represent less than 2% of the overall volume of sediment to be dredged and are not representative of the cobalt concentrations in the material that will be dredged. Therefore, the noncancer hazard index for cobalt is likely overestimated.

4. Toxicity Values. Significant uncertainty may be associated with the derivation of

RfDs, CSFs, and all toxicity values. Toxicity values based on human epidemiological studies are not available for most chemicals, and those human studies that are available generally lack exposure data and are confounded by exposure to multiple chemicals, recall bias, and lifestyle issues. Laboratory animal studies are used to derive most toxicity values and the practice of extrapolating from effects in animals to predict human toxic response is a major source of uncertainty in risk assessment.

5. Potential for Overestimation within Exposure Scenarios. The SCA is planned to

be closely managed and maintained. Consequently, the assumed potential for exposure to sediments is hypothetical and may overestimate risks, particularly for young children who would be unlikely to come into contact with this material.

6. Exposure Assumptions. There is considerable uncertainty regarding the likelihood

of exposure to a given medium of concern. It is unknown whether all of the exposure pathways modeled will ever be actually complete or whether the individuals evaluated will actually be exposed to COPCs. Exposure estimates used to calculate risks and hazards may also be relatively uncertain. Many of the exposure parameter values applied are default values determined by USEPA rather than site-specific values. As such, risk estimates based on these exposure parameters generally represent conservative estimates.

7. Inhalation of Volatilized Chemicals. The air concentrations used as a starting point

in deriving offsite air estimates are all assumed to be the lower of either the NYSDEC’s DAR-1 guidance value (“Guidelines for Control of Toxic Ambient Air Contaminants’) (NYSDEC, 2007), and EPA’s Regional Screening levels (RSL) (US EPA, 2010a) for industrial settings after adjustment for a 5-year exposure duration for all chemicals identified in the sediments as volatile. Both bench scale testing of

8

sediments and testing of volatilization from sediments suggest that instead approximately half of the chemicals would actually be volatilized from sediments. This assumption contributes to a cumulative air exposure risk estimate that is higher than what would likely occur.

8. Application of the Highest Receptor Location in Air Estimates. Work zone

perimeter concentrations were used to estimate offsite air concentrations. The resulting estimates were evaluated, and the highest yearly average concentration was applied. The selection of this maximum value means that all other yearly average concentrations would be lower and thus exposure estimates for most people would be lower than this value.

9. Tentatively Identified Compounds (TICs). Standard analytical protocols for

Superfund sites require that an extensive list of chemicals be analyzed and reported. This list is known as the Target Compound List (TCL)/Target Analyte List (TAL) for organic chemicals and metals, respectively. The list was developed with consideration of the chemicals most commonly found at sites. Other chemicals that are not routinely analyzed for or reported might be present at sites. These chemicals are known as tentatively identified compounds (TICs), since neither the identity nor the concentration can be reported with certainty. The TICs could present a contribution to risk that has not been quantified, however, they typically do not contribute significantly to the overall risk or hazard at a site. The extensive database of more than 33,000 samples collected at the site provides a robust data set that supports the basis for the risks identified in this assessment.

Conclusions This supplemental risk assessment applied conservative exposure assumptions to evaluate potential risks associated with the operation of the SCA and to address community concerns. Estimates were designed to represent two hypothetical future scenarios: 1) exposure to contaminants that could migrate from the site in air during the operation of the SCA and 2) exposure to sediments within the SCA post-closure if the SCA were to fail, sediments were released, and people would come onto Wastebed 13 and come into contact with the sediment on or near the SCA. Both of these potential future scenarios were intended to represent the reasonable maximum exposure potential and both assume individuals of all ages could be exposed. As such, these risk estimates are likely higher than risks that would likely be experienced by most receptors. All resulting risk estimates and target organ-specific hazard indices were within levels identified by EPA as acceptable. The finding of acceptable risk estimates through application of these health protective assumptions, indicates that the SCA will not result in unacceptable risks for the surrounding community. Nevertheless, the SCA will be closely monitored to ensure that sediments are managed with care and secured appropriately and that offsite migration of chemicals in air is limited or prevented. This HHRA can also be used as a tool for risk managers during the implementation of the remedy and management of the SCA. The exposure scenarios evaluated in this HHRA are future

9

hypothetical situations, so the outcome can be used to help assess the effectiveness of the remedy. For example, the air concentrations modeled in the residential areas are based on work zone perimeter concentrations that reflect the maximum annual average concentrations for all chemicals. It is highly unlikely that every volatile chemical would be present at that concentration for a one year period of time to result in that exposure scenario. During remedy implementation, if monitored air concentrations indicate a trend towards chemicals reaching this maximum annual average concentration for a sustained period of time, risk managers can modify site operations to reduce these concentrations so that the actual risks are much lower than those estimated here. Risk managers can also use this HHRA to assist in managing risks in the unlikely event of a failure of the SCA. If sediments are released as a result of a failure of the SCA, measures would be implemented to address the release in accordance with site management plans (e.g.. Spill Contingency Plan) to be developed. The measures may include additional sampling and characterization, for example, speciating potential risk-driving chemicals such as mercury and chromium, to ensure that the actual exposures that would occur once the interim controls are implemented are not greater than the exposures in the scenarios defined in this HHRA.

10

1. Introduction

A Baseline Human Health Risk Assessment (HHRA) (TAMS 2002a) and Baseline Ecological Risk Assessment (BERA) (TAMS 2002b) were performed for the Lake Bottom Subsite of the Onondaga Lake Site in 2002. The HHRA estimated risks exceeding acceptable levels related to consumption of fish that had accumulated contaminants from sediments and surface water. The BERA estimated potential for chemical waste in the lake to produce adverse ecological effects to ecological receptors present in and near the lake. A remedial action was proposed to address these risks and, after public review and comment, a remedy was selected. The remediation includes dredging sediments from the lake and placing them in a Sediment Consolidation Area (SCA) located near the lake. In response to a recent request from the community and elected officials, EPA has prepared this supplemental HHRA to identify any risks posed by implementing the remedial action for the lake. This remedy includes hydraulically dredging sediments from the lake, piping the water/sediment mixture up to Wastebed 13 and into geotextile tubes, collecting and treating the water that drains from the geotextile tubes, and encapsulating the geotextile tubes containing sediments in a lined cell on the wastebed, which will then be capped, maintained, and monitored to ensure that it is protective of human health and the environment. The risk assessment provided here evaluates two possible means people could be exposed to chemicals from sediments. The ways people could be exposed are called exposure scenarios and include: 1. Offsite exposure to chemicals that might volatilize from sediments and water during

sediment management and dewatering in the SCA and migrate beyond the SCA. 2. Onsite exposure to chemicals in sediments in the SCA if somehow the sediment containment

system was to fail and people were to come into contact with the sediments. Although this scenario is unlikely due to the design and engineering of the SCA, this hypothetical exposure, which evaluates potential risks during the time between a hypothetical failure and when the materials are again secured, was included at the request of the community. This assessment was conducted using the assumptions typically used to estimate residential exposures and is a very health-protective approach because in the unlikely event of a failure of the SCA, any potential exposure that might occur would require people coming onto Wastebed 13 and contacting sediments on or near the SCA. The exposure scenario requires that these individuals would need to contact the sediments daily for the 45 day period after the release. During this period, engineering controls such as additional fencing and/or cover material would be implemented to mitigate exposures and corrective actions would be initiated

This introduction provides a discussion of the HHRA for Onondaga Lake to provide the background for the planned remediation, then briefly describes the basis used to conduct the risk assessment for the SCA. 1.1. Background on Onondaga Lake HHRA A Baseline Human Health Risk Assessment (HHRA) and Baseline Ecological Risk Assessment (BERA) were performed for the Lake Bottom Subsite of the Onondaga Lake Site in 2002. The reports were developed as part of the remedial investigation, as required by the National Oil and

11

Hazardous Substances Pollution Contingency Plan (NCP) (55 Fed. Reg. 8665-8865, March 1990) (US EPA, 1990), which states that the baseline risk assessment should “characterize the current and potential threats to human health and the environment that may be posed by contaminants migrating to ground water or surface water, releasing to air, leaching through soil, remaining in the soil, and bioaccumulating in the food chain” (Section 300.430(d)(4)). The main purpose for conducting the baseline risk assessments is to identify the potential baseline risks to human health and the environment posed by the site and for the risk managers to evaluate whether these potential baseline risks warrant a remedial action. In the baseline risk assessments for the Lake Bottom Subsite, risks were identified above the acceptable levels defined in the NCP to ecological receptors and to people who eat fish from the lake (TAMS, 2002a, and TAMS, 2002b.). These risks were sufficient to warrant a remedial action. The remedy that was proposed, presented to the public and then selected includes dredging sediment from the lake and placing it in a sediment consolidation area located near the lake. Wastebed 13 was identified as the preferred location following completion of the siting evaluation in September 2006 as documented in the NYSDEC’s October 2006 Fact Sheet. Following review of public comments, Wastebed 13 in the Town of Camillus was selected as the location for the SCA (NYSDEC, NYSDOH, and US EPA, 2010). In considering the supplemental risk assessment presented in this document, it is important to clarify the purpose and outcome of the baseline risk assessments for Onondaga Lake. The purpose was to determine the potential risks posed by the Lake site as it currently exists and to evaluate whether these risks justify a remedial action. The reports identified that the only unacceptable risks as defined by the NCP were to ecological receptors, and to people consuming fish caught from Onondaga Lake. Although direct contact (inadvertently ingesting small amounts of sediment or having sediment contact the skin) with lake sediments to humans was evaluated in the HHRA, exposure to these lake sediments did not result in unacceptable cancer risks1 or noncancer hazards2. 1.2. Overview of the Sediment Containment Area

The Onondaga Lake Bottom Subsite of the Onondaga Lake Site includes the contaminated surface water and sediments in the 4.5-square mile lake. Industrial waste and municipal sewage have been discharged to the lake for over 100 years. Mercury contamination is found throughout

1 In an HHRA, exposures are evaluated based on the potential risk of developing cancer and the potential for non-cancer health hazards. The likelihood of an individual developing cancer is expressed as a probability. For example, a 10-4 cancer risk means a “one-in-ten-thousand excess cancer risk,” or one additional cancer may be seen in a population of 10,000 people as a result of exposure to site contaminants under the conditions explained in the Exposure Assessment of the HHRA. Current federal Superfund guidelines for acceptable exposures are “generally concentration levels that represent an excess upper bound cancer to an individual of between 10-4 to 10-6” (40 CFR § 300.430[e][2][A][2]) (corresponding to a one-in-ten-thousand to a one-in-a-million excess cancer risk). The 10-6 risk is used as the point of departure for determining remediation goals and cancer risk estimates greater than 10-4 represent unacceptable risks. 2 For non-cancer health effects, a “hazard quotient” (HQ) is calculated for each contaminant. An HQ represents the ratio of the estimated exposure to the corresponding reference doses (RfDs). The sum of the HQs is termed the “hazard index” (HI). The key concept for a non-cancer HI is that a “threshold level” (measured as an HQ or HI of 1) exists, at or below which non-cancer health effects are not expected to occur.

12

the lake. Other contaminants present in lake sediments include benzene, toluene, ethylbenzene, xylenes, chlorinated benzenes, polycyclic aromatic hydrocarbons, PCBs, and polychlorinated dibenzo-p-dioxins/polychlorinated dibenzofurans. A Record of Decision (ROD) selecting a remedy for the Lake Bottom subsite was issued in July 2005. The selected remedy includes dredging up to an estimated 2.65 million cubic yards of contaminated sediments, isolation capping of an estimated 425 acres in the littoral zone (water depths ranging from 0 to 30 feet), thin layer capping of an estimated 154 acres, an oxygenation pilot study for the water in the deeper portion of the lake, and monitored natural recovery in the profundal zone (water depths exceeding 30 feet). While the most highly contaminated materials will be treated and/or disposed of off-site, the dredged sediment will be placed in a nearby SCA. Wastewater generated by the dredging/sediment handling processes as a result of dewatering of the sediments at the SCA will be treated prior to being discharged back to the lake (NYSDEC and US EPA, 2005). A draft Explanation of Significant Differences which describes a change to a portion of the remedy required by the ROD in the southwest portion of the lake was approved in December 2006. The change was necessary to ensure the stability of the adjacent causeway and the adjacent area which includes a portion of I-690, and is supported by extensive sampling of the area which indicates that the pure phase chemical contamination is significantly less extensive than estimated in the ROD. In October 2006, NYSDEC made available for public review and comment an SCA Siting Evaluation Report which assessed Solvay Wastebeds 1 – 15 and B as potential locations for the SCA based on accessibility, estimated capacity, current and future site use, geotechnical considerations, and distance from residences (Parsons, 2006). Based on the evaluation results, Wastebed 13 in the Town of Camillus was selected as the location for the SCA (see Figure 1 in Appendix A). This decision was documented in the Consent Decree between Honeywell and NYSDEC for the Lake Bottom subsite (United States District Court, Northern District of New York, 2007) (89-CV-815). The dewatering method for the dredged material presumed during the development and issuance of the ROD was a large open settling basin; however, the local community raised concerns pertaining to potential odor generation using this dewatering method. In response to these community concerns, an extensive evaluation comparing the geotextile tube and settling basin dewatering methods based on 10 site-specific dewatering objectives was conducted (Parsons and Geosyntec, 2009). Based on this evaluation, it was determined that there are many site-specific benefits of using geotextile tubes as compared to settling basins. These benefits include:

• The potential for significantly reduced odors and emissions; • Primary containment of the dredged sediments within the geotextile tubes; • Reduction in required berm height and preloading requirements as compared to an open

settling basin, thereby reducing scale of construction activities and associated truck traffic and noise levels;

• Reduction in required footprint as compared to an open settling basin because of lower SCA perimeter dike height, thereby reducing the visibility of the SCA and related construction activities;

13

• Improved ability to maintain geotechnical stability and SCA liner integrity as a result of the lower hydraulic head and flexibility related to tube placement as compared to an open settling basin; and

• Reduction in time to closure.

As a result of this evaluation, use of geotextile tubes for dewatering has been incorporated into the design documents. The design presently calls for the construction of a 72-acre SCA within Wastebed 13 to store the approximately 2.2 million cubic yards of material to be dredged from the lake. The layout, as shown on Figure 2 in Appendix A, was developed to maintain the buffer zones requested from the community (i.e., a 500-ft buffer along the western boundary of Wastebed 13) and an additional 200-ft buffer zone from the northern boundary of Wastebed 13. For dewatering the Onondaga Lake sediment, it is anticipated that geotextile tubes 80 to 90 ft in circumference and 200 to 300 ft in length would be used within the lined SCA. Slurry would be pumped into the tubes via ports along the top of the tubes, and the filtrate (water) would drain through the openings of the geotextile. Solids would be retained within the geotextile tubes (Parsons, 2009). 1.3 Method Used to Calculate Risks for the SCA

Consistent with EPA risk assessment guidance, a four-step process was utilized for assessing potential human health risks for the SCA:

1. Hazard Identification – identifies the contaminants of potential concern associated with site-related contaminants based on several factors such as toxicity, frequency of occurrence and concentration.

2. Exposure Assessment – estimates the magnitude of actual and/or potential human exposures, the frequency and duration of these exposures and the exposure pathways (i.e., ingesting contaminated sediment) under future exposure scenarios and under the reasonable maximum exposure anticipated.

3. Toxicity Assessment – determines the types of adverse health effects associated with chemical exposures, and the relationship between magnitude of exposure (dose) and severity of adverse effects (response).

4. Risk Characterization – summarizes and combines outputs of the exposure and toxicity assessments to provide a quantitative assessment of site-related risks and hazards, and presents a discussion of the uncertainties of the process.

As indicated above, for the SCA, two potential exposure scenarios were identified: exposures to contaminants in offsite air; and exposure to contaminants in sediments in the hypothetical case of a breach of the SCA. These two exposure scenarios are discussed in detail in Section 2.

2. Conceptual Site Model and Human Exposure Pathways

As stated above, this HHRA was developed in response to concerns raised by the community and elected officials that exposures to site-related contaminants might occur during the process during which the dredged sediments are pumped into the geotextile tubes and once the geotextile tubes are encapsulated in place in the SCA located in Wastebed 13. This section describes the

14

possible sources, receptors, and exposure pathways included in the HHRA through development of a conceptual site model and identification of possible human exposure pathways. The conceptual site model and exposure pathways are also summarized in Table 1 of Appendix B. A conceptual site model was developed to evaluate potential exposure pathways for chemicals of concern within SCA sediments during remediation and in the unlikely event of a future failure of the SCA. Exposure pathways are defined as the course a chemical takes from a source to an exposed receptor. Exposure pathways consist of the following four elements: 1) a source; 2) a mechanism of release, retention, or transport of a chemical to a given medium (e.g., air, water, soil); 3) a point of human contact with the medium (i.e., exposure point); and 4) a route of exposure at the point of contact (e.g., incidental ingestion, dermal contact). If any of these elements is missing, the pathway is considered incomplete (i.e., it does not present a means of exposure) (US EPA, 1989). A conceptual site model examines the range of potential exposure pathways and identifies those that are present and that may be important for human receptors, and it eliminates those pathways that are either incomplete or that constitute negligible exposures (i.e., exposures consistent with background or below a risk-based threshold). Exposure pathways are grouped into exposure scenarios as described above. Two scenarios were identified to characterize the potential exposures associated with the following: 1) placement and dewatering of sediment in the SCA and 2) exposures that could occur in the event of hypothetical failure of the SCA. The first scenario evaluates potential exposures that could occur during remediation. Specifically, chemicals may volatilize from sediments and from the water seeping from the geotextile tubes. Volatilized chemicals then could migrate from the SCA towards nearby residents. The second scenario assesses potential onsite exposures to sediments that may be released from a hypothetical failure of the SCA. 2.1. Potential Human Exposure Pathways Related to Sediment Remediation

Inhalation of volatilized chemicals is the only exposure pathway considered potentially complete during the disposal and dewatering of sediment in the SCA. The sediments, in the form of a slurry (a mixture of approximately 90% water and 10% solids) will be transported to the SCA via a double-walled pipe and pumped into the geotextile tubes in place in the SCA. Due to the engineering of this process, there is no reasonably anticipated exposure of sediments to the community or to people other than those workers trained and skilled in the handling of this type of material. Remediation workers will be protected from physical and chemical hazards through implementation and adherence to a site health and safety plan. Once the wet sediment material is pumped into the geotextile tubes, filtrate (water) will drain from the slurry, leaving behind dewatered sediments that will ultimately remain inside the geotextile tubes. The water will be collected, treated appropriately within an enclosed on-site water treatment plant to remove any site-related contaminants that are dissolved in the water, and discharged to the Onondaga County Metropolitan Wastewater Treatment Plant (METRO) plant for further treatment and thus there are no complete exposure pathways related to the geotextile tube filtrate water from this process. During the process by which the water is drained from the geotextile tubes and collected, certain chemicals may volatilize into the air and be dispersed into surrounding air, migrating from the SCA and resulting in lower concentrations in surrounding

15

areas. These potential air concentrations are evaluated within this risk assessment. The list of chemicals that are being evaluated in this scenario is discussed in Section 3.2. The populations near the SCA and who could potentially be exposed to these volatile chemicals include people living in neighborhoods near the facility and workers at the Town of Camillus’ construction and demolition landfill at Wastebed 15, which is shown in Figure 2 of Appendix A. However, because the workers are exposed less frequently (fewer days per year and fewer hours per day), their potential exposures would be less than exposures evaluated in this HHRA (i.e, for residents living in the neighborhoods near the SCA). Therefore, offsite workers are evaluated qualitatively, meaning that their potential risks will be discussed in the risk characterization section of this HHRA. The residents who live near the SCA will be evaluated quantitatively, meaning that their potential risks will be conservatively estimated, based on an assumed hypothetical exposure to the sediment chemicals that could volatilize into the air. For this evaluation, all chemicals present in sediments that could volatilize and that had toxicity values were included in the risk assessment. This list is further discussed in Section 3.2. In order to conservatively estimate the maximum concentration of these chemicals that might migrate into the community, concentrations at the work zone perimeter were used as a starting point. These concentrations, which are health-based air concentrations developed by either EPA or NYSDEC to be protective of commercial exposures over a 5 year duration, are discussed in detail below and in Appendix C. Control measures will be implemented to ensure that these criteria are met. An air dispersion model was then used to estimate the chemical concentrations at community receptors assuming the maximum allowable concentrations were present at the site boundary. A significant level of conservatism is built into this approach because it assumes that site boundary concentrations are at the maximum level for all chemicals, when in reality the worst case is that one or a select few chemicals may be approaching this level while the vast majority of chemical concentrations would be significantly below the criteria. This dispersion modeling is discussed more thoroughly in Appendix C. The maximum allowable site perimeter concentrations used in this modeling are based on criteria established by NYSDEC’s DAR-1 guidance (“Guidelines for the Control of Toxic Ambient Air Contaminants”), and US EPA’s Regional Screening Levels (RSL) (NYSDEC, 2007) (US EPA, 2010a), which are typically developed for both industrial and residential settings. The use of the DAR-1 and RSL levels establish risk-based concentrations designed to protect public health from effects which may be associated with long-term (70 year) exposure to these contaminants. These criteria have been modified to account for the duration of the project (5 years). For the purposes of this evaluation, the lower (more conservative) of either the modified DAR-1 values or the industrial RSLs have been selected as the maximum allowable site perimeter concentrations. These concentrations will form the basis of air quality criteria that will be specified for this project, and enforced at the workplace perimeter and for this reason they provide a good means to evaluate the potential exposures that could occur during the course of the sediment remediation. The air exposure scenario considered that offsite residents of all ages could be exposed by breathing contaminants in the air. Figure 3 of Appendix A shows the location where the sediment consolidation area is located, and the two boundaries utilized for this modeling analysis. This figure shows the work zone perimeter (green line), at which the above referenced

16

maximum allowable site perimeter concentrations will be enforced, as well as the nearest residential receptor boundary (blue line). This line was developed using aerial photographs, maps, and driving reconnaissance, and identifies the closest residential location in every direction from the SCA. For this analysis, residential locations (houses) as well as publicly accessible non-residential areas (e.g., parks, churches, etc.) were included. The modeling analysis calculated the average concentration along every point of this receptor boundary, assuming these maximum allowable concentrations were present at the site boundary. As an additional level of conservatism, the HHRA used the highest modeled air concentrations along this receptor boundary to estimate risks to people in this scenario. Table 3-2 in Appendix B shows the modeled air concentrations used in the risk assessment. 2.2. Potential Human Exposure Pathways Related to Hypothetical Release of Sediment at the SCA Once the dredging is complete, the sediments have drained, and the SCA is properly closed, a post-closure monitoring plan will be implemented to ensure ongoing maintenance of the facility and to confirm that the facility remains intact and no material has been released. The community has asked that a risk assessment be performed to determine what the potential risks to the community would be in the event that the SCA containment somehow failed and there was potential for people in the community to enter the site to come into contact with the sediments. This HHRA evaluates the risks to the community in the hypothetical scenario in which there is potential for people of all ages to be exposed to sediment at chemical concentrations in the SCA for a 45 day period before sediments were again contained. The 45 day response time to repair any damage to the SCA is considered the maximum amount of time it would take to make necessary repairs, recover any sediments released within the berm, and perform sampling to ensure no material had migrated from the point of release. It should be noted that interim measures such as fencing, cover material, or other engineering controls would be implemented shortly after any release to limit the potential for any exposure, and it is not likely that people would be in daily contact with sediments released from the SCA for the entire 45 day period; however, this is evaluated to provide a very conservative estimate of the hypothetical exposure. This exposure scenario considered that adult, adolescent (ages 6 – 16 years) and child (ages 6 years and under) residents could potentially be exposed by daily incidental ingestion of and dermal contact with contaminated sediments at the SCA during the 45 day period for the hypothetical release from the SCA. However, note that such a scenario is considered unlikely for the following reasons: the SCA contains four layers of containment (geotextile tubes, a lined cell, a newly constructed berm, and the final cover system) one or more of which would need to fail, and some way for individuals to access the sediment at the SCA, where the nearest residence is approximately 1500 feet from the SCA, which is shown on Figure 3 in Appendix A. Such exposure potential is considered unlikely, but is evaluated here for hypothetical purposes.

3. Hazard Identification

This section outlines the data used in the risk assessment, how data were collected, the criteria

17

for selecting the chemicals of potential concern (COPCs), and the calculation of the exposure point concentrations (EPCs).

3.1. Data Collection and Evaluation

This HHRA is focused on implementation of the remedy in the Consent Decree (United States District Court, Northern District of New York, 2007) (89-CV-815) for the Lake Bottom sediments: dredging the sediments and placing them in the SCA in Wastebed 13. Many samples have been collected from the areas of the lake that will be dredged during the remedial investigation and the pre-design investigation. Dredging will occur to varying depths of up to 4 meters, so only cores that represented these depths were included. In total, 329 locations from the areas targeted for dredging are used in this HHRA, including samples collected at 290 locations over the past 5 years as part of the pre-design investigation. Although many other samples have been collected, this subset of the data is considered the most representative because they were collected from areas to dredged and are most representative of the entire depth of the area to be dredged. Figure 4 in Appendix A illustrates the sample locations that are included in this HHRA. Samples were taken from different lengths along the depth of the core sample and sediments will be extensively mixed during remediation (e.g, they will be pumped as a slurry), so a length weighted average (LWA) approach was used to estimate the concentration of each contaminant for each sample location. In summary, the LWA concentration for each chemical was calculated by averaging the concentration detected for each core segment for the length of the core up to the anticipated depth of dredging for that location. These length weighted average concentrations were then combined to derive exposure point concentrations for each contaminant of potential concern. A detailed explanation of the procedure for calculating the LWA and a list of all sample locations are included in Appendix D. It should also be noted that not all chemicals were analyzed in every sample. Some samples were only analyzed for a smaller suite of chemicals. There are several reasons for this, such as certain areas were identified in the remedial investigation as not containing specific types of chemicals, so there was no need to look for them in those areas during the pre-design investigation. The LWA concentrations are representative concentrations for each sediment core from areas that will be dredged. It should be noted that all analytical methods used were approved by EPA and NYSDEC and followed proper quality assurance/quality control procedures.

3.2. Criteria for Selecting COPCs

Table 2-1 in Appendix B summarizes the analytical data for concentrations for sediments to be dredged that were used to determine the COPCs for the scenario that evaluates exposure to sediments in this risk assessment. Table 2-1 includes sediment data from the areas to be dredged. Table 2-2 identifies the chemicals that are included in the scenario to evaluate inhalation of volatile chemicals migrating from the SCA. Essential nutrients calcium, magnesium, potassium, and sodium were not evaluated due to low toxicity, which is consistent with EPA Region 2 protocols. Consistent with EPA guidance for risk assessment, chemicals identified as tentatively identified compounds (TICs) were not included in the analysis. TICs are chemicals for which either the identity or the concentration cannot be accurately reported. EPA guidance also recommends that chemicals that are detected infrequently (i.e., detections in less

18

than 5% of the samples) should not be included. This approach assures that the data used are of sufficient quality to be appropriate for risk assessment (US EPA, 1989).

In Table 2-1, the maximum detected concentration for each chemical for all sample locations was compared to the corresponding risk-based screening value for residential soils, from the Regional Screening Level (RSL) table (US EPA, 2010a). The RSL values represent an excess cancer risk of one in one million (1x10-6) or a hazard quotient of 1. The non-cancer hazard quotients from the RSL table were adjusted to 0.1 prior to comparison to account for potential exposures to multiple chemicals. If the maximum LWA concentration of the chemical exceeded its respective RSL value, the chemical was retained for quantitative analysis. If the concentration of a chemical was below its respective RSL value, that chemical was determined unlikely to cause adverse effects and was not included for quantitative analysis in the HHRA.

The RSL for methyl mercury was used to screen mercury in order to be health-protective; this is a health-protective approach since very little of the mercury in the sediments is present in the form of methyl mercury. Based on data from the RI, the maximum percentage of methyl mercury was 1.4% of the total mercury (TAMS, 2002c). However, all mercury was assumed to be methyl mercury for this HHRA. The RSL for hexavalent chromium was used to screen chromium in order to be health-protective; this is a health-protective approach since most of the chromium in the sediments is likely present in less toxic toxic forms. Based on data collected in 2008, 21 sediment samples from 7 locations near the Crucible Pump Station were analyzed for hexavalent chromium and no hexavalent chromium was detected (Environmental Data Services, Inc., 2008). Only four of the 7 locations were in areas to be dredged. Since the data confirmed no hexavalent chromium was present in these four locations, chromium results from these four location are evaluated as trivalent chromium. It should be noted that the maximum detected trivalent chromium concentration from these four locations of 4,830 mg/kg (from location OL-VC-20139) is less than the RSL value for trivalent chromium of 12,000 mg/kg (adjusted by 0.1 as described above) and therefore trivalent chromium is not included in this HHRA. All other chromium results were conservatively assumed to be hexavalent chromium for this HHRA. For lead, the screening values recommended by EPA of 400 mg/kg for residential was used without adjustment and was compared with an average concentration in sediments; this is consistent with EPA guidance, as the value of 400 mg/kg is not based on a hazard quotient and so no adjustment is needed (US EPA, 2003b). Table 2-2 identifies the chemicals that are identified as COPCs for the exposure scenario that evaluates volatilization from sediments, including water draining from the geotextile tubes, and inhalation of airborne site-related chemicals. The list of COPCs was generated differently than the list for sediments. All chemicals identified in sediments were considered for this scenario. Chemicals were included if they had each of the following: 1) chemicals that are considered volatile (i.e., having a molecular weight less than 200 g/mole and a Henry’s Law Constant greater than 1 x 10-5 atm-m3/mole [US EPA, 2010b]) were retained based on the categorization provided within the EPA RSL tables; 2) chemicals identified as volatile were retained for evaluation in the risk assessment if they had a toxicity value for use in risk assessment.

19

This method provides a health protective means to determine which chemicals to evaluate for the volatilization pathway. The volatilization pathway was considered in evaluating the remedial operations for Onondaga Lake sediments. Specifically, a series of three tests, including wind tunnel testing, was run to identify the chemicals that are likely to have the potential to volatilize. A detailed discussion of the testing can be found in Appendix E. Table 2-2 of Appendix B shows the identification of COPCs for the volatilization pathway. As indicated there, only some of the chemicals that were identified as COPCs were detected in the wind tunnel testing. The more comprehensive set of chemicals were included here (i.e., all chemicals in sediments that were identified as volatile and had toxicity values for use in risk assessment) in order to provide a more conservative estimate of potential risks. Polychlorinated biphenyls (PCBs) were not included in the assessment of inhalation of volatile chemicals migrating from the SCA since there were no detections of PCBs in the Phase I PDI wind tunnel air samples and based on the concentrations of PCBs in the sediments that will be dredged, PCBs are not expected to contribute significantly to the risks associated with inhalation of volatile chemicals.

3.3. Calculation of the Exposure Point Concentration

Exposure point concentrations (EPCs) for sediments were calculated using chemical analyses of sediment samples for materials that will be deposited in the SCA. The EPCs for the volatilization of chemicals from sediments were estimated through dispersion modeling. Sediment EPCs were calculated for chemicals with concentrations that exceeded their screening values in Table 2-1 of Appendix B (i.e., sediment COPCs) using ProUCL, version 4.0 (US EPA, 2007). The EPC is the 95% Upper Confidence Limit (UCL) on the arithmetic mean of a LWA chemical concentration, and provides a 95% level of confidence that the true mean will not be greater. It is based upon the distribution of the data. The ProUCL program tests the normal, lognormal, and gamma distributions of each data set and recommends the appropriate statistic using parametric and non-parametric statistical methods. If analytical data indicated a non-detect result for a chemical, a value of ½ of the detection limit was used in calculating the UCL. For chemicals with a data set that is too small to calculate this statistical upperbound average concentration, the maximum detected concentration was used. The only chemicals for which this happened were 1,2,3/4,5-tetrachlorobenzene and pentachlorobenzene. The EPCs for the sediments are shown in Table 3-1 of Appendix B, and the ProUCL outputs showing all of the statistics for each chemical, can be found in Appendix F.

For the exposure scenario that considered inhalation of airborne volatile chemicals, EPCs were estimated through dispersion modeling. As explained in Section 2.1, maximum allowable air concentrations at the work zone perimeter were identified, based on consideration of air quality values set by USEPA and NYSDEC. These criteria (DAR-1 values) and regional screening values are risk-based concentrations designed to protect against adverse human health effects which may be associated with long-term exposure to these contaminants. It should be noted that workers at the SCA facility are skilled and trained in the handling of this type of material and are protected by OSHA and by adherence to a site health and safety plan. These maximum allowable site-perimeter concentrations were fixed at the site perimeter, and an air dispersion model was then used to estimate the chemical concentrations at community receptors. A detailed

20

explanation of the modeling can be found in Appendix C, while Table 3-2 in Appendix B presents the EPCs for air.

4. Exposure Assessment

The exposure assessment evaluates pathways by which people are or can be exposed to the contaminants of concern in different media (e.g., sediment, soil, air). This exposure assessment considers only hypothetical future exposure scenarios. The quantification of exposure is based on factors including, but not limited to, the concentrations that people are or can be exposed to, the potential frequency (number of days per year), and the duration of exposure (number of years). The exposure assessment is based on the maximum site-specific parameters that can reasonably be expected at the site, which is termed the reasonable maximum exposure or RME. The goal of this HHRA is to estimate the RME expected to occur during operation of the SCA and if the SCA sediments somehow were released in the future and there would be potential for area residents to enter onto Wastebed 13, come onto or near the SCA, and come into contact with these materials. In other words, the RME is the greatest exposure that is reasonably expected to occur. As a result, the risk assessment provides upper-bound estimates of the risks and hazards for people living near the SCA facility using health-protective exposure assumptions so that these risks and hazards are not underestimated. The exposure assumptions for each receptor can be found in Tables 4-1 through 4-4 in Appendix B. Following is a description of the exposure parameters used for each receptor in this assessment.

4.1. Exposure Assumptions

Child Resident The child resident (up to 6 years old) could be exposed to contaminants in the air and in the sediments. When evaluating the scenario that considers potential exposure to sediments that will be placed in the SCA, the child resident is assumed to be coming onto Wastebed 13 and contacted sediments at or near the SCA, with exposure through dermal contact and incidental ingestion. The child is assumed to have an exposed skin surface area of 2,800 cm2, which includes head, forearms, hands, lower legs, and feet. For the soil to skin adherence factor (the factor that relates how much soil sticks to the skin and is available for absorption across the skin), a value of 0.2 mg/cm2 is used. The child is assumed to weigh 15 kg (approximately 33 pounds). The clean up of any sediment released is assumed to be 45 days. These exposure assumptions are shown in Table 4-2 of Appendix B (US EPA, 1989; US EPA, 1991a; US EPA, 1997a; US EPA, 2002a; US EPA, 2004a; US EPA, 2009). Adolescent Resident: Appendix B, Table 4-3 provides a summary of the exposure terms used to estimate exposures for adolescents ages 6-16. As indicated there, the surface area ranges from 2949 to 5386 cm2 which includes head, forearms, hands, and lower legs. A soil to skin adherence factor of 0.2 mg/cm2 is used for ages 6-12 and a factor of 0.07 mg/cm2 is used for ages 13-16. The adolescent is assumed to weigh from 22 to 58 kg. Because the exposure time is 45 days within one year the exposure estimate is a single average estimate for all of the ages from age 6 to age 16 years (Appendix B, Table 4.3) (US EPA, 1989; US EPA, 1991a; US EPA, 1997a; US EPA, 2002a; US EPA, 2004a; US EPA, 2009).

21

Adult Resident The adult resident (greater than 16 years old) could be exposed to contaminants in the air and in the sediments. When evaluating the scenario that looks at exposure to sediments after a release at the SCA, the adult resident is assumed to be on Wastebed 13, be at or near the SCA, and be exposed through dermal contact and incidental ingestion. The adult is assumed to have an exposed skin surface area of 5,700 cm2, which includes head, forearms, hands, and lower legs. For the soil to skin adherence factor value of 0.07 mg/cm2 is used. The adult is assumed to weigh 70 kg (approximately 155 pounds). The cleanup of any sediment released is assumed to be 45 days. These exposure assumptions are shown in Table 4-1 of Appendix B (US EPA, 1989; US EPA, 1991a; US EPA, 1997a; US EPA, 2002a; US EPA, 2004a; US EPA, 2009). Inhalation Exposure Inhalation exposure is evaluated differently from exposure through dermal contact and ingestion. While exposure through these two pathways requires each age-specific population to be assessed independently, the evaluation of inhalation does not. When age-specific populations are exposed through similar scenarios, such as residential scenarios, certain toxicological considerations need to be addressed when quantifying this exposure. Therefore, adults and children residents are evaluated together. As shown in Table 4-4 of Appendix B, inhalation exposure is assumed to occur over 350 days per year for 5 years, the length of time estimated for the dredging to last (US EPA, 2009). Age-Dependent Adjustment Factors (ADAFs) Certain carcinogenic chemicals are known to act through a mutagenic mode of action for carcinogenicity. This means that some ages are particularly susceptible to the carcinogenic potential of these chemicals, and this increased susceptibility must be accounted for when quantifying the risks. In this assessment no chemicals included in the inhalation exposure are known to act through this mode of action. For the sediment exposure scenario, only PAHs are considered to have this mode of action. To account for the mode of action of PAHs, Table 4-3 was created showing the age-adjusted exposure parameters used for the age groupings requiring adjustment to the cancer risk calculations. These age groups include children ages 0-2 and 2-6; adolescents ages 6-16; and adults ages 16 and up. Because the exposure period is less than one year, the exposure estimates used in this assessment are presented as an average exposure over each of these age periods (Appendix B, Table 4-3) (US EPA, 2005a; US EPA, 2005b).

4.2. Estimating Exposure

Dermal Exposure to Soil To calculate dermal exposure to soil, Exhibit 1-3 in RAGS, Part E, Supplemental Guidance for Dermal Risk Assessment (EPA, 2004a) was followed. Cancer risks and non-cancer hazards for arsenic, cadmium, dioxins (as TCDD Equivalents), PCBs, hexachlorobenzene, benzo(a)pyrene and other PAHs were calculated using the dermal absorption factors in Exhibit 3-4 of RAGS, Part E (EPA, 2004a). Incidental Ingestion of Soil The incidental ingestion pathway was assessed following the exposure model presented in Exhibit 6-14 in RAGS, PART A, Volume I Human Health Evaluation Manual (EPA, 1989). Cancer risks and non-cancer hazards were estimated for all chemicals for which toxicity information was available.

22

Inhalation of Volatiles For estimating the cancer risk and non-cancer hazard from inhalation of volatiles, Equations 10 and 11 presented in RAGS, Part F, Supplemental Guidance for Inhalation Risk Assessment were used when inhalation toxicity values were available (EPA, 2009). Mutagenic Mode of Action As stated in Section 4.1, when carcinogenic chemicals are identified as acting through a mutagenic mode of action, quantification of risk from exposure to these chemicals must address the susceptibility of certain populations of certain ages. This approach, using the Age Dependent Adjustment Factors (ADAFs), follows the Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (EPA, 2005b). In summary, a 10-fold adjustment to the toxicity of a chemical is necessary when assessing risks to children in the first 2 years of life, a 3-fold adjustment is used to account for the susceptibility of children aged 2 to less than 16, while no adjustment is needed for exposures that occur to people over the age of 16. These factors are shown in Tables 6-1 and 6-2 in Appendix B. This is discussed in more detail in Section 5.2. 5. Toxicity Assessment The toxicity assessment conservatively estimates the types of adverse health effects potentially associated with exposures to contaminants at the site and the relationship between the magnitude of exposure (dose) and severity of adverse effects (response). In December 2003, EPA’s Office of Solid Waste and Emergency Response (OSWER) issued a directive outlining the hierarchy of toxicity values to be used for risk assessment purposes. Values that come from the Integrated Risk Information System (IRIS), which represents EPA’s consensus database for cancer and non-cancer toxicity information, belong in Tier I of the hierarchy. Tier II is the Provisional Peer-Reviewed Toxicity Values (PPRTV). Tier III includes other sources of toxicity information such as California EPA, the Agency for Toxic Substances and Diseases (ATSDR), and the Health Effects Assessment Summary Table (HEAST). For this assessment, IRIS values were used when they were available. PPRTVs were used in the absence of IRIS values if they were available. All toxicity values from Tier III have been approved by the EPA Office of Research and Development, National Center for Environmental Assessment (NCEA), Superfund Technical Support Center. (US EPA, 2003a). 5.1. Health Effects Criteria for Non-Carcinogens Tables 5-1 and 5-2 in Appendix B provide data on non-cancer health effects associated with the COPCs. The toxicity values presented are the oral reference dose (RfD), the absorbed RfD for dermal exposure, and the inhalation reference concentrations (RfC). The non-cancer health endpoint (i.e., the target organ) associated with the chemical can also be found on these tables. 5.2. Health Effects Criteria for Carcinogens Tables 6-1 and 6-2 in Appendix B provide dose-response information in the form of the cancer slope factor for the ingestion, dermal contact, and inhalation routes. The weight of evidence

23

(WOE) for each chemical, which is used to characterize the extent to which the available human epidemiology and animal studies indicate that a chemical may cause cancer in humans, is also shown (US EPA, 1989). The WOE is categorized into six groups: (A) Known Human Carcinogen (B-1) Probable Human Carcinogen – based on limited evidence of carcinogenicity in humans and sufficient evidence of carcinogenicity in animals; (B-2) Probable Human Carcinogen – based on sufficient evidence of carcinogenicity in animals; (C) Possible Human Carcinogen; (D) Not classifiable as a human carcinogen; and

(E) Evidence chemical is not a carcinogen in humans. The EPA 2005 Cancer Guidelines, however, provide an update to the original 1986 Cancer Guidelines and subsequent updates. In summary, the 2005 Cancer Guidelines emphasize the value of understanding the biological changes that the chemical can cause and how these changes might lead to the development of cancer (US EPA, 2005a). They also discuss methods to evaluate and use such information, including information about an agent's postulated mode of action, or the series of steps and processes that lead to cancer formation. Mode of action data, when available and of sufficient quality, may be useful in drawing conclusions about the potency of an agent, its potential effects at low doses, whether findings in animals are relevant to humans, and which populations or life stages may be particularly susceptible. In the absence of mode-of-action information, default options are available to allow the risk assessment to proceed. The 2005 Guidelines recommend that an agent's human carcinogenic potential be described in a weight-of-evidence narrative rather than the previously identified letter categories. The narrative summarizes the full range of available evidence and describes any conditions associated with conclusions about an agent's hazard potential. The following are the five recommended standard hazard descriptors:

• carcinogenic to humans • likely to be carcinogenic to humans • suggestive evidence of carcinogenic potential • inadequate information to assess carcinogenic potential • not likely to be carcinogenic to humans

EPA is evaluating the carcinogenic weight of evidence of chemicals through the IRIS chemical process. The requirements for in-depth analysis of mode-of-action data and the review process does not allow the equating of a chemical evaluated under the old letter system classification with the 2005 Classification narrative; rather, a full analysis of the data is required. (US EPA, 2005a) The 2005 Cancer Guidelines also include Supplemental Guidance on the evaluation of early lifetime exposures. For example, where data are available on mutagenic mode of action for carcinogenesis, the Supplemental Guidance provides procedures for developing chemical-specific potency factors that account for early life susceptibility. In most cases, these data do not

24

exist and standard age-dependent adjustment factors can be applied to account for early life susceptibility. Because chemical-specific toxicity data on early life susceptibility are not available for most chemicals (vinyl chloride being the exception), cancer risks from the COPCs in this HHRA that are known to be carcinogenic by mutagenic mode of action (benzo(a)anthracene, benzo(a)pyrene, benzo(b)fluoranthene, benzo(k)fluoranthene, dibenzo(a,h)anthracene, and indeno(1,2,3-cd)pyrene) were calculated using the general age-dependent adjustment factors recommended in the Supplemental Guidance. They are: a 10-fold adjustment to the toxicity value for ages 0 – <2 years; a 3-fold adjustment to the toxicity value for ages 2 – <16 years; and no adjustment to the toxicity value for ages 16 years and older. See Section 6 for a discussion of where these adjustments are presented in the HHRA (US EPA, 2005b). 6. Risk Characterization

In risk characterization, quantitative exposure estimates and toxicity factors are combined to calculate numerical estimates of potential health risk. In this section, potential cancer and noncancer health risks are estimated assuming exposure to chemicals detected in site media. As described in Section 4, Exposure Assessment, potential risks are estimated for two hypothetical scenarios related to chemicals in sediments that will be placed in the SCA:

1. Potential future exposures through air that could occur during remediation: This scenario evaluated risks associated with inhalation of COPCs that could volatilize from sediment and water during the period of sediment management and dewatering and be transported in air to offsite residents.

2. Potential future exposures that could occur in the event of hypothetical failure of the

SCA. This scenario evaluated potential risks associated with unintentional (incidental) ingestion and skin (dermal) contact with sediments in the event of a failure of the SCA and people coming onto the SCA and into contact with sediments. Adults, adolescents, and children are evaluated in this scenario.

There are no complete exposure scenarios under current conditions since the sediment dredging activities have not started, and once these activities begin, measures will be in place to limit future exposures to acceptable levels. Thus, the hypothetical future exposure scenarios evaluated here provide a conservative means to evaluate potential risks posed by COPCs in sediment to be placed in the SCA. The risk characterization methods described in RAGS (US EPA, 1989b) were used to calculate RME excess lifetime cancer risks for carcinogens and hazard indices for contaminants with noncancer health effects. These methods and the results of the risk characterization are described below. Tables in Appendix B show detailed results of the risk calculations for each exposure pathway, including exposure point concentrations and intakes calculated for the reasonable maximum exposure scenarios, toxicity values used in risk estimates, and potential risk estimates for each COPC in each exposure pathway.

25

6.1. Risk Characterization for Carcinogens 6.1.1. Methods Quantifying total excess lifetime cancer risk requires calculating risks associated with exposure to individual carcinogens and aggregating risks associated with simultaneous exposure to multiple carcinogenic chemicals. A cancer risk estimate for a single carcinogen is calculated by multiplying the intake by its carcinogenic slope factor (CSF) for oral or dermal risk estimates or by the inhalation unit risk (IUR) for inhalation risks (US EPA, 1989; US EPA, 2009):

Cancer Risk = Intake x (CSF or IUR) A 1×10–6 cancer risk represents a one in one million additional probability that an individual may develop cancer over a 70-year lifetime as a result of exposure under the conditions and scenarios evaluated. Because cancer risks are assumed to be additive, risks associated with simultaneous exposure to more than one carcinogen in a given medium are aggregated to determine a total cancer risk for each exposure pathway. Total cancer risks for each pathway are then summed for reasonable combinations of exposure pathways to determine the total cancer risk for the population of concern. The likelihood that actual risks are greater than estimated risks is very low because of the conservative assumptions used to develop cancer risk estimates. The findings presented here are compared with the range of acceptable risk levels cited in the NCP (U.S. EPA 1990b). The NCP states that risk levels in the range of 10–4 to 10–6 and lower are considered to be within the range of acceptable risks for Superfund sites. 6.1.2. Quantification of Carcinogenic Risks Carcinogenic risk estimates were calculated for children, adolescents, and adults in the RME scenarios as the probability of additional cancers associated with the exposure pathways evaluated. Based on the exposure assumptions and toxicity values described above, Appendix B Table 8-1 provides a summary of risk estimates for all complete exposure pathways in the RME scenarios. This table also provides a summary of COPCs accounting for the majority of the risk estimates in each pathway. As described in Sections 4 and 5, carcinogenic PAHs are evaluated using four age groups (0-2 years, 2-6 years, 6-16 years, and over 16 years) so that modifying factors can be applied to account for assumed additional potency related to time of exposure. Detailed risk estimates are provided in Appendix B Tables 7-1 through 7-4 for all chemicals except PAHs and in Appendix B Supplement A Tables 7-1 through 7-4 for PAHs. In Table 8-1, risk estimates for carcinogenic chemicals other than PAHs are also shown so that a total cumulative risk can be presented for each of the four age groups considered for carcinogenic PAHs. Cancer risk estimates for an adult or child exposure to carcinogenic chemicals other than PAHs (Appendix B Tables 7-1 and 7-2) were adjusted to reflect the exposure estimates within the four age groups estimated for PAHs. This adjustment was made by taking the ratio of the exposure estimate (chronic daily intake or [CDI]) for children ages 0-2 or 2-6 over the CDI for children ages 0-6 and multiplying that product times the total cancer risk

26

estimate for chemicals other than PAHs. Cancer risk estimates for adolescents’ exposure to chemicals other than PAHs were calculated as the ratio of the adolescent CDI over that for adults. The cumulative risk estimates for all carcinogens in each of the four age groups is shown in Table 8-1 in Appendix B. As indicated in Table 8-1, no exposure pathways result in risk estimates greater than the 10-4 risk level that is the upper end of the EPA target risk range.

• Inhalation – Adults and Children: The hypothetical future inhalation cancer risk

estimates for offsite adults and children were 4×10-6 with primary contributors to risk being ethylbenzene and naphthalene based on application of the California EPA unit risks for these chemicals. The risk estimates associated with the remaining carcinogenic chemicals combined were 2 ×10-6 (Appendix B Table 7-3).

• Contact with Sediments – Adults: The hypothetical future cancer risk estimates

associated with exposure to sediments (oral and dermal exposure routes) was 1 ×10-6 for all carcinogenic chemicals combined (Appendix B Table 8-1).

• Contact with Sediments – Children: The hypothetical future cancer risk estimate

associated with exposure to sediments (oral and dermal routes) was 3×10-5 for children ages zero to 2 and 1 ×10-5 for children ages 2-6 (Appendix B Table 8-1), with the primary contributors to risk being hexavalent chromium and the carcinogenic PAHs .

• Contact with Sediments – Adolescents: The hypothetical future cancer risk estimate

associated with exposure to sediments (oral and dermal routes) was 3×10-6 for adolescents ages 6 to 16 (Appendix B Table 8-1) also associated primarily with hexavalent chromium and the carcinogenic PAHs.

6.2. Quantification of Hazard Indices for Effects other than Cancer 6.2.1. Methods Unlike carcinogenic effects, other potential adverse health effects are not expressed as a probability. Instead, these effects are expressed as the ratio of the estimated exposure over a specified time period to the RfD or RfC derived for a similar exposure period (e.g., Chronic Daily Intake:chronic RfD or RfC). This ratio is termed a hazard quotient (US EPA, 1989; US EPA, 2009):

HQ = Intake ÷ (RfD or RfC) If the Chronic Daily Intake, or CDI, exceeds the RfD or RfC (i.e., the hazard quotient is greater than 1), there may be concern for noncancer adverse health effects, and as this quotient increase, the potential for noncancer health effects increases. Exposures resulting in a hazard quotient less than or equal to 1 are very unlikely to result in noncancer adverse health effects.

27

In initial risk calculations, hazard quotients for individual COPCs are summed for each exposure pathway to derive a hazard index. Hazard indices for each exposure pathway are then summed to determine the total hazard index for each population of concern. In the event a total hazard index exceeds 1, the hazard index is segregated by primary target organs because adding hazard quotients of compounds that do not affect the same target organ could overestimate the potential for adverse effects. Consistent with the latest RAGS guidance (U.S. EPA 1998b), hazard quotients for individual chemicals that share the same critical effect or primary target organ as reported in IRIS (U.S. EPA 2010c) or other resources used to derive toxicity values are summed across exposure pathways to determine a total hazard index for that target organ. If the hazard index for a particular target organ exceeds 1, the hazard index for each target organ can be further evaluated by identifying the mode of action on the target organ. In this step, separate hazard indices for each mode of action for each target organ are calculated.

6.2.2. Quantification of Noncarcinogenic Risks Detailed risk estimates are provided in Appendix B Tables 7-1 through 7-3 and Appendix B, Table 8-2 (inhalation exposure) and Table 8-3 (sediment exposure) show the hazard indices (HI) for COPCs summed based on the primary target organs (or systems) that they affect. Tables 8-2 and 8-3 also show the primary chemicals contributing to the total hazard index for each pathway-endpoint combination. For adult exposure to sediments (see Table 7-2), none of the scenarios had hazard quotients or combined hazard indices greater than 1.0. Although the total HI (approximately 2) for child exposure to sediments (ingestion and dermal contact) slightly exceeds 1 (as shown in Table 7-1), the HI values for each endpoint do not exceed 1 (as shown in Table 8-3). In addition, although the total HI (approximately 4) for the inhalation exposure exceeds 1 (as shown in Table 7-3), the HI values for all receptors for each target organ do not exceed 1 (as shown in Table 8-2). 6.3. Off-Site Workers As stated previously, off-site workers are not quantitatively evaluated in this HHRA. Exposure from inhalation of volatile chemicals that may migrate from the SCA are quantitatively evaluated for the nearby residents, assuming a standard residential exposure typically used in risk assessments. This exposure scenario assumes that residents are exposed for 350 days per year for the duration of the project, 5 years. Off-site workers who might be working at Wastebed 15 or in other nearby areas would be exposed less frequently, up to 250 days per year, which assumes 5 days per week for 50 weeks during the year. Under this scenario, the off-site worker would have a lower exposure and therefore the cancer risks and noncancer hazards would be proportionally less. Since the risks estimated for the residents fall within the acceptable risk range, the risks and hazards to the off-site worker are also in this range. 6.4. Uncertainty Assessment Key uncertainties in the risk assessment should be considered in order to better place the risk estimates within context. Estimates provided here are hypothetical and were derived as a means to evaluate plans for the SCA and better limit potential future risks. However, many of the

28

assumptions applied here tend to overestimate potential future site risks. These include the following:

6.4.1. Hexavalent Chromium. Chromium can exist is several forms, and two forms are most commonly considered in an HHRA, the trivalent form and the hexavalent form. Results for samples collected from 4 locations near the Crucible Lake Pump Station site that were analyzed for both total and hexavalent chromium did not identify any hexavalent chromium; these results were therefore considered to be in the trivalent form, which is significantly less harmful, and were ultimately screened out and not considered further in this HHRA. All other chromium results were analyzed only for total chromium, and were conservatively considered to be in the hexavalent form, which is more toxic. Under this conservative assumption, hexavalent chromium risk estimates were responsible for approximately 90% of the cancer risk estimates for ingestion of sediments. This is likely an overestimate of the true risk, since hexavalent chromium is reasonably assumed only to be a small percentage of the total chromium in the sediments, and the limited data which are available indicate that the chromium is not present as hexavalent chromium. In the unlikely event of a failure of the SCA containment and a release of the sediments, measures would be implemented to address the release in accordance with site management plans (e.g., Spill Contingency Plan) to be developed. The measures may include additional sampling and characterization, including speciating potential risk-driving chemicals such as chromium.

6.4.2. Methyl mercury. Mercury is another chemical that can exist in several forms. However, these forms were not analyzed in most sediment samples, and mercury was reported as total mercury. For the hypothetical sediment exposure pathway in this HHRA, all mercury risk estimates were calculated assuming mercury was present as methyl mercury in sediments, which is the most conservative approach when assessing oral and dermal exposures. As stated previously, existing data indicate that the maximum percentage of methyl mercury is 1.4%. Nevertheless, all hazard indices for mercury for both the sediment and the inhalation exposure pathways were within acceptable levels. As stated in Section 3.2, this HHRA assumed conservatively that all mercury is present in the more toxic form. In the unlikely event of a release of sediments, measures would be implemented to address the release in accordance with site management plans (e.g., Spill Contingency Plan) to be developed. The measures may include additional sampling and characterization, including speciating potential risk-driving chemicals such as mercury. 6.4.3. Cobalt. One exposure scenario estimates potential risk from direct contact from sediments in the hypothetical event of a failure of the SCA and individuals come into contact with the sediments. Under this scenario, cobalt contributes approximately 19% to the overall hazard index to the child from exposure to sediments (the total noncancer hazard index to the adult are less than 1, so no further evaluation is necessary). The sediment data set for cobalt consists of 133 samples collected from 37 locations within the area to be dredged. A length weighted average (LWA) was calculated for each location and these 37 LWAs were used to calculate the exposure point concentration for cobalt that was used in the risk assessment. Of these 37 locations, two locations near the Crucible Lake Pump Station site have significantly higher LWA cobalt concentrations than all other locations in the area to be dredged and strongly influence the overall

29

exposure point concentration. These two locations represent less than 2% of the overall volume of sediment to be dredged and are not representative of the cobalt concentrations in the material that will be dredged. Therefore, the noncancer hazard index for cobalt is likely overestimated. 6.4.4. Toxicity Values. Significant uncertainty may be associated with the derivation of RfDs and CSFs. Toxicity values based on human epidemiological studies are not available for most chemicals, and those human studies that are available generally lack exposure data and are confounded by exposure to multiple chemicals, recall bias, and lifestyle issues. Laboratory animal studies are used to derive most toxicity values and the practice of extrapolating from effects in animals to predict human toxic response is a major source of uncertainty in risk assessment. RfD development is a health-protective and conservative process, which uses a No Observable Adverse Effect Level (NOAEL) or a Lowest Observable Adverse Effect Level (LOAEL) from an animal study, divided by a series of 3- or 10-fold Uncertainty Factors (UFs). The UFs are intended to account for differences between humans and laboratory animals, variation in sensitivity within the human population, differences between subchronic and chronic exposures, use of a LOAEL versus a NOAEL, and the strength of the toxicology database for a particular chemical. The combination of several UFs results in RfDs that are several orders of magnitude lower than the doses that produce minimal or no effects in animals. CSFs may also highly conservative and contain multiple sources of uncertainty, including the methods of extrapolation from high doses to low doses and from animals to humans. In addition, genetic constitution, diet, occupational and home environments, activity patterns, and other cultural factors influence human susceptibility to cancer. To compensate for this uncertainty, CSFs generally represent the 95% UCL on the probability of a carcinogenic response at a certain dose rate over a lifetime. Many chemicals do not have peer-reviewed toxicity values available. For example, many of the PAHs do not have RfDs available to assess non-cancer health hazards. Many Class C carcinogens do not have SFs derived. This lack of toxicity information underestimates the actual risk. 6.4.5. Potential for Overestimation within Exposure Scenarios. The SCA is planned to be closely managed and maintained. Consequently, the assumed potential for exposure to sediments is hypothetical and may overestimate risks, particularly for young children (due to their increased susceptibility from exposure to chemicals acting through a mutagenic mode of action) who would be unlikely to come into contact with this material. All risk estimates for sediments for adults are within 1×10-6. 6.4.6. Exposure Assumptions. There is considerable uncertainty regarding the likelihood of exposure to a given medium of concern. It is unknown whether all of the exposure pathways modeled will ever be actually complete or whether the individuals evaluated will actually be exposed to COPCs. For example, for the sediment exposure scenario to be complete, children, adolescents, and adults must enter the site and gain

30

access to sediments released from the SCA. Although this is theoretically possible, there is no evidence that such activities would actually take place.

Exposure estimates used to calculate risks and hazards may also be relatively uncertain. Many of the exposure parameter values applied are default values determined by USEPA (1989, 1991, 1997a), or rely on professional judgment rather than site-specific values. As such, risk estimates based on these exposure parameters generally represent conservative estimates. In particular, the RME scenario relies heavily on guidance documents that may not have the most current and accurate information on exposures. For example, exposure to sediments released from the SCA is assumed to occur every day for the 45 day period it would take for the sediments to be remediated, and assuming any exposure at the SCA is consistent with the types of exposures typical in residential scenarios. Due to the distance from the SCA to the residences (the nearest residence is approximately 1500 feet from the SCA) and the likelihood of interim measures such as additional fencing, cover material, or other engineering controls to be implemented, this type of exposure is not anticipated. As a result, risks and hazards predicted under the RME scenario may potentially overestimate risks and hazards at the site. In the unlikely event of a failure of the SCA and a release of sediments, it is recommended that the actual exposures that would occur, once the interim controls are implemented, be evaluated to ensure that they are not greater than the exposures in the scenarios defined in this HHRA. 6.4.7. Inhalation of Volatilized Chemicals. The air concentrations used as a starting point in deriving offsite air estimates are all assumed to be the lower of either the DAR-1 number or the industrial RSLs after adjustment for a 5-year exposure duration for all chemicals identified in the sediments as volatile. Both bench scale testing of sediments and testing of volatilization from sediments suggest that instead approximately half of the chemicals would actually be volatilized from sediments. This assumption results in a cumulative air exposure risk estimate that is higher than what would likely occur.

6.4.8. Application of the Highest Receptor Location in Air Estimates. Workplace perimeter concentrations were used to estimate offsite air concentrations. The resulting estimates were evaluated, and the highest yearly average concentration was applied. The selection of this maximum value means that all other yearly average concentrations would be lower and thus exposure estimates for most people would be lower than this value. 6.4.9. Tentatively Identified Compounds (TICs). Standard analytical protocols for Superfund sites require that an extensive list of chemicals be analyzed and reported. This list is known as the Target Compound List (TCL)/Target Analyte List (TAL) for organic chemicals and metals, respectively. The list was developed with consideration of the chemicals most commonly found at sites. Other chemicals that are not routinely analyzed for or reported might be present at sites. These chemicals are known as tentatively identified compounds (TICs), since neither the identity nor the concentration can be reported with certainty. The TICs could present a contribution to risk that has not been quantified, however, they typically do not contribute significantly to the overall risk or hazard at a site. The extensive database of more than 33,000 samples collected at the

31

site provides a robust data set that supports the basis for the risks identified in this assessment.

7. Conclusions This supplemental risk assessment applied conservative exposure assumptions to evaluate potential risks associated with the establishment of the SCA and to address community concerns. Estimates were designed to represent two hypothetical future scenarios: 1) exposure to contaminants that could migrate from the site in air during operation of the SCA and 2) exposure to sediments within the SCA post-closure if the SCA were to fail, sediments were released and people would come onto Wastebed 13 and contact the sediments on or near the SCA. Both of these potential future scenarios were intended to represent the reasonable maximum exposure potential and both assume individuals of all ages could be exposed. As such, these risk estimates are likely higher than risks that would likely be experienced by most receptors. All resulting risk estimates and hazard indices were within levels identified by EPA as acceptable. The finding of acceptable risk estimates through application of these health protective assumptions, indicates that the plans for the SCA will not result in unacceptable risks for the surrounding community. Nevertheless, the SCA will be closely monitored to ensure that sediments are managed with care and secured appropriately and that offsite migration of chemicals in air is limited or prevented. This HHRA can also be used as a tool for risk managers during the implementation of the remedy and management of the SCA. The exposure scenarios evaluated in this HHRA are future hypothetical situations, so the outcome can be used to help assess the effectiveness of the remedy. For example, the air concentrations modeled in the residential areas are based on work zone perimeter concentrations that reflect the maximum annual average concentrations for all chemicals. It is highly unlikely that every volatile chemical would be present at that concentration for a one year period of time to result in that exposure scenario. During remedy implementation, if monitored air concentrations indicate a trend towards chemicals reaching this maximum annual average concentration for a sustained period of time, risk managers can modify site operations to reduce these concentrations so that the actual risks are much lower than those estimated here. Risk managers can also use this HHRA to assist in managing risks in the unlikely event of a failure of the SCA. If sediments are released as a result of a failure of the SCA, measures would be implemented to address the release in accordance with site management plans (e.g.. Spill Contingency Plan) to be developed. The measures may include additional sampling and characterization, for example, speciating potential risk-driving chemicals such as mercury and chromium, to ensure that the actual exposures that would occur once the interim controls are implemented are not greater than the exposures in the scenarios defined in this HHRA.

32

33

8. References ATSDR (Agency for Toxic Substances and Disease Registry). 2000. Toxicological Profile for Polychlorinated Biphenyls (PCBs). U.S. Department of Health and Human Services. Public Health Service. Dourson, M.L., and J.F. Stara. 1983. Regulatory history and experimental support of uncertainty (safety) factors. Reg. Toxicol. Pharmacol. 3:224-238. Environmental Data Services, Inc. 2008. Data Usability Summary Report, Onondaga Lake, Syracuse, NY, for Chemtech Sample Delivery Group Z3768, Hexavalent Chromium. Prepared for Earth Tech under contract to NYSDEC. August 28, 2008. New York State Department of Environmental Conservation. 2007. DAR-1 AGC/SGC Tables. September 10, 2007. New York State Department of Environmental Conservation, New York State Department of Health, and United States Environmental Protection Agency Region 2. 2010, Frequently Asked Questions (FAQs), Onondaga Lake Dredging Project Sediment Consolidation Area (SCA) at Wastebed 13. April. New York State Department of Environmental Conservation and United States Environmental Protection Agency Region 2. 2005. Record of Decision. Onondaga Lake Bottom Subsite of the Onondaga Lake Superfund Site. July. New York State, State Environmental Board. 2006. 6 NYCRR Subparts 375-1 through 375-4 and Subpart 375-6. Available at http://www.dec.ny.gov/chemical/34189.html. Accessed on November 21, 2007. Parsons. 2006. Onondaga Lake Sediment Consolidation Area (SCA) Siting Evaluation. Prepared for Honeywell, Syracuse, New York. September. Parsons. 2009. SCA Dewatering Evaluation Report. Prepared for Honeywell, Morristown, NJ, February. Parsons and Geosyntec. 2009. Draft Onondaga Lake Sediment Consolidation Area Civil and Geotechnical Initial Design Submittal. August. TAMS. 2002a. Onondaga Lake Human Health Risk Assessment. Original document prepared by Exponent, Bellevue, Washington, for Honeywell, East Syracuse, New York. Revision prepared by TAMS, New York, New York and YEC, Valley Cottage, New York, for New York State Department of Environmental Conservation, Albany, New York. December. TAMS. 2002b. Onondaga Lake Baseline Ecological Risk Assessment. Original document prepared by Exponent, Bellevue, Washington, for Honeywell, East Syracuse, New York. Revision prepared by TAMS, New York, New York and YEC, Valley Cottage, New York, for New York State Department of Environmental Conservation, Albany, New York. December.

34

TAMS. 2002c. Onondaga Lake Remedial Investigation. Original document prepared by Exponent, Bellevue, Washington, for Honeywell, East Syracuse, New York. Revision prepared by TAMS, New York, New York and YEC, Valley Cottage, New York, for New York State Department of Environmental Conservation, Albany, New York. December. United States District Court, Northern District of New York. 2007. State of New York and Denise M. Sheehan against Honeywell International, Inc. Consent Decree Between the State of New York and Honeywell International, Inc. Senior Judge Scullin. Dated October 11, 2006. File January 4, 2007. US EPA. 1989. Risk Assessment Guidance for Superfund (RAGS), Volume 1, Human Health Evaluation Manual, Part A. US Environmental Protection Agency, Office of Emergency and Remedial Response, Washington, DC. EPA/540/1-89/002. US EPA. 1990. National Priorities List for Uncontrolled Hazardous Waste Sites. 55 Federal Register Vol 55, No. 50. March 1990. US EPA. 1991a. Human Health Evaluation Manual, Supplemental Guidance: “Standard Default Exposure Factors.” OSWER Directive 9285.6-03. US EPA. 1991b. RAGS Volume 1, Human Health Evaluation Manual, Part B, Development of Risk-based Preliminary Remediation Goals. US EPA, OSWER, Washington, DC. 9285.7-01B. US EPA. 1991c. Role of Baseline Risk Assessment in Superfund Remedy Selection Decisions. OSWER Directive 9355.0-30. USEPA. 1992a. Final Guidelines for Exposure Assessment: Notice. 57 Federal Register 104:22888-22938. U.S. Environmental Protection Agency, Washington, DC. May 29. USEPA. 1992b. Supplemental Guidance to RAGS: Calculating the Concentration Term. OSWER 9285.7-081. U.S. Environmental Protection Agency. May. USEPA. 1993. Provisional Guidelines for Quantitative Risk Assessment of Polycyclic Aromatic Hydrocarbons. USEPA/600/R-93/089. U.S. Environmental Protection Agency, Office of Research and Development, Washington, DC. July. USEPA. 1995a. Policy for Risk Characterization at the U.S. Environmental Protection Agency. U.S. Environmental Protection Agency, Office of the Administrator, Washington, DC. March. US EPA. 1995b. Land Use in the CERCLA Remedy Selection Process. OSWER Directive 9355.7-04. US EPA. 1997a. Exposure Factors Handbook: Volume I (General Factors) & III (Activity Factors). EPA/600/P-95/002Fa and c.

35

USEPA. 1997b. Health Effects Assessment Summary Tables (HEAST). FY 1997 Update. USEPA-540-R-97-036. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response. July. US EPA. 1998. Clarification to the 1994 Revised Interim Soil Lead Guidance for CERCLA sites and RCR Corrective Action Facilities. OSWER Directive 9200.4-27P. US EPA. 2001. RAGS Volume 1, Human Health Evaluation Manual, Part D, Standardized Planning, Reporting, and Review of Superfund Risk Assessments. US EPA, OSWER, Washington, DC. OSWER 9285.7-47. US EPA 2002a. Child-Specific Exposure Factors Handbook (Interim Report) 2002. U.S. Environmental Protection Agency, Office of Research and Development, National Center for Environmental Assessment, Washington Office, Washington, DC, EPA-600-P-00-002B, 2002 US EPA. 2002b. Supplemental Guidance for Developing Soil Screening Levels for Superfund Sites. OSWER 9355.4-24. US EPA. 2002c. Calculating the Upper Confidences Limits for Exposure Point Concentrations at Hazardous Waste Sites. OSWER 9285.6-10. US EPA. 2003a. Human Health Toxicity Values in Superfund Risk Assessments. OSWER Directive 9285.7-53. US EPA. 2003b. Recommendations of the Technical Review Workgroup for Lead for and Approach to Assessing Risks Associated with Adult Exposure to Lead in Soil. EPA 540-R-03-001. US EPA. 2004a. RAGS Volume 1, Human Health Evaluation Manual, Part E, Supplemental Guidance for Dermal Risk Assessment. US EPA, OSWER, Washington, DC. OSWER 9285.7-02EP. US EPA. 2004b. Region 9 Preliminary Remediation Goals. http://www.epa.gov/region09/waste/sfund/prg/index.htm US EPA. 2004c. Integrated Risk Information System. www.epa.gov/iris. US EPA. March 2005a. Guidelines for Carcinogen Risk Assessment. EPA/630/P-03/001F. US EPA. March 2005b. Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens. US EPA. April 2007. ProUCL User’s Guide Version 4.0. Prepared for US EPA by Lockheed Martin. EPA/600/R-07/038.

36

US EPA. January 2009. RAGS Volume 1, Human Health Evaluation Manual, Part F, Supplemental Guidance for Inhalation Risk Assessment. US EPA, OSWER, Washington, DC. OSWER 9285.7-82. US EPA. May 2010a. Regional Screening Levels. Generic tables. http://www.epa.gov/reg3hwmd/risk/human/rb-concentration_table/Generic_Tables/index.htm US EPA. May 2010b. Regional Screening Levels. Users Guide. http://www.epa.gov/reg3hwmd/risk/human/rb-concentration_table/usersguide. US EPA. May 2010c. Integrated Risk Information System (IRIS). http://www.epa.gov/iris/. Last updated May 21, 2010. Accessed May 25, 2010.

Appendix A

Figures

Wastebed 13

Wastebed 11 Wastebeds

9-10

Wastebed14

Wastebed 12

Wastebed 15

NFIGURE 1

Sediment Consolidation AreaOnondaga Lake Bottom Subsite

R di l D i

LATITUDE: N 43° 5’ 57”LONGITUDE: W 76° 10’ 41”

Syracuse

SITE LOCATION MAPNew YorkQuadrangle

SOURCE: U.S.G.S. SYRACUSE WEST QUADRANGLE

Remedial Design

Dredge Area A

Dredge Area B

Dredge Area C

Dredge Area D

Dredge Area E

Figure 4. Sample LocationsDredge Area Length-Weighted Averages

LegendSample LocationsDredge Area (RA-A)Dredge Area (RA-B)Dredge Area (RA-C)Dredge Area (RA-D)Dredge Area (RA-E)

0 10.5Miles

Note: Not all parameters were analyzed at each location.

SCA Perimeter

Work zone Perimeter

Site Perimeter fence

Golden Golden MeadowsMeadows

Site Perimeter fence

~200 ft BufferAirpo

rt Road

Treatment ~500 ft Buffer

Treatment Plant And Process Areas

Active Town of Camillus C&D

Landfill

Greenfield Greenfield VillageVillage

Thomas/ Thomas/ James AveJames Ave

gg

West ColonyWest Colony SCA BUFFER ZONES AND 

FIGURE 2

Not to ScaleNot to Scale

Starlight Starlight EstatesEstates

West Colony West Colony PointPoint SURROUNDING 

COMMUNITIES

tu695

tu5

SCA

AREA

Wastebeds 1-8

Nine Mile Creek

Interbed

Wastebed 11

Wastebeds 9 & 10

Retention Ponds

Wastebeds 12-15

GeddesBrook

WATER TREATMENT PLANT

AND PROCESS AREAS

GOLDEN MEADOWS

STARLIGHT ESTATES

GREENFIELD VILLAGE

WEST COLONY POINT

THOMAS/

JAMES AVE

¥FIGURE 3

0 1,500 3,000750

Feet

This document was developed in color. Reproduction in B/W may not represent the data as intended. Aerial Image: April 2009, NYS GIS Clearinghouse

LEGEND

NEAREST RESIDENTIAL RECEPTOR BOUNDARY

WORK ZONE PERIMETER

RECEPTOR COMMUNITY

DREDGE TRANSPORT PIPELINE

SCA AREA

WATER TREATMENT PROCESS AREAS

DISPERSION MODELING

ANALYSIS BOUNDARIES

Appendix B

RAGS Part D Tables

Table 1. Selection of exposure pathwaysOnondaga Lake sediment containment area (SCA)

Scenario Medium Exposure Exposure Receptor Receptor Exposure Type of Rationale for Selection or Exclusion

Timeframe Medium Point Population Age Route Analysis of Exposure Pathway

Future SCA Sediments Surface "Soil" Surface "Soil" Resident Adult Ingestion Quantitative In the event of a release, residents may be exposed to contaminants.

Dermal Quantitative In the event of a release, residents may be exposed to contaminants.

Inhalation None Inahalation exposure is expected to be minimal relative to ingestion and dermal.

Future SCA Sediments Surface "Soil" Surface "Soil" Resident Adolescent Ingestion Quantitative In the event of a release, residents may be exposed to contaminants.

Age 6- 16 Dermal Quantitative In the event of a release, residents may be exposed to contaminants.

Inhalation None Inahalation exposure is expected to be minimal relative to ingestion and dermal.

Future SCA Sediments Surface "Soil" Surface "Soil" Resident Child Ingestion Quantitative In the event of a release, residents may be exposed to contaminants.

Age 0 - 6 Dermal Quantitative In the event of a release, residents may be exposed to contaminants.

Inhalation None Inahalation exposure is expected to be minimal relative to ingestion and dermal.

Future Outdoor Air Outdoor air Offsite areas Resident Adult Inhalation Quantitative Residents may be exposed to volatiles relaased from the SCA.

Future Outdoor Air Outdoor air Offsite areas Worker Adult Inhalation Qualitative Workers may be exposed to volatiles relaased from the SCA.

Future Outdoor Air Outdoor air Offsite areas Resident Adolescent Age 6 - 16 Inhalation Quantitative Residents may be exposed to volatiles

relaased from the SCA.

Future Outdoor Air Outdoor air Offsite areas Resident Child Age 0 - 6 Inhalation Quantitative Residents may be exposed to volatiles

relaased from the SCA.

Page 1 of 32

Table 2-1. Occurrence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SedimentsExposure Medium: SedimentsExposure Point: Sediments in SCA

Part 375 List Parameter

Minimum Detected

Value

Maximum Detected

Value

Concen-tration Units

Location of Maximum

ConcentrationDetection Frequency

Frequency of

detection

Frequency of

detection >5%

Tentatively Identified

Compound Volatile?Range of

Detection Limits

Concentration Used for

ScreeningCoPC Flag

Rationale\ Substance Deletion or Selection

INORGANIC ANALYTESAluminum 551 17200 mg/kg S322 133/133 100% Yes No No - 17000 7700 N Yes ASLAntimony 0.25 15.4 mg/kg P53 89/133 67% Yes No No 0.24 - 2.5 15.0 3.1 N Yes ASL

Yes Arsenic 0.23 33.6 mg/kg S337 128/133 96% Yes 0.193 - 0.42 34.0 0.39 C Yes ASLYes Barium 20.1 22600 mg/kg S341 133/133 100% Yes - 23000 1500 N Yes ASLYes Beryllium 0.037 0.93 mg/kg S322 116/133 87% Yes 0.031 - 0.54 0.93 16 N No BSLYes Cadmium 0.077 36 mg/kg P23 122/150 81% Yes 0.036 - 1.1 36.0 7.0 N Yes ASLYes Chromiumc 2.1 6310 mg/kg S337 144/144 100% Yes - 6300 0.29 C d Yes ASL

Cobalt 0.29 179 mg/kg OL-VC-20139 129/133 97% Yes No No 0.18 - 0.63 180 2.3 N Yes ASLYes Copper 1.6 753 mg/kg S337 150/150 100% Yes - 750 310 N Yes ASLYes Cyanide 0.94 28 mg/kg S309 46/109 42% Yes 0.63 - 1.6 28.0 160 N No BSLYes Lead 0.54 4390 mg/kg S337 150/150 100% Yes - 4400 400 N e Yes ASL

Magnesium 2530 39800 mg/kg S345 150/150 100% Yes No No - 40000 N/A No NUTYes Manganese 83.8 1790 mg/kg OL-VC-20139 133/133 100% Yes - 1800 180 N f Yes ASLYes Mercury 0.0086 163 mg/kg OL-VC-70134 808/966 84% Yes 0.0048 - 0.5 160 0.78 N g Yes ASLYes Nickel 1.75 2090 mg/kg OL-VC-20139 149/150 99% Yes 2.2 - 2.2 2100 150 N h Yes ASLYes Selenium 0.53 5.9 mg/kg S337 81/133 61% Yes 0.16 - 1.2 5.9 39 N No BSLYes Silver 0.11 6.5 mg/kg S337 81/133 61% Yes 0.098 - 1.1 6.5 39 N No BSL

Thallium 0.313533 2.9 mg/kg S342 22/133 17% Yes No No 0.16 - 1.8 2.9 N/A No NTXV di 0 858 279 /k OL VC 20139 129/133 97% Y N N 0 76 1 3 280 0 55 N Y ASL

Screening Toxicity Values b

Vanadium 0.858 279 mg/kg OL-VC-20139 129/133 97% Yes No No 0.76 - 1.3 280 0.55 N Yes ASLYes Zinc 5.2 710 mg/kg S337 150/150 100% Yes - 710 2300 N No BSL

PESTICIDES/PCDDFS/PCBSYes 2-Butanone 0.002 0.73 mg/kg OL-STA-20008-VC 44/116 38% Yes 0.011 - 41 0.73 2780 N No BSLYes Aldrin 0.00438 0.00506 mg/kg S342 2/89 2% No 0.001 - 0.0496 0.0051 0.029 C No IFDYes Alpha-BHC 0.00138 0.013 mg/kg S313 22/97 23% Yes 0.001 - 0.0061 0.013 0.08 C No BSLYes Alpha-Chlordane 0.00112 0.0148 mg/kg S352 17/105 16% Yes 0.001 - 0.0061 0.015 1.6 C i No BSLYes Beta-BHC 0.00105 0.11 mg/kg P1 16/89 18% Yes 0.001 - 0.0049 0.11 0.27 C No BSL

Constituents of Chlordane (alp 0.00108 0.0504 mg/kg S314 31/108 29% Yes No No 0.001 - 0.0061 0.050 N/A No NTXYes DDD 0.0013 0.016 mg/kg S312 8/94 9% Yes 0.001 - 0.012 0.016 2.0 C No BSLYes DDE 0.00108 0.034 mg/kg S314 35/92 38% Yes 0.001 - 0.012 0.034 1.4 C No BSLYes DDT 0.0013 0.0883 mg/kg S313 38/82 46% Yes 0.001 - 0.012 0.088 1.7 C No BSLYes Delta-BHC 0.00119 0.00484 mg/kg S309 15/51 29% Yes 0.001 - 0.0061 0.0048 0.08 C j No BSLYes Dibenzofuran 0.068 81 mg/kg S313 54/127 43% Yes 0.033 - 6.1 81 7.8 N Yes ASLYes Dieldrin 0.00105 0.0412 mg/kg S325 34/102 33% Yes 0.001 - 0.012 0.041 0.030 C Yes ASLYes Endosulfan I 0.00113 0.00905 mg/kg S309 16/104 15% Yes 0.001 - 0.0061 0.0091 37 N k No BSLYes Endosulfan II 0.00157 0.00223 mg/kg S312 2/96 2% No 0.001 - 0.012 0.0022 37 N k No IFDYes Endosulfan Sulfate 0.00111 0.0314 mg/kg S313 14/75 19% Yes 0.001 - 0.012 0.031 37 N k No BSLYes Endrin 0.00112 0.0084 mg/kg S313 11/96 11% Yes 0.001 - 0.012 0.0084 1.8 N No BSL

Endrin aldehyde 0.00177 0.00177 mg/kg S318 1/97 1% No No No 0.001 - 0.012 0.0018 N/A No IFDEndrin ketone 0.00115 0.049 mg/kg S313 6/98 6% Yes No No 0.001 - 0.012 0.049 1.8 N No BSL

Page 2 of 32

Table 2-1. Occurrence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SedimentsExposure Medium: SedimentsExposure Point: Sediments in SCA

Part 375 List Parameter

Minimum Detected

Value

Maximum Detected

Value

Concen-tration Units

Location of Maximum

ConcentrationDetection Frequency

Frequency of

detection

Frequency of

detection >5%

Tentatively Identified

Compound Volatile?Range of

Detection Limits

Concentration Used for

ScreeningCoPC Flag

Rationale\ Substance Deletion or Selection

Screening Toxicity Values b

Yes Gamma-BHC (Lindane) 0.00699 0.0186 mg/kg S344 2/96 2% No 0.001 - 0.0061 0.019 0.52 C No IFDYes Heptachlor 0.00251 0.0183 mg/kg S309 6/108 6% Yes 0.001 - 0.0061 0.018 0.11 C No BSL

Heptachlor epoxide 0.00126 0.0519 mg/kg S314 24/97 25% Yes No No 0.001 - 0.0061 0.052 0.053 C* No BSLMethoxychlor 0.0011 0.0131 mg/kg S309 15/92 16% Yes No No 0.001 - 0.061 0.013 31 N No BSLToxaphene – – mg/kg – 0/105 0% No No No 0.001 - 0.61 ND 0.44 C* No IFD

ata Dioxins (as TCDD equivalents 6.0E-07 0.00023 mg/kg S346 8/8 100% Yes No No - 0.00023 4.5E-06 C Yes ASLYes PCBs 0.0052 23 mg/kg OL-VC-20135 315/792 40% Yes 0.0041 - 0.51 23 0.11 N l Yes ASL

ORGANIC ANALYTESYes 1,1,1-Trichloroethane 0.0027 0.017 mg/kg S313 2/148 1% No 0.003 - 42 0.017 874 N No IFD

1,1,2,2-Tetrachloroethane – – mg/kg – 0/148 0% No No Yes 0.003 - 42 ND 0.56 C* No IFD1,1,2-Trichloroethane – – mg/kg – 0/148 0% No No Yes 0.003 - 42 ND 1.1 C* No IFD

Yes 1,1-Dichloroethane 0.0027 0.0027 mg/kg S313 1/148 1% No 0.003 - 42 0.0027 3.3 C No IFDYes 1,1-Dichloroethene – – mg/kg – 0/148 0% No 0.003 - 42 ND 24 N No IFD

1,2,3,4-Tetrachlorobenzene 0.0028 30 mg/kg P3 17/17 100% Yes No -- - 30 N/A No NTX1,2,3-Trichlorobenzene 0.0037 240 mg/kg P15 35/851 4% No No Yes 0.00068 - 1400 240 4.9 N No IFD1,2,4-Trichlorobenzene 0.0016 440 mg/kg OL-VC-10050 182/979 19% Yes No Yes 0.00068 - 1400 440 6.2 N Yes ASL

Yes 1,2,4-Trimethylbenzene 0.017 58 mg/kg S309 10/10 100% Yes - 58 6.2 N Yes ASL1,2-Dibromo-3-chloropropane – – mg/kg – 0/22 0% No No Yes 0.11 - 21 ND 0.0054 C* No IFD1,2-Dibromoethane – – mg/kg – 0/22 0% No No Yes 0.054 - 10 ND 0.034 C* No IFD

Y 1 2 Di hl b 0 00018 4200 /k OL VC 10097 496/979 51% Y 0 0059 59 4200 191 N Y ASLYes 1,2-Dichlorobenzene 0.00018 4200 mg/kg OL-VC-10097 496/979 51% Yes 0.0059 - 59 4200 191 N Yes ASLYes 1,2-Dichloroethane – – mg/kg – 0/148 0% No 0.003 - 42 ND 0.43 C No IFD

1,2-Dichloroethene (total) – – mg/kg – 0/18 0% No No Yes 0.011 - 5.3 ND 70 N No IFD1,2-Dichloropropane 0.012 0.012 mg/kg S313 1/147 1% No No Yes 0.003 - 42 0.012 0.89 C* No IFD1,3,5-Trichlorobenzene 0.00045 1100 mg/kg P15 73/848 9% Yes No -- 0.00068 - 1500 1100 N/A No NTX

Yes 1,3,5-Trimethylbenzene 0.016 29 mg/kg S309 6/6 100% Yes - 29 78 N No BSLYes 1,3-Dichlorobenzene 0.00021 130 mg/kg OL-VC-10102 161/978 16% Yes 0.0059 - 1500 130 191 N m No BSLYes 1,4-Dichlorobenzene 0.00062 12000 mg/kg OL-VC-10097 554/979 57% Yes 0.0059 - 59 12000 2.4 C Yes ASL

2,4,5-Trichlorophenol – – mg/kg – 0/124 0% No No No 0.17 - 59 ND 610 N No IFD2,4,6-Trichlorophenol – – mg/kg – 0/124 0% No No No 0.077 - 59 ND 6.1 N No IFD2,4-Dichlorophenol – – mg/kg – 0/124 0% No No No 0.077 - 59 ND 18 N No IFD2,4-Dimethylphenol 0.077 1.6 mg/kg S349 4/126 3% No No No 0.077 - 59 1.6 120 N No IFD2,4-Dinitrophenol – – mg/kg – 0/124 0% No No No 0.19 - 300 ND 12 N No IFD2,4-Dinitrotoluene – – mg/kg – 0/127 0% No No No 0.077 - 59 ND 1.6 C* No IFD2,6-Dinitrotoluene – – mg/kg – 0/127 0% No No No 0.077 - 59 ND 6.1 N No IFD2-Chloronaphthalene – – mg/kg – 0/127 0% No No Yes 0.033 - 18 ND 630 N No IFD2-Chlorophenol 0.23 0.23 mg/kg S337 1/124 1% No No Yes 0.033 - 18 0.23 39 N No IFD2-Hexanone 0.0045 0.0045 mg/kg S311 1/148 1% No No Yes 0.007 - 83 0.0045 21 N No IFD

Yes 2-Methylphenol 0.03 2.3 mg/kg S349 6/127 5% No 0.033 - 18 2.3 306 N No IFD2-Nitroaniline – – mg/kg – 0/127 0% No No No 0.067 - 45 ND 61 N No IFD2-Nitrophenol – – mg/kg – 0/124 0% No No -- 0.077 - 59 ND N/A No IFD

Page 3 of 32

Table 2-1. Occurrence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SedimentsExposure Medium: SedimentsExposure Point: Sediments in SCA

Part 375 List Parameter

Minimum Detected

Value

Maximum Detected

Value

Concen-tration Units

Location of Maximum

ConcentrationDetection Frequency

Frequency of

detection

Frequency of

detection >5%

Tentatively Identified

Compound Volatile?Range of

Detection Limits

Concentration Used for

ScreeningCoPC Flag

Rationale\ Substance Deletion or Selection

Screening Toxicity Values b

3,3'-Dichlorobenzidine – – – 0/127 0% No No No 0.033 - 18 ND 1.1 C No IFD3-Nitroaniline – – mg/kg – 0/127 0% No No -- 0.19 - 300 ND N/A No IFD4,6-Dinitro-2-methylphenol – – mg/kg – 0/124 0% No No No 0.19 - 300 ND 0.49 N No IFD4-Bromophenyl phenyl ether – – mg/kg – 0/127 0% No No -- 0.033 - 18 ND N/A No IFD4-Chloro-3-methylphenol – – mg/kg – 0/124 0% No No No 0.077 - 59 ND 610 N No IFD4-Chloroaniline 0.08 0.44 mg/kg S337 5/127 4% No No No 0.033 - 18 0.44 2.4 C* No IFD4-Chlorophenyl phenyl ether – – mg/kg – 0/127 0% No No -- 0.033 - 18 ND N/A No IFD4-Ethyltoluene 7.9 43 mg/kg P1 2/2 100% Yes No -- - 43 N/A No NTX4-Methyl-2-pentanone 0.0045 0.0045 mg/kg S311 1/148 1% No No Yes 0.007 - 83 0.0045 530 N No IFD

Yes 4-Methylphenol 0.1 4.7 mg/kg S349 20/127 16% Yes 0.033 - 18 4.7 31 N No BSL4-Nitroaniline – – mg/kg – 0/127 0% No No No 0.19 - 300 ND 24 C* No IFD4-Nitrophenol – – mg/kg – 0/123 0% No No -- 0.19 - 300 ND N/A No IFD

Yes Acetone 0.012 3 mg/kg S344 44/107 41% Yes 0.013 - 57 3.0 6130 N No BSLYes Benzene 0.00028 160 mg/kg OL-VC-10054 704/978 72% Yes 0.0012 - 110 160 1.1 C Yes ASL

Benzyl alcohol – – mg/kg – 0/109 0% No No No 0.17 - 59 ND 610 N No IFDBis(2-chloroethoxy)methane – – mg/kg – 0/127 0% No No No 0.033 - 18 ND 18 N No IFDBis(2-chloroethyl)ether – – mg/kg – 0/126 0% No No Yes 0.033 - 18 ND 0.21 C* No IFDBis(2-chloroisopropyl)ether – – mg/kg – 0/109 0% No No -- 0.033 - 11 ND N/A No IFDBis(2-ethylhexyl)phthalate 0.03 5.5 mg/kg S322 42/127 33% Yes No No 0.036 - 18 5.5 35 C* No BSLB di hl th /k 0/148 0% N N Y 0 003 42 ND 0 27 C* N IFDBromodichloromethane – – mg/kg – 0/148 0% No No Yes 0.003 - 42 ND 0.27 C* No IFDBromoform – – mg/kg – 0/148 0% No No No 0.003 - 42 ND 61 C* No IFDBromomethane – – mg/kg – 0/148 0% No No Yes 0.0062 - 83 ND 0.73 N No IFDButylbenzyl phthalate 0.13 0.13 mg/kg S337 1/126 1% No No No 0.033 - 18 0.13 260 C* No IFDCarbazole 0.048 19 mg/kg S313 33/127 26% Yes No -- 0.033 - 18 19 N/A No NTXCarbon disulfide 0.001 2.37 mg/kg S345 41/148 28% Yes No Yes 0.004 - 42 2.4 82 N No BSL

Yes Carbon tetrachloride – – mg/kg – 0/141 0% No 0.003 - 42 ND 0.25 C No IFDYes Chlorobenzene 0.00043 6000 mg/kg OL-VC-10097 587/978 60% Yes 0.005 - 59 6000 29 N Yes ASL

Chlorodibromomethane 3.75 3.75 mg/kg S337 1/146 1% No No Yes 0.003 - 42 3.8 0.68 C* No IFDChloroethane – – mg/kg – 0/148 0% No No Yes 0.0062 - 83 ND 1500 N No IFD

Yes Chloroform – – mg/kg – 0/148 0% No 0.003 - 42 ND 0.30 C No IFDChloromethane – – mg/kg – 0/148 0% No No Yes 0.0062 - 83 ND 12 N No IFD

Yes cis-1,2-Dichloroethene 0.001 0.057 mg/kg S314 7/129 5% Yes 0.003 - 42 0.057 78 N No BSLcis-1,3-Dichloropropene – – mg/kg – 0/148 0% No No Yes 0.003 - 42 ND 1.7 C* No IFDCyclohexane 0.036 1.4 mg/kg S342 2/23 9% Yes No Yes 0.11 - 21 1.4 700 N No BSLDichlorodifluoromethane – – mg/kg – 0/22 0% No No Yes 0.11 - 21 ND 180 N No IFDDiethyl phthalate 1.8 1.8 mg/kg S337 1/127 1% No No No 0.033 - 18 1.8 4900 N No IFDDimethyl phthalate – – mg/kg – 0/126 0% No No -- 0.033 - 18 ND N/A No IFDDi-n-butyl phthalate 0.032 0.21 mg/kg S337 6/127 5% No No No 0.033 - 18 0.21 610 N No IFDDi-n-octyl phthalate – – mg/kg – 0/127 0% No No -- 0.033 - 18 ND N/A No IFD

Yes Ethylbenzene 0.00061 72 mg/kg OL-VC-10044 532/978 54% Yes 0.00014 - 1500 72 5.4 C Yes ASL

Page 4 of 32

Table 2-1. Occurrence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SedimentsExposure Medium: SedimentsExposure Point: Sediments in SCA

Part 375 List Parameter

Minimum Detected

Value

Maximum Detected

Value

Concen-tration Units

Location of Maximum

ConcentrationDetection Frequency

Frequency of

detection

Frequency of

detection >5%

Tentatively Identified

Compound Volatile?Range of

Detection Limits

Concentration Used for

ScreeningCoPC Flag

Rationale\ Substance Deletion or Selection

Screening Toxicity Values b

Yes Hexachlorobenzene 0.0062 20 mg/kg P3 97/144 67% Yes 0.001 - 18 20 0.30 C Yes ASLHexachlorobutadiene 0.066 0.066 mg/kg S304 1/127 1% No No No 0.033 - 18 0.066 6.1 N No IFDHexachlorocyclopentadiene – – mg/kg – 0/120 0% No No No 0.077 - 59 ND 37 N No IFDHexachloroethane – – mg/kg – 0/127 0% No No No 0.033 - 18 ND 6.1 N No IFDIsophorone 0.34 0.34 mg/kg S337 1/127 1% No No No 0.033 - 18 0.34 510 C* No IFDIsopropylbenzene 0.063 5.2 mg/kg S309 17/23 74% Yes No Yes 0.57 - 0.89 5.2 210 N No BSLMethyl acetate – – mg/kg – 0/1 0% No No Yes 1.2 - 1.2 ND 7800 N No IFD

Yes Methyl Tert-Butyl Ether – – mg/kg – 0/22 0% No 0.054 - 10 ND 43 C No IFDMethylcyclohexane 0.088 6 mg/kg S342 10/24 42% Yes No -- 1.1 - 21 6.0 N/A No NTX

Yes Methylene chloride 0.00525 7.2 mg/kg S347 18/148 12% Yes 0.004 - 83 7.2 11 C No BSLn-Heptane 3.7 3.7 mg/kg S342 1/1 100% Yes No -- - 3.7 N/A No NTXn-Hexadacane 0.077 1.4 mg/kg S337 4/4 100% Yes No -- - 1.4 N/A No NTXn-Hexane 1.4 1.4 mg/kg S342 1/1 100% Yes No Yes - 1.4 57 N No BSLNitrobenzene – – mg/kg – 0/126 0% No No Yes 0.033 - 18 ND 4.8 C* No IFDn-Nitrosodimethylamine – – mg/kg – 0/109 0% No No No 0.17 - 59 ND 0.0023 C* No IFDn-Nitroso-di-n-propylamine – – mg/kg – 0/127 0% No No No 0.033 - 18 ND 0.069 C* No IFDn-Nitrosodiphenylamine – – mg/kg – 0/127 0% No No No 0.033 - 18 ND 99 C* No IFDn-Nonadecane 0.18 0.18 mg/kg S304 1/1 100% Yes No -- - 0.18 N/A No NTXPentachlorobenzene 0.0045 7.1 mg/kg P3 16/17 94% Yes No No 0.0054 - 0.0054 7.1 4.9 N Yes ASL

Y P t hl h l /k 0/124 0% N 0 19 300 ND 3 0 C N IFDYes Pentachlorophenol – – mg/kg – 0/124 0% No 0.19 - 300 ND 3.0 C No IFDYes Phenol 0.0046 14 mg/kg OL-VC-10083A 538/835 64% Yes 0.005 - 18 14.0 1830 N No BSLYes sec-Butylbenzene 0.024 89 mg/kg P1 14/14 100% Yes - 89 N/A No NTX

Styrene 0.052 10 mg/kg S344 43/148 29% Yes No Yes 0.003 - 42 10 630 N No BSLYes Tetrachloroethene 0.002 0.12 mg/kg S344 6/148 4% No 0.003 - 42 0.12 0.55 C No IFDYes Toluene 0.000083 230 mg/kg S341 645/978 66% Yes 0.0012 - 540 230 497 N No BSLYes trans-1,2-Dichloroethene 0.0027 0.0029 mg/kg S314 2/129 2% No 0.003 - 42 0.0029 15 N No IFD

trans-1,3-Dichloropropene – – mg/kg – 0/144 0% No No Yes 0.003 - 42 ND 1.7 C* No IFDYes Trichloroethene 0.0017 0.004 mg/kg S351 2/148 1% No 0.003 - 42 0.0040 2.8 C No IFD

Trichlorofluoromethane – – mg/kg – 0/22 0% No No Yes 0.054 - 10 ND 790 N No IFDUnres. comb. of 1,2,3/4,5 0.013 22 mg/kg P3 16/17 94% Yes No -- 0.0057 - 0.0057 22 1.8 N Yes ASL

Yes Vinyl chloride 0.0041 0.013 mg/kg S313 2/148 1% No 0.0062 - 83 0.013 0.06 C No IFDYes Xylenes, total 0.00025 822 mg/kg OL-VC-20170 700/978 72% Yes 0.0024 - 22 820 63 N Yes ASL

PAHs2-Methylnaphthalene 0.034 100.5 mg/kg S346 93/127 73% Yes No Yes 0.033 - 0.54 100 31 N Yes ASL

Yes Acenaphthene 0.0019 85 mg/kg S313 596/928 64% Yes 0.0026 - 18 85 344 N No BSLYes Acenaphthylene 0.002 25 mg/kg OL-VC-10038 624/928 67% Yes 0.0026 - 18 25 172 N n No BSLYes Anthracene 0.0006 95 mg/kg S313 684/927 74% Yes 0.0043 - 18 95 1720 N No BSLYes Benz[a]anthracene 0.0017 100 mg/kg S313 703/928 76% Yes 0.0026 - 18 100 0.15 C Yes ASLYes Benzo[a]pyrene 0.002 65 mg/kg S313 680/928 73% Yes 0.0026 - 18 65 0.015 C Yes ASLYes Benzo[b]fluoranthene 0.0011 57 mg/kg S313 683/928 74% Yes 0.0043 - 18 57 0.15 C Yes ASL

Page 5 of 32

Table 2-1. Occurrence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SedimentsExposure Medium: SedimentsExposure Point: Sediments in SCA

Part 375 List Parameter

Minimum Detected

Value

Maximum Detected

Value

Concen-tration Units

Location of Maximum

ConcentrationDetection Frequency

Frequency of

detection

Frequency of

detection >5%

Tentatively Identified

Compound Volatile?Range of

Detection Limits

Concentration Used for

ScreeningCoPC Flag

Rationale\ Substance Deletion or Selection

Screening Toxicity Values b

Yes Benzo[ghi]perylene 0.0018 35 mg/kg S313 658/928 71% Yes 0.0026 - 18 35 172 N n No BSLYes Benzo[k]fluoranthene 0.0022 60 mg/kg S313 609/928 66% Yes 0.0026 - 18 60 1.5 C Yes ASLYes Chrysene 0.003 100 mg/kg S313 705/928 76% Yes 0.0026 - 18 100 15 C Yes ASLYes Dibenz[a,h]anthracene 0.0013 17 mg/kg S313 522/928 56% Yes 0.0026 - 18 17.0 0.015 C Yes ASLYes Fluoranthene 0.0025 250 mg/kg S313 740/928 80% Yes 0.0044 - 6.1 250 229 N Yes ASLYes Fluorene 0.0018 460 mg/kg OL-VC-70112 367/928 40% Yes 0.0026 - 6.1 460 229 N Yes ASLYes Indeno[1,2,3-cd]pyrene 0.0013 38 mg/kg S313 652/928 70% Yes 0.0043 - 18 38 0.15 C Yes ASLYes Naphthalene 0.0013 2900 mg/kg OL-VC-10065 683/962 71% Yes 0.00068 - 1.5 2900 3.6 C Yes ASLYes Phenanthrene 0.0025 380 mg/kg S313 758/928 82% Yes 0.0043 - 6.1 380 N/A No NTXYes Pyrene 0.0012 150 mg/kg S313 741/928 80% Yes 0.0044 - 6.1 150 172 N No BSL

Page 6 of 32

Table 2-1. Occurrence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SedimentsExposure Medium: SedimentsExposure Point: Sediments in SCA

Notes: All results reported as dry weight data from Onondaga Lake sediments Rationale Codes:

to be dredged and provided by Parsons Corporation and Anchor QEA. Selection Reason:For the purposes of screening, field replicates have been averaged. ASL - above screening levels

– - either no detected or undetected values HIST - infrequent detection but associated historicallyARAR - applicable or relevant and appropriate requirement Deletion Reason:

C - carcinogenic based on a cancer risk of 1´10–6 BKG - below or consistent with background levels

CoPC - chemical of potential concern BSL - below screening levelN - noncarcinogenic based on hazard quotient of 0.1 IFD - infrequent detection

N/A - not available NTX - no toxicity information

ND - not detected NUT - essential nutrientPAH - polycyclic aromatic hydrocarbonRSL - Regional Screening Level TBC - to be considered

a These Eastern USA background values are from New York State TAGM 4046, Table 4 - Heavy Metals - Recommended soil cleanup objectives.b Screening toxicity values for soil are the RSLs taken from U.S. EPA Region IX (2010). RSLs correspond to 1´10–6 or a hazard quotient of 0.1, whichever is lower.c The chromium results in the table do not include the 4 locations for which hexavalent chromium data are available. See Section 3.2. d This default carcinogenic screening value for chromium is that for chromium (VI).e This default non-carcinogenic screening value for lead has not been adjusted. Lead screened based on mean concentration consistent with use of mean in lead model.f This default non-carcinogenic screening value for manganese is that for ingestion of water.

g This default non-carcinogenic screening value for mercury is that for methylmercury.h This default non-carcinogenic screening value for nickel is that for soluble salts.i This default carcinogenic screening value for alpha-chlordane is that for chlordane.j This default carcinogenic screening value for delta-BHC is that for alpha-BHC.k This default non-carcinogenic screening value for the endosulfan I, endosulfan II, and endosulfan sulfate is that for endosulfan.l This default non-carcinogenic screening value for PCBs is that for Aroclor-1254.m This default non-carcinogenic screening value for 1,3-dichlorobenzene is that for 1,2-dichlorobenzene.n This default non-carcinogenic screening value for value for acenaphthalene and benzo[ghi]perylene is that for pyrene.

Page 7 of 32

Table 2-2. Occurence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe:FutureMedium: Sediments to be placed in SCAExposure Medium: AirExposure Point: Air

Parameter

Minimum Detected

Value

Maximum Detected

Value

Concen- tration Units

Location of Maximum

ConcentrationDetection Frequency

Range of Detection Limits

Concentration Used for

Screening Volatile?

Has Tox

Value?

Detected in Wind

Tunnel Testing?

Potential ARAR/ TBC Value

Potential ARAR/ TBC

SourceCoPC Flag

Rationale\ Substance Deletion or Selection

INORGANIC ANALYTESArsenic No No No No NVBarium No No No No NVBeryllium No No No No NVCadmium No No No No NVChromium No No No No NVCopper No No No No NVCyanide Yes Yes No Yes CNLeadd No No No No NVManganese No No No No NVMercury Yes Yes No Yes Elem HgNickel No No No No NVSelenium No No No No NVSilver No No No No NVZinc No No No No NVPESTICIDES/PCBS/PCDDFS2-Butanone Yes Yes Yes Yes VAldrin No No No No NVAlpha-BHC No No No No NVAlpha-Chlordane No No No No NVBeta-BHC No No No No NVDDD No No No No NVDDE No No No No NVDDT No No No No NVDelta-BHC No No No No NVDibenzofuran Yes No No No NVDieldrin No No No No NVEndosulfan I No No No No NVEndosulfan II No No No No NVEndosulfan Sulfate No No No No NVEndrin No No No No NVGamma-BHC (Lindane) No No No No NVHeptachlor No No No No NVPCBs No No No No NVDioxins (as TCDD equivalents) No Yes No No NVORGANIC ANALYTES1,1,1-Trichloroethane Yes Yes Yes Yes V1,1-Dichloroethane Yes Yes No Yes V1,1-Dichloroethene Yes Yes No Yes V1,2,3-Trichlorobenzene Yes No Yes No NV1,2,4-Trichlorobenzene Yes Yes Yes Yes V1,2,4-Trimethylbenzene Yes Yes Yes Yes V1,2-Dichlorobenzene Yes Yes Yes Yes V1,2-Dichloroethane Yes Yes No Yes V

Page 8 of 32

Table 2-2. Occurence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe:FutureMedium: Sediments to be placed in SCAExposure Medium: AirExposure Point: Air

Parameter

Minimum Detected

Value

Maximum Detected

Value

Concen- tration Units

Location of Maximum

ConcentrationDetection Frequency

Range of Detection Limits

Concentration Used for

Screening Volatile?

Has Tox

Value?

Detected in Wind

Tunnel Testing?

Potential ARAR/ TBC Value

Potential ARAR/ TBC

SourceCoPC Flag

Rationale\ Substance Deletion or Selection

1,3,5-Trichlorobenzene -- No Yes No NTX1,3,5-Trimethylbenzene Yes No Yes No NTX1,3-Dichlorobenzene Yes No No No NTX1,4-Dichlorobenzene Yes Yes Yes Yes V2-Methylphenol No No No No NV4-Methylphenol No No No No NVAcetone Yes Yes No Yes VBenzene Yes Yes Yes Yes VCarbon tetrachloride Yes Yes No Yes VChlorobenzene Yes Yes Yes Yes VChloroform Yes Yes Yes Yes Vcis-1,2-Dichloroethene Yes No No No NTXEthylbenzene Yes Yes Yes Yes ASLHexachlorobenzene No Yes Yes No NVMethyl Tert-Butyl Ether Yes Yes No Yes VMethylene chloride Yes Yes Yes Yes VPentachlorophenol No No No No NVPhenol No No Yes No NVsec-Butylbenzene Yes No No No NTXTetrachloroethene Yes Yes Yes Yes VToluene Yes Yes Yes Yes Vtrans-1,2-Dichloroethene Yes Yes No Yes VTrichloroethene Yes Yes Yes Yes VVinyl chloride Yes Yes No Yes VXylenes, total Yes Yes Yes Yes VPAHsAcenaphthene Yes No No No NTXAcenaphthylene No No No No NTXAnthracene Yes No No No NTXBenz[a]anthracene No No No No NTXBenzo[a]pyrene No No No No NTXBenzo[b]fluoranthene No No No No NTXBenzo[ghi]perylene No No No No NTXBenzo[k]fluoranthene No No No No NTXChrysene No No No No NTXDibenz[a,h]anthracene No No No No NTXFluoranthene No No No No NTXFluorene Yes No Yes No NTXIndeno[1,2,3-cd]pyrene No No No No NTXNaphthalene Yes Yes Yes Yes VPhenanthrene No No No No NTXPyrene Yes No No No NTX

Page 9 of 32

Table 2-2. Occurence, distribution, and selection of contaminants of potential concernOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SedimentsExposure Medium: SedimentsExposure Point: Sediments in SCA

Notes: CoPC - chemical of potential concern Rationale Codes:PAH - polycyclic aromatic hydrocarbon Selection Reason:

V - chemical is volatile or was identified in Wind Tunnel TestingElem Hg - elemental mercury can volatilize

CN - specific forms of cyanide can volatilizeDeletion Reason:

NV - chemical is not volatileNTX - no toxicity information

Page 10 of 32

Table 3-1. Medium specific exposure point concentration summaryOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: Sediment to be placed in SCA mg/kg 0.000001Exposure Medium: Sediment ng/kgExposure Point: Sediment

UnitsArithmetic

Mean 95% UCL

Maximum Detected

ValueMaximum Qualifier

EPC Units

Medium EPC Value Medium EPC Statistic

Medium EPC Rationale

Aluminum mg/kg 4,000 4,700 13,000 mg/kg 4,700 95% KM (BCA) UCL AAntimony mg/kg 2.0 4.7 12 mg/kg 4.7 97.5% KM (Chebyshev) UCL A

1 Arsenic mg/kg 6.3 10 27 mg/kg 10 95% KM (Chebyshev) UCL A2 Barium mg/kg 600 1,200 2,800 mg/kg 1,200 95% KM (Chebyshev) UCL A3 Cadmium mg/kg 3.6 7.2 21 mg/kg 7.2 95% KM (Chebyshev) UCL A4 Chromium mg/kg 170 790 3,700 mg/kg 790 97.5% KM (Chebyshev) UCL A

Cobalt mg/kg 13 48 180 mg/kg 48 97.5% KM (Chebyshev) UCL A5 Copper mg/kg 84 150 560 mg/kg 150 95% KM (Chebyshev) UCL A6 Manganese mg/kg 350 580 1,800 mg/kg 580 95% KM (Chebyshev) UCL A7 Mercury mg/kg 9.0 15 160 mg/kg 15 97.5% KM (Chebyshev) UCL A8 Nickel mg/kg 110 350 2,100 mg/kg 350 95% KM (Chebyshev) UCL A9 Vanadium mg/kg 24 78 280 mg/kg 78 97.5% KM (Chebyshev) UCL A9 Dibenzofuran mg/kg 2.0 4.8 16 mg/kg 4.8 95% KM (Chebyshev) UCL A

10 Dieldrin mg/kg 0.0045 0.005 0.024 mg/kg 0.005 95% KM (t) UCLDioxins (as TCDD equivalents) mg/kg 0.000075 0.00013 0.00023 mg/kg 0.00013 95% Student's-t UCL A

10 PCBs mg/kg 0.59 1.3 23 mg/kg 1.3 97.5% KM (Chebyshev) UCL A1,2,3/4,5 Tetrachlorobenzene mg/kg 6.9 21 21 mg/kg 21 Maximum detected B1,2,4-Trichlorobenzene mg/kg 6.8 17 380 mg/kg 17 97.5% KM (Chebyshev) UCL A

11 1,2,4-Trimethylbenzene mg/kg 9 19 28 mg/kg 19 95% KM (t) UCL A12 1,2-Dichlorobenzene mg/kg 15 38 1100 mg/kg 38 97.5% KM (Chebyshev) UCL A13 1,4-Dichlorobenzene mg/kg 25 84 3000 mg/kg 84 97.5% KM (Chebyshev) UCL A14 Benzene mg/kg 2.8 4.7 46 mg/kg 4.7 97.5% KM (Chebyshev) UCL A15 Chlorobenzene mg/kg 19 52 1500 mg/kg 52 97.5% KM (Chebyshev) UCL A16 Ethylbenzene mg/kg 4.3 12 380 mg/kg 12 97.5% KM (Chebyshev) UCL A17 Hexachlorobenzene mg/kg 1.2 6.4 19 mg/kg 6.4 99% KM (Chebyshev) UCL A

Pentachlorobenzene mg/kg 1.9 6.7 6.7 mg/kg 6.7 Maximum detected B22 Xylenes, total mg/kg 28 45 310 mg/kg 45 97.5% KM (Chebyshev) UCL A

2-Methylnaphthalene mg/kg 6.4 13 37 mg/kg 13 95% KM (Chebyshev) UCL A23 Benz[a]anthracene mg/kg 2.3 4 31 mg/kg 4.0 97.5% KM (Chebyshev) UCL A24 Benzo[a]pyrene mg/kg 1.8 3.2 32 mg/kg 3.2 97.5% KM (Chebyshev) UCL A25 Benzo[b]fluoranthene mg/kg 1.9 3.2 28 mg/kg 3.2 97.5% KM (Chebyshev) UCL A26 Benzo[k]fluoranthene mg/kg 0.88 1.5 17 mg/kg 1.5 97.5% KM (Chebyshev) UCL A27 Chrysene mg/kg 2.3 3.9 30 mg/kg 3.9 97.5% KM (Chebyshev) UCL A28 Dibenz[a,h]anthracene mg/kg 0.40 0.58 5.6 mg/kg 0.58 97.5% KM (Chebyshev) UCL A

Fluoranthene mg/kg 5 8.2 53 mg/kg 8.2 97.5% KM (Chebyshev) UCL A29 Fluorene mg/kg 3.2 10 340 mg/kg 10 97.5% KM (Chebyshev) UCL A30 Indeno[1,2,3-cd]pyrene mg/kg 0.92 1.5 14 mg/kg 1.5 97.5% KM (Chebyshev) UCL A31 Naphthalene mg/kg 83 130 820 mg/kg 130 97.5% KM (Chebyshev) UCL ANotes:

NotEPC - exposure point concentrationUCL - 95 percent upper confidence limit on the mean concentration

A - Medium EPC value per ProUCL 4.0 recommendation .B - Maximum detected concentration used for EPC concentration.

Duplicate sample results were averaged in calculations.Arithmetic mean was calculated using the full detection limit for undetected results and UCL calculations took into account most likely distribution as well as various detection limits. Because of this some UCLs are lower than arithmetic averages.Statistical assessment of data was conducted using ProUCL 4.0 on length-weighted data for Onondaga Lake Sediments provided by Parsons Corporation and Anchor QEA.

Reasonable Maximum ExposureSediment data length-weighted averages

Chemical of Potential Concern

Page 11 of 32

Table 3-2. Medium specific exposure point concentration summaryOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: Sediment to be placed in SCAExposure Medium: SedimentExposure Point: Air

UnitsArithmetic

Mean95% UCL

Maximum Detected

ValueMaximum Qualifier EPC Units

Medium EPC Valuea

Medium EPC

StatisticMedium EPC

Rationale

Mercury elementalb mg/m3 mg/m3 0.000067Cyanide (hydrogen cyanide) mg/m3 mg/m3 0.000672-Butanone mg/m3 mg/m3 1.11,1,1-Trichloroethane mg/m3 mg/m3 0.221,1-Dichloroethane mg/m3 mg/m3 0.00201,1-Dichloroethene mg/m3 mg/m3 0.0161,2,4-Trichlorobenzene mg/m3 mg/m3 0.00191,2,4-Trimethylbenzene mg/m3 mg/m3 0.00681,2-Dichlorobenzene (ortho) mg/m3 mg/m3 0.0801,2-Dichloroethane mg/m3 mg/m3 0.000121,3,5-Trimethylbenzene mg/m3 mg/m3 0.0641,4-Dichlorobenzene mg/m3 mg/m3 0.00029Acetone mg/m3 mg/m3 6.2Benzene mg/m3 mg/m3 0.00042Carbon tetrachloride mg/m3 mg/m3 0.00022Chlorobenzene mg/m3 mg/m3 0.024Chloroform mg/m3 mg/m3 0.00014Ethylbenzene mg/m3 mg/m3 0.0054Methyl Tert-Butyl Ether mg/m3 mg/m3 0.012Methylene chloride mg/m3 mg/m3 0.0069Naphthalene mg/m3 mg/m3 0.00040Tetrachloroethene mg/m3 mg/m3 0.00055Toluene mg/m3 mg/m3 1.1trans-1,2-Dichloroethene mg/m3 mg/m3 0.014Trichloroethene mg/m3 mg/m3 0.0016Vinyl chloride mg/m3 mg/m3 0.00074Xylenes mg/m3 mg/m3 0.022

Notes:EPC - exposure point concentration

a Estimated offsite residential concentrations derived by dispersion modeling and selecting the maximum yearly average concentration converted to mg/m3.b Elemental mercury was not detected in sediments, which instead had inorganic mercury forms. Hypothetical estimates here based on elemental mercury.

Reasonable Maximum Exposure

Chemical of Potential Concern

Page 12 of 32

Table 4-1. Values used for daily intake calculationsOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SCA SedimentsExposure Medium: Surface 'Soil'Exposure Point: Surface 'soil'Receptor Population: ResidentialReceptor Age: Adult

Exposure Parameter Parameter Definition Units RME RME Intake Equation/

Route Code Value Rationale/ Model NameReference

Ingestion CS Chemical concentration in Sediment/soil mg/kg see Tables 2-1, 3-1 see Tables 2-1, 3-1 Chronic Daily Intake (CDI) (mg/kg-day) =CF Conversion factor kg/mg 0.000001 -- CS x CF x IR x FI x EF x ED / (BW x AT)IR Ingestion rate mg soil/day 100 USEPA 1997 RME CDI:FI Fraction ingested -- 1 a For carcinogens (CS)*:EF Exposure frequency days/yr 45 a 2.5E-09ED Exposure duration yrs 1 a For non carcinogens (CS)*:BW Body weight kg 70 USEPA 1997 1.8E-07AT Averaging time - carcinogens days 25,550 USEPA 1989AT Averaging time - noncarcinogens days 365 USEPA 1989

Dermal CS Chemical concentration in Sediment/soil mg/kg see Tables 2-1, 3-1 see Tables 2-1, 3-1 Chronic Daily Intake (CDI) (mg/kg-day) =CF Conversion factor kg/mg 0.000001 -- CS x CF x SA x AF x EF x ED x ABSSA Skin surface area available for contact cm2/event 5,700 USEPA 2004 / (BW x AT)AF Sediment/soil-to-skin adherence factor mg/cm2 0.07 USEPA 2004 RME CDI:EF Exposure frequency days/yr 45 a For carcinogens (CS* ABS)*:ED Exposure duration yrs 1 a 1.0E-08BW Body weight kg 70 USEPA 1997 Non carcinogens (CS*ABS)*:AT Averaging time - carc days 25,550 USEPA 1989 7.0E-07AT Averaging time - nonc days 365 USEPA 1989

ABS Dermal absorption factor unitless chemical specific USEPA 1997

Notes: -- - not applicableCT - central tendency

RME - reasonable maximum exposurea Based on best professional judgment regarding the number of days that sediment might remain exposed.

Page 13 of 32

Table 4-2. Values used for daily intake calculationsOnondaga Lake sediment containment area (SCA)

Scenario Timeframe: FutureMedium: SCA SedimentsExposure Medium: Surface 'Soil'Exposure Point: Surface soilReceptor Population: ResidentialReceptor Age: Child

Exposure Parameter Parameter Definition Units RME RME Intake Equation/

Route Code Value Rationale/ Model NameReference

Ingestion CS Chemical concentration in Sediment/soil mg/kg see Tables 2-1, 3-1 see Tables 2-1, 3-1 Chronic Daily Intake (CDI) (mg/kg-day) =CF Conversion factor kg/mg 0.000001 -- CS x CF x IR x FI x EF x ED / (BW x AT)IR Ingestion rate mg soil/day 200 USEPA 1997 RME CDI:FI Fraction ingested -- 1 a For carcinogens (CS)*:EF Exposure frequency days/yr 45 a 2.3E-08ED Exposure duration yrs 1 a For non carcinogens (CS)*:BW Body weight kg 15 USEPA 1997 1.6E-06AT Averaging time - carcinogens days 25,550 USEPA 1989AT Averaging time - noncarcinogens days 365 USEPA 1989

Dermal CS Chemical concentration in Sediment/soil mg/kg see Tables 2-1, 3-1 see Tables 2-1, 3-1 Chronic Daily Intake (CDI) (mg/kg-day) =CF Conversion factor kg/mg 0.000001 -- CS x CF x SA x AF x EF x ED x ABSSA Skin surface area available for contact cm2/event 2,800 USEPA 2004 / (BW x AT)AF Sedimen/soil-to-skin adherence factor mg/cm2 0.2 USEPA 2004 RME CDI:EF Exposure frequency days/yr 45 a For carcinogens (CS* ABS)*:ED Exposure duration yrs 1 a 6.6E-08BW Body weight kg 15 USEPA 1997 Non carcinogens (CS*ABS)*:AT Averaging time - carc days 25,550 USEPA 1989 4.6E-06AT Averaging time - nonc days 365 USEPA 1989

ABS Dermal absorption factor unitless chemical specific USEPA 1997

Notes: -- - not applicableCT - central tendency

RME - reasonable maximum exposurea Based on best professional judgment regarding the number of days that sediment might remain exposed.

Page 14 of 32

Table 4-3. Values used for daily intake calculations for carcinogenic PAHsOnondaga Lake sediment containment area (SCA)

Age(year)

Exposure Duration

(ED) (years)

Exposure Frequency

(EF) (Days)

Body Weight1

(BW) (kg)

Soil Intake Rate2

(IR) (mg/day)

Soil Adherence

Factor3 (AF) (mg/cm2-

event)

Exposed Surface Area4

(SA) (cm2)

Total Surface

Area (cm2) Age

Group

Soil Ingestion CDI

(mg/kg-day)Dermal CDI (mg/kg-day)

Ingestion Chronic Daily

Intake (mg/kg-day)

Dermal Chronic Daily Intake (mg/kg-day)

0 1 45 9.1 200 0.2 2,625 5,910 0-<2 yrs Sum 5 7.0E-08 1.8E-07 3.9E-08 1.0E-071 1 45 11.3 200 0.2 2,571 5,910 0-<2 yrs Average 6 3.5E-08 9.09E-08 3.1E-08 8.0E-082 1 45 13.3 200 0.2 2,434 5,910 2-<6 yrs Sum 5 8.8E-08 2.5E-07 2.6E-08 6.4E-083 1 45 15.3 200 0.2 2,893 6,565 2-<6 yrs Average 6 2.2E-08 6.3E-08 2.3E-08 6.7E-084 1 45 17.4 200 0.2 3,175 7,185 2.0E-08 6.4E-085 1 45 19.7 200 0.2 3,255 7,860 1.8E-08 5.8E-086 1 45 22.6 100 0.2 2,949 8,545 7.8E-09 4.6E-087 1 45 24.9 100 0.2 3,182 9,265 7.1E-09 4.5E-088 1 45 28.1 100 0.2 3,434 10,000 6-<16 yrs Sum 5 4.9E-08 3.2E-07 6.3E-09 4.3E-089 1 45 31.5 100 0.2 3,657 10,650 6-<16 yrs Average 6 4.9E-09 3.2E-08 5.6E-09 4.1E-08

10 1 45 36.3 100 0.2 3,819 11,750 4.9E-09 3.7E-0811 1 45 41.1 100 0.2 4,111 12,650 4.3E-09 3.5E-0812 1 45 45.3 100 0.2 4,453 13,700 3.9E-09 3.5E-0813 1 45 50.4 100 0.07 4,916 14,750 3.5E-09 1.2E-0814 1 45 56 100 0.07 5,205 15,800 3.1E-09 1.1E-0815 1 45 58.1 100 0.07 5,386 16,350 3.0E-09 1.1E-08

16 1 45 62.6 100 0.07 5,534 16,800 16-<30 yrs Sum 5 3.6E-08 1.4E-07 2.8E-09 1.1E-0817 1 45 63.2 100 0.07 5,641 17,150 16-<30 yrs Average 6 2.60E-09 1.03E-08 2.8E-09 1.1E-0818 1 45 65.1 100 0.07 5,700 18,000 2.7E-09 1.1E-0819 1 45 66 100 0.07 5,700 18,000 2.7E-09 1.1E-0820 1 45 67.2 100 0.07 5,700 18,000 2.6E-09 1.0E-0821 1 45 67 2 100 0 07 5 700 18 000 2 6E-09 1 0E-08

Mutagenic mode of action calculations for PAHsAge-adjusted exposure factors (RME)

Age-Adjusted CDIs for Cancer

21 1 45 67.2 100 0.07 5,700 18,000 2.6E-09 1.0E-0822 1 45 67.2 100 0.07 5,700 18,000 2.6E-09 1.0E-0823 1 45 67.2 100 0.07 5,700 18,000 2.6E-09 1.0E-0824 1 45 67.2 100 0.07 5,700 18,000 2.6E-09 1.0E-0825 1 45 71.5 100 0.07 5,700 18,000 2.5E-09 9.8E-0926 1 45 71.5 100 0.07 5,700 18,000 2.5E-09 9.8E-0927 1 45 71.5 100 0.07 5,700 18,000 2.5E-09 9.8E-0928 1 45 71.5 100 0.07 5,700 18,000 2.5E-09 9.8E-0929 1 45 71.5 100 0.07 5,700 18,000 2.5E-09 9.8E-09

Equations: Chronic Daily Intake (CDI) (mg/kg-day) =Ingestion algorithm: Dermal algorithm: CF = Conversion Factor mg to kg 0.000001CS x CF x IR x FI x EF x ED / (BW x AT) CS x CF x SA x AF x EF x ED / BW x AT AT = Averaging time - Cancer 25550

References:1 EPA 1997. Exposure Factors Handbook. Tables 7-2 (adults) and 7-3 (children), mean. Values are mean of male and female. Source: National Center of Health Statistics (NCHS) 1987.2 EPA 1991. Standard Default Exposure Factors. Default for resident child and adult.3 EPA 2004. Risk Assessment Guidance for Superfund Volume I: Human Health Evaluation Manual (Part E, Supplemental Guidance for Dermal Risk Assessment).Recommended default adherence factor for a child resident (0.2) and adult resident (0.07). For older children, the geometric mean weighted adherence factor forchildren playing in wet soil was used for children 6 - 12, as an estimate of a high-end soil contact activity (see Exhibit 3-3).4 EPA 2004. Risk Assessment Guidance for Superfund Volume I: Human Health Evaluation Manual (Part E, Supplemental Guidance for Dermal Risk Assessment)Calculated from Exhibit C-1 - Body Part-Specific Surface Area Calculations (Children). Data from Exposure Factors Handbook, Tables 6-6, 6-7 and 6-8.Surface area of head, forearms, hands, lower legs and feet (for child <6 years); feet excluded from surface area calculation for >6 years.Surface area for >18 is recommended default for adult resident (EPA 2004).5 Summed CDI for each age group is used in calculating average CDI.6 Average CDI for each age group is used in calculating risk estimates.

Page 15 of 32

Table 4-4. Values used for daily intake calculationsOnondaga Lake sediment containment area (SCA) air emissions

Scenario Timeframe: Current/FutureMedium: Outdoor airExposure Medium: Outdoor airExposure Point: Offsite areasReceptor Population: Offsite residentReceptor Age: Adult and Child

Exposure Parameter Parameter Definition Units RME RME Intake Equation/

Route Code Value Rationale/ Model NameReference

Ingestion Chronic Daily Intake (CDI) (mg/m3) =CA Chemical concentration in aira mg/m3 see Tables 2-2, 3-2 see Tables 2-2, 3-2 CA x EF x ED / ATFI Fraction from source -- 1 U.S. EPA 1997 RME CDI:EF Exposure frequency days/year 350 U.S. EPA 1997 For carcinogens (CA)*:ED Exposure duration years 5 U.S. EPA 1997 6.8E-02ATc Averaging time - carcinogens days 25,550 U.S. EPA 1989 For non carcinogens (CA)*:ATn Averaging time - noncarcinogens days 1,825 U.S. EPA 1989 9.6E-01

Notes: -- - not applicableCT - central tendency

RME - reasonable maximum exposurea Inhalation rate is assumed in toxicity values.

Page 16 of 32

Table 5-1. Non-cancer toxicity data - oral/dermalOnondaga Lake sediment containment area (SCA) sediments

Chemical Oral Absorption Primary Combinedof Potential Chronic/ Target Uncertainty/Modifying

Concern Subchronic Value Units Value Units Organ Factors Source Date

Aluminum chronic 1.0E+00 mg/kg-d >50% 1.0E+00 mg/kg-d CNS 100 PPRTV 10/23/2006Arsenic chronic 3.0E-04 mg/kg-d 95% 2.9E-04 mg/kg-d Skin 3/1 IRIS 2/1/2003Antimony chronic 4.0E-04 mg/kg-d 15% 6.0E-05 mg/kg-d Blood 1000/1 IRIS 12/3/2002Barium chronic 2.0E-01 mg/kg-d 7% 1.4E-02 mg/kg-d Kidney 300/1 IRIS 7/11/2005Cadmium (food) chronic 1.0E-03 mg/kg-d 2.50% 2.5E-05 mg/kg-d Kidney 10/1 IRIS 1/2/1998Chromium (as hexavalent) chronic 3.0E-03 mg/kg-d 1.3% - 2.5% 7.5E-05 mg/kg-d NOAEL 300/3 IRIS 10/28/2003Cobalt chronic 3.0E-04 mg/kg-d >50% 3.0E-04 mg/kg-d Thyroid 3000 PPRTV 8/25/2008Copper chronic 4.0E-02 mg/kg-d >50% 4.0E-02 mg/kg-d GI NA HEAST 1997Manganese chronic 1.4E-01 mg/kg-d 4% 5.6E-03 mg/kg-d CNS 1/1 IRIS 12/3/2002Mercury (as mercuric chloride) chronic 3.0E-04 mg/kg-d 7% 2.1E-05 mg/kg-d Immune 1000/1 IRIS 12/3/2002Nickel (as soluble salts) chronic 2.0E-02 mg/kg-d 4% 8.0E-04 mg/kg-d Whole Body 300/1 IRIS 12/10/1998Vanadium chronic 9.0E-03 mg/kg-d 3% 2.3E-04 mg/kg-d Hair cystine 100/1 IRIS 12/3/2002Dibenzofuran chronic ND mg/kg-d >50% 1.0E-03 mg/kg-d NA NA IRIS 12/3/2002Dioxin (as TCDD equivalents) chronic 1.0E-09 mg/kg-d 50% - 83% 1.00E-09 mg/kg-d CNS 100 ATSDR 9/1/2008PCBs (as Aroclor 1254) chronic 2.0E-05 mg/kg-d 80% - 96% 2.0E-05 mg/kg-d Immune 300/1 IRIS 4/1/1997Dieldrin chronic 5.0E-05 mg/kg-d 5.0E-05 mg/kg-d Liver 100/1 IRIS 3/7/20051,2,4-Trichlorobenzene chronic 1.0E-02 mg/kg-d >50% 1.0E-02 mg/kg-d Kidney 1000/1 IRIS 12/3/20021,2,4-Trimethylbenzene chronic ND mg/kg-d >50% ND mg/kg-d NA NA IRIS 3/30/20101,2-Dichlorobenzene chronic 9.0E-02 mg/kg-d >50% 9.0E-02 mg/kg-d NOAEL 1000/1 IRIS 1/12/20001,4-Dichlorobenzene chronic 7.0E-02 mg/kg-d >50% 7.0E-02 mg/kg-d Liver 100 ATSDR 7/1/20061,2,3,4/5 Tetrachlorobenzene chronic 3.0E-04 mg/kg-d >50% 3.0E-04 mg/kg-d Kidney 1000/1 IRIS 12/3/2002Pentachlorobenzene chronic 8.0E-04 mg/kg-d >50% 8.0E-04 mg/kg-d Liver, Kidney 10000/1 IRIS 10/28/2003Benzene chronic 4.0E-03 mg/kg-d >50% 4.0E-03 mg/kg-d Blood 300/1 IRIS 10/1/2008Chlorobenzene chronic 2.0E-02 mg/kg-d >50% 2.0E-02 mg/kg-d Liver 1000/1 IRIS 10/28/2003Ethylbenzene chronic 1.0E-01 mg/kg-d >50% 1.0E-01 mg/kg-d Liver, Kidney 1000/1 IRIS 12/10/1998Hexachlorobenzene chronic 8.0E-04 mg/kg-d >50% 8.0E-04 mg/kg-d Liver 100/1 IRIS 10/28/2003Methylene chloride chronic 6.0E-02 mg/kg-d >50% 6.0E-02 mg/kg-d Liver 100/1 IRIS 7/30/2003Pentachlorophenol chronic 3.0E-02 mg/kg-d 76% - 100% 3.0E-02 mg/kg-d Liver, Kidney 100/1 IRIS 1/2/1998Tetrachloroethene chronic 1.0E-02 mg/kg-d >50% 1.0E-02 mg/kg-d Liver 1000/1 IRIS 1/2/1998Vinyl chloride chronic 3.0E-03 mg/kg-d >50% 3.0E-03 mg/kg-d Liver 30/1 IRIS 10/28/2003Xylenes (total) chronic 2.0E-01 mg/kg-d >50% 2.0E-01 mg/kg-d Whole Body 1000/1 IRIS 2/21/20032-Methyl napthalene chronic 4.0E-03 mg/kg-d >50% 4.0E-03 mg/kg-d Lung 1000/1 IRIS 3/5/2007Benz[a]anthracene chronic ND mg/kg-d 58% - 89% ND mg/kg-d NA NA IRIS 11/1/1994Benzo[a]pyrene chronic ND mg/kg-d 58% - 89% ND mg/kg-d NA NA IRIS 12/10/1998Benzo[b]fluoranthene chronic ND mg/kg-d 58% - 89% ND mg/kg-d NA NA IRIS 11/1/1994Benzo[k]fluoranthene chronic ND mg/kg-d 58% - 89% ND mg/kg-d NA NA IRIS 11/1/1994Chrysene chronic ND mg/kg-d 58% - 89% ND mg/kg-d NA NA IRIS 11/1/1994Dibenz[a,h]anthracene chronic ND mg/kg-d 58% - 89% ND mg/kg-d NA NA IRIS 11/1/1994Fluorene chronic 4.0E-02 mg/kg-d 58% - 89% 4.0E-02 mg/kg-d Blood 3000/1 IRIS 4/1/1997Fluoranthene chronic 4.0E-02 mg/kg-d 58% - 89% 4.0E-02 mg/kg-d Liver, Kidney, Blood 3000/1 IRIS 4/1/1997Indeno[1,2,3-cd]pyrene chronic ND mg/kg-d 58% - 89% ND mg/kg-d NA NA IRIS 11/1/1994Naphthalene chronic 2.0E-02 mg/kg-d 58% - 89% 2.0E-02 mg/kg-d Whole Body 3000/1 IRIS 9/1/2002Notes:

(1) USEPA 2001bNA - not availableN/A - not applicableIRIS - Integrated Risk Information SystemHEAST - Health Effects Assessment Summary TablesATSDR - Agency for Toxic Substatnces and Disease Registrymg/kg-d - milligrams per kilogram per day

*Conversion Factors: 5 mg/L (5 ppm) given as 0.350 mg/kg/day in the discussion section of the critical study

Efficiency for Dermal (1)

Oral RfD Absorbed RfD for Dermal RfD:Target Organs

Page 17 of 32

Table 5-2. Non-cancer toxicity data - inhalationOnondaga Lake sediment containment area (SCA) sediments

Chemical Chronic/ Primary Combinedof Potential Subchronic Target Uncertainty/Modifying

Concern Value Units Value Units Organ Factors Source Date

Aluminum chronic 5.00E-03 mg/m3 NA CNS 300 PPRTV 10/23/2006Arsenic chronic 1.5E-05 mg/m3 NA Developmental NA Cal EPA 3/1/2009Antimony chronic 2.00E-04 mg/m3 NA Respiratory 300 IRIS 12/3/2002Barium chronic 5.0E-04 mg/m3 NA Developmental 1000/1 HEAST 7/31/1997Cadmium (food) chronic 1.0E-05 mg/m3 NA Kidney 3/3 ATSDR 9/1/2008Chromium (hexavalent particulate) chronic 1.0E-04 mg/m3 NA Respiratory 300/1 IRIS 10/28/2003Cobalt chronic 6.00E-06 mg/m3 NA Respiratory 300 PPRTV 8/25/2008Copper chronic ND mg/m3 NA NA NA IRIS 2/10/1998Manganese chronic 5.0E-05 mg/m3 NA CNS 1000/1 IRIS 12/3/2002Mercury (elemental) chronic 3.0E-04 mg/m3 NA CNS NA Cal EPA 3/1/2009Nickel (as soluble salts) chronic 9.0E-05 mg/m3 NA Respiratory 1/30/1900 ATSDR 8/1/2005Vanadium chronic ND NA NA NA PPRTV 9/30/2009Dibenzofuran chronic ND NA NA NA PPRTV 6/11/2007PCBs (as Aroclor 1254) chronic ND NA NA NA IRIS 4/1/1997Dieldrin chronic ND NA NA NA IRIS 3/7/20051,2,4-Trimethylbenzene chronic 0.007 mg/m3 NA Blood 3000/1 PPRTV 6/11/20071,2-Dichlorobenzene chronic 0.2 mg/m3 NA Whole Body 1000 HEAST 6/19/19051,4-Dichlorobenzene chronic 0.8 mg/m3 NA Liver 100/1 IRIS 1/12/20001,1-Dichloroethane chronic ND NA NA NA PPRTV 9/27/20061,1-Dichloroethene chronic 0.2 mg/m3 NA Liver 30/1 IRIS 6/22/20051,2,3,4/5 Tetrachlorobenzene chronic ND NA NA NA IRIS 12/3/2002Pentachlorobenzene chronic ND NA NA NA IRIS 10/28/20031,2-Dichloroethane chronic 2.4 mg/m3 NA Liver 90 ATSDR 5/11/2001trans-1,2-Dichloroethene chronic 0.06 mg/m3 NA Respiratory 3000 provisional 3/1/2006Acetone chronic 31 mg/m3 NA CNS 100 ATSDR 5/1994Benzene chronic 0.03 mg/m3 NA Blood 300/1 IRIS 10/1/20082-Butanone chronic 5 mg/m3 NA Developmental (skeletal) 300/1 IRIS 9/26/2003Carbon tetrachloride chronic 0.19 mg/m3 NA Liver 30 ATSDR 9/2005Chlorobenzene chronic 0.05 mg/m3 NA Liver 1000 PPRTV 12/1/2003Chloroform chronic 0.098 mg/m3 NA Liver 100 ATSDR 3/19/1997Cyanide (hydrogen cyanide) chronic 0.003 mg/m3 NA CNS/thyroid 1000/1 IRIS 1/9/20022-Methyl napthalene chronic ND NA NA NA IRIS 3/5/2007Ethylbenzene chronic 1 mg/m3 NA Developmental 300/1 IRIS 12/10/1998Toluene chronic 5 mg/m3 NA CNS 10/1 IRIS 4/3/2007Hexachlorobenzene chronic ND NA NA NA IRIS 10/28/2003Methylene chloride chronic 1 mg/m3 NA Liver 30 ATSDR 9/1/2000Methyl Tert-Butyl Ether chronic 3 mg/m3 NA Liver 100 IRIS 12/10/1998Pentachlorophenol chronic ND NA NA NA IRIS 1/2/1998Tetrachloroethene chronic 0.27 mg/m3 NA CNS 100 ATSDR 9/1/1997Trichlorobenzene, 1,2,4- chronic 0.002 mg/m3 NA Urinary 3000 PPRTV 6/19/2009Trichloroethane, 1,1,1- chronic 5 mg/m3 NA Blood 100/1 IRIS 9/28/2007Trichloroethene chronic ND NA NA NA IRIS 6/7/2004Trimethylbenzene, 1,3,5- chronic ND NA NA NA PPRTV 4/22/2009Vinyl chloride chronic 0.1 mg/m3 NA Liver 30/1 IRIS 10/28/2003Xylenes (total) chronic 0.1 mg/m3 NA CNS 300/1 IRIS 2/21/2003Benz[a]anthracene chronic ND NA NA NA IRIS 11/1/1994Benzo[a]pyrene chronic ND NA NA NA IRIS 12/10/1998Benzo[b]fluoranthene chronic ND NA NA NA IRIS 11/1/1994Benzo[k]fluoranthene chronic ND NA NA NA IRIS 11/1/1994Chrysene chronic ND NA NA NA IRIS 11/1/1994Dibenz[a,h]anthracene chronic ND NA NA NA IRIS 11/1/1994Fluorene chronic ND NA NA NA IRIS 4/1/1997Fluoranthene chronic ND NA NA NA IRIS 4/1/1997Indeno[1,2,3-cd]pyrene chronic ND NA NA NA IRIS 11/1/1994Naphthalene chronic 3.0E-03 mg/m3 NA Respiratory 3000/1 IRIS 9/1/2002Notes:

(1) RfD calculated from RfC based on an adult inhalation rate of 20 m3/d and a body weight of 70 kg (RfD = RfC * 20 m3/d * (1/70 kg))(2) Elemental Mercury was used as a surrogate.

ATSDR - Agency for Toxic Substances and Disease Registry ND - Not DevelopedHEAST - Health Effects Assessment Summary Tables PPRTV - Provisional Peer Reviewed Toxicity ValuesIRIS - Integrated Risk Information System RfD - Reference doseNA - Not Applicable

Inhalation RfC Extrapolated RfD RfC:Target Organs

Page 18 of 32

Table 6-1. Cancer toxicity data - oral/dermalOnondaga Lake sediment containment area (SCA) sediments

Chemical Oral Absorption Weight of Evidence/of Potential Efficiency for Dermal Cancer Guideline

Concern Value Units Value Units Description Source Date

Aluminum NA NA Not Evaluated PPRTV 10/23/2006Arsenic 1.5E+00 (mg/kg-day)-1 95% 1.5E+00 (mg/kg-day)-1 A IRIS 2/1/2003Antimony NA Not Evaluated PPRTV 7/29/2008Barium NA NA NA (mg/kg-day)-1 D IRIS 7/11/2005Cadmium (food) NA NA NA Not Evaluated IRIS 1/2/1998Chromium (as hexavalent) 5.0E-01 (mg/kg-day)-1 NA NA (mg/kg-day)-1 Likely NJDEP 7/1/1905Cobalt NA Not Evaluated PPRTV 8/25/2008Copper NA NA NA D IRIS 12/10/1998Manganese NA NA NA D IRIS 12/3/2002Mercury (as mercuric chloride) NA NA NA C IRIS 12/3/2002Nickel (as soluble salts) NA NA NA Not Evaluated IRIS 12/10/1998Vanadium NA NA NA Not Evaluated IRIS 12/3/2002Dibenzofuran NA NA NA D IRIS 12/3/2002Dioxins (as TCDD Equivalents) 1.3E+05 (mg/kg-day)-1 50% - 83% 1.3E+05 (mg/kg-day)-1 B2 CalEPA 5/1/2009PCBs 2.0E+00 (mg/kg-day)-1 80 - 96% 2.0E+00 (mg/kg-day)-1 B2 IRIS 9/4/2007Dieldrin 16 (mg/kg-day)-1 16 (mg/kg-day)-1 B2 IRIS 3/7/20051,2,4-Trimethylbenzene NA NA NA Not Evaluated IRIS 4/30/20101,2-Dichlorobenzene NA NA NA D IRIS 1/12/20001,4-Dichlorobenzene 5.4E-03 (mg/kg-day)-1 NA 5.4E-03 (mg/kg-day)-1 2B (IARC) CalEPA 5/1/2009Benzene 5.5E-02 (mg/kg-day)-1 >50% 5.5E-02 (mg/kg-day)-1 A IRIS 1/9/2000Chlorobenzene NA NA NA D IRIS 10/28/2003Ethylbenzene 1.1E-02 (mg/kg-day)-1 NA 1.1E-02 (mg/kg-day)-1 Not Classified CalEPA May-09Hexachlorobenzene 1.6E+00 (mg/kg-day)-1 >50% 1.6E+00 (mg/kg-day)-1 B2 IRIS 10/28/20031,2,3,4/5 Tetrachlorobenzene NA Not Evaluated IRIS 12/3/2002Pentachlorobenzene NA D IRIS 10/28/2003Methylene chloride 7.5E-03 (mg/kg-day)-1 >50% 7.5E-03 (mg/kg-day)-1 B2 IRIS 7/30/2003Pentachlorophenol 1.2E-01 (mg/kg-day)-1 76% - 100% (mg/kg-day)-1 B2 IRIS 1/2/1998Tetrachloroethene 5.4E-01 (mg/kg-day)-1 >50% 5.4E-01 (mg/kg-day)-1 2A (IARC) CalEPA 5/1/20091,2,4-Trichlorobenzene NA D IRIS 12/3/2002Vinyl chloride 1.4E+00 (mg/kg-day)-1 >50% 1.4E+00 (mg/kg-day)-1 A IRIS 10/28/2003Xylenes (total) NA Data are inadequate IRIS 2/21/20032-Methyl napthalene NA Data are inadequate IRIS 3/5/2007Benz[a]anthracene 7.3E-01 (mg/kg-day)-1 1 7.3E-01 (mg/kg-day)-1 B2 NCEA 1 11/1/1994Benzo[a]pyrene 7.3E+00 (mg/kg-day)-1 1 7.3E+00 (mg/kg-day)-1 B2 IRIS 1 12/10/1998Benzo[b]fluoranthene 7.3E-01 (mg/kg-day)-1 1 7.3E-01 (mg/kg-day)-1 B2 NCEA 1 11/1/1994Benzo[k]fluoranthene 7.3E-02 (mg/kg-day)-1 1 7.3E-02 (mg/kg-day)-1 B2 NCEA 1 11/1/1994Chrysene 7.3E-03 (mg/kg-day)-1 1 7.3E-03 (mg/kg-day)-1 B2 NCEA 1 11/1/1994Dibenz[a,h]anthracene 7.3E+00 (mg/kg-day)-1 1 7.3E+00 (mg/kg-day)-1 B2 NCEA 1 11/1/1994Fluorene NA NA NA D IRIS 4/1/0997Fluoranthene NA NA D IRIS 4/1/0997Indeno[1,2,3-cd]pyrene 7.3E-01 (mg/kg-day)-1 1 7.3E-01 (mg/kg-day)-1 B2 NCEA 1 11/1/1994Naphthalene NA NA NA C IRIS 9/1/2002Notes: U.S. EPA weight of evidence group:

CalEPA - California Environmental Protection Agency A - Human carcinogen C - Possible human carcinogenHEAST - Health Effects Assessment Summary Tables B1 - Probable human carcinogen - indicates that limited human D - Not classifiable as a human carcinogenIRIS - Integrated Risk Information System data are available E - Evidence of noncarcinogenicityNA - not available B2 - Probable human carcinogen - indicates sufficient evidenceNCEA - National Center for Environmental Assessment in animals and inadequate or no evidence in humans

1 This chemical operates with a mutagenic mode of action (USEPA 2005). Chemical-specific data are not available, thus, USEPA (2005) default age-dependant adjustment factors (ADAF) will be applied to the slope factor as follows Age Age ADAF

0-<2 102-<16 3

16-<30 1

Oral Cancer Slope Factor Absorbed Cancer Slope Factor Oral Cancer Slope Factorfor Dermal

Page 19 of 32

Table 6-2. Cancer toxicity data - inhalationOnondaga Lake sediment containment area (SCA) sediments

Chemical Unit Risk Weight of Evidence/of Potential Cancer Guideline

Concern Value Units Value Units Description Source Date

Aluminum ND NA Not Evaluated PPRTV 10/23/2006Arsenic 4.3E+00 per mg/m3 NA A IRIS 2/1/2003Antimony ND NA Not Evaluated IRIS 12/3/2002Barium ND NA D IRIS 7/11/2005Cadmium 1.8E+00 per mg/m3 NA B1 IRIS 1/2/1998Chromium (as hexavalent) 8.4E-05 per mg/m3 NA A IRIS 10/28/2003Cobalt ND NA Likely to be carcinogenic to humans PPRTV 8/25/2008Copper NA NA D IRIS 2/10/1998Manganese ND NA D IRIS 12/3/2002Mercury (as mercuric chloride) ND NA C IRIS 12/3/2002Nickel (as soluble salts) ND NA Not Evaluated IRIS 12/10/1998Vanadium ND NA Not Evaluated IRIS 12/3/2002Dibenzofuran ND NA D IRIS 12/3/2002PCBs 1.0E-01 per mg/m3 NA B2 IRIS 9/4/2007Dieldrin 4.6E+00 per mg/m3 NA B2 IRIS 3/7/20051,2,4-Trimethylbenzene ND NA Not Evaluated IRIS 4/30/20101,2-Dichlorobenzene ND NA D IRIS 1/12/20001,4-Dichlorobenzene 1.1E-02 per mg/m3 NA 2B (IARC) CalEPA 20091,1-Dichloroethane 1.6E-03 per mg/m3 NA C CalEPA 19921,1-Dichloroethene ND NA C IRIS 8/13/20021,2,3,4/5 Tetrachlorobenzene ND NA Not evaluated IRIS 12/3/2002Pentachlorobenzene NA D IRIS 10/28/20031,2-Dichloroethane 2.6E-02 per mg/m3 NA B2 IRIS 1/1/1991trans-1,2-Dichloroethene ND NA Not Evaluated IRIS 2/9/2004Acetone ND NA Data are inadequate IRIS 7/31/2003Benzene 7.8E-03 per mg/m3 NA A IRIS 1/9/20002-Butanone ND NA Data are inadequate IRIS 9/26/2003Carbon tetrachloride 6.0E-03 per mg/m3 NA Likely to be carcinogenic to humans IRIS 3/31/2010Chlorobenzene NA NA D IRIS 10/28/2003Chloroform 2.3E-02 per mg/m3 NA B2 IRIS 10/19/2001Cyanide (hydrogen cyanide, 74-90- ND NA Not Evaluated IRIS 1/9/20022-Methyl napthalene NA Data are inadequate IRIS 3/5/2007Ethylbenzene 2.5E-03 per mg/m3 NA Not Classified CalEPA 11/14/07Toluene ND NA Data are inadequate IRIS 9/23/2005Hexachlorobenzene 4.6E-01 per mg/m3 NA B2 IRIS 10/28/2003Methylene chloride 4.7E-04 per mg/m3 NA B2 IRIS 7/30/2003Methyl Tert-Butyl Ether 2.6E-04 per mg/m3 NA 3 (IARC) CalEPA 3/1/1999Pentachlorophenol 5.1E-03 per mg/m3 NA B2 CalEPA 7/1/1905Tetrachloroethene 5.9E-03 per mg/m3 NA 2A (IARC) CalEPA 7/1/1905Trichlorobenzene, 1,2,4- ND NA D IRIS 3/1/1991Trichloroethane, 1,1,1- ND NA Data are inadequate IRIS 9/28/2007Trichloroethene 2.0E-03 per mg/m3 NA 2A (IARC) CalEPA 7/1/1905Trimethylbenzene, 1,3,5- ND NA Data are inadequate PPRTV 4/22/2009Vinyl chloride 8.8E-03 per mg/m3 NA A IRIS 10/28/2003Xylenes (total) NA NA Data are inadequate IRIS 2/21/2003Benz[a]anthracene 1.1E-01 per mg/m3 NA B2 CalEPA (1) 11/1/1994Benzo[a]pyrene 1.1E+00 per mg/m3 NA B2 CalEPA (1) 12/10/1998Benzo[b]fluoranthene 1.1E-01 per mg/m3 NA B2 CalEPA (1) 11/1/1994Benzo[k]fluoranthene 1.1E-01 per mg/m3 NA B2 CalEPA (1) 11/1/1994Chrysene 1.1E-02 per mg/m3 NA B2 CalEPA (1) 11/1/1994Dibenz[a,h]anthracene 1.2E+00 per mg/m3 NA B2 CalEPA (1) 11/1/1994Fluoranthene NA D IRIS 4/1/1997Fluorene NA NA Data are inadequate IRIS 4/1/1997Indeno[1,2,3-cd]pyrene 1.1E-01 per mg/m3 NA B2 CalEPA (1) 11/1/1994Naphthalene 3.4E-02 per mg/m3 NA A CalEPA 8/3/2004Dioxins (as TCDD Equivalents) 38000 NA B2 CalEPA (1) 2003Notes:

CalEPA - California Environmental Agency NA U.S. EPA weight of evidence group:HEAST - Health Effects Assessment Summary Tables A - Human carcinogen C - Possible human carcinogenIRIS - Integrated Risk Information System B1 - Probable human carcinogen - indicates that limited D - Not classifiable as a human carcinogeNA - Not Available human data are available E - Evidence of noncarcinogenicityNCEA - National Center for Environmental Assessment B2 - Probable human carcinogen - indicates sufficient ND - Not Developed evidence in animals and inadequate or no evidence in humans

(1) This chemical operates with a mutagenic mode of action (USEPA 2005). Chemical-specific data are not available, thus, USEPA (2005) default age-dependant adjustment factors (ADAF)will be applied to the slope factor as follows:

Age Age ADAF

0-<2 10

2-<16 3

16-<30 1

Unit Risk: Inhalation CSF

per mg/m3

Inhalation Cancer Slope Factor

Page 20 of 32

Scenario Timeframe: Future

Receptor Population: Residential

Receptor Age: Child Ages 0-6

Medium Chemical of EPC Cancer Risk Calculations Non-Cancer Hazard CalculationsPotential Concern Value Units Intake/Exposure Concentration CSF/Unit Risk Cancer Risk Intake/Exposure Concentration RfD/RfC Hazard Quotient

Value Units Value Units Value Units Value Units

Aluminum 4,700 mg/kg 1.1E-4 mg/kg-day NA -- -- 7.7E-3 mg/kg 1.0 mg/kg-day 0.008Sediment Surface soils Surface soils Ingestion Antimony 4.7 mg/kg 1.1E-7 mg/kg-day NA -- -- 7.7E-6 mg/kg 0.0004 mg/kg-day 0.02Containment Area Arsenic 10 mg/kg 2.3E-7 mg/kg-day 1.5 (mg/kg-day)-1 4E-7 1.6E-5 mg/kg 0.0003 mg/kg-day 0.05Sediments Barium 1,200 mg/kg 2.8E-5 mg/kg-day NA -- -- 2.0E-3 mg/kg 0.20 mg/kg-day 0.01

Cadmium 7.2 mg/kg 1.7E-7 mg/kg-day NA -- -- 1.2E-5 mg/kg 0.001 mg/kg-day 0.01Chromium (as Cr VI) 790 mg/kg 1.9E-5 mg/kg-day 0.5 (mg/kg-day)-1 9E-6 1.3E-3 mg/kg 0.003 mg/kg-day 0.4Cobalt 48 mg/kg 1.1E-6 mg/kg-day NA -- -- 7.9E-5 mg/kg 0.0003 mg/kg-day 0.3Copper 150 mg/kg 3.5E-6 mg/kg-day NA -- -- 2.5E-4 mg/kg 0.04 mg/kg-day 0.006Manganese 580 mg/kg 1.4E-5 mg/kg-day NA -- -- 9.5E-4 mg/kg 0.14 mg/kg-day 0.007Mercury (as mercuric chloride) 15 mg/kg 3.5E-7 mg/kg-day NA -- -- 2.5E-5 mg/kg 0.000 mg/kg-day 0.08Nickel (as soluble salts) 350 mg/kg 8.2E-6 mg/kg-day NA -- -- 5.8E-4 mg/kg 0.02 mg/kg-day 0.03Vanadium 78 mg/kg 1.8E-6 mg/kg-day NA -- -- 1.3E-4 mg/kg 0.01 mg/kg-day 0.01Dibenzofuran 4.8 mg/kg 1.1E-7 mg/kg-day NA -- -- 7.9E-6 mg/kg ND mg/kg-day --Dieldrin 0.005 mg/kg 1.2E-10 mg/kg-day 16 (mg/kg-day)-1 2E-9 8.2E-9 mg/kg 0.00005 mg/kg-day 0.0002Dioxins (as TCDD Equivalents) 1.3E-04 mg/kg 3.0E-12 mg/kg-day 130000 (mg/kg-day)-1 4E-7 2.1E-10 mg/kg 0.000000001 mg/kg-day 0.2PCBs (Aroclor 1254 noncancer) 1.3 mg/kg 3.1E-8 mg/kg-day 2.0 (mg/kg-day)-1 6E-8 2.1E-6 mg/kg 0.00002 mg/kg-day 0.11,2,3/4,5 Tetrachlorobenzene 21 mg/kg 4.9E-7 mg/kg-day NA -- -- 3.5E-5 mg/kg 0.0003 mg/kg-day 0.11,2,4-Trichlorobenzene 17 mg/kg 3.9E-7 mg/kg-day NA -- -- 2.8E-5 mg/kg 0.010 mg/kg-day 0.0031,2,4-Trimethylbenzene 19 mg/kg 4.5E-7 mg/kg-day NA -- -- 3.1E-5 mg/kg ND -- --1,2-Dichlorobenzene 38 mg/kg 8.9E-7 mg/kg-day NA -- -- 6.2E-5 mg/kg 0.09 mg/kg-day 0.00071,4-Dichlorobenzene 84 mg/kg 2.0E-6 mg/kg-day 0.01 (mg/kg-day)-1 1E-8 1.4E-4 mg/kg 0.07 mg/kg-day 0.002Benzene 4.7 mg/kg 1.1E-7 mg/kg-day 0.055 (mg/kg-day)-1 6E-9 7.7E-6 mg/kg 0.004 mg/kg-day 0.002Chlorobenzene 52 mg/kg 1.2E-6 mg/kg-day NA -- -- 8.5E-5 mg/kg 0.02 mg/kg-day 0.004Ethylbenzene 12 mg/kg 2.8E-7 mg/kg-day 0.011 (mg/kg-day)-1 3E-9 2.0E-5 mg/kg 0.1 mg/kg-day 0.0002Hexachlorobenzene 6.4 mg/kg 1.5E-7 mg/kg-day 1.6 (mg/kg-day)-1 2E-7 1.1E-5 mg/kg 0.001 mg/kg-day 0.01Pentachlorobenzene 6.7 mg/kg 1.6E-7 mg/kg-day 0.011 (mg/kg-day)-1 2E-9 1.1E-5 mg/kg 0.0008 mg/kg-day 0.01Xylenes, total 45 mg/kg 1.1E-6 mg/kg-day NA -- -- 7.4E-5 mg/kg 0.2 mg/kg-day 0.00042-Methylnaphthalene 13 mg/kg 3.1E-7 mg/kg-day NA -- -- 2.1E-5 mg/kg 0.004 mg/kg-day 0.01Benz[a]anthracene 4.0 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 6.6E-6 mg/kg ND -- --Benzo[a]pyrene 3.2 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 5.3E-6 mg/kg ND -- --Benzo[b]fluoranthene 3.2 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 5.3E-6 mg/kg ND -- --Benzo[k]fluoranthene 1.5 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 2.5E-6 mg/kg ND -- --Chrysene 3.9 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 6.4E-6 mg/kg ND -- --Dibenz[a,h]anthracene 0.58 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 9.5E-7 mg/kg ND -- --Fluoranthene 8.2 mg/kg 1.9E-7 mg/kg-day -- -- 1.3E-5 mg/kg 0.04 mg/kg-day 0.0003Fluorene 10 mg/kg 2.3E-7 mg/kg-day NA -- -- 1.6E-5 mg/kg 0.04 mg/kg-day 0.0004Indeno[1,2,3-cd]pyrene 1.5 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 2.5E-6 mg/kg ND -- --Naphthalene 130 mg/kg a mg/kg-day a -- 2.1E-4 mg/kg 0.02 mg/kg-day 0.01

Exp. Route Total 1E-5 1.4Dermal Arsenic 10 mg/kg 2.0E-8 mg/kg-day 1.5 (mg/kg-day)-1 3E-8 1.4E-6 mg/kg 0.0003 mg/kg-day 0.005

Cadmium 7.2 mg/kg 4.7E-10 mg/kg-day NA -- -- 3.3E-8 mg/kg 0.00003 mg/kg-day 0.001Dieldrin 0.005 mg/kg 3.3E-11 mg/kg-day 16 (mg/kg-day)-1 5E-10 2.3E-9 mg/kg 0.00002 mg/kg-day 0.0001Dioxins (as TCDD Equivalents) 1.3E-04 mg/kg 1.2E-12 mg/kg-day 130000 (mg/kg-day)-1 2E-7 8.3E-11 mg/kg 0.000000001 mg/kg-day 0.08PCBs (Aroclor 1254 noncancer) 1.3 mg/kg 8.5E-9 mg/kg-day 2.0 (mg/kg-day)-1 2E-8 8.4E-7 mg/kg 0.00002 mg/kg-day 0.04Hexachlorobenzene 6.4 mg/kg 4.2E-8 mg/kg-day 1.6 (mg/kg-day)-1 7E-8 2.9E-6 mg/kg 0.0008 mg/kg-day 0.004Benz[a]anthracene 4.0 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 2.4E-6 mg/kg ND -- --Benzo[a]pyrene 3.2 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 1.9E-6 mg/kg ND -- --Benzo[b]fluoranthene 3.2 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 1.9E-6 mg/kg ND -- --Benzo[k]fluoranthene 1.5 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 9.0E-7 mg/kg ND -- --Chrysene 3.9 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 2.3E-6 mg/kg ND -- --Dibenz[a,h]anthracene 0.58 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 3.5E-7 mg/kg ND -- --Fluoranthene 8.2 mg/kg 7.0E-8 mg/kg-day NA 4.9E-6 mg/kg 0.04 mg/kg-day 0.0001Fluorene 10 mg/kg 8.5E-8 mg/kg-day NA -- -- 6.0E-6 mg/kg 0.04 mg/kg-day 0.0001Indeno[1,2,3-cd]pyrene 1.5 mg/kg a mg/kg-day a (mg/kg-day)-1 -- 9.0E-7 mg/kg ND -- --Naphthalene 130 mg/kg 1.1E-6 mg/kg-day NA -- -- 7.8E-5 mg/kg 0.02 mg/kg-day 0.004

Exp. Route Total 3E-7 0.1

1E-5 2

Exposure Medium Total 1E-5 2

Medium Total 1E-5 2

Total of Receptor Risks Across All Media 1E-5 Total of Receptor Hazards Across All Media 2

a See Supplement A for mutagenic mode of action cancer risk calculations for PAHs.

Table 7-1. Calculation of chemical cancer risks and non-cancer hazardsReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Exposure Route

Exposure Point Total

Exposure Medium

Exposure Point

Page 21 of 32

Scenario Timeframe: Future

Receptor Population: Residential

Receptor Age: Adult

Medium Chemical of EPC Cancer Risk Calculations Non-Cancer Hazard CalculationsPotential Concern Value Units Intake/Exposure Concentration CSF/Unit Risk Cancer Risk Intake/Exposure Concentration RfD/RfC Hazard Quotient

Value Units Value Units Value Units Value Units

Aluminum 4,700 mg/kg 1.2E-5 mg/kg-day NA -- -- 8.3E-4 mg/kg 1.0 mg/kg-day 0.001Sediment Surface soils Surface soils Ingestion Antimony 4.7 mg/kg 1.2E-8 mg/kg-day NA -- -- 8.3E-7 mg/kg 0.0004 mg/kg-day 0.002Containment Area Arsenic 10 mg/kg 2.5E-8 mg/kg-day 1.5 (mg/kg-day)-1 4E-8 1.8E-6 mg/kg 0.0003 mg/kg-day 0.006Sediments Barium 1,200 mg/kg 3.0E-6 mg/kg-day NA -- -- 2.1E-4 mg/kg 0.20 mg/kg-day 0.001

Cadmium 7.2 mg/kg 1.8E-8 mg/kg-day NA -- -- 1.3E-6 mg/kg 0.001 mg/kg-day 0.001Chromium (as Cr VI) 790 mg/kg 2.0E-6 mg/kg-day 0.5 (mg/kg-day)-1 1E-6 1.4E-4 mg/kg 0.003 mg/kg-day 0.05Cobalt 48 mg/kg 1.2E-7 mg/kg-day NA -- -- 8.5E-6 mg/kg 0.0003 mg/kg-day 0.03Copper 150 mg/kg 3.8E-7 mg/kg-day NA -- -- 2.6E-5 mg/kg 0.04 mg/kg-day 0.0007Manganese 580 mg/kg 1.5E-6 mg/kg-day NA -- -- 1.0E-4 mg/kg 0.14 mg/kg-day 0.0007Mercury (as mercuric chloride) 15 mg/kg 3.8E-8 mg/kg-day NA -- -- 2.6E-6 mg/kg 0.000 mg/kg-day 0.009Nickel (as soluble salts) 350 mg/kg 8.8E-7 mg/kg-day NA -- -- 6.2E-5 mg/kg 0.02 mg/kg-day 0.003Vanadium 78 mg/kg 2.0E-7 mg/kg-day NA -- -- 1.4E-5 mg/kg 0.01 mg/kg-day 0.002Dibenzofuran 4.8 mg/kg 1.2E-8 mg/kg-day NA -- -- 8.5E-7 mg/kg ND mg/kg-day --Dieldrin 0.005 mg/kg 1.3E-11 mg/kg-day 16 (mg/kg-day)-1 2E-10 8.8E-10 mg/kg 0.00005 mg/kg-day 0.00002Dioxins (as TCDD Equivalents) 1.3E-04 mg/kg 3.2E-13 mg/kg-day 130000 (mg/kg-day)-1 4E-8 2.3E-11 mg/kg 0.000000001 mg/kg-day 0.02PCBs (Aroclor 1254 noncancer) 1.3 mg/kg 3.3E-9 mg/kg-day 2.0 (mg/kg-day)-1 7E-9 2.3E-7 mg/kg 0.00002 mg/kg-day 0.011,2,3/4,5 Tetrachlorobenzene 21 mg/kg 5.3E-8 mg/kg-day NA -- -- 3.7E-6 mg/kg 0.0003 mg/kg-day 0.011,2,4-Trichlorobenzene 17 mg/kg 4.2E-8 mg/kg-day NA -- -- 3.0E-6 mg/kg 0.010 mg/kg-day 0.00031,2,4-Trimethylbenzene 19 mg/kg 4.8E-8 mg/kg-day NA -- -- 3.3E-6 mg/kg ND -- --1,2-Dichlorobenzene 38 mg/kg 9.6E-8 mg/kg-day NA -- -- 6.7E-6 mg/kg 0.09 mg/kg-day 0.000071,4-Dichlorobenzene 84 mg/kg 2.1E-7 mg/kg-day 0.01 (mg/kg-day)-1 1E-9 1.5E-5 mg/kg 0.07 mg/kg-day 0.0002Benzene 4.7 mg/kg 1.2E-8 mg/kg-day 0.055 (mg/kg-day)-1 7E-10 8.3E-7 mg/kg 0.004 mg/kg-day 0.0002Chlorobenzene 52 mg/kg 1.3E-7 mg/kg-day NA -- -- 9.2E-6 mg/kg 0.02 mg/kg-day 0.0005Ethylbenzene 12 mg/kg 3.0E-8 mg/kg-day 0.011 (mg/kg-day)-1 3E-10 2.1E-6 mg/kg 0.1 mg/kg-day 0.00002Hexachlorobenzene 6.4 mg/kg 1.6E-8 mg/kg-day 1.6 (mg/kg-day)-1 3E-8 1.1E-6 mg/kg 0.001 mg/kg-day 0.001Pentachlorobenzene 6.7 mg/kg 1.7E-8 mg/kg-day 0.011 (mg/kg-day)-1 2E-10 1.2E-6 mg/kg 0.0008 mg/kg-day 0.001Xylenes, total 45 mg/kg 1.1E-7 mg/kg-day NA -- -- 7.9E-6 mg/kg 0.2 mg/kg-day 0.000042-Methylnaphthalene 13 mg/kg 3.3E-8 mg/kg-day NA -- -- 2.3E-6 mg/kg 0.004 mg/kg-day 0.001Benz[a]anthracene 4.0 mg/kg a mg/kg-day a -- -- 7.0E-7 mg/kg ND -- --Benzo[a]pyrene 3.2 mg/kg a mg/kg-day a -- -- 5.6E-7 mg/kg ND -- --Benzo[b]fluoranthene 3.2 mg/kg a mg/kg-day a -- -- 5.6E-7 mg/kg ND -- --Benzo[k]fluoranthene 1.5 mg/kg a mg/kg-day a -- -- 2.6E-7 mg/kg ND -- --Chrysene 3.9 mg/kg a mg/kg-day a -- -- 6.9E-7 mg/kg ND -- --Dibenz[a,h]anthracene 0.58 mg/kg a mg/kg-day a -- -- 1.0E-7 mg/kg ND -- --Fluoranthene 8.2 mg/kg 2.1E-8 mg/kg-day NA -- -- 1.4E-6 mg/kg 0.04 mg/kg-day 0.00004Fluorene 10 mg/kg 2.5E-8 mg/kg-day NA -- -- 1.8E-6 mg/kg 0.04 mg/kg-day 0.00004Indeno[1,2,3-cd]pyrene 1.5 mg/kg a mg/kg-day a -- -- 2.6E-7 mg/kg ND -- --Naphthalene 130 mg/kg a mg/kg-day a -- 2.3E-5 mg/kg 0.02 mg/kg-day 0.001

Exp. Route Total 1E-6 0.2Dermal Arsenic 10 mg/kg 3.0E-9 mg/kg-day 1.5 (mg/kg-day)-1 5E-9 2.1E-7 mg/kg 0.0003 mg/kg-day 0.0007

Cadmium 7.2 mg/kg 7.2E-11 mg/kg-day NA -- -- 5.1E-9 mg/kg 0.00003 mg/kg-day 0.0002Dieldrin 0.005 mg/kg 5.0E-12 mg/kg-day 16 (mg/kg-day)-1 8E-11 3.5E-10 mg/kg 0.00002 mg/kg-day 0.00002Dioxins (as TCDD Equivalents) 1.3E-04 mg/kg 1.8E-13 mg/kg-day 130000 (mg/kg-day)-1 2E-8 1.3E-11 mg/kg 0.000000001 mg/kg-day 0.01PCBs (Aroclor 1254 noncancer) 1.3 mg/kg 1.8E-9 mg/kg-day 2.0 (mg/kg-day)-1 4E-9 1.3E-7 mg/kg 0.00002 mg/kg-day 0.006Hexachlorobenzene 6.4 mg/kg 6.4E-9 mg/kg-day 1.6 (mg/kg-day)-1 1E-8 4.5E-7 mg/kg 0.0008 mg/kg-day 0.0006Benz[a]anthracene 4.0 mg/kg a mg/kg-day a -- -- 3.7E-7 mg/kg ND -- --Benzo[a]pyrene 3.2 mg/kg a mg/kg-day a -- -- 2.9E-7 mg/kg ND -- --Benzo[b]fluoranthene 3.2 mg/kg a mg/kg-day a -- -- 2.9E-7 mg/kg ND -- --Benzo[k]fluoranthene 1.5 mg/kg a mg/kg-day a -- -- 1.4E-7 mg/kg ND -- --Chrysene 3.9 mg/kg a mg/kg-day a -- -- 3.6E-7 mg/kg ND -- --Dibenz[a,h]anthracene 0.58 mg/kg a mg/kg-day a -- -- 5.3E-8 mg/kg ND -- --Fluoranthene 8.2 mg/kg 1.1E-8 mg/kg-day NA -- -- 7.5E-7 mg/kg 0.04 mg/kg-day 0.00002Fluorene 10 mg/kg 1.0E-8 mg/kg-day NA -- -- 9.1E-7 mg/kg 0.04 mg/kg-day 0.00002Indeno[1,2,3-cd]pyrene 1.5 mg/kg a mg/kg-day a -- -- 1.4E-7 mg/kg ND -- --Naphthalene 130 mg/kg a mg/kg-day a -- -- 1.2E-5 mg/kg 0.02 mg/kg-day 0.0006

Exp. Route Total 4E-8 0.02Exposure Point Total 1E-6 0.2

Exposure Medium Total 1E-6 0.2Medium Total 1E-6 0.2

Total of Receptor Risks Across All Media 1E-6 Total of Receptor Hazards Across All Media 0.2

Exposure Route

Table 7-2. Calculation of chemical cancer risks and non-cancer hazardsReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Exposure Medium

Exposure Point

Page 22 of 32

Scenario Timeframe: FutureReceptor Population: Offsite residentReceptor Age: Adult and Child

Medium Chemical of EPC Cancer Risk Calculations Non-Cancer Hazard CalculationsPotential Concern Value Units Intake/Exposure Concentration CSF/Unit Risk Intake/Exposure Concentration RfD/RfC

Inhalation Value Units Value Units Value Units Value UnitsAir Air Chemicals detected in wind tunnel tests

1,4-Dichlorobenzene 0.00029 mg/m3 2.0E-5 mg/m3 1.1E-02 (mg/m3) -1 2E-7 2.8E-4 mg/m3 0.8 mg/m3 0.0004Area Sediments 1,2-Dichlorobenzene (ortho) 0.080 mg/m3 5.5E-3 mg/m3 ND -- -- 7.7E-2 mg/m3 0.2 mg/m3 0.4

1,2,4-Trichlorobenzene 0.0019 mg/m3 1.3E-4 mg/m3 ND -- -- 1.9E-3 mg/m3 0.002 mg/m3 0.9Benzene 0.00042 mg/m3 2.8E-5 mg/m3 7.8E-03 (mg/m3) -1 2E-7 4.0E-4 mg/m3 0.03 mg/m3 0.01Chlorobenzene 0.024 mg/m3 1.7E-3 mg/m3 NA -- -- 2.3E-2 mg/m3 0.05 mg/m3 0.5Ethylbenzene based on CalEPA IUR 0.0054 mg/m3 3.7E-4 mg/m3 2.5E-03 (mg/m3) -1 9E-7 5.2E-3 mg/m3 1.0 mg/m3 0.005Naphthalene (based on CalEPA IUR) 0.00040 mg/m3 2.7E-5 mg/m3 3.4E-02 (mg/m3) -1 9E-7 3.8E-4 mg/m3 0.003 mg/m3 0.1Toluene 1.1 mg/m3 7.6E-2 mg/m3 ND -- -- 1.1E+0 mg/m3 5.0 mg/m3 0.2Xylenes 0.022 mg/m3 1.5E-3 mg/m3 NA -- -- 2.1E-2 mg/m3 0.10 mg/m3 0.2

Additional volatile chemicals detected in sediments that have inhalation toxicity values -- --Cyanide (hydrogen cyanide, 74-90-8) 0.00067 mg/m3 4.6E-5 mg/m3 ND -- -- 6.4E-4 mg/m3 0.003 mg/m3 0.2Mercury (elemental) 0.000067 mg/m3 4.6E-6 mg/m3 ND -- -- 6.4E-5 mg/m3 0.0003 mg/m3 0.21,1,1-Trichloroethane 0.22 mg/m3 1.5E-2 mg/m3 ND -- -- 2.1E-1 mg/m3 5.0 mg/m3 0.041,1-Dichloroethane 0.0020 mg/m3 1.4E-4 mg/m3 1.6E-03 (mg/m3) -1 2E-7 1.9E-3 mg/m3 ND -- --1,1-Dichloroethene 0.016 mg/m3 1.1E-3 mg/m3 ND -- -- 1.5E-2 mg/m3 0.20 mg/m3 0.071,2-Dichloroethane 0.00012 mg/m3 8.5E-6 mg/m3 2.6E-02 (mg/m3) -1 2E-7 1.2E-4 mg/m3 2.4 mg/m3 0.000051,2,4-Trimethylbenzene 0.0068 mg/m3 4.7E-4 mg/m3 ND -- -- 6.5E-3 mg/m3 0.007 mg/m3 0.92-Butanone 1.1 mg/m3 7.6E-2 mg/m3 ND -- -- 1.1E+0 mg/m3 5.0 mg/m3 0.2Acetone 6.2 mg/m3 4.3E-1 mg/m3 ND -- -- 6.0E+0 mg/m3 31 mg/m3 0.2Carbon tetrachloride 0.00022 mg/m3 1.5E-5 mg/m3 6.0E-03 (mg/m3) -1 9E-8 2.1E-4 mg/m3 0.19 mg/m3 0.001Chloroform 0.00014 mg/m3 9.6E-6 mg/m3 2.3E-02 (mg/m3) -1 2E-7 1.4E-4 mg/m3 0.098 mg/m3 0.001Methyl Tert-Butyl Ether 0.012 mg/m3 8.5E-4 mg/m3 2.6E-04 (mg/m3) -1 2E-7 1.2E-2 mg/m3 3.0 mg/m3 0.004Methylene chloride 0.0069 mg/m3 4.7E-4 mg/m3 4.7E-04 (mg/m3) -1 2E-7 6.6E-3 mg/m3 1.0 mg/m3 0.007Tetrachloroethene 0.00055 mg/m3 3.8E-5 mg/m3 5.9E-03 (mg/m3) -1 2E-7 5.3E-4 mg/m3 0.27 mg/m3 0.002trans-1,2-Dichloroethene 0.014 mg/m3 9.6E-4 mg/m3 ND -- -- 1.3E-2 mg/m3 0.06 mg/m3 0.2Trichloroethene 0.0016 mg/m3 1.1E-4 mg/m3 2.0E-03 (mg/m3) -1 2E-7 1.6E-3 mg/m3 ND -- --Vinyl chloride 0.00074 mg/m3 5.0E-5 mg/m3 8.8E-03 (mg/m3) -1 4E-7 7.1E-4 mg/m3 0.10 mg/m3 0.007

Exp. Route Total 4E-6 4Exposure Point Total 4E-6 4

Exposure Medium Total 4E-6 4Medium Total 4E-6 4

Total of Receptor Risks Across All Media 4E-6 Total of Receptor Hazards Across All Media 4

Sediment Containment

Table 7-3. Calculation of chemical cancer risks and non-cancer hazardsReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Hazard Quotient

Cancer Risk

Exposure Medium

Exposure Point

Exposure Route

Page 23 of 32

CancerPercent

ContributionScenario / Exposure Pathway Risk by PathwayInhalation of Volatiles from Sediment Containment Area (SCA) Adult, Adolescent, Child

Offsite Migration to AirOffsite residents

Outdoor Air 4E-6 100%Direct Contact with Surface Soil (SCA sediments)

Hypothetical 45 day residence on SCAChemicals other than PAHs Child Ages 0-2

Ingestion of Surface Soil 2E-5 49%Dermal Contact with Surface Soil 4E-7 1%

2E-5Carcinogenic PAHs

Ingestion of Surface Soil 1E-5 38% PAHDermal Contact with Surface Soil 4E-6 13% PAH

Risk sum for children ages 0-2 3E-5 100%Chemicals other than PAHs Child Ages 2-6

Ingestion of Surface Soil 1E-5 74% Chromium VIDermal Contact with Surface Soil 3E-7 2%

1E-5Carcinogenic PAHs

Ingestion of Surface Soil 2E-6 17% PAHDermal Contact with Surface Soil 8E-7 6% PAH

Risk sum for children ages 2-6 1E-5 100%Chemicals other than PAHs Child Ages 6-16

Ingestion of Surface Soil 2E-6 67%Dermal Contact with Surface Soil 1E-7 4%

2E-6Carcinogenic PAHs (adolescent- ages 6-16)

Ingestion of Surface Soil 5E-7 16% PAHDermal Contact with Surface Soil 4E-7 13% PAH

Risk sum for children ages 6-16 3E-6 100%Chemicals other than PAHs Adult

Ingestion of Surface Soil 1E-6 86%Dermal Contact with Surface Soil 4E-8 3% Dioxins, hexachlorobenzene, arsenic

1E-6Carcinogenic PAHs (adult)

Ingestion of Surface Soil 9E-8 7% PAHDermal Contact with Surface Soil 5E-8 4% PAH

Risk sum for adults 1E-6 100%Notes: Risk estimates for carcinogens other than PAHs were calculated for the age groups by taking the ratio of the CDI for that age group over the CDI

for children (for age groups 0-2 and 2-6) or over the CDI for adults (for adolescents). (See below)

PAH - Polycyclic aromatic hydrocarbonsRatios used to adjust CDIs for chemicals other than PAHs

Ingestion Dermal ContactRatio of chronic daily intake for 0-2 child / 0-6 child 1.5 1.4Ratio of chronic daily intake for 2-6 child / 0-6 child 0.9 1.0Ratio of chronic daily intake for 6-16 adolescent / adult 2.0 3.2

Table 8-1. Summary of total excess lifetime cancer risks for reasonable maximum exposure scenarios

Chemicals with Primary Contribution to Riskfor each Pathway

Chromium VI

Dioxins, hexachlorobenzene, arsenic

Dioxins, hexachlorobenzene, arsenic

Dioxins, hexachlorobenzene, arsenic

Chromium VI

Ethylbenzene, naphthalene

Chromium VI

Page 24 of 32

Table 8-2. Summary of total hazard indices for reasonable maximum exposure scenarios air exposure to volatiles

HazardReceptor/Exposure Pathway Index

Inhalation of Volatiles: Sediment Containment Area (SCA) Soils

Offsite residents

Adult, Adolescent, and Child Chemicals with Primary Contribution to Hazard Index

Outdoor Air Hazard Indices by EndpointHazard Index for Blood Endpoint 1 1,2,4-TrimethylbenzeneHazard Index for CNS Endpoint 1 Mercury, toluene, acetone, xylenes, cyanide

Hazard Index for Developmental Endpoint: 0.2 2-ButanoneHazard Index for Liver 0.6 Chlorobenzene

Hazard Index for Respiratory 0.4 trans 1,2-Dichloroethene, naphthaleneHazard Index for Whole Body 0.4 1,2-Dichlorobenzene

Hazard Index for All Other Endpoints 0.9 1,2,4-Trichlorobenzene

Table 8-3. Summary of total hazard indices for reasonable maximum exposure scenarios sediment direct contact

HazardReceptor/Exposure Pathway Index

Direct contact: Sediment Containment Area (SCA) Soils PercentContribution Chemicals with Primary Contribution

Residents - Future Child Ages 1-6 by Pathway to Hazard IndexSurface Soil Hazard Indices by Endpoint

Ingestion of Surface SoilHazard Index for Blood Endpoint 0.02 AntimonyHazard Index for CNS Endpoint 0.2 Dioxins

Hazard Index for GI Endpoint 0.006 CopperHazard Index for Hair Cystine Endpoint 0.01 Vanadium

Hazard Index for Immune System Endpoint: 0.2 PCBs (Aroclor 1254, noncancer), mercuryHazard Index for Liver 0.03

Hazard Index for Kidney 0.2 1,2,3/4,5 TetrachlorobenzeneHazard Index for Whole Body 0.04 Nickel (as soluble salts)

Hazard Index for Skin 0.05 ArsenicHazard Index for Thyroid 0.3 CobaltHazard Index for NOAEL 0.4 Chromium (as Cr VI)

Hazard Index for Lung 0.005 2-MethylnaphthaleneTotal Percent Contribution: Ingestion of Surface Soil 91% Chromium VI, cobalt, dioxins

Dermal Contact with Surface SoilHazard Index for Blood Endpoint 0.0003 FluoreneHazard Index for CNS Endpoint 0.08 Dioxins

Hazard Index for Immune System Endpoint: 0.04 PCBs (Aroclor 1254 noncancer)Hazard Index for Liver 0.004 Hexachlorobenzene

Hazard Index for Kidney 0.001 CadmiumHazard Index for Whole Body 0.004 Naphthalene

Hazard Index for Skin 0.005 ArsenicTotal Percent Contribution: Dermal Contact with Surface Soil 9% Dioxin, PCBs

Total for Future Residents - Surface Soil: 100%Note:

PCB - Polychlorinated biphenyls

Pentachlorobenzene, hexachlorobenzene

Page 25 of 32

Supplement A

Table 7-1. Calculation of chemical cancer risksReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Scenario Timeframe: Future

Receptor Population: Residential

Receptor Age: Child Ages 0-2

Medium Chemical of EPC Cancer Risk CalculationsPotential Concern Value Units Intake/Exposure Concentration CSF/Unit Risk Cancer Risk

Surface soils Surface soilsIngestion Value Units Value Units

Benz[a]anthracene 4.0 mg/kg 1.4E-7 mg/kg-day 7.3 (mg/kg-day)-1 1E-6

Benzo[a]pyrene 3.2 mg/kg 1.1E-7 mg/kg-day 73 (mg/kg-day)-1 8E-6

Benzo[b]fluoranthene 3.2 mg/kg 1.1E-7 mg/kg-day 7.3 (mg/kg-day)-1 8E-7

Benzo[k]fluoranthene 1.5 mg/kg 5.2E-8 mg/kg-day 0.73 (mg/kg-day)-1 4E-8

Chrysene 3.9 mg/kg 1.4E-7 mg/kg-day 0.073 (mg/kg-day)-1 1E-8

Dibenz[a,h]anthracene 0.58 mg/kg 2.0E-8 mg/kg-day 73 (mg/kg-day)-1 1E-6

Indeno[1,2,3-cd]pyrene 1.5 mg/kg 5.2E-8 mg/kg-day 7.3 (mg/kg-day)-1 4E-7

Exp. Route Total 1E-5

Dermal Benz[a]anthracene 4.0 mg/kg 4.7E-8 mg/kg-day 7.3 (mg/kg-day)-1 3E-7

Benzo[a]pyrene 3.2 mg/kg 3.8E-8 mg/kg-day 73 (mg/kg-day)-1 3E-6

Benzo[b]fluoranthene 3.2 mg/kg 3.8E-8 mg/kg-day 7.3 (mg/kg-day)-1 3E-7

Benzo[k]fluoranthene 1.5 mg/kg 1.8E-8 mg/kg-day 0.73 (mg/kg-day)-1 1E-8

Chrysene 3.9 mg/kg 4.6E-8 mg/kg-day 0.073 (mg/kg-day)-1 3E-9

Dibenz[a,h]anthracene 0.58 mg/kg 6.9E-9 mg/kg-day 73 (mg/kg-day)-1 5E-7

Indeno[1,2,3-cd]pyrene 1.5 mg/kg 1.8E-8 mg/kg-day 7.3 (mg/kg-day)-1 1E-7

Exp. Route Total 4E-6

2E-5

Exposure Medium Total 2E-5

Medium Total 2E-5

Total of Receptor Risks Across All Media 2E-5

Sediment Containment Area Sediments

Exposure Point Total

Exposure Medium

Exposure Point

Exposure Route

Page 26 of 32

Supplement A

Table 7-2. Calculation of chemical cancer risksReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Scenario Timeframe: Future

Receptor Population: Residential

Receptor Age: Child Ages 2-6

Medium Chemical of EPC Cancer Risk CalculationsPotential Concern Value Units Intake/Exposure Concentration CSF/Unit Risk Cancer Risk

Surface soils Surface soilsIngestion Value Units Value Units

Benz[a]anthracene 4.0 mg/kg 8.8E-8 mg/kg-day 2.2 (mg/kg-day)-1 2E-7

Benzo[a]pyrene 3.2 mg/kg 7.0E-8 mg/kg-day 22 (mg/kg-day)-1 2E-6

Benzo[b]fluoranthene 3.2 mg/kg 7.0E-8 mg/kg-day 2.2 (mg/kg-day)-1 2E-7

Benzo[k]fluoranthene 1.5 mg/kg 3.3E-8 mg/kg-day 0.22 (mg/kg-day)-1 7E-9

Chrysene 3.9 mg/kg 8.5E-8 mg/kg-day 0.022 (mg/kg-day)-1 2E-9

Dibenz[a,h]anthracene 0.58 mg/kg 1.3E-8 mg/kg-day 22 (mg/kg-day)-1 3E-7

Indeno[1,2,3-cd]pyrene 1.5 mg/kg 3.3E-8 mg/kg-day 2.2 (mg/kg-day)-1 7E-8

Exp. Route Total 2E-6

Dermal Benz[a]anthracene 4.0 mg/kg 3.3E-8 mg/kg-day 2.2 (mg/kg-day)-1 7E-8

Benzo[a]pyrene 3.2 mg/kg 2.6E-8 mg/kg-day 22 (mg/kg-day)-1 6E-7

Benzo[b]fluoranthene 3.2 mg/kg 2.6E-8 mg/kg-day 2.2 (mg/kg-day)-1 6E-8

Benzo[k]fluoranthene 1.5 mg/kg 1.2E-8 mg/kg-day 0.22 (mg/kg-day)-1 3E-9

Chrysene 3.9 mg/kg 3.2E-8 mg/kg-day 0.022 (mg/kg-day)-1 7E-10

Dibenz[a,h]anthracene 0.58 mg/kg 4.8E-9 mg/kg-day 22 (mg/kg-day)-1 1E-7

Indeno[1,2,3-cd]pyrene 1.5 mg/kg 1.2E-8 mg/kg-day 2.2 (mg/kg-day)-1 3E-8

Exp. Route Total 8E-7

3E-6

Exposure Medium Total 3E-6

Medium Total 3E-6

Total of Receptor Risks Across All Media 3E-6

Exposure RouteSediment

Containment Area Sediments

Exposure Medium

Exposure Point

Exposure Point Total

Page 27 of 32

Supplement A

Table 7-3. Calculation of chemical cancer risksReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Scenario Timeframe: Future

Receptor Population: Residential

Receptor Age: 6-16

Medium Chemical of EPC Cancer Risk CalculationsPotential Concern Value Units Intake/Exposure Concentration CSF/Unit Risk Cancer Risk

Surface soils Surface soilsIngestion Value Units Value Units

Benz[a]anthracene 4.0 mg/kg 2.0E-8 mg/kg-day 2.19 (mg/kg-day)-1 4E-8

Benzo[a]pyrene 3.2 mg/kg 1.6E-8 mg/kg-day 21.9 (mg/kg-day)-1 3E-7

Benzo[b]fluoranthene 3.2 mg/kg 1.6E-8 mg/kg-day 2.19 (mg/kg-day)-1 3E-8

Benzo[k]fluoranthene 1.5 mg/kg 7.4E-9 mg/kg-day 0.219 (mg/kg-day)-1 2E-9

Chrysene 3.9 mg/kg 1.9E-8 mg/kg-day 0.0219 (mg/kg-day)-1 4E-10

Dibenz[a,h]anthracene 0.58 mg/kg 2.9E-9 mg/kg-day 21.9 (mg/kg-day)-1 6E-8

Indeno[1,2,3-cd]pyrene 1.5 mg/kg 7.4E-9 mg/kg-day 2.19 (mg/kg-day)-1 2E-8

Exp. Route Total 5E-7

Dermal Benz[a]anthracene 4.0 mg/kg 1.6E-8 mg/kg-day 2.2 (mg/kg-day)-1 4E-8

Benzo[a]pyrene 3.2 mg/kg 1.3E-8 mg/kg-day 22 (mg/kg-day)-1 3E-7

Benzo[b]fluoranthene 3.2 mg/kg 1.3E-8 mg/kg-day 2.2 (mg/kg-day)-1 3E-8

Benzo[k]fluoranthene 1.5 mg/kg 6.2E-9 mg/kg-day 0.22 (mg/kg-day)-1 1E-9

Chrysene 3.9 mg/kg 1.6E-8 mg/kg-day 0.022 (mg/kg-day)-1 4E-10

Dibenz[a,h]anthracene 0.58 mg/kg 2.4E-9 mg/kg-day 22 (mg/kg-day)-1 5E-8

Indeno[1,2,3-cd]pyrene 1.5 mg/kg 6.2E-9 mg/kg-day 2.2 (mg/kg-day)-1 1E-8

Exp. Route Total 4E-7

Exposure Point Total 9E-7

Exposure Medium Total 9E-7

Medium Total 9E-7

Total of Receptor Risks Across All Media 9E-7

Exposure PointSediment

Containment Area Sediments

Exposure Route

Exposure Medium

Page 28 of 32

Supplement A

Table 7-4. Calculation of chemical cancer risksReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Scenario Timeframe: Future

Receptor Population: Residential

Receptor Age: Adult

Medium Chemical of EPC Cancer Risk CalculationsPotential Concern Value Units Intake/Exposure Concentration CSF/Unit Risk Cancer Risk

Surface soils Surface soilsIngestion Value Units Value Units

Benz[a]anthracene 4.0 mg/kg 1.0E-8 mg/kg-day 0.73 (mg/kg-day)-1 8E-9

Benzo[a]pyrene 3.2 mg/kg 8.3E-9 mg/kg-day 7.3 (mg/kg-day)-1 6E-8

Benzo[b]fluoranthene 3.2 mg/kg 8.3E-9 mg/kg-day 0.73 (mg/kg-day)-1 6E-9

Benzo[k]fluoranthene 1.5 mg/kg 3.9E-9 mg/kg-day 0.073 (mg/kg-day)-1 3E-10

Chrysene 3.9 mg/kg 1.0E-8 mg/kg-day 0.0073 (mg/kg-day)-1 7E-11

Dibenz[a,h]anthracene 0.58 mg/kg 1.5E-9 mg/kg-day 7.3 (mg/kg-day)-1 1E-8

Indeno[1,2,3-cd]pyrene 1.5 mg/kg 3.9E-9 mg/kg-day 0.73 (mg/kg-day)-1 3E-9

Exp. Route Total 9E-8

Dermal Benz[a]anthracene 4.0 mg/kg 5.4E-9 mg/kg-day 0.73 (mg/kg-day)-1 4E-9

Benzo[a]pyrene 3.2 mg/kg 4.3E-9 mg/kg-day 7.3 (mg/kg-day)-1 3E-8

Benzo[b]fluoranthene 3.2 mg/kg 4.3E-9 mg/kg-day 0.73 (mg/kg-day)-1 3E-9

Benzo[k]fluoranthene 1.5 mg/kg 2.0E-9 mg/kg-day 0.073 (mg/kg-day)-1 1E-10

Chrysene 3.9 mg/kg 5.2E-9 mg/kg-day 0.007 (mg/kg-day)-1 4E-11

Dibenz[a,h]anthracene 0.58 mg/kg 7.8E-10 mg/kg-day 7.3 (mg/kg-day)-1 6E-9

Indeno [1,2,3-cd]pyrene 1.5 mg/kg 2.0E-9 mg/kg-day 0.73 (mg/kg-day)-1 1E-9

Exp. Route Total 5E-8

Exposure Point Total 1E-7

Exposure Medium Total 1E-7

Medium Total 1E-7

Total of Receptor Risks Across All Media 1E-7

Exposure Medium

Exposure Point

Exposure RouteSediment

Containment Area Sediments

Page 29 of 32

Supplement B

Table B-1. Dermal absorption factors for soil and sedimentDermal

AbsorptionFactor

Compound (DA) Volatile?b

InorganicsArsenic 0.03 a NoBarium 0.01 a NoCadmium (soil food) 0.001 NoChromium -- NoCopper -- NoManganese -- NoMercury (total) -- NoNickel -- No

Pesticides/PCBsDibenzofuran -- YesDioxins (as TCDD equivalents) 0.03 a NoPCBs 0.14 a NoDieldrin 0.10 a NoAroclor 1254 0.14 a No

Organics1,2,4-Trimethylbenzene -- Yes1,2-Dichlorobenzene -- Yes1,4-Dichlorobenzene -- YesBenzene -- YesChlorobenzene -- YesEthylbenzene -- YesHexachlorobenzene 0.10 a NoXylenes, total -- Yes

PAHs -all 0.13 a

Benz[a]anthracene 0.13 a NoBenzo[a]pyrene 0.13 a NoBenzo[b]fluoranthene 0.13 a NoBenzo[k]fluoranthene 0.13 a NoChrysene 0.13 a NoDibenz[a,h]anthracene 0.13 a NoFluoranthene 0.13 a NoFluorene 0.13 a NoIndeno[1,2,3-cd]pyrene 0.13 a NoNaphthalene 0.13 a No

Notes:-- - not available

b Volatile chemicals are not evaluated for dermal pathway (EPA 2007).

a Absorption factors from EPA 2007 dermal guidance: Exhibit 3-4.

Page 30 of 32

Supplement B

Table B-2. Detailed hazard indices by endpoint for air exposure pathwayReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Endpoint ChemicalHazard Quotients

and IndicesBlood

Benzene 0.011,1,1-Trichloroethane 0.041,2,4-Trimethylbenzene 0.9

Sum of blood: 1.0CNS

Mercury (elemental) 0.2Toluene 0.2Acetone 0.2Tetrachloroethene 0.002Xylenes 0.2Cyanide (hydrogen cyanide, 74-90-8) 0.2

Sum of CNS: 1.0Developmental

Ethylbenzene 0.012-Butanone 0.2

Sum of developmental: 0.2Liver

Chlorobenzene 0.51,4-Dichlorobenzene 0.00041,2-Dichloroethane 0.000051,1-Dichloroethene 0.07Carbon tetrachloride 0.001Chloroform 0.001Methylene chloride 0.007Methyl Tert-Butyl Ether 0.004Vinyl chloride 0.007

Sum of liver: 0.6Respiratory

Naphthalene 0.1trans-1,2-Dichloroethene 0.2

Sum of respiratory: 0.4Whole body

1,2-Dichlorobenzene (ortho) 0.4Sum of whole body: 0.4

Other1,2,4-Trichlorobenzene 0.9

Sum of all others: 0.9

Page 31 of 32

Supplement B

Table B-3. Detailed hazard indices by endpoint for child's hypothetical sediment exposure pathwayReasonable maximum exposure

Onondaga Lake sediment containment area (SCA) sediments

Endpoint Chemical

Hazard Quotients and Indices Ingestion

Hazard Quotients and Indices Dermal

Blood Antimony 0.02 --Benzene 0.002 --Fluoranthene 0.0003 0.0001Fluorene 0.0004 0.0001

0.02 0.0003Sum of blood: 0.02

CNS Aluminum 0.008 --Manganese 0.007 --Dioxins (as TCDD Equivalents) 0.2 0.08

0.2 0.08Sum of CNS: 0.3

Gastrointestinal Copper 0.006 --Sum gastrointestinal: 0.006

Hair cystine Vanadium 0.01 --Sum of hair cystine: 0.01

Immune Mercury (as mercuric chloride) 0.08 --PCBs (Aroclor 1254 noncancer) 0.1 0.04

0.2 0.04Sum of immune: 0.2

Liver 1,4-Dichlorobenzene 0.002 --Chlorobenzene 0.004 --Hexachlorobenzene 0.01 0.004Dieldrin 0.0002 0.0001Fluoranthene 0.0003 0.0001Pentachlorobenzene 0.01 --Ethylbenzene 0.0002 --

0.03 0.004Sum of liver: 0.04

Kidney Barium 0.010 --Cadmium 0.01 0.0011,2,3/4,5 Tetrachlorobenzene 0.1 --1,2,4-Trichlorobenzene 0.003 --Fluoranthene 0.0003 0.0001Pentachlorobenzene 0.01 --Ethylbenzene 0.0002 --

0.2 0.001Sum of kidney: 0.2

Whole body Nickel (as soluble salts) 0.03 --Xylenes, total 0.0004 --Naphthalene 0.01 0.004

0.04 0.004Sum of whole body: 0.04

Skin Arsenic 0.05 0.0050.05 0.005

Sum of skin: 0.06

Thyroid Cobalt 0.3 --Sum of thyroid: 0.3

NOAEL Chromium (as Cr VI) 0.4 --1,2-Dichlorobenzene 0.0007 --

Sum of NOAEL: 0.4

Lung 2-Methylnaphthalene 0.005 --Sum of lung: 0.005

Page 32 of 32

Appendix C

Air Dispersion Model Documentation

DISPERSION MODELING ANALYSIS METHODOLOGY

This report summarizes the methodology that was utilized to assess the dispersion of contaminants from the sediment consolidation area (SCA) located on Wastebed 13 into the surrounding areas. For this assessment, USEPA’s AERMOD dispersion model was utilized, using site specific weather data that has been collected from a meteorological station that was installed on Wastebed 13 in 2005. A modeling protocol document, which describes the basic approach and methodology for air dispersion analyses to assess potential ambient impacts of air emissions, was prepared by Honeywell in 2008, and submitted to NYSDEC and USEPA for review. This document was reviewed and approved in June, 2008. Modeling completed in support of this risk assessment was conducted in compliance with the approved protocol, with the following exceptions:

• More recent versions of the model and model software were utilized, which were unavailable at the time the protocol was prepared.

• The model receptor grid contained within the protocol was updated to reflect a refined analysis grid, and to add a model analysis line delineating the nearest receptors.

• For the purposes of this modeling, a unit emission rate modeling scenario was been utilized, to calculate the dispersion factor between the maximum work zone perimeter concentration, and the maximum nearest receptor concentration. This dispersion factor was then applied to the maximum allowable site-perimeter concentrations to calculate a maximum receptor concentration. This methodology is not described in the protocol.

• The footprint of the SCA has been updated from its delineation within the protocol, to more accurately reflect the current design.

As described in the 3rd bullet above, the model was utilized to calculate a dispersion factor. The dispersion factor calculated as part of this modeling analysis represents the minimum amount of dilution to the air leaving the work zone perimeter as it travels to the nearest residential areas. In the worst-case scenario, the concentration of site-related contaminants in the air at the nearest residential neighborhood will be a minimum of 4.5 times lower (the dispersion factor) than the concentration at the site perimeter.

The table below presents the assumed maximum allowable site perimeter annual average concentrations (Column F) that are not anticipated to be exceeded at the work zone perimeter during the 5-year operational period. The values in Column F were derived from the more conservative (lower) of the USEPA (Column D) and NYSDEC (Column E) air quality criteria considered applicable to the project. The NYSDEC criteria are guidelines (not established ambient air quality standards) intended to protect the general public from adverse health effects that may be induced by exposure to ambient air contaminants (NYSDEC DAR-1, http://www.dec.ny.gov/chemical/30681.html).

The corresponding predicted maximum annual average residential receptor concentrations (i.e., the highest annual average concentrations that would be found over the course of a year in the residential neighborhoods) are in Column G. The values in Column G are calculated by dividing the values in Column F by the “dispersion factor” of 4.51.

Column H presents EPA’s Regional Screening Levels for a residential setting, which are USEPA’s risk-based maximum allowable concentrations for the public to be exposed to in a residential setting on a daily basis. The EPA screening levels (SLs) are developed using risk assessment guidance from the EPA Superfund program and can be used for Superfund sites. SLs are considered by the Agency to be protective for humans (including sensitive groups) over a lifetime. They are used for site “screening” and as initial cleanup goals, if applicable. Generally, at sites where contaminant concentrations fall below SLs, no further action or study is warranted (to address the air pathway) under the Superfund program. Chemical concentrations above the SL would not automatically designate a site as “dirty” or trigger a response action; however, exceeding a SL suggests that further evaluation of the potential risks by site contaminants is appropriate (USEPA, http://www.epa.gov/reg3hwmd/risk/human/rb-concentration_table/faq.htm#FAQ1).

For all volatile chemicals, the values in Column G (worst-case concentrations in the residential neighborhoods) are lower than the residential RSLs in Column H, indicating that the worst-case concentrations in the air that would potentially exist at the nearest residential receptor would not cause adverse health impacts.

PARSONS

AIR DISPERSION MODELING PROTOCOL FOR ONONDAGA LAKE

Prepared For:

5000 Brittonfield Parkway East Syracuse, NY 13057

Prepared By:

Parsons

290 Elwood Davis Road, Suite 312 Liverpool, New York 13088

Phone: (315) 451-9560 Fax: (315) 451-9570

MAY 2008

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

I

TABLE OF CONTENTS Page

EXECUTIVE SUMMARY ........................................................................................ ES-1

SECTION 1 PROJECT OVERVIEW AND SITE DESCRIPTION........................ 1-1

1.1 PROJECT SUMMARY..................................................................................... 1-1

1.2 STUDY AREA .................................................................................................. 1-2

1.3 ONSITE METEOROLOGICAL MONITORING PROGRAM........................ 1-2

SECTION 2 MODEL AND MODEL OPTIONS....................................................... 2-1

2.1 AERMOD MODEL........................................................................................... 2-1

2.2 METEOROLOGICAL DATA PRE-PROCESSOR.......................................... 2-2 2.2.1 AERMET................................................................................................. 2-2 2.2.2 AERSURFACE ....................................................................................... 2-2

2.3 AERMAP TERRAIN PRE-PROCESSOR........................................................ 2-3

SECTION 3 MODEL INPUT DATA.......................................................................... 3-1

3.1 SOURCE CHARACTERIZATION AND EMISSION RATES....................... 3-1 3.1.1 Source Identification ............................................................................... 3-1 3.1.2 Source Characterization........................................................................... 3-1 3.1.3 Source Emission Rates ............................................................................ 3-2

3.2 METEOROLOGICAL DATA INPUTS ........................................................... 3-3 3.2.1 Meteorological Data Sources .................................................................. 3-3 3.2.2 AERMET Pre-Processing........................................................................ 3-4 3.2.3 Surface Characteristics ............................................................................ 3-6 3.2.4 Transport Wind Speed and Direction ...................................................... 3-8 3.2.5 Temperature............................................................................................. 3-9 3.2.6 Atmospheric Stability.............................................................................. 3-9 3.2.7 Mixing Heights...................................................................................... 3-10 3.2.8 Sounding Data ....................................................................................... 3-11 3.2.9 AERMOD Meteorological Input Files .................................................. 3-11

3.3 MODEL RECEPTOR GRID........................................................................... 3-11

SECTION 4 DISPERSION MODELING RESULTS REPORTING...................... 4-1

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

II

TABLE OF CONTENTS (CONTINUED)

LIST OF FIGURES

Figure 1-1 Met Tower Location Map

Figure 1-2 SB-13 Tower 2006 Wind Rose

Figure 1-3 SB-13 Tower 2007 Wind Rose

Figure 1-4 Lakeshore/Willis Ave 2007 Wind Rose

Figure 3-1 Receptor Grid for Modeling

Figure 3-2 Receptor Grid for Modeling on Basemap

Figure 3-3 SCA Base Model Receptor Grid

Figure 3-4 SMU Base Model Receptor Grid

LIST OF TABLES

Table 1-1 Meteorological Monitoring Program - Parameters Measured

Table 3-1 Site Specific and NWS Data Station

Table 3-2 Site-Specific Seasons for Use in AERSURFACE

Table 3-3 Seasonal Values of Micrometeorological Parameters

Table 3-4A Seasonal Average Micrometeorological Parameters Around WB-13 Tower in 2006 By Wind Sector

Table 3-4B Seasonal Average Micrometeorological Parameters Around WB-13 Tower in 2007 By Wind Sector

Table 3-4C Seasonal Average Micrometeorological Parameters Around Lakeshore Tower in 2007 By Wind Sector

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

ES-1

EXECUTIVE SUMMARY

This modeling protocol document describes the basic approach and methodology for air dispersion modeling analyses to assess potential ambient impacts of air emissions from activities associated with the remediation of Onondaga Lake.

Specifically, the modeling approach for the Onondaga Lake project is intended to provide a comprehensive analysis of potential air impacts associated with emissions that may be generated by the various remediation activities. These activities include, but are not necessarily limited to, sediment dredging, active management of the materials at the sediment consolidation area (SCA), and dewatering of the sediments at the SCA. Additional activities which may generate emissions, such as slurry piping and SCA effluent treatment, will also be assessed. These activities may require control measures, however, it may not be necessary to incorporate these activities into the model.

Additional emission sources and/or activities may be identified beyond those specifically anticipated for the remediation of Onondaga Lake. To ensure that the potential impacts from these additional sources are differentiated from potential impacts caused by Lake remedial activities, it may be necessary to quantify their impacts using dispersion modeling or other evaluation methods such as monitoring. This would ensure that any mitigative strategies for the Lake activities are appropriately designed, and potential impacts from other sources can be considered when developing air quality criteria and monitoring plans for the Lake remediation. Should additional sources associated with other Honeywell remediation projects be identified for which dispersion modeling is determined to be appropriate, modeling analyses for these activities would be conducted consistent with the protocol established in this document or other protocols approved by DEC.

The potential emission sources to be analyzed, and the actual compounds and quantities thereof that may be emitted, are being established and documented in detail as separate study components of the overall design program. Specific contaminants and their estimated flux rates are not discussed in this protocol. Once the sources, compounds, and estimated emission rates are finalized, this information will be separately documented and used to develop the source and emissions data files input to the modeling analysis. Additionally, air quality goals are also being developed as a separate component of the overall investigation. Specific goals recommendations are not included in this report.

The results of the modeling analyses will be used to address four key objectives:

1. Estimate the magnitudes, extent, and spatial and temporal variability of predicted ambient air concentrations (dispersion estimates) of project-related air emissions for the identified Chemicals of Interest (COIs).

2. Identify those sources and emissions having the greatest potential impacts.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

ES-2

3. Design of best management practices, control systems and operations strategies to reduce air emissions resulting from remedial activities to levels less than applicable short-term and long-term ambient air standards, and established threshold levels.

4. Document and communicate the potential impacts of air emissions of COIs and odors from the remedial activities to the community through a series of public outreach and educational programs.

These objectives will be met by conducting a comprehensive series of dispersion modeling analyses to estimate the potential ambient impacts of remediation activity emissions. The analyses will be based on the potential sources identified and on the emission rates and remediation activity schedules as developed via separate remediation design and study components. The analyses will incorporate site-specific meteorological data, and will ultimately account for background air quality levels to be measured as part of an anticipated future data collection activity.

Key components of the actual modeling approach include:

• Use of USEPA’s AERMOD model; run in a sequential mode, to assess potential impacts for various averaging periods and in all terrain regimes.

• Use of an onsite meteorological database as the basis for assembling hourly data records suitable for input to AERMOD. The onsite databases will be supplemented by other surface weather observation data from Syracuse (Hancock) International Airport and upper air sounding data from Buffalo, New York.

• Use of AERMOD-related processing programs, including AERSURFACE, AERMET and AERMAP, respectively, to determine land use characteristics and associated boundary layer meteorological parameters, to process the meteorological input data, and to develop the model receptor grid and terrain heights for input to the model.

• Assessment of potential emissions from various remediation activities and scenarios, primarily dredging operations in the lake and management of dredged sediment in the SCA. Other potential emitting activities or sources, such as sediment transfer points or water runoff collection and treatment facilities, will be identified once the final design is completed and included in the analysis.

Dispersion model estimates will be determined for an extensive modeling receptor grid that encompasses the lake, the SCA, and the surrounding environs. Following development of an appropriate set of air quality goals and odor thresholds, the model estimates will be compared to the goals to ensure the final design results in acceptable impacts for all appropriate general and identified sensitive receptor locations. The evaluation of impacts will be conducted as a separate portion of the overall design, and is not addressed in this protocol.

A key component of the analysis is accounting for the variability in potential source locations and emission rates over time. Potential emission sources, including dredging locations, dredge area sizes, containment area size and shape, as well as corresponding emission rates, will likely vary from one activity location to another. Emission rates of various compounds may also vary as a function of wind speed and of air and water temperature variations throughout the year.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

ES-3

Shorter term variations in emissions may be a function of work activity levels from day to night and from weekdays to weekends. In the SCA, emission rates may vary with time as individual containment cells are filled and the volume of dredged material deposited increases.

The modeling approach has been set up such that all of these factors can be accounted for in the set up and execution of the modeling analyses via a “source management” approach to characterizing the source inputs, executing discrete model runs for different sources and source groups, and in integrating and analyzing model results. Although the maximum short-term estimated concentrations of emitted compounds are likely to result from a specific source at a specific time, the long-term average impacts will be represented by the combined results for all sources, emission rates, and variations therein that occur through out a full year of activity.

The following sections describe in more detail the modeling approach and how the specific model inputs will be developed.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

1-1

SECTION 1

PROJECT OVERVIEW AND SITE DESCRIPTION

1.1 PROJECT SUMMARY

Onondaga Lake is a 4.6-square-mile lake located just northwest of the city of Syracuse in central New York State. Over 200 years of heavy industrial activity and population growth on shores of the lake and its nearby tributaries have impacted the quality of the lake ecosystem. As a result of the presence of hazardous substances or hazardous wastes, the lake has been identified as a federal Superfund site on the USEPA National Priority List.

In 1992, AlliedSignal (now known as Honeywell) entered into a consent decree with the State of New York to initiate a remedial investigation (RI) and feasibility study (FS) for Onondaga Lake. The Final FS was submitted to New York State Department of Environmental Conservation (NYSDEC) in November 2004 (Parsons, 2004). On July 1, 2005, NYSDEC issued the Record of Decision (ROD) for the Lake, and the Final Consent Decree was issued on January 4, 2007.

The remedy specified in the ROD and CD includes the dredging of 2,653,000 cubic yards (cy) of sediment from the littoral zone (nearshore areas) of the lake and placement of a sediment cap over portions of the lake bottom. These activities will take place in each of several identified Sediment Management Units (SMUs) in the lake, though most of the dredging will occur in the in-lake waste deposit (ILWD), which largely exists in SMU 1 and 2. Sediment dredged from these areas will be deposited in an SCA located to the west of the lake in an area known as Wastebed 13.

As discussed in more detail in Section 2.0, several activities and areas have been identified as potential sources of air emissions that will be the focus of the modeling analyses. These include:

• the removal and onshore management of lake bottom debris;

• the area in and around the sediment dredging operations within the SMUs;

• conveyance of the dredged material to the SCA;

• management of dredged sediment in the SCA; and

• treatment of water generated by the dredging/sediment handling processes as a result of dewatering of the sediments in the SCA.

Separate sampling, analysis, and emission modeling studies are being conducted to identify the specific COIs and the quantities of each that may volatilize to the air from these activities.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

1-2

1.2 STUDY AREA

Figure 1-1 is an aerial photograph of Onondaga Lake and surrounding areas that has been marked to show the key activity areas of interest for the assessment of potential air impacts from project-related emissions.

Most of the dredging activities will be conducted in the southeastern portion of the lake, in SMUs 1, 2, 6, and 7. Material dredged from the lake will be transferred via enclosed pipeline to the SCA at Wastebed 13, which is located about 3.5 miles inland from the southern lakeshore.

Wastebed 13 is part of a large system that was originally utilized as settling basins for the disposal of Solvay waste, and it received the waste material from 1981 to 1986. Wastebed 13 occupies approximately 163 acres and is located in the Town of Camillus, Onondaga County, New York. It is bordered to the north by Ninemile Creek and CSX Railroad tracks; to the west by an Onondaga County Garage property, a formal gravel excavation owned by Honeywell, and a few residential properties; and to the east and south by Wastebeds 12 and 14, respectively. The basin complex of 12-15, of which 13 is a part of, is on property owned by Honeywell, with access restricted by a fence and gated entrance.

The lake surface is at a base elevation of 362.82 ft above Mean Sea Level (MSL). Terrain heights within the first mile of the lakeshore range up to 450 ft MSL. Base elevation in the Wastebed 13 area is about 400 ft MSL with some hills rising to 600 ft MSL within 1 kilometer (km) to the southeast. With the buildup of material deposited over the years, the average elevation of the top of Wastebed 13 is approximately 65 ft above the immediate surrounding offsite grade. It has been covered with soil and planted with various forms of vegetation, including shrubs and stands of trees. The wastebed area is located away from the built-up areas of Syracuse and surrounding towns, but is situated within short distances of nearby residential areas and school properties in Camillus. The wastebed is located about 2.5 miles west of the State Fairgrounds.

1.3 ONSITE METEOROLOGICAL MONITORING PROGRAM

As part of the Phase I and Phase II Pre-Design Investigations, an onsite meteorological monitoring program was initiated to support the dispersion modeling analyses and other analytical activities related to the overall Onondaga Lake remediation project. The monitoring program consists of two 10-meter towers instrumented to provide the requisite data for use in the AERMOD model.

The primary goals of the meteorological monitoring program are to (1) collect onsite data directly representative of the potential emission sources and surrounding areas, (2) to obtain measurements of all of the key parameters that are suitable for input to the AERMOD model, and (3) collect measurements using monitoring instrumentation having operating and performance specifications consistent with all US Environmental Protection Agency (USEPA) monitoring requirements and guidelines.

The locations of the two meteorological monitoring sites are shown in Figure 1-1. The first, or primary, Tower Site #1 – designated as the Wastebed 13 site – is situated directly atop the

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

1-3

wastebed complex and along the southeastern edge of Wastebed 13. The tower is sited in an openly exposed area such that measurements are considered representative of the conditions prevailing across the entire wastebed. This site was installed and became operational on December 1, 2005. Thus, the first two full calendar years of data for the Wastebed 13 site are available for the years January through December 2006 and January through December 2007.

The secondary Tower Site #2 – designated as the Lakeshore/Willis Avenue Site – is located near the southern lakeshore. The tower is sited in an openly exposed area just off the western side of Route 690 where it intersects with Willis Avenue. Measurements at this site are considered as representative of the southern lake area where most of the dredging activities will take place. This site was installed and became operational on December 1, 2006. The first full calendar year of data at this site is available for the period January through December 2007.

The towers are variously instrumented at surface, 2-meter, and 10-meter heights to obtain measurements of wind speed, wind direction, temperature and delta-temperature, relative humidity, vertical wind speed, stability parameters, solar and net radiation, barometric pressure, and precipitation. Table 1-1 provides a listing of the various parameters and measurement heights for the instrumentation installed on each tower.

The two towers are basically instrumented the same, with two exceptions. First, Net Radiation (which is the preferred radiation parameter in the hierarchy of inputs to AERMET) is measured at Wastebed 13 while Solar Radiation is measured at Lakeshore/Willis Ave. When designing Tower #2 (Lakeshore/Willis Ave), it was believed that Net Radiation values would not vary significantly over the relatively short distance and comparable surface settings of the two sites. Solar radiation is also not expected to vary between the two sites; therefore, it was determined that collection of both Net and Solar Radiation measurements (one parameter at each site) would enhance the overall project database provided by the combined monitoring systems.

Secondly, precipitation measurements are collected at Site #1, but not collected at Site #2. As with the radiation parameters, precipitation measurements are not expected to vary between the two sites; thus, one set of measurements was deemed sufficient for the study area.

The sites are powered via a solar power-charged battery system and are each equipped with a digital data acquisition system and cell-phone based modems for remote data communications via the internet. All sensor outputs are scanned at 1-second intervals, with readings used to calculate 5-minute, hourly, and daily averaged values for each parameter.

Complete details of the meteorological monitoring program, including detailed site descriptions, monitoring equipment specifications, and the routine operating and quality assurance procedures employed to ensure collection of a valid, accurate, and complete database, are contained in the Honeywell – Onondaga Lake Meteorological Monitoring Program SOP/QA Manual (Parsons, 2008).

Data from the two sites are downloaded and subject to preliminary verification 2 to 3 times per week. The databases are subjected to more rigorous quality assurance validation checks and

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

1-4

final processing in monthly blocks, which then yield the annual databases used to process the model input files.

Data reports for each site, which include hourly data values for each parameter, plus daily and monthly data summaries, are submitted to NYSDEC on a calendar quarter basis, along with updated tables listing the Data Recovery Rates for each measured and calculated parameter.

For reference purposes, Figures 1-2, 1-3, and 1-4 present wind roses depicting the annual combined frequency distributions of wind speed and direction measurements recorded during 2006 and 2007 at Wastebed 13 and during 2007 at Site #2. The 2006 and 2007 onsite data for Site #1 show that winds from the southwest through northwest were prevalent throughout most of each year, while winds from the north through east to south were recorded much less frequently. At Site #2, there is a much more prevalent peak of winds from the due west, offset by reduced frequencies of both southwesterly and northwesterly winds. There is also a slightly higher frequency of winds from the southeast, albeit at relatively low speeds, at Site #2 than seen for Site #1.

Data collected during each calendar year at the onsite meteorological towers will be combined with concurrent offsite surface weather observation data from Syracuse International Airport and upper air sounding data from Buffalo and input to the AERMOD pre-processor program AERMET to develop complete onsite databases for use in the dispersion modeling analyses of remediation activity air emissions. The details of the data bases generated using the onsite data are presented in Section 3.2.

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

FIGURE 1-1

MET TOWER LOCATION MAP

Tower #1

Tower #2

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

FIGURE 1-2

SB-13 TOWER 2006 WIND ROSE

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

FIGURE 1-3

SB-13 TOWER 2007 WIND ROSE

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

FIGURE 1-4

WILLIS TOWER 2007 WIND ROSE

HONEYWELL ONONDAGA LAKE REMEDIATION PROJECTAIR DISPERSION MODELING PROTOCOL

TOWER NO. 1 - Wastebed 13 Site - Primary Tower - Start-Up December 1, 2005

Ground-Based Measurements

Precipitation

2-Meter Level Measurements

TemperatureRelative HumidityDew Point Temperature (computed)Net RadiationBarometeric Pressure

10-Meter Level Measurements

Horizontal Wind SpeedHorizontal Wind DirectionStandard Deviation of Horizontal WD or Sigma-Theta (computed)Vertical Wind SpeedStandard Deviation of Vertical Wind Speed (Sigma w) -computedTemperatureDelta Temperature (10m - 2m)

TOWER NO. 2 - Lakeshore/Willis Avenus Site - Secondary Tower - Start-Up December 1, 2006

2-Meter Level Measurements

TemperatureRelative HumidityDew Point Temperature (computed)Solar RadiationBarometric Pressure

10-Meter Level Measurements

Horizontal Wind SpeedHorizontal Wind DirectionStandard Deviation of Horizontal WD or Sigma-Theta (computed)Vertical Wind SpeedStandard Deviation of Vertical Wind Speed (Sigma w) -computedTemperatureDelta Temperature (10m - 2m)

Air Dispersion Modeling Protocol for Onondaga LakeTABLE 1-1

Meteorological Monitoring Program - Parameters Measured

P:\Honeywell -SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Modeling Protocol\Tables 1-1, 3-3.xls5/12/2008 1 of 1 PARSONS

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

2-1

SECTION 2

MODEL AND MODEL OPTIONS

2.1 AERMOD MODEL

The dispersion modeling analysis will be conducted using the USEPA’s AERMOD dispersion model. This model has recently been developed and formally approved by USEPA as the replacement for the previous ISCST3 model for regulatory and other impact analysis purposes.

AERMOD is a current state-of-the-art dispersion model for assessment of pollutant concentrations from a variety of source types, including point, area, and volume sources, and from both surface-based and elevated source release heights. AERMOD is appropriate for modeling in all terrain regimes by implementing USEPA guidance for assessing impacts in simple, complex, and intermediate terrain.

AERMOD is a steady-state plume model, using Gaussian distributions in the vertical and horizontal for stable conditions, and in the horizontal for convective conditions. The vertical concentration distribution for convective conditions results from an assumed bi-Gaussian probability density function of the vertical velocity.

AERMOD’s advantages include the ability to model impacts in simple, complex, and “intermediate" terrain. Intermediate terrain in this context refers to receptors that are above the source release height, but below the plume height predicted by AERMOD model algorithms. AERMOD also uses an arbitrarily large number of meteorological data levels to create profiles of wind, temperature, and turbulence that can vary with height.

Surface parameters such as roughness length, albedo, and Bowen Ratio, which have a large influence on atmospheric boundary layer dispersion conditions, can be selected by direction sector and by month to provide a more accurate characterization of the modeling domain than predecessor models such as ISC3.

AERMOD is considered to be the model that will provide the most representative and realistic estimates of impacts of emissions from Onondaga Lake remediation activity sources.

The most up-to-date Windows-based BEE-LINE Software version of AERMOD will be used for the modeling. (Current versions are BEEST version 9.72 with USEPA AERMOD version 07026). The current version of AERMOD also includes the most recent generation of building downwash algorithms known as “PRIME” (Plume Rise Model Enhancement). In general, “AERMOD-PRIME” will not be invoked for this analysis due to the area and/or volume source characterization of the remediation activities and locations and the lack of major building structures in the activity area. However, if the project ultimately includes any operational or control system structures with release vents or stacks below Good Engineering Practice (GEP)

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

2-2

stack height, then the locations and dimensions of these and any nearby structures will be input to the model and the PRIME algorithms will be invoked.

Model options corresponding to the regulatory default settings within the model that will be used in this analysis include the use of:

• the parameter DFAULT in the MODELOPT record on the Control Pathway;

• elevated terrain algorithms requiring input of terrain height data;

• buoyancy induced dispersion;

• an iterative approach to estimate stable boundary layer plume rise;

• rural dispersion coefficients based on examination of land use patterns and characteristics in the vicinity of the source and meteorological measurement location;

• calm hour processing routines;

• missing-data processing routines; and

• sequential date checking.

2.2 METEOROLOGICAL DATA PRE-PROCESSOR

2.2.1 AERMET

The AERMOD Meteorological Pre-processor (AERMET) is the companion program of AERMOD that is used to pre-process the meteorological data required to run the model. AERMET pre-processing utilizes surface meteorological data from site-specific (onsite) data collection programs and offsite stations such as those run by the National Weather Service (NWS), along with NWS sounding data from upper air measurement locations, to calculate the planetary boundary layer parameters required by AERMOD to perform dispersion and transport computations.

For this modeling analysis, the most recent version of AERMET will be used to pre-process the meteorological data used as input to the model. (Current AERMET version is 06341 and is compatible with the current version of AERMOD - 07026).

Additional details describing the meteorological databases and the AERMET pre-processing procedures used for this analysis are provided in Section 3.2 (Meteorological Data Inputs).

2.2.2 AERSURFACE

USEPA’s recently issued updated version of the AERSURFACE pre-processor is a program that utilizes digital maps of land use and cover to help define the land use characteristics of the study area and to assign land-use specific values of associated key micrometeorological parameters that are then incorporated into the meteorological data processing via AERMET. Values for the key micrometeorological variables can be defined on a seasonal basis. In addition, the surface roughness values can be defined for different directional sectors from the meteorological measurement source to account for varying land use and cover patterns throughout the study area.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

2-3

AERSURFACE assigns albedo and Bowen ratio values based on a simple unweighted average land cover type over a representative domain, which by default is a 10 by 10 km region centered on the measurement location. For surface roughness, AERSURFACE assigns values based on an inverse-distance weighted average land cover type defined by wind direction sector out to 1 km from the measurement site. The radius can be varied between 0.1 and 5.0 km, however, 1 km is the recommended default value. Wind direction sectors as small as 30 degrees (0-30, 30-60, 60-90, 330-360) can be employed for determining surface roughness values. Additional details describing the use of the AERSURFACE pre-processing procedures used for this analysis are provided in Sections 3.2.2 and 3.2.3 (AERMET pre-processing and Surface Characteristics).

2.3 AERMAP TERRAIN PRE-PROCESSOR

The AERMOD Terrain Pre-processor (AERMAP) is the companion program of AERMOD that is used to pre-process the model receptor grid. AERMAP pre-processing routines utilize electronic digital elevation model (DEM) files corresponding to 7.5-minute US Geological Survey (USGS) topographic map quadrangles, along with a project-specific model receptor grid, to calculate the parameters required by AERMOD to handle air flow through complex terrain. The primary parameter calculated by AERMAP is the height scale (hc), or hill height. The height scale is used to calculate the critical dividing streamline height (Hcrit) for each receptor point based on the controlling terrain feature for the receptor. In addition, AERMAP optionally computes receptor elevations from the DEM file data.

For this modeling analysis, the most current version of AERMAP (version 06341_or greater) will be used to pre-process the receptor model grid. The latest versions of AERMAP contain the USGS approved NADCON 2.1 program that converts North American Datum (NAD) of 1927 to NAD of 1983.

AERMAP will be run with the “TERRHGTS/EXTRACTED” keywords selected, which allows interpolation of model receptor point elevations directly from the DEM data.

Additional details on the AERMAP pre-processing procedures used for this analysis are provided in Section 3.3 (Model Receptor Grid).

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-1

SECTION 3

MODEL INPUT DATA

3.1 SOURCE CHARACTERIZATION AND EMISSION RATES

3.1.1 Source Identification

As discussed in Section 1, several remedial activities have been identified as potential sources of emissions. These activities include removal and management of debris, dredging of the lake sediments, conveyance of the dredged material, management of the dredged material in the SCA, and treatment of the SCA effluent water. Potential compounds emitted and their associated flux rates are being evaluated as a separate part of the overall investigation, and will be documented in a separate report.

Additional emission sources and/or activities may be identified beyond those specifically anticipated for the remediation of Onondaga Lake. To ensure that the potential impacts from these additional sources are differentiated from potential impacts caused by lake remedial activities, it may be necessary to quantify their impacts using dispersion modeling or other evaluation methods such as monitoring. This would ensure that any mitigative strategies for the lake activities are appropriately designed, and potential impacts from other sources can be considered when developing air quality criteria and monitoring plans for the lake remediation. Should additional sources associated with other Honeywell remediation projects be identified for which dispersion modeling is determined to be appropriate, modeling analyses for these activities would be conducted consistent with the protocol established in this document or other protocols approved by DEC.

3.1.2 Source Characterization

Depending on the nature of the activity, emission sources will be represented as either area sources, or point sources. Area sources will likely consist of activities such as the dredging and SCA operations, while activities such as the water treatment facility, that include a vent or stack through which emissions are exhausted, will be input to the model as a point source.

Area sources are specified in terms of dimensions and/or coordinates defining the locations and sizes of square, rectangular, circular, or irregularly shaped polygonal areas. Emissions may be presumed to be evenly emitted over the entire area of each defined source, or varying rates may be determined for multiple sub units of the source. The locations, sizes, and shapes of area sources are specified and input to the AERMOD model in one of three ways:

1. The horizontal dimensions (length and width) defining a square or rectangular emitting area, along with the coordinates (UTM coordinates derived from USGS maps) defining the location of the SW corner of the rectangular or square area, or

2. The diameter of a circular shaped area source, along with the UTM coordinates of the centerpoint of the area, or

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-2

3. The UTM coordinates of a series of points defining the outline of an irregular, or polygon shaped emitting area.

For point sources, each source will be input with a single set of coordinates denoting the actual location of each stack or vent.

For area sources, the AERMOD model uses the dimensions of rectangular and circular area sources to directly calculate the actual total size of the area (e.g., square meters) of the source. For irregularly shaped polygon sources, designated as Area-Poly sources, the AERMOD model has a subroutine to integrate the sizes of subcells in the area based on the input coordinates in order to determine the overall size of the area source.

Unlike industrial exhaust stacks, emissions from area sources typically have no vertical momentum or relatively warm/hot temperatures that would otherwise impart a vertical velocity or buoyancy to an emitted air stream. Emissions will be presumed to be released at ambient temperature.

3.1.3 Source Emission Rates

The actual emission rates to be input to the model for each source are being evaluated as a separate component of the overall investigation. In general, emission rates for the specified COIs will be provided for model input in terms of mass per unit time per unit area, such as grams per second per square meter (g/sec/m2) for area sources, and mass per unit time (g/sec) for point sources.

The emission rate data, including variations, will be used to help populate a Source Identification matrix that will account for all of the sources and all of the time, location, and activity level variations in emissions from all identified sources.

In order to appropriately estimate both short-term peak and long-term average ambient concentrations and levels of air contaminants accounting for the various factors that will affect actual emission rates, two sets of model runs will be made for each source or set of sources. Maximum short-term (e.g.; 1-hour) impacts will be determined based on modeling the maximum anticipated hourly emission rate(s) for each source for all hours in the meteorological data record. This will ensure capturing the maximum predicted impact since it cannot otherwise be specified at what time or under what meteorological dispersions the maximum impact(s) may occur.

For assessment of long-term (i.e., annual average) impacts, the source data and emission rates inputs to the model will be set up to explicitly account for the scheduling of the various dredging and SCA operational activities. This includes specifying time of year and the duration that certain areas of each SMU will undergo dredging, the anticipated dredging rates and SCA deposition volumes, the sizes of each area to be dredged, and the number and size of the SCA cells to be filled, along with anticipated daily and weekly work schedules (e.g., if there are no activities on weekends). This will be accomplished by preparing an external hourly emission rate file based on these variables that will be used as the emission rate input file to AERMOD.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-3

To ensure the schedule assumed for the model matches current schedule and design assumptions made by other groups, the Emission and Odor Technical Work Group will continue to coordinate with other work groups to ensure updated schedule assumptions are continuingly used for modeling.

3.2 METEOROLOGICAL DATA INPUTS

Sequential (i.e., hour-by-hour) air quality dispersion modeling with AERMOD requires the input of suitable meteorological data. These data include hourly values of variables including wind speed, wind direction, ambient surface temperature, stability parameters, and mixing height that are directly measured from both onsite (i.e., local) and offsite (i.e., regional) representative sources or derived from the directly measured variables. These data are assembled and then “pre-processed” into a single database suitable for direct input to the model. This component of the model input preparation will be accomplished via use of the USEPA meteorological preprocessing program AERMET, which is the companion “meteorological pre-processor” to AERMOD. This section addresses the various elements of the meteorological database to be assembled, the basis for the selection or derivation of a particular data value or set of values, and the pre-processing routine.

3.2.1 Meteorological Data Sources

The meteorological database will be assembled from the following three types of data sources:

• Onsite Surface Data: Two Site-Specific Meteorological Towers

• Offsite Surface Data: NWS Syracuse, NY Surface Station

• Upper Air Data: NWS Buffalo, NY Upper Air Station

Table 3-1 presents summary information on the monitoring stations that will be used in the meteorological pre-processing.

Table 3-1

Site-Specific and NWS Data Stations

Data Station Identifier/WBAN Latitude Longitude Elevation Format

Camillus, New York (Site-Specific)

Wastebed 13 Station 43.071 N 76.255 W 476 ft User Defined

Geddes, New York (Site-Specific)

Lakeshore/Willis Ave 43.068 N 76.201 W 387 ft User Defined

Syracuse, New York (Surface) SYR/14771 43.117 N 76.100 W 410 ft ISHD/SCRAM

Buffalo, New York (Upper Air) BUF/14733 42.933 N 78.733 W 705 ft FSL

Data from the two onsite Onondaga Lake meteorological monitoring stations are being collected specifically in support of the ambient air quality modeling and data analysis activities comprising the overall air quality impact assessment. These two stations, the first (Wastebed 13) installed near Wastebed 13 in December 2005; and the other (Lakeshore/Willis Ave) installed at

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-4

the southern lakeshore near SMU’s 1, 6, and 7 in December 2006, are intended to provide model input data directly representative of the key identified emission sources in these two areas.

As described in Section 1.3, the towers are equipped with instrumentation at the surface and 10-meter levels for measuring the specific parameters to be used as input to AERMOD, including wind speed, wind direction, temperature and temperature profiles, wind turbulence, solar/net radiation, and barometric pressure. Data collection at both of these stations is expected to continue through the end of the remediation program. Modeling analyses conducted during the course of the program will utilize data for as many calendar years as are available for processing at the time the modeling is conducted.

When modeling emissions from the remedial activities within the lake, data from the Lakeshore/Willis Ave Site (Site #2) at the south end of the lake will be used as input. When modeling emissions from the SCA, databases comprised of surface data from the Wastebed 13 station (Site #1) will be used. When multiple sources are modeled in both areas, a decision may be made as to which data source best represents the overall dispersion conditions for the area. This decision will further consider the extent to which emissions from one source area or another are dominating the overall magnitude of predicted impacts, and/or which portions of the receptor grid are of greatest interest.

The onsite meteorological data will be supplemented with National Weather Service (NWS) aviation surface weather observations recorded at the Hancock International Airport in Syracuse. Data from this NWS station will be included in the meteorological pre-processing for two purposes: (1) to provide supplemental data for AERMOD that is not collected via the onsite monitoring program; and (2) to fill in any missing data gaps that might occur in the onsite data records. The primary parameters utilized from Syracuse Airport (i.e., those that will not be available from the onsite program) are hourly observations of cloud ceiling height and sky cover. These parameters are mandatory for inclusion as input to the AERMET pre-processor.

Upper air sounding data from the NWS Buffalo, NY upper air measurement station will also be included in the AERMET pre-processing. Parameters used as input include the morning soundings of atmospheric pressure, dry bulb temperature, dew-point temperature, wind speed, and wind direction. These parameters are also mandatory for inclusion as input to the AERMET pre-processor.

The following sections describe the AERMET pre-processing in more detail, the specific sources for each meteorological parameter required for the air quality modeling analysis, the data substitution methodology used in the event of missing data, and the frequency with which various data sources were used to develop the meteorological database for modeling.

3.2.2 AERMET Pre-Processing

The meteorological data from the three meteorological data sources noted above will be pre-processed via the AERMOD Meteorological Pre-Processor (AERMET). AERMET pre-processing utilizes surface meteorological data from site-specific (onsite) data collection programs and offsite stations such as those run by the NWS, along with NWS sounding data

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-5

from upper air measurement locations, to calculate the planetary boundary layer parameters required by AERMOD to perform dispersion and transport computations.

AERMET version 06341 or greater will be used to pre-process the meteorological data used as input to the model. The AERMET preprocessing of the data from the three sources selected for this analysis is performed in three stages that are further described as follows:

Stage 1

• In the first step of pre-processing, the onsite meteorological data will be processed. Electronic files of onsite meteorological database will be converted into a format suitable for input to AERMET and converting the data measurement units into those required by AERMET processed the data. One file will be created for each year for each of the two measurement stations. Each file will contain only the data collected from the individual station except for solar and net radiation data. Net radiation data from Wastebed-13 and solar radiation data from Lakeshore/Willis Ave will be included in all input files as these data are considered to be representative of the entire local study area and will be used by AERMET if provided.

• The second step of pre-processing involves the surface data from Syracuse Airport. This data is generally available from the National Climatic Data Center (NCDC) in the Integrated Surface Hourly Data (TD3505 - ISHD) format with fixed length records. This format is suitable for input to AERMET.

• The third step of pre-processing consists of processing the upper air morning sounding data from the NWS Buffalo, NY Station. The sounding data will be obtained from NCDC in the FSL format, which allows it to be directly input to AERMET.

• In each of the first three steps of AERMET pre-processing, all data will be checked against quality assurance criteria contained within the AERMET program. Automatic corrections and substitutions are employed as appropriate and as recommended in the AERMET User’s Guide.

Stage 2

• The files generated from the three steps described above are then merged into a single file for the final step of the meteorological pre-processing.

Stage 3

• Utilizing USEPA’s AERSURFACE tool, land use characteristics are defined and micrometeorological parameters objectively selected within a representative domain centered on the meteorological measurement location within the overall study area. This is performed individually for each measurement site for each full year of data collection in order to account for the year to year variability in snow cover and annual precipitation amounts that would affect surface moisture conditions. A discussion of the process followed and surface values determined is presented in Section 3.2.3 below.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-6

• The final step of the meteorological pre-processing utilizes the merged data file to create two files suitable for direct use in AERMOD. The two files are the surface data file (“.SFC”), which contains the planetary boundary layer parameters used to compute dispersion conditions, and the atmospheric profile data file (“.PFL”), which contains the multi-level wind and temperature data used to compute plume rise and transport conditions.

• If the data being pre-processed does not represent a full calendar year, it will be necessary to develop a “single calendar year” model input database. To do this, the last X months of the first year in which data is collected (e.g., August – December) will be placed at the end of the January through July portion of the record and the dates in the file will be changed to reflect the current year.

3.2.3 Surface Characteristics

AERMOD modeling analyses require the input of certain surface conditions that influence boundary layer parameter estimates at the primary meteorological measurement location. These surface conditions are quantified by three characteristics, namely: surface roughness length, Bowen ratio and surface albedo, which relate to the height of obstacles to the wind flow, the amount of moisture at the surface, and the reflectivity of the surface, respectively.

The surface roughness length is related to the height of obstacles to the wind flow and is, in theory, the height at which the mean horizontal wind speed is zero. For this modeling analysis, the surface roughness will be objectively determined via the AERSURFACE tool and will reflect terrain and land use characteristics in a subset of the study area out to 1 km from the onsite meteorological measurement locations, as well as any seasonal variations, in accordance with guidance contained in the AERSURFACE User’s Guide (EPA-454/B-08-001, January 2008).

The daytime Bowen ratio is an indicator of moisture present in the surface surrounding the meteorological measurement location. It is the ratio of the sensible heat flux to the latent heat flux and is used for determining planetary boundary layer parameters for convective conditions. Information in the AERSURFACE User’s Guide suggests that the Bowen ratio be determined as a function of the time of year, the average land-use type over a 10 by 10 km domain centered on the measurement locations, and whether dry, wet or average moisture conditions are prevalent in the area under consideration. For this analysis, average moisture conditions will be used, but only after confirming against a current 30-year climatological record for the Syracuse area on a year-by-year basis. If a particular data collection year or season is determined to be abnormally wet or dry, then the Bowen ratios for wet or dry moisture conditions will be used as appropriate.

The albedo is the fraction of total incident solar radiation reflected by the ground surface back into space without absorption. Typical values range from 0.1 for thick deciduous forests to 0.90 for fresh snow. Appropriate seasonal values for the albedo for processing the meteorological database will determined as described herein. Information in the AERSURFACE User’s Guide suggests that the albedo value be determined as a function of the time of year and the average land-use type over a 10 by 10 km domain centered on the measurement locations.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-7

Because surface characteristics vary as a function of time (i.e., seasonality) and land-use, AERMOD requires the input of surface characteristic data based on user–defined time periods (annual, seasonal or monthly) and directional sectors from a minimum of 1 to a maximum of 12 30-degree sectors (for surface roughness only). For the purpose of determining appropriate surface characteristic values for the two meteorological towers in the Onondaga Lake study area, AERMET pre-processing will employ the use of 12 monthly time periods and 12 30-degree directional sectors within a 1 km radius of the tower locations for surface roughness and within a 10 by 10 km domain centered on the tower locations for albedo and Bowen ratio. Specifying the use of AERSURFACE in this fashion will accurately account for major differences in seasonal conditions and differential land use patterns throughout the study area.

Monthly time periods are used so that site-specific seasons can be defined for every year being processed. Consistent with AERSURFACE guidance, the site-specific seasons will correspond to annual vegetative growth cycles as well as whether continuous snow over exists for one or more months. This approach allows use of the surface characteristic values contained in the AERSURFACE User’s Guide – Appendix A, Tables A-1 through A-3. The site-specific seasons for 2006 and 2007 are defined in Table 3-2 as follows:

Table 3-2 Site-Specific Seasons for Use in AERSURFACE

Seasons Months in 2006 # of Months Months in 2007 # of Months

Late Fall or Winter w/ No Snow

Dec, Jan 2 Not Used --

Winter w/ Snow Feb, Mar 2 Dec to Mar 4

Spring Apr, May 2 Apr, May 2

Summer Jun to Sep 4 Jun to Sep 4

Autumn Oct, Nov 2 Oct, Nov 2

The basis for using the Late Fall/Winter with no snow category in 2006 is confirmation via historical weather observations that little to no continuous snow cover existed in the Syracuse area in January and December of 2006. In contrast, all the winter months in 2007 had significant continuous snow cover.

The basis for using a four-month summer season that includes September, is the fact that we have observed no drop-off in vegetative cover at the meteorological tower locations through the end of September.

The maximum number of directional sectors (12) will be employed because land-use type varies appreciably within the environs of the study area

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-8

To objectively define the land use characteristics and micrometeorological parameters of the Onondaga Lake study area, USEPA’s AERSURFACE tool (January 2008 release) will be applied to a digital mapping version of land use and cover based on U.S. Geological Survey (USGS) National Land Cover Data 1992 archives (NLCD92) for the Syracuse, NY area. The NLCD92 archive provides land cover data at a spatial resolution of 30 meters and mapped using an Albers Conic Equal Area projection. The land cover data is based on a 21-category classification scheme.

The objective analysis resolves fractional land use/cover data into predominant land use/cover data utilizing a grid cell process for a domain extending 5 km in all directions from the center of the study area (10 by 10 km grid centered on measurement locations). AERSURFACE computes the inverse-distance weighted geometric mean land cover type for an upwind distance of 1 km to determine the associated surface roughness within twelve 30-degree wind direction sectors (0-30, 30-60, 60-90, 330-360). Albedo and Bowen ratio determinations are based on simple unweighted arithmetic and geometric means, respectively, within the entire 10 by 10 km domain.

For reference purposes, Table 3-3 summarizes the seasonal input values for different land use categories and seasons used in the AERSURFACE tool.

For modeling the SCA emissions, AERSURFACE has been run with these seasonal input values and with a centroid corresponding to the coordinates of the Wastebed 13 meteorological station. The results are presented in Table 3-4, which summarizes the seasonal average output values for albedo, Bowen Ratio, and surface roughness within each of the 12 30 degree wind sectors extending 3 km out from the study area (Wastebed 13 station location). These are the values input to AERMET to process with the rest of the meteorological data.

3.2.4 Transport Wind Speed and Direction

The primary source of reference wind speed will be the 10-meter level data measured at the onsite towers. If any 10-meter wind speed data values are missing, data from the other onsite tower will be substituted. If data from both towers are missing, then data from Syracuse Airport will be substituted. However, if onsite wind data were missing for only one or two hours, data for the missing hour(s) will be interpolated from onsite data for the hour before and the hour after the missing hour(s). Wind speed is a scalar quantity, and for short periods (one or two hours), it is considered appropriate to fill in the missing onsite wind speed data by linear interpolation. If three or more consecutive hours of wind data were missing from the onsite records, then wind data from either the other onsite tower or from Syracuse Airport will be substituted.

The primary source of transport wind direction will be the 10-meter level data measured at the onsite meteorological towers. If onsite wind data are missing for one or two hours, data for the missing hour(s) will be substituted with data from the other tower or will be interpolated from onsite data for the hour before and the hour after the missing hour(s).

Since wind direction is a vector quantity, direct linear interpolation to fill in missing hourly values is not strictly appropriate. Rather, a multi-step interpolation scheme is applied for wind

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-9

direction data. Short periods of missing onsite wind direction data are filled in the following manner.

First, the onsite wind data for the hour preceding and the hour following the missing hour(s) are resolved into orthogonal easterly and northerly (u and v) components. Values of easterly and northerly wind components for the missing hour(s) are then calculated through linear interpolation based on the hours preceding and following the missing period. Finally, the interpolated orthogonal components for each of the missing hours are recombined into a vector wind, and the direction of the resultant vector wind is used for the missing hour.

If three or more consecutive hours of wind direction data are missing from both of the onsite data records, then wind data from Syracuse Airport will be used. Standard USEPA meteorological preprocessing procedures are used to randomize any offsite wind direction data that are not reported to the nearest degree.

3.2.5 Temperature

Ambient temperature values will be based on 2-meter level measurements from the onsite meteorological towers. If onsite temperature data are missing from both towers for periods between one and four hours, the gaps were filled by linear interpolation based on the hour preceding and following the period of missing data. Longer gaps will be filled in with data from Syracuse Airport.

3.2.6 Atmospheric Stability

Atmospheric stability parameters such as the surface friction velocity (u∗), sensible heat flux (H); Monin-Obukhov length (L, a stability parameter relating u∗ to H) and mixing heights are computed each hour based on the best available data as determined by a series of hierarchal algorithms contained in AERMET.

Unlike the meteorological input requirements for the ISC model, AERMOD does not require that Pasquill-Gifford stability classes be determined. Rather, as defined in AERMET, the atmosphere is considered unstable if the flux of sensible heat is upward at the surface and the time of day is approximately between sunrise and sunset. Specifically, atmospheric conditions are defined as being unstable when L < 0. Otherwise, the atmosphere is considered stable (L>0).

The sections directly below briefly describe the calculations used to compute the stability parameters based on the onsite and offsite data available for this modeling analysis.

Unstable Atmosphere

When atmospheric conditions are determined to be unstable (i.e., when L < 0), both the convective and mechanical mixing heights are calculated. To get to this point, AERMET first estimates the sensible heat flux (H) hour-by-hour by estimating it from the Net Radiation measurements and Bowen ratio. The computations are performed by AERMET following the equations presented in the AERMET User’s Guide (Section 5.4.3).

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-10

Once the heat flux is known, the u∗ and Monin-Obukhov length (L) for the convective boundary layer (CBL) are computed via an iterative algorithm in AERMET whereby u∗ and L change with each iteration, but the heat flux remains constant. The iteration continues until consecutive values of L differ by 1% or less.

Lastly, the mixing heights are calculated. The convective mixing height (Zic) is estimated by a formulation that is based on a one-dimensional (height) energy balance approach. In this approach, the heat flux in the CBL at the surface, and entrained from the stable air aloft, leads to vertical mixing, a rise in the base of the elevated temperature inversion, and an increase of the energy of the boundary layer air. The convective mixing height created by this dynamic is computed from the potential temperature distribution (θ(z)) from the morning sounding and the estimated heat flux (H). AERMET restricts the growth of the convective mixing height to 4,000 meters.

The mechanical mixing height (Zim) is determined from the diagnostic expression:

Zim = 2,300 (u∗3/2).

Stable Atmosphere

When atmospheric conditions are determined to be stable (i.e., when L > 0), only the mechanical mixing height is calculated. To get to this point, AERMET estimates the sensible heat flux (H) for the stable atmosphere hour-by-hour by using estimates of the surface friction velocity (u∗) and a temperature scale (θ∗). The surface friction velocity and temperature scale are computed directly from NWS cloud cover data and onsite wind speed and surface temperature. The computations are performed by AERMET following the equations presented in the AERMET User’s Guide (Section 5.4.4). Because the heat flux may become unrealistically large in the case of strong winds, a limit of –64 W/m2 is placed on the heat flux value for any hour.

Once the heat flux is known, the Monin-Obukhov length (L) for the stable boundary layer (SBL) is directly computed using the surface friction velocity, ambient temperature and heat flux. No iterations are performed.

Lastly, the mixing height is calculated. The mechanical mixing height (Zim) is determined from the diagnostic expression Zim = 2,300 (u∗3/2).

3.2.7 Mixing Heights

The calculations above produce a continuous record of mechanical mixing heights, along with a record of convective mixing heights that is restricted to daytime hours of upward heat flux. Once these mixing heights are computed, the mechanical mixing heights are subjected to an algorithm in AERMET that smoothes all hours - both those determined to be stable and unstable - so that the effect of any large hour-to-hour fluctuations of the surface friction velocity on the mixing height is minimized.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-11

3.2.8 Sounding Data

Calculation of the convective mixing height is based on the twice-daily (morning and afternoon) upper-air sounding data from Buffalo, NY. When pre-processing the meteorological database, AERMET retrieves data up to and including the first measurement level above 5,000 meters.

For the occasional cases where the sounding is lost at a level well below 5,000 meters, there may not be adequate upper air data to allow AERMET to compute the convective mixing height for late afternoon hours. When this situation exists, AERMET extends the sounding to 5,000 meters by computing the potential temperature gradient for the upper 500 meters of the existing sounding data and extending this to the 5,000-meter level.

If a sounding is missing, a value equal to the monthly average morning or afternoon mixing height, as appropriate, is substituted.

3.2.9 AERMOD Meteorological Input Files

The meteorological database, including all of the parameters developed as described above for each hour in each yearly record via the AERSURFACE and AERMET processing routines, are created as one boundary layer file (*.sfc) and one atmospheric profile file (*.pfl) for subsequent input to AERMOD.

3.3 MODEL RECEPTOR GRID

With the exception of any stacks or vents associated with a wastewater treatment facility or other emissions control devices, virtually all of the sources currently anticipated to be modeled are surfaced-based and will not have any significant upward exhaust momentum or buoyancy (as an industrial stack would have). Therefore, it is expected that maximum concentrations of source emissions will occur very near to the sources themselves and then decrease with distance. The potential for maximum impacts to occur somewhat farther away from the sources will exist for cases where additive impacts from multiple sources at different locations may occur at more distance model receptor points. This condition is most likely to occur in the estimation of long-term (i.e., annual) average impacts at a given model receptor, which will represent the combined impacts of several sources that may be emitting anytime during the year.

As a result, the model receptor grid for the modeling analyses is focused on establishing a fairly dense grid of model receptor points around the SCA and in and around the dredging areas at the lake. Since the potential magnitude of estimated concentrations is not yet known, it is not yet possible to definitively determine how far the final grid should extend.

However, an initial grid extending out to a distance of 3-4 km to the north, east, and south from the lake shoreline, and 5 km to the west of Wastebed 13, has been developed to ensure that the magnitude and variability of estimated concentrations of modeled air contaminants are adequately described, and to demonstrate that estimated concentrations at the outer edges of the grid are well below any applicable ambient air quality standard and/or threshold values.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-12

Based on the above, Figure 3-1 depicts a plot of the initial Cartesian Coordinate model receptor grid system developed for this analysis. The overall grid encompasses an area that is 16 km (east-west) x 12 km (north-south) roughly centered on a mid-point between the southwestern lakeshore and the SCA. The grid is comprised of discrete receptors and is defined by a series of nested sub-grids of varying density and also by strings of property line model receptors surrounding the source areas.

There are a total of 12,625 receptor points in the overall grid. Figure 3-2 is a similar plot of the model receptor grid superimposed over a topographic map section of the area that shows the grid’s geographical coverage

Nested Grids

The innermost grid consists of model receptors placed at 50-meter intervals out to distances of at least 1 km in each direction from the N-E-S-most edges of the lake dredging area and from the S-W-N-most boundaries of Wastebed 13, as well as covering the entire area in between these two potential emission source locations. This grid includes analysis points placed on the lake surface that are outside of the dredging work areas.

The next sub-grids have receptors placed at 250-meter intervals out to distances roughly 3-4 km from each of the source areas, followed by model receptor points at 500-meter spacing intervals out to the outer edges of the grid.

Property Line Model Receptors

A series of property line model receptor points are placed at 25-meter intervals around the entire wastebed boundary and around the initial SMU (i.e., SMU1) scheduled for dredging operations. These are seen in the close-up views of the model receptor grid in Figures 3-3 and 3-4. For this analysis, the property lines define the areas representing Honeywell property in the wastebed area with restricted (fenced and gated) access, and work areas in the dredge zone that will be cordoned off during dredging operations.

As seen in Figure 3-2, the Wastebed area consists of four separate Wastebeds, designated as Wastebeds 12, 13 (the SCA), 14, and 15, plus a Lagoon and Open Area. This entire wastebed area is fenced in with access controlled via a locked gate. Honeywell has, however, granted limited access to a local municipality for depositing waste in a portion of Wastebed 15. Thus, this portion of the wastebed has been excluded from the site property and receptor points have been placed in this area.

Additional receptors representing specifically identified locations may be added to the grid. First, any sensitive receptors, such as schools, hospitals, outdoor recreation areas, or other major public facilities may be represented by discrete receptor points if not deemed to be adequately represented by one or more of the primary receptor points. The locations of any such receptors will be jointly determined by members of the Emissions and Odors Work Group (i.e., determined Honeywell and NYSDEC technical staff representatives).

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-13

Second, during the course of model execution, additional receptors may be added with a denser spacing interval in areas of maximum predicted concentrations if needed to better define the point(s) of maximum impact and/or concentration gradients in critical impact areas. Given the density and coverage of the initial grid, the need for such additional receptors is unlikely.

Once some initial modeling results are available and examined in detail, the overall size and spacing of the receptor grid may be modified to eliminate unnecessary receptors if it is determined that no resolution of results is lost. (Given the large total number of receptors on the initial grid, reducing the overall numbers of receptors would substantially reduce computer run times.)

Workplace Model Receptors

A series of onsite worker exposure model receptors are also included in the base grid to directly assess levels of air emissions potential experienced by personnel working directly onsite at the SCA and in-lake areas. Modeled impacts at these analysis points will be used to analyze short-term impacts to onsite workers. Analysis point spacing is assumed to be approximately 50 meters for these areas.

Note: For various modeling objectives, model runs may be executed using portions of the overall grid. For example, modeling of worker exposure levels will be limited to the portion of the grids covering the direct work activity areas. In contrast, offsite impact analyses will exclude the work area receptors.

AERMAP Pre-Processing

In addition to the X-Y coordinates of each designated model receptor point on the grid, the model input includes the terrain height for each point. The terrain elevations have been defined using the AERMAP Terrain Pre-processor of AERMOD. AERMAP pre-processing utilizes digital elevation model (DEM) files corresponding to 7.5-minute topographic quadrangles, along with a project-specific receptor grid, to calculate the following parameters:

• A height scale (hc) for each receptor point. The height scale, or hill height, is required as input to the AERMOD model. The height scale is used to calculate the critical dividing streamline height (Hcrit) for each receptor based on the controlling terrain for the receptor. These are also calculated based on the data contained in the nine DEM files.

• The terrain elevation of each receptor extracted from the DEM files corresponding to the 7.5-minute series USGS topographic quadrangles defining the modeling analysis domain.

For this analysis, AERMAP is used to calculate both the required height scale data, as well as the model receptor elevations. The AERMAP preprocessing of the model receptor grid for this analysis can be further described as follows:

• AERMAP version 06341 or greater will be used to preprocess the model receptor grid described above.

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

3-14

• Receptor elevation data will be extracted by AERMAP. Receptor elevations via AERMAP are actual terrain heights.

• 7.5-minute DEM files corresponding to the nine topographic quadrangles containing and surrounding the Onondaga Lake study area will be used as input. These nine quads also define the modeling domain. The nine quads (DEM files) to be used for the analysis are:

Syracuse West Syracuse East Jamesville South Onondaga Marcellus Camillus Baldwinsville Brewerton Cicero

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

FIGURE 3-1

BASE MODEL RECEPTOR GRID

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

FIGURE 3-2

BASE RECEPTOR GRID

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

SCA BASE MODEL RECEPTOR GRID

Grid in meters (1 in. = 400 m)

N

Wastebed 13

Wastebed 14

Wastebed12

Wastebed 15

Lagoon

Open AreaMet Tower

Location

NE Cell

SE Cell

SW Cell

NW Cell

FIGURE 3-3

290 ELWOOD DAVIS RD, SUITE 312, LIVERPOOL, NY 13088 PHONE: (315) 451-9560

PARSONS

ONONDAGA LAKE

SYRACUSE, NEW YORK

P:\HON-SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Figures

FIGURE 3-4

SMU 1 BASE MODEL RECEPTOR GRID

Lakeshore

SMU1

“Fenceline”Met Tower

Location

N

Grid in meters (1 in. = 300 m)

HONEYWELL ONONDAGA LAKE REMEDIATION PROJECTAIR DISPERSION MODELING PROTOCOL

Season Land Use Classification (EPA MRLC) Albedo Bowen Ratio Surface Roughnes

Water 0.1 0.1 0.001Low Intensity Residential 0.45 0.5 0.5High Intensity Residential 0.35 0.5 1High Intensity Commercial/Industrial 0.35 0.5 0.8Pasture/Hay 0.6 0.5 0.01Row Crops 0.6 0.5 0.01Parks, Lawns, Golf Courses 0.6 0.5 0.005

Winter Evergreen Forest 0.35 0.5 1.3Mixed Forest 0.42 0.5 0.9Deciduous Forest 0.5 0.5 0.5Woody Wetlands 0.3 0.5 0.5Emergent Wetlands 0.3 0.5 0.1Barren; Quarries, Strip Mines, Gravel Pits 0.6 0.5 0.3Barren; Bare Rock and Sand 0.6 0.5 0.05Barren; Transitional 0.45 0.5 0.2

Water 0.1 0.1 0.001Low Intensity Residential 0.16 0.8 0.52High Intensity Residential 0.18 1.5 1High Intensity Commercial/Industrial 0.18 1.5 0.8Pasture/Hay 0.14 0.3 0.03Row Crops 0.14 0.3 0.03Parks, Lawns, Golf Courses 0.15 0.3 0.015

Spring Evergreen Forest 0.12 0.7 1.3Mixed Forest 0.14 0.7 1.15Deciduous Forest 0.16 0.7 1Woody Wetlands 0.14 0.2 0.7Emergent Wetlands 0.14 0.1 0.2Barren; Quarries, Strip Mines, Gravel Pits 0.2 1.5 0.3Barren; Bare Rock and Sand 0.2 1.5 0.05Barren; Transitional 0.18 1 0.2

Water 0.1 0.1 0.001Low Intensity Residential 0.16 0.8 0.54High Intensity Residential 0.18 1.5 1High Intensity Commercial/Industrial 0.18 1.5 0.8Pasture/Hay 0.2 0.5 0.15Row Crops 0.2 0.5 0.2Parks, Lawns, Golf Courses 0.15 0.5 0.02

Summer Evergreen Forest 0.12 0.3 1.3Mixed Forest 0.14 0.3 1.3Deciduous Forest 0.16 0.3 1.3Woody Wetlands 0.14 0.2 0.7Emergent Wetlands 0.14 0.1 0.2Barren; Quarries, Strip Mines, Gravel Pits 0.2 1.5 0.3Barren; Bare Rock and Sand 0.2 1.5 0.05Barren; Transitional 0.18 1 0.2

Water 0.1 0.1 0.001Low Intensity Residential 0.16 1 0.54High Intensity Residential 0.18 1.5 1High Intensity Commercial/Industrial 0.18 1.5 0.8Pasture/Hay 0.2 0.7 0.15Row Crops 0.2 0.7 0.2Parks, Lawns, Golf Courses 0.15 0.7 0.015

Autumn Evergreen Forest 0.12 0.8 1.3Mixed Forest 0.14 0.9 1.3Deciduous Forest 0.16 1 1.3Woody Wetlands 0.14 0.2 0.7Emergent Wetlands 0.14 0.1 0.2Barren; Quarries, Strip Mines, Gravel Pits 0.2 1.5 0.3Barren; Bare Rock and Sand 0.2 1.5 0.05Barren; Transitional 0.18 1 0.2

TABLE 3-3Air Dispersion Modeling Protocol for Onondaga LakeSeasonal Values of Micrometeorological Parameters

P:\Honeywell -SYR\443584 - Operations\09 Reports\9.01 - Dispersion Modeling Protocol\Modeling Protocol\Tables 1-1, 3-3.xls5/12/2008 1 of 1 PARSONS

AIR DISPERSION MODELING PROTOCOL FOR ONONDAGA LAKE

DRAFT FINAL

p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\table 3-4.doc Page 1 of 6

Table 3-4A Seasonal Average Micrometeorological Parameters

Around WB-13 Tower in 2006 by Wind Sector

Season Land Cover Albedo Bowen Ratio* Surface Roughness

Late Autumn 0.17 0.75 0.338 Winter 0.45 0.43 0.268 Spring 0.15 0.51 0.490

Summer 0.16 0.44 0.765

Sector 1 000 to 030

Autumn 0.16 0.74 0.761

Late Autumn 0.17 0.75 0.254 Winter 0.45 0.43 0.195 Spring 0.15 0.51 0.346

Summer 0.16 0.44 0.630

Sector 2 030 to 060

Autumn 0.16 0.74 0.630

Late Autumn 0.17 0.75 0.184 Winter 0.45 0.43 0.132 Spring 0.15 0.51 0.269

Summer 0.16 0.44 0.564

Sector 3 060 to 090

Autumn 0.16 0.74 0.561

Late Autumn 0.17 0.75 0.284 Winter 0.45 0.43 0.216 Spring 0.15 0.51 0.427

Summer 0.16 0.44 0.736

Sector 4 090 to 120

Autumn 0.16 0.74 0.734

Late Autumn 0.17 0.75 0.190 Winter 0.45 0.43 0.137 Spring 0.15 0.51 0.274

Summer 0.16 0.44 0.565

Sector 5 120 to 150

Autumn 0.16 0.74 0.564

Late Autumn 0.17 0.75 0.182 Winter 0.45 0.43 0.132 Spring 0.15 0.51 0.248

Summer 0.16 0.44 0.541

Sector 6 150 to 180

Autumn 0.16 0.74 0.541

Late Autumn 0.17 0.75 0.262 Winter 0.45 0.43 0.199 Spring 0.15 0.51 0.382

Summer 0.16 0.44 0.692

Sector 7 180 to 210

Autumn 0.16 0.74 0.692

AIR DISPERSION MODELING PROTOCOL FOR ONONDAGA LAKE

DRAFT FINAL

p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\table 3-4.doc Page 2 of 6

Table 3-4A (cont) Seasonal Average Micrometeorological Parameters

Around WB-13 Tower in 2006 by Wind Sector

Season Land Cover Albedo Bowen Ratio* Surface Roughness

Late Autumn 0.17 0.75 0.368 Winter 0.45 0.43 0.310 Spring 0.15 0.51 0.481

Summer 0.16 0.44 0.678

Sector 8 210 to 240

Autumn 0.16 0.74 0.677

Late Autumn 0.17 0.75 0.255 Winter 0.45 0.43 0.199 Spring 0.15 0.51 0.339

Summer 0.16 0.44 0.547

Sector 9 240 to 270

Autumn 0.16 0.74 0.537

Late Autumn 0.17 0.75 0.132 Winter 0.45 0.43 0.090 Spring 0.15 0.51 0.196

Summer 0.16 0.44 0.437

Sector 10 270 to 300

Autumn 0.16 0.74 0.430

Late Autumn 0.17 0.75 0.250 Winter 0.45 0.43 0.184 Spring 0.15 0.51 0.402

Summer 0.16 0.44 0.744

Sector 11 300 to 330

Autumn 0.16 0.74 0.739

Late Autumn 0.17 0.75 0.273 Winter 0.45 0.43 0.205 Spring 0.15 0.51 0.433

Summer 0.16 0.44 0.743

Sector 12 330 to 360

Autumn 0.16 0.74 0.736

For WB-13 2006 input to AERMET, the seasons were defined as follows: Late Fall: Dec, Jan 2 months Winter: Feb-Mar 2 months Spring: Apr-May 2 months Summer: Jun-Sep 4 months Fall: Oct-Nov 2 months

AIR DISPERSION MODELING PROTOCOL FOR ONONDAGA LAKE

DRAFT FINAL

p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\table 3-4.doc Page 3 of 6

Table 3-4B Seasonal Average Micrometeorological Parameters

Around WB-13 Tower in 2006 by Wind Sector

Season Land Cover Albedo Bowen Ratio* Surface Roughness

Late Autumn -- -- -- Winter 0.45 0.43 0.268 Spring 0.15 0.51 0.490

Summer 0.16 0.44 0.765

Sector 1 000 to 030

Autumn 0.16 0.74 0.761

Late Autumn -- -- -- Winter 0.45 0.43 0.195 Spring 0.15 0.51 0.346

Summer 0.16 0.44 0.630

Sector 2 030 to 060

Autumn 0.16 0.74 0.630

Late Autumn -- -- -- Winter 0.45 0.43 0.132 Spring 0.15 0.51 0.269

Summer 0.16 0.44 0.564

Sector 3 060 to 090

Autumn 0.16 0.74 0.561

Late Autumn -- -- -- Winter 0.45 0.43 0.216 Spring 0.15 0.51 0.427

Summer 0.16 0.44 0.736

Sector 4 090 to 120

Autumn 0.16 0.74 0.734

Late Autumn -- -- -- Winter 0.45 0.43 0.137 Spring 0.15 0.51 0.274

Summer 0.16 0.44 0.565

Sector 5 120 to 150

Autumn 0.16 0.74 0.564

Late Autumn -- -- -- Winter 0.45 0.43 0.132 Spring 0.15 0.51 0.248

Summer 0.16 0.44 0.541

Sector 6 150 to 180

Autumn 0.16 0.74 0.541

Late Autumn -- -- -- Winter 0.45 0.43 0.199 Spring 0.15 0.51 0.382

Summer 0.16 0.44 0.692

Sector 7 180 to 210

Autumn 0.16 0.74 0.692

AIR DISPERSION MODELING PROTOCOL FOR ONONDAGA LAKE

DRAFT FINAL

p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\table 3-4.doc Page 4 of 6

Table 3-4B (cont) Seasonal Average Micrometeorological Parameters

Around WB-13 Tower in 2006 by Wind Sector

Season Land Cover Albedo Bowen Ratio* Surface Roughness

Late Autumn -- -- -- Winter 0.45 0.43 0.310 Spring 0.15 0.51 0.481

Summer 0.16 0.44 0.678

Sector 8 210 to 240

Autumn 0.16 0.74 0.677

Late Autumn -- -- -- Winter 0.45 0.43 0.199 Spring 0.15 0.51 0.339

Summer 0.16 0.44 0.547

Sector 9 240 to 270

Autumn 0.16 0.74 0.537

Late Autumn -- -- -- Winter 0.45 0.43 0.090 Spring 0.15 0.51 0.196

Summer 0.16 0.44 0.437

Sector 10 270 to 300

Autumn 0.16 0.74 0.430

Late Autumn -- -- -- Winter 0.45 0.43 0.184 Spring 0.15 0.51 0.402

Summer 0.16 0.44 0.744

Sector 11 300 to 330

Autumn 0.16 0.74 0.739

Late Autumn -- -- -- Winter 0.45 0.43 0.205 Spring 0.15 0.51 0.433

Summer 0.16 0.44 0.743

Sector 12 330 to 360

Autumn 0.16 0.74 0.736

For WB-13 2007 input to AERMET, the seasons were defined as follows: Late Fall: Not Used 0 months Winter: Dec-Mar 4 months Spring: Apr-May 2 months Summer: Jun-Sep 4 months Fall: Oct-Nov 2 months

AIR DISPERSION MODELING PROTOCOL FOR ONONDAGA LAKE

DRAFT FINAL

p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\table 3-4.doc Page 5 of 6

Table 3-4C Seasonal Average Micrometeorological Parameters Around Lakeshore Tower in 2007 by Wind Sector

Season Land Cover Albedo Bowen Ratio* Surface Roughness

Late Autumn -- -- -- Winter 0.40 0.41 0.002 Spring 0.16 0.65 0.003

Summer 0.16 0.60 0.003

Sector 1 000 to 030

Autumn 0.16 0.81 0.003

Late Autumn -- -- -- Winter 0.40 0.41 0.002 Spring 0.16 0.65 0.002

Summer 0.16 0.60 0.002

Sector 2 030 to 060

Autumn 0.16 0.81 0.002

Late Autumn -- -- -- Winter 0.40 0.41 0.002 Spring 0.16 0.65 0.002

Summer 0.16 0.60 0.002

Sector 3 060 to 090

Autumn 0.16 0.81 0.002

Late Autumn -- -- -- Winter 0.40 0.41 0.073 Spring 0.16 0.65 0.099

Summer 0.16 0.60 0.110

Sector 4 090 to 120

Autumn 0.16 0.81 0.108

Late Autumn -- -- -- Winter 0.40 0.41 0.292 Spring 0.16 0.65 0.428

Summer 0.16 0.60 0.484

Sector 5 120 to 150

Autumn 0.16 0.81 0.461

Late Autumn -- -- -- Winter 0.40 0.41 0.564 Spring 0.16 0.65 0.638

Summer 0.16 0.60 0.664

Sector 6 150 to 180

Autumn 0.16 0.81 0.652

Late Autumn -- -- -- Winter 0.40 0.41 0.697 Spring 0.16 0.65 0.733

Summer 0.16 0.60 0.746

Sector 7 180 to 210

Autumn 0.16 0.81 0.739

AIR DISPERSION MODELING PROTOCOL FOR ONONDAGA LAKE

DRAFT FINAL

p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\table 3-4.doc Page 6 of 6

Table 3-4C (cont) Seasonal Average Micrometeorological Parameters Around Lakeshore Tower in 2007 by Wind Sector

Season Land Cover Albedo Bowen Ratio* Surface Roughness

Late Autumn -- -- -- Winter 0.40 0.41 0.547 Spring 0.16 0.65 0.625

Summer 0.16 0.60 0.651

Sector 8 210 to 240

Autumn 0.16 0.81 0.639

Late Autumn -- -- -- Winter 0.40 0.41 0.339 Spring 0.16 0.65 0.462

Summer 0.16 0.60 0.510

Sector 9 240 to 270

Autumn 0.16 0.81 0.488

Late Autumn -- -- -- Winter 0.40 0.41 0.431 Spring 0.16 0.65 0.613

Summer 0.16 0.60 0.693

Sector 10 270 to 300

Autumn 0.16 0.81 0.676

Late Autumn -- -- -- Winter 0.40 0.41 0.343 Spring 0.16 0.65 0.495

Summer 0.16 0.60 0.560

Sector 11 300 to 330

Autumn 0.16 0.81 0.541

Late Autumn -- -- -- Winter 0.40 0.41 0.010 Spring 0.16 0.65 0.011

Summer 0.16 0.60 0.012

Sector 12 330 to 360

Autumn 0.16 0.81 0.012

For Lake/Willis Ave 2007 input to AERMET, the seasons were defined as follows: Late Fall: Not Used 0 months Winter: Dec-Mar 4 months Spring: Apr-May 2 months Summer: Jun-Sep 4 months Fall: Oct-Nov 2 months

AIR DISPERSION MODELING PROTOCOL

FOR ONONDAGA LAKE

Parsons p:\honeywell -syr\443584 - operations\09 reports\9.01 - dispersion modeling protocol\modeling protocol\revised to nysdec\ambient dispersion modeling protocol final.doc March 12, 2008

4-1

SECTION 4

DISPERSION MODELING RESULTS REPORTING

Dispersion model runs will be set up and executed as discrete runs for each appropriate combination of sources and/or source groups, operating scenario(s) defining dredging and SCA activities, air contaminant, model receptor grid, and meteorological data record. As appropriate, source groups will be designated within each run in order to segregate the impacts of the different sources or groups of sources modeled, and for the variations in emission rates for each source as a function of time and/or activity level (this includes sources being turned “off” if no emissions are anticipated for a particular time period.).

The dispersion estimates from the modeling analyses will be summarized in a comprehensive series of tables and graphs that present dispersion estimates for the above runs as follows:

• for each individual source and source group;

• for all sources combined;

• for each year in the meteorological data input file (if more than 1 year of onsite data is available and used in the analysis);

• for each contaminant and averaging period (short-term and long-term) for which an ambient air standard or threshold has been established; and

• for model receptor grid.

These modeling results will then be used to:

• identify those sources and emissions having the greatest potential impacts;

• design of best management practices, control systems and operation strategies to reduce air emissions resulting from remedial activities to levels less than applicable short-term and long-term ambient air standards, and established threshold levels; and

• document and communicate the potential impacts of air emissions of COIs and odors from the remedial activities to the community through a series of public outreach and educational programs.

To evaluate the impact from the remedial activities, a set of site-specific air goals will be developed for the project. The goals will address risks associated with exposures to both onsite (workers) and offsite receptors (general public). Development of these goals and comparison of the estimated ambient air impacts to the goals will be conducted as a separate portion of the overall design process.

Appendix D

Length Weighted Average Procedure and Sample Locations

Procedure for Calculation of Length-Weighted Average Concentrations

A length-weighted average approach was used to calculate concentrations considered representative of the dredge material from Onondaga Lake. The process for calculating the average involved the following steps: 1) data processing and 2) calculation of length-weighted average concentrations at each core location. These steps are described briefly below.

The project database was queried for samples within the dredge prism (i.e., the area where sediment will be removed from the lake bottom) from both the remedial investigation (RI) and pre-design investigation (PDI). The table below summarizes the number of locations from each year of sampling. A complete list of locations is provided in Table D-1.

PHASE  # LOCATIONS 

PDI Phase 1 (2005)  65 PDI Phase 2 (2006)  127 PDI Phase 3 (2007)  9 PDI Phase 4 (2008)  21 PDI Phase 5 (2009)  69 RI (1992)  10 RI (2000)  29 

Total Number of Locations =  330 Note: Not all locations were sampled for each analyte. 

Data ignored in this analysis included cores having a length less than 50 percent of the dredge cut thickness (which were considered non-representative of the dredge material), samples fully beneath the dredge cut, data that failed to meet data quality criteria during validation, and data without coordinates. Total xylene was not reported during the RI, and was therefore calculated as the sum of m- & p-xylene and o-xylene, using one-half the method detection limit (MDL) for non-detect (ND) sample results. Replicates collected during the PDI were processed using the maximum of the two samples. Replicates collected during the RI were processed, using the following rules:

• if both samples detect, use average • if one detect & one non-detect, use average with ND at one-half MDL • If both samples non-detect, use lowest MDL

For dioxin/furans, one-half MDL was used for non-detect sample results and toxic equivalents (TEQs) were calculated using the 2005 World Health Organization humans/mammalian toxic equivalency factors for dioxin/furans relative to 2,3,7,8-TCDD. Because dioxin data are limited, the locations were not restricted to core length greater than 50 percent of the dredge cut. Dioxin data were only available for a total of eight locations in three remediation areas within the dredge prism (i.e., two from RA-A, four from RA-D (Center), and two from RA-E).

Using the processed data, length-weighted average concentrations (LWAs) were then calculated at each core location. Section lengths used in the LWA calculation were based on the section end-depth minus the section start-depth, except for samples from core intervals that “straddled” the dredge cut, for which

the section end-depth was set equal to the dredge cut depth. Using these lengths for all sections within a core, the LWA concentration over the dredge prism thickness at that location was calculated as follows:

∑

 

where i is the individual sample section within a core location.

The length-weighted average concentrations for each core location were then used to calculate the mean concentration and 95 percent upper confidence limit (UCL) on the mean concentration for each contaminant within the dredge prism using ProUCL 4.0 software.

 

Table D‐1.  Locations included in the length‐weighted average (LWA) calculation.Location Group Name Location ID Easting Northing Year Sampled Total Core Length (ft) Dredge Depth (ft)

PDI Phase 1 OL‐SB‐60004‐VC 928702.86 1120572.51 2005 13.2 4.28

PDI Phase 1 OL‐SB‐60005‐VC 928807.22 1119933.09 2005 13.2 3.34

PDI Phase 1 OL‐SB‐60006‐VC 928676.36 1119541.02 2005 13.2 9.91

PDI Phase 1 OL‐SB‐60007‐VC 928698.34 1119195.42 2005 13.2 5.68

PDI Phase 1 OL‐SB‐60008‐VC 928646.8 1118910.34 2005 13.2 5.21

PDI Phase 1 OL‐SB‐60010‐VC 928579.48 1118356.42 2005 13.2 5.05

PDI Phase 1 OL‐SB‐60011‐VC 928296.95 1117842.33 2005 13.2 5.03

PDI Phase 1 OL‐SB‐60012‐VC 928157.47 1117573.85 2005 13.2 5.08

PDI Phase 1 OL‐SB‐60013‐VC 928031.29 1117304.78 2005 13.2 3.39

PDI Phase 1 OL‐SB‐60014‐VC 927755.82 1117202.15 2005 13.2 5.18

PDI Phase 1 OL‐SB‐60015‐VC 927606 1117016.9 2005 13.2 5.25

PDI Phase 1 OL‐SB‐70001‐VC 927065.4 1116285.51 2005 19.6 3.28

PDI Phase 1 OL‐SB‐70002‐VC 927021.33 1116473.36 2005 19.8 3.55

PDI Phase 1 OL‐SB‐70003‐VC 927305.1 1116492.03 2005 19.7 3.30

PDI Phase 1 OL‐SB‐70004‐VC 927225.11 1116590.38 2005 19.3 4.25

PDI Phase 1 OL‐STA‐10002‐VC 924291.63 1118311.97 2005 13.2 1.30

PDI Phase 1 OL‐STA‐10003‐VC 924907.56 1118153.61 2005 13.2 10.18

PDI Phase 1 OL‐STA‐10004‐VC 924950.9 1118356.41 2005 13.2 7.44

PDI Phase 1 OL‐STA‐10005‐VC 925576.04 1117898.15 2005 13.2 10.00

PDI Phase 1 OL‐STA‐10006‐VC 925796.04 1118267.11 2005 13.2 3.11

PDI Phase 1 OL‐STA‐10007‐VC 926138.49 1117646.82 2005 13.2 10.20

PDI Phase 1 OL‐STA‐10008‐VC 923319.57 1118082.96 2005 19.4 3.45

PDI Phase 1 OL‐STA‐10009‐VC 924174.44 1117950.72 2005 18.7 5.53

PDI Phase 1 OL‐STA‐10010‐VC 924773.69 1117686.08 2005 19.5 8.80

PDI Phase 1 OL‐STA‐10011‐VC 925301.95 1117369.9 2005 19.3 5.62

PDI Phase 1 OL‐STA‐10012‐VC 925924.34 1116800.55 2005 19.7 9.87

PDI Phase 1 OL‐STA‐10013‐VC 923909.42 1118383.6 2005 13.2 3.42

PDI Phase 1 OL‐STA‐10015‐VC 925443.6 1118225.72 2005 13.2 6.82

PDI Phase 1 OL‐STA‐10016‐VC 925898.56 1117871.97 2005 13.2 12.67

PDI Phase 1 OL‐STA‐10017‐VC 926104 1118245.6 2005 12.6 3.28

PDI Phase 1 OL‐STA‐10018‐VC 923779.99 1117842.07 2005 19.2 5.72

PDI Phase 1 OL‐STA‐10019‐VC 923856.72 1118120.13 2005 19.1 5.50

PDI Phase 1 OL‐STA‐10020‐VC 924394.2 1117726.18 2005 19.8 8.75

PDI Phase 1 OL‐STA‐10021‐VC 924467.87 1117945.12 2005 19.3 5.38

PDI Phase 1 OL‐STA‐10022‐VC 924536.07 1118152.34 2005 19.7 1.28

PDI Phase 1 OL‐STA‐10023‐VC 925019.68 1117452.38 2005 19.7 5.50

PDI Phase 1 OL‐STA‐10024‐VC 925237.66 1117861.9 2005 19.6 9.90

PDI Phase 1 OL‐STA‐10025‐VC 925488.14 1117211.28 2005 18.5 9.90

PDI Phase 1 OL‐STA‐10026‐VC 925702.21 1117571.58 2005 19.7 9.90

PDI Phase 1 OL‐STA‐20002‐VC 922578.09 1118665.45 2005 36.8 10.62

PDI Phase 1 OL‐STA‐20003‐VC 922643.33 1118494.74 2005 40.2 2.97

PDI Phase 1 OL‐STA‐20005‐VC 922826.08 1118448.78 2005 36.6 8.91

PDI Phase 1 OL‐STA‐20006‐VC 922874.8 1118261.1 2005 40.2 6.30

PDI Phase 1 OL‐STA‐20008‐VC 923030.84 1118223.48 2005 36.5 6.59

PDI Phase 1 OL‐STA‐20009‐VC 922522.9 1118611.1 2005 40.1 1.98

PDI Phase 1 OL‐STA‐20011‐VC 922701.28 1118547.15 2005 36.8 9.90

PDI Phase 1 OL‐STA‐20012‐VC 922766 1118395 2005 36.6 6.56

PDI Phase 1 OL‐STA‐20014‐VC 922919.31 1118321.86 2005 36.7 6.59

PDI Phase 1 OL‐STA‐20015‐VC 922985.55 1118145.43 2005 33.5 6.54

PDI Phase 1 OL‐STA‐30010‐VC 918258.98 1124185.79 2005 13.2 4.14

PDI Phase 1 OL‐STA‐30012‐VC 919124.6 1123593.04 2005 13.2 4.90

PDI Phase 1 OL‐STA‐30013‐VC 919404.09 1123299.06 2005 13.2 4.23

PDI Phase 1 OL‐STA‐30017‐VC 920559.97 1121166.67 2005 13.0 2.06

PDI Phase 1 OL‐STA‐30018‐VC 920912.28 1120707.1 2005 13.0 3.74

PDI Phase 1 OL‐STA‐30019‐VC 921110.3 1120319.7 2005 13.2 4.86

PDI Phase 1 OL‐STA‐40001‐VC 915404.7 1125692.7 2005 13.2 2.53

PDI Phase 1 OL‐STA‐40002‐VC 916064.26 1125857.26 2005 13.2 2.50

PDI Phase 1 OL‐STA‐60016‐VC 928828.31 1120195.85 2005 13.2 5.48

PDI Phase 1 OL‐STA‐60017‐VC 928659.52 1120228.61 2005 13.2 3.85

PDI Phase 1 OL‐STA‐60018‐VC 928435.12 1118118.19 2005 13.2 13.69

PDI Phase 1 OL‐STA‐60019‐VC 928293.14 1118199.33 2005 13.2 13.54

PDI Phase 1 OL‐STA‐70005‐VC 926798.39 1116226.07 2005 19.6 4.67

PDI Phase 1 OL‐STA‐70006‐VC 926798.06 1116417 2005 19.8 3.69

PDI Phase 1 OL‐STA‐70007‐VC 927512.3 1116660.37 2005 19.8 3.27

PDI Phase 1 OL‐STA‐70008‐VC 927417.02 1116761.59 2005 19.8 3.72

PDI Phase 2 OL‐VC‐10038 923415.06 1117947.64 2006 18.9 5.97

PDI Phase 2 OL‐VC‐10039 923575.4 1118088.4 2006 18.0 9.90

PDI Phase 2 OL‐VC‐10040 923587.3 1118351.9 2006 13.2 12.53

PDI Phase 2 OL‐VC‐10041 923642.1 1117900.7 2006 19.7 5.93

PDI Phase 2 OL‐VC‐10041A 923643.9 1117900 2006 19.8 6.11

PDI Phase 2 OL‐VC‐10042 923697.1 1118505 2006 19.8 8.11

PDI Phase 2 OL‐VC‐10042A 923698.1 1118507.2 2006 19.8 8.11

PDI Phase 2 OL‐VC‐10044 923964.8 1118365.5 2006 19.0 3.49

PDI Phase 2 OL‐VC‐10046 924008.6 1118047.3 2006 19.2 5.51

PDI Phase 2 OL‐VC‐10046A 924010.7 1118045 2006 18.1 5.51

Location Group Name Location ID Easting Northing Year Sampled Total Core Length (ft) Dredge Depth (ft)

PDI Phase 2 OL‐VC‐10048 924158.4 1118168.8 2006 19.7 8.90

PDI Phase 2 OL‐VC‐10048A 924160.5 1118168.3 2006 17.9 8.80

PDI Phase 2 OL‐VC‐10049 924167.4 1117989.7 2006 19.6 5.50

PDI Phase 2 OL‐VC‐10049A 924167.8 1117991 2006 19.8 5.50

PDI Phase 2 OL‐VC‐10050 925985.1 1117816.4 2006 19.4 13.35

PDI Phase 2 OL‐VC‐10051 926006.9 1117854.7 2006 19.8 13.28

PDI Phase 2 OL‐VC‐10052 925963.8 1117852.6 2006 19.8 13.14

PDI Phase 2 OL‐VC‐10053 924313 1117722 2006 19.2 8.70

PDI Phase 2 OL‐VC‐10053A 924313.3 1117723.8 2006 19.8 8.80

PDI Phase 2 OL‐VC‐10054 924273.2 1117823.8 2006 18.8 8.77

PDI Phase 2 OL‐VC‐10055 924347.9 1118435.4 2006 18.9 3.76

PDI Phase 2 OL‐VC‐10056 924315.8 1118096.5 2006 19.2 5.50

PDI Phase 2 OL‐VC‐10057 924432.4 1118236.3 2006 19.5 5.31

PDI Phase 2 OL‐VC‐10057A 924433.9 1118237.8 2006 19.7 5.31

PDI Phase 2 OL‐VC‐10058 924442.8 1117838 2006 19.8 5.51

PDI Phase 2 OL‐VC‐10058A 924444.6 1117839.5 2006 17.4 5.50

PDI Phase 2 OL‐VC‐10059 924590.4 1117726.6 2006 19.3 8.80

PDI Phase 2 OL‐VC‐10060 924676.3 1117861.4 2006 18.0 5.64

PDI Phase 2 OL‐VC‐10062 924727.1 1117993.7 2006 19.2 5.62

PDI Phase 2 OL‐VC‐10062A 924727.2 1117996.1 2006 19.0 5.62

PDI Phase 2 OL‐VC‐10064 924771.7 1117684.3 2006 19.8 8.80

PDI Phase 2 OL‐VC‐10064A 924774.2 1117684.7 2006 19.8 8.80

PDI Phase 2 OL‐VC‐10065 924792.2 1117697.1 2006 19.8 8.80

PDI Phase 2 OL‐VC‐10066 924773.4 1117659 2006 19.8 8.80

PDI Phase 2 OL‐VC‐10067 924750.2 1117697.5 2006 19.1 8.80

PDI Phase 2 OL‐VC‐10068 924860.7 1117921.7 2006 19.6 5.50

PDI Phase 2 OL‐VC‐10069 924950.9 1117812.2 2006 19.8 5.22

PDI Phase 2 OL‐VC‐10070 924919.9 1117455.2 2006 19.1 5.45

PDI Phase 2 OL‐VC‐10070A 924918.5 1117457 2006 19.2 5.59

PDI Phase 2 OL‐VC‐10071 925183.7 1118398.9 2006 19.8 2.73

PDI Phase 2 OL‐VC‐10072 925154.1 1118217.6 2006 19.8 9.88

PDI Phase 2 OL‐VC‐10073 925124.4 1118025.1 2006 19.8 9.98

PDI Phase 2 OL‐VC‐10073A 925112.8 1118024.8 2006 19.8 10.00

PDI Phase 2 OL‐VC‐10074 925137.1 1117740.6 2006 19.8 5.29

PDI Phase 2 OL‐VC‐10075 925124.1 1117274 2006 18.9 5.51

PDI Phase 2 OL‐VC‐10076 925240.3 1117526.4 2006 19.8 5.50

PDI Phase 2 OL‐VC‐10077 925345.5 1118080 2006 17.9 10.06

PDI Phase 2 OL‐VC‐10078 925415.9 1117846 2006 19.8 10.00

PDI Phase 2 OL‐VC‐10078A 925416.6 1117845.2 2006 19.8 10.00

PDI Phase 2 OL‐VC‐10079 925420.7 1117654.2 2006 18.9 9.92

PDI Phase 2 OL‐VC‐10080 925396.3 1116980.6 2006 19.3 7.90

PDI Phase 2 OL‐VC‐10080A 925394.2 1116982.2 2006 19.3 7.90

PDI Phase 2 OL‐VC‐10081 925496 1117441 2006 19.8 9.90

PDI Phase 2 OL‐VC‐10081A 925496.8 1117443.5 2006 19.6 9.98

PDI Phase 2 OL‐VC‐10082 925614.9 1118354.3 2006 19.8 0.84

PDI Phase 2 OL‐VC‐10083 925633.8 1118077.4 2006 19.8 9.89

PDI Phase 2 OL‐VC‐10083A 925631.9 1118076.4 2006 19.8 9.89

PDI Phase 2 OL‐VC‐10084 925660.7 1117489.5 2006 19.8 9.85

PDI Phase 2 OL‐VC‐10085 925641.9 1117135 2006 19.8 9.94

PDI Phase 2 OL‐VC‐10086 925660.6 1116945.8 2006 18.0 9.51

PDI Phase 2 OL‐VC‐10086A 925659.4 1116944.4 2006 18.8 9.51

PDI Phase 2 OL‐VC‐10087 925592.6 1116769.1 2006 19.0 5.02

PDI Phase 2 OL‐VC‐10088 925722.1 1117764.5 2006 19.8 9.87

PDI Phase 2 OL‐VC‐10089 925742.9 1117288.4 2006 19.7 9.87

PDI Phase 2 OL‐VC‐10090 925905.2 1118132.3 2006 19.8 9.30

PDI Phase 2 OL‐VC‐10091 925945.2 1117683.1 2006 19.6 13.20

PDI Phase 2 OL‐VC‐10092 925874.7 1117470.8 2006 19.0 9.90

PDI Phase 2 OL‐VC‐10093 925873.5 1116983.5 2006 19.8 9.85

PDI Phase 2 OL‐VC‐10094 925868.4 1116632.3 2006 19.8 6.07

PDI Phase 2 OL‐VC‐10095 925972.7 1118333.9 2006 19.6 1.02

PDI Phase 2 OL‐VC‐10095A 925974.7 1118335.5 2006 19.1 1.02

PDI Phase 2 OL‐VC‐10096 925984.7 1117842.4 2006 19.8 13.40

PDI Phase 2 OL‐VC‐10096A 925983.8 1117840.7 2006 17.4 12.62

PDI Phase 2 OL‐VC‐10097 926020.4 1117273.4 2006 18.4 13.20

PDI Phase 2 OL‐VC‐10097A 926019.5 1117275.1 2006 18.5 13.20

PDI Phase 2 OL‐VC‐10098 926041.7 1117287 2006 19.8 13.20

PDI Phase 2 OL‐VC‐10099 926020.2 1117250.5 2006 19.8 13.11

PDI Phase 2 OL‐VC‐10100 925996.8 1117287.3 2006 19.1 13.28

PDI Phase 2 OL‐VC‐10101 926082.6 1117979 2006 19.8 8.71

PDI Phase 2 OL‐VC‐10102 926067.6 1117485 2006 17.2 13.20

PDI Phase 2 OL‐VC‐10103 926131.8 1117084.3 2006 18.0 13.19

PDI Phase 2 OL‐VC‐10103A 926133.5 1117084.4 2006 18.0 13.19

PDI Phase 2 OL‐VC‐10104 926107.6 1116687.1 2006 19.8 9.97

PDI Phase 2 OL‐VC‐10104A 926107.1 1116685.6 2006 18.2 9.97

PDI Phase 2 OL‐VC‐10105 926187.2 1116859.4 2006 19.8 9.87

PDI Phase 2 OL‐VC‐10106 926169.9 1116498.4 2006 17.5 7.34

Location Group Name Location ID Easting Northing Year Sampled Total Core Length (ft) Dredge Depth (ft)

PDI Phase 2 OL‐VC‐10107 926342.7 1116313.7 2006 17.5 3.98

PDI Phase 2 OL‐VC‐20069 921376.75 1119611.81 2006 13.2 1.55

PDI Phase 2 OL‐VC‐20070 921495.5 1119473.16 2006 13.2 4.12

PDI Phase 2 OL‐VC‐20071 921703.69 1119428.66 2006 13.2 4.00

PDI Phase 2 OL‐VC‐20072 922398.5 1118947.4 2006 19.8 2.95

PDI Phase 2 OL‐VC‐20075 922723.1 1118763.6 2006 17.2 9.87

PDI Phase 2 OL‐VC‐20078 923175.9 1118583.7 2006 19.8 2.61

PDI Phase 2 OL‐VC‐20079 923077.2 1118450.4 2006 19.8 9.49

PDI Phase 2 OL‐VC‐20079A 923079.3 1118448.3 2006 19.8 9.95

PDI Phase 2 OL‐VC‐20080 921606.78 1119415.12 2006 13.2 4.08

PDI Phase 2 OL‐VC‐20081 921665.04 1119359.43 2006 13.2 3.04

PDI Phase 2 OL‐VC‐20082 921635.02 1119386.98 2006 13.2 4.10

PDI Phase 2 OL‐VC‐30043 918489.4 1124086.72 2006 19.4 3.29

PDI Phase 2 OL‐VC‐40035 914699 1125978.2 2006 19.8 8.05

PDI Phase 2 OL‐VC‐40036 914971.57 1126012.57 2006 17.3 4.18

PDI Phase 2 OL‐VC‐40037 915058.26 1126276.44 2006 19.8 2.44

PDI Phase 2 OL‐VC‐40038 915146.1 1125835.82 2006 19.8 3.41

PDI Phase 2 OL‐VC‐40039 915408.88 1126343.52 2006 19.8 4.75

PDI Phase 2 OL‐VC‐40040 915413.8 1125964.16 2006 19.8 2.28

PDI Phase 2 OL‐VC‐40041 915733.76 1125806.35 2006 19.3 2.50

PDI Phase 2 OL‐VC‐60059 927880.63 1121126.21 2006 9.9 0.40

PDI Phase 2 OL‐VC‐60068 928330.16 1119471.31 2006 18.7 3.98

PDI Phase 2 OL‐VC‐60069 928237.73 1118853.37 2006 19.6 4.66

PDI Phase 2 OL‐VC‐60070 927621.13 1117647.53 2006 20.0 3.48

PDI Phase 2 OL‐VC‐70016 926210 1118264.3 2006 19.2 3.30

PDI Phase 2 OL‐VC‐70017 926250.8 1117925.9 2006 19.8 6.75

PDI Phase 2 OL‐VC‐70018 926309.2 1117598.1 2006 19.6 9.72

PDI Phase 2 OL‐VC‐70021 926381.7 1117330.6 2006 19.8 9.66

PDI Phase 2 OL‐VC‐70021A 926379.5 1117330 2006 19.8 9.92

PDI Phase 2 OL‐VC‐70023 926348.6 1116853.6 2006 19.5 9.91

PDI Phase 2 OL‐VC‐70024 926641.2 1116857.8 2006 18.5 1.12

PDI Phase 2 OL‐VC‐70024A 926638.4 1116858 2006 19.0 1.12

PDI Phase 2 OL‐VC‐70025 926720.1 1116331.36 2006 13.2 4.24

PDI Phase 2 OL‐VC‐70026 926928.11 1116387.43 2006 13.2 3.75

PDI Phase 2 OL‐VC‐70027 926946.39 1116245.77 2006 13.2 3.39

PDI Phase 2 OL‐VC‐70028 927228.87 1116367.79 2006 13.2 3.38

PDI Phase 2 OL‐VC‐70029 927423.17 1116559.92 2006 13.2 3.30

PDI Phase 2 OL‐VC‐70030 927657.89 1116809.31 2006 13.2 3.27

PDI Phase 2 OL‐VC‐70032 926402.5 1117342.1 2006 19.8 9.70

PDI Phase 2 OL‐VC‐70033 926380.3 1117304.8 2006 19.8 9.84

PDI Phase 2 OL‐VC‐70034 926358.6 1117342.6 2006 19.2 9.85

PDI Phase 3 OL‐VC‐40132 915292.22 1125765.45 2007 8.0 2.52

PDI Phase 3 OL‐VC‐40133 915583.27 1125637.41 2007 8.6 2.54

PDI Phase 3 OL‐VC‐40134 915758.14 1126025.8 2007 8.6 1.50

PDI Phase 3 OL‐VC‐40135 915930.27 1125732.85 2007 8.0 2.54

PDI Phase 3 OL‐VC‐40136 916146.6 1126059.4 2007 7.1 2.75

PDI Phase 3 OL‐VC‐60114 928122.31 1121075.08 2007 9.1 0.32

PDI Phase 3 OL‐VC‐60115 928200.96 1120799.33 2007 9.9 1.47

PDI Phase 3 OL‐VC‐60116 928270.17 1120406.27 2007 9.0 0.49

PDI Phase 3 OL‐VC‐60117 928609 1120059.82 2007 7.8 3.96

PDI Phase 4 OL‐VC‐20135 921241.52 1120105.9 2008 9.6 0.40

PDI Phase 4 OL‐VC‐20139 921417.45 1119752.59 2008 8.9 1.38

PDI Phase 4 OL‐VC‐20140 921481.46 1119699.63 2008 7.5 3.51

PDI Phase 4 OL‐VC‐20141 921539.47 1119588.96 2008 10.0 3.82

PDI Phase 4 OL‐VC‐20144 922288.2 1119193.88 2008 4.0 2.19

PDI Phase 4 OL‐VC‐20147 922310.4 1118811.83 2008 9.0 3.90

PDI Phase 4 OL‐VC‐40205 914556.22 1126119.97 2008 9.2 2.37

PDI Phase 4 OL‐VC‐40206 914751.87 1126235.46 2008 7.4 0.43

PDI Phase 4 OL‐VC‐40207 915230.25 1126093.81 2008 7.0 3.00

PDI Phase 4 OL‐VC‐40209 915229.79 1125666.7 2008 7.8 2.54

PDI Phase 4 OL‐VC‐40210 915769.42 1125685.08 2008 8.8 2.53

PDI Phase 4 OL‐VC‐40211 915778.62 1125592.12 2008 7.2 2.42

PDI Phase 4 OL‐VC‐60195 928400.17 1120427.14 2008 3.8 2.23

PDI Phase 4 OL‐VC‐60196 928530.29 1120261.38 2008 3.9 2.86

PDI Phase 4 OL‐VC‐60200 928540.23 1118702.07 2008 6.0 5.13

PDI Phase 4 OL‐VC‐60201 928280.99 1118495.22 2008 7.0 4.30

PDI Phase 4 OL‐VC‐60202 928418.04 1118347.49 2008 7.1 5.25

PDI Phase 4 OL‐VC‐70112 926681.28 1116565.21 2008 9.4 2.06

PDI Phase 4 OL‐VC‐70113 926914.79 1116657.99 2008 7.9 1.91

PDI Phase 4 OL‐VC‐70114 927140.77 1116812.85 2008 5.0 3.04

PDI Phase 4 OL‐VC‐70115 927417.08 1116954.21 2008 8.1 4.08

PDI Phase 5 OL‐VC‐20161 921067.05 1120169.36 2009 12.0 2.52

PDI Phase 5 OL‐VC‐20162 921113.26 1120164.11 2009 12.0 4.48

PDI Phase 5 OL‐VC‐20163 921189.92 1120057.72 2009 11.5 1.18

PDI Phase 5 OL‐VC‐20166 921457.21 1119815.17 2009 6.0 0.30

PDI Phase 5 OL‐VC‐20167 921386.81 1119689.38 2009 6.0 1.53

Location Group Name Location ID Easting Northing Year Sampled Total Core Length (ft) Dredge Depth (ft)

PDI Phase 5 OL‐VC‐20168 921443.7 1119536 2009 6.0 3.90

PDI Phase 5 OL‐VC‐20169 921532.46 1119460.67 2009 6.0 3.69

PDI Phase 5 OL‐VC‐20170 921648.82 1119362.11 2009 6.0 3.94

PDI Phase 5 OL‐VC‐20176 921646.81 1119613.46 2009 6.0 3.40

PDI Phase 5 OL‐VC‐20179 922484.74 1118822.37 2009 6.0 0.35

PDI Phase 5 OL‐VC‐20180 922391.41 1118731.78 2009 6.0 4.09

PDI Phase 5 OL‐VC‐30109 918305.28 1124293.61 2009 8.0 2.02

PDI Phase 5 OL‐VC‐30110 918363.35 1124163.33 2009 8.2 4.08

PDI Phase 5 OL‐VC‐30111 918139.01 1124189.81 2009 7.1 4.07

PDI Phase 5 OL‐VC‐30112 918232.5 1124163.05 2009 7.0 4.72

PDI Phase 5 OL‐VC‐30113 918308.65 1124109.15 2009 8.0 4.81

PDI Phase 5 OL‐VC‐30122 919295.33 1123499.67 2009 8.0 2.62

PDI Phase 5 OL‐VC‐30123 919197.96 1123519.25 2009 8.0 4.39

PDI Phase 5 OL‐VC‐30124 919240.39 1123446.94 2009 8.0 4.90

PDI Phase 5 OL‐VC‐30125 919312.46 1123360.76 2009 8.0 4.30

PDI Phase 5 OL‐VC‐40215 914634.81 1126082.85 2009 10.0 2.51

PDI Phase 5 OL‐VC‐40216 914583.23 1125992.9 2009 8.8 1.78

PDI Phase 5 OL‐VC‐40217 914684.74 1125907.94 2009 9.2 3.41

PDI Phase 5 OL‐VC‐40218 914795.06 1125972.42 2009 10.0 8.08

PDI Phase 5 OL‐VC‐40219 914782.46 1126045.13 2009 10.0 5.77

PDI Phase 5 OL‐VC‐40220 914821.03 1125803.68 2009 10.0 6.84

PDI Phase 5 OL‐VC‐40221 914855.67 1125921.86 2009 10.0 6.60

PDI Phase 5 OL‐VC‐40222 914895.4 1125922.95 2009 10.0 6.51

PDI Phase 5 OL‐VC‐40223 914936.43 1125934.31 2009 10.0 0.37

PDI Phase 5 OL‐VC‐40224 915151.24 1125747.75 2009 4.0 2.56

PDI Phase 5 OL‐VC‐40225 915111.69 1125644.76 2009 3.8 2.54

PDI Phase 5 OL‐VC‐40227 915288.36 1125582.95 2009 4.0 2.54

PDI Phase 5 OL‐VC‐40228 915312.85 1125645.9 2009 4.0 2.54

PDI Phase 5 OL‐VC‐40229 915639.62 1125599.59 2009 4.0 2.54

PDI Phase 5 OL‐VC‐40230 915656.73 1125541.7 2009 4.0 2.54

PDI Phase 5 OL‐VC‐40231 915774.58 1125547.97 2009 4.0 2.53

PDI Phase 5 OL‐VC‐40232 915897.62 1125597.62 2009 4.0 2.60

PDI Phase 5 OL‐VC‐40233 915907.37 1125654.62 2009 4.0 2.52

PDI Phase 5 OL‐VC‐40234 915912.89 1125950.78 2009 4.0 1.75

PDI Phase 5 OL‐VC‐40236 916320.52 1126308.68 2009 4.0 2.11

PDI Phase 5 OL‐VC‐40251 915236.01 1125853.2 2009 8.0 4.15

PDI Phase 5 OL‐VC‐40252 915621.48 1125848.58 2009 7.1 2.50

PDI Phase 5 OL‐VC‐40253 915859.32 1125858.07 2009 8.0 2.64

PDI Phase 5 OL‐VC‐50072 927388.27 1121707.69 2009 4.0 0.56

PDI Phase 5 OL‐VC‐60229 928889.24 1120046.72 2009 6.0 3.35

PDI Phase 5 OL‐VC‐60230 928852.32 1119831.48 2009 4.8 3.27

PDI Phase 5 OL‐VC‐60231 928752.09 1119595.96 2009 7.5 5.12

PDI Phase 5 OL‐VC‐60231A 928752.09 1119596.00 2009 10.8 5.12

PDI Phase 5 OL‐VC‐60232 928785.48 1119304.95 2009 7.6 3.18

PDI Phase 5 OL‐VC‐60233 928775.12 1119047.99 2009 5.3 0.25

PDI Phase 5 OL‐VC‐60234 928533.2 1118197.81 2009 6.0 3.58

PDI Phase 5 OL‐VC‐60235 928270.26 1117711.46 2009 6.0 4.76

PDI Phase 5 OL‐VC‐60236 928105.35 1117362.06 2009 5.8 1.51

Location Group Name Location ID Easting Northing Year Sampled Total Core Length (ft) Dredge Depth (ft)

PDI Phase 5 OL‐VC‐60237 927833.03 1116937.47 2009 6.0 3.41

PDI Phase 5 OL‐VC‐60242 927832.36 1117484.03 2009 6.0 5.25

PDI Phase 5 OL‐VC‐60243 927699.41 1117611.67 2009 6.0 4.44

PDI Phase 5 OL‐VC‐60244 927503.98 1117753.74 2009 6.0 1.91

PDI Phase 5 OL‐VC‐60246 928068.98 1118668.02 2009 6.0 3.90

PDI Phase 5 OL‐VC‐60247 928540.08 1119301.6 2009 6.0 5.10

PDI Phase 5 OL‐VC‐60248 928318.39 1119294.65 2009 6.0 4.29

PDI Phase 5 OL‐VC‐60249 928017.8 1119270.65 2009 6.0 0.60

PDI Phase 5 OL‐VC‐60250 928530.42 1119805.43 2009 6.1 4.24

PDI Phase 5 OL‐VC‐60251 928290.67 1119816.91 2009 6.0 1.44

PDI Phase 5 OL‐VC‐60253 928499.87 1120424.41 2009 8.0 3.25

PDI Phase 5 OL‐VC‐70126 927626.56 1116655.21 2009 6.0 1.14

PDI Phase 5 OL‐VC‐70134 927024.11 1116551.99 2009 6.0 3.36

PDI Phase 5 OL‐VC‐70135 927449.17 1116829.05 2009 10.0 3.45

PDI Phase 5 OL‐VC‐70136 926894.19 1116835.03 2009 6.0 2.25

PDI Phase 5 OL‐VC‐70137 927211.47 1117156.44 2009 6.0 2.77

RI P Stations P1 926536.1547 1116418.769 1992 3.0 3.25

RI P Stations P15 925987.4391 1117839.707 1992 6.9 13.43

RI P Stations P16 928588.1579 1120322.542 1992 3.0 3.38

RI P Stations P22 924677.7631 1118376.243 1992 4.9 3.54

RI P Stations P23 925399.8959 1118314.382 1992 4.9 3.90

RI P Stations P3 926635.9285 1116857.442 1992 4.9 1.12

RI P Stations P36 921622.9492 1119518.004 1992 3.0 4.02

RI P Stations P53 919166.8172 1123677.959 1992 3.0 2.35

RI P Stations P6 927545.225 1117217.357 1992 3.0 5.16

RI P Stations P81 915209.683 1126366.728 1992 3.0 4.55

RI Sediment S304 915420.7452 1126195.545 2000 23.9 2.14

RI Sediment S307 921647.6066 1119488.162 2000 19.7 4.06

RI Sediment S309 923490.9926 1118177.918 2000 25.9 10.30

RI Sediment S311 925289.5712 1117120.209 2000 19.7 5.48

RI Sediment S312 926023.4914 1117275.523 2000 23.0 13.20

RI Sediment S313 926500.8476 1116312.347 2000 26.2 3.57

RI Sediment S314 926449.13 1116583.136 2000 26.2 3.06

RI Sediment S316 927404.8505 1117542.506 2000 26.2 2.20

RI Sediment S318 928074.2206 1118022.524 2000 26.2 3.87

RI Sediment S321 928456.9945 1118749.207 2000 26.2 5.25

RI Sediment S322 928435.6388 1120176.231 2000 26.2 2.17

RI Sediment S325 921147.9913 1120082.98 2000 6.6 3.38

RI Sediment S337 922798.8208 1118488.805 2000 6.6 9.87

RI Sediment S341 923910.8037 1117794.277 2000 6.6 9.14

RI Sediment S342 923950.711 1118375.321 2000 6.6 3.50

RI Sediment S343 924526.7252 1117697.59 2000 6.6 5.54

RI Sediment S344 925066.3757 1118146.603 2000 6.6 10.29

RI Sediment S345 924923.8845 1117453.342 2000 6.6 5.45

RI Sediment S346 924923.347 1117238.349 2000 6.6 5.82

RI Sediment S347 925223.6023 1117467.861 2000 6.6 5.47

RI Sediment S348 925641.2545 1116868.21 2000 6.6 7.27

RI Sediment S349 925683.0979 1116633.667 2000 6.6 4.96

RI Sediment S350 926012.0921 1116488.238 2000 6.6 4.55

RI Sediment S351 926467.9819 1116245.098 2000 6.6 3.98

RI Sediment S352 926824.6478 1116216.053 2000 6.6 3.39

RI Sediment S353 927268.0303 1116583.493 2000 6.6 4.25

RI Sediment S358 914359.7908 1126256.253 2000 1.0 0.41

RI Sediment S359 914825.9 1125844.7 2000 1.0 7.68

RI Sediment S360 915765.8506 1125666.085 2000 1.0 2.53

Note: Not all parameters were analyzed at each location.

Appendix E

Air Quality Bench Testing Summary

ONONDAGA LAKE AIR QUALITY BENCH TESTING

This report summarizes the pre-design investigation (PDI) activities conducted by Honeywell related to potential air quality impacts during the execution of the remedy described in the ROD issued for the Onondaga Lake project by NYSDEC and USEPA. Multiple studies have been performed in several phases of PDI work to evaluate what potential compounds in the sediment need to be considered for air quality analysis. All PDI activities were conducted under NYSDEC oversight, and in accordance with NYSDEC approved workplans. These workplans are available in document repositories. This memo focuses on the selection and identification process of these compounds, and reports on the detection frequency of each compound during the various phases.

In Phase I of the PDI, initial wind tunnel testing assessed potential emissions from an open basin, which was the initial sediment dewatering concept. Nineteen compounds (or groups of compounds) plus odor parameters were identified for analysis in the wind tunnel study, as presented in Table 1. This list of compounds was developed by selecting compounds known to exist in the lake sediment, and that also appear on USEPA’s Hazardous Air Pollutant (HAP) list. Based on the results of the testing, it was recommended, and agreed to by NYSDEC that eleven compounds be targeted for evaluation in any future testing. Table 2 summarizes the compounds targeted for future testing, and the detection frequency during Phase 1 testing.

The Phase II/III PDI Odorant Characterization Study identified which compounds had the greatest impact on odor, and developed correlations between chemical concentrations and various odor parameters. This testing evaluated headspace air samples and targeted several compound groups known to be odorous. The compounds tested during this phase of testing are presented in Table 3. Table 4 presents the detection frequency during the Phase II/III Odor Characterization Study for those compounds that were targeted for future testing (Table 2) in Phase I.

The Phase III PDI wind tunnel evaluation evaluated different odor and air emission mitigation techniques. Compounds targeted for this phase of investigation were based on results from Phase I (Table 2), and additional odorant compounds identified in Phase II/III. Table 5 presents the compounds which were evaluated, and their detection frequency during Phase III PDI wind tunnel testing. Statistics presented in this memo do not include samples collected during analyses in which mitigation techniques were utilized, as these tests may not be representative of the full emissions potential of the sediment.

The laboratories utilized in the various phases of investigation analyzed and reported additional compounds of interest. These additional compounds, and their detection frequency in the various phases of investigation, are summarized in Table 6.

PDI Activity Work Plan Summary Report Phase 1 Wind Tunnel

Appendix D – Onondaga Lake PDI: Air Emission and Odor Work Plan (Parsons, 2005)

Wind Tunnel Testing Report (Service Engineering, 2008)

Phase II/III Odor Characterization Study

Onondaga Lake PDI: Phase II Work Plan Addendum 5 (Parsons, 2006)

Onondaga Lake PDI: Phase II & III Odorant Characterization and Analysis Summary Report (Parsons et. al,. 2008)

Phase III Wind Tunnel Testing

Onondaga Lake PDI: Phase III Addendum 7 Work Plan, Air Emissions and Odors (O’Brien & Gere, 2008)

Onondaga Lake PDI: Phase III Addendum 7 Summary Report, Air Emissions and Odors (O’Brien & Gere, 2009)

Table 1

Phase 1 PDI Wind Tunnel Study: Targeted Compounds

VOCs PAHs Benzene Flourene Chlorobenzene Naphthalene Ethylbenzene Phenanthrene Dichlorobenzenes Pyrene 1,2,3 Trichlorobenzenes Other Compounds 1,2,4 Trichlorobenzenes PCBs 1,3,5 Trichlorobenzenes Phenol Hexachlorobenzene (HCB) Mercury Toluene Hydrogen Sulfide Xylenes (o-,m-,p-) Mercaptans Odor Parameters Detection Thresholds (OUs) Recognition Thresholds (OUs)

Table 3 Phase II/III PDI Odor Characterization Study: Targeted Compounds

VOCs Aldehydes Sulfur Gases Ethanol Propionaldehyde Dimethyl Sulfide 2-Butanone Buthyraldehyde Carbon Disulfide Benzene Benzaldehyde n-Propyl Mercaptan Chlorobenzene Isovaleraldehyde Ethyl Methyl Sulfide Ethylbenzene Valeraldehyde Thiophene Dichlorobenzenes n-Hexaldehyde Diemethyl Disulfide Trichlorobenzenes 2,5-Dimethybenzaldehyde 3-Methylthiophene Hexachlorobenzene (HCB) Amines 2-Ethylthiophene Toluene Dimethylamine Diethyl Disulfide Xylenes (o-,m-,p-) Trimethylamine Styrene Diethylamine Nonane Buthylamine Cumene Ethyltoluene Trimethylbenzene Dichlorobenzene Naphthalene

Table 2 Phase 1 PDI Wind Tunnel Study: Compounds

Targeted for Any Future Evalautions VOCs Benzene 15 detects in 15 tests Chlorobenzene 4 detects in 15 tests Ethylbenzene 15 detects in 15 tests Dichlorobenzenes 4 detects in 15 tests 1,2,3 Trichlorobenzenes 2 detects in 15 tests 1,2,4 Trichlorobenzenes 3 detects in 15 tests 1,3,5 Trichlorobenzenes 1 detects in 15 tests Hexachlorobenzene (HCB) 1 detects in 15 tests Toluene 12 detects in 15 tests Xylenes (o-,m-,p-) 15 detects in 15 tests PAHs Naphthalene 10 detects in 15 tests

Table 4

Phase II/III PDI Odor Characterization Study: Targeted Compounds Detected

Benzene 12 detects in 14 samples Chlorobenzene 13 detects in 14 samples 1,2 Dichlorobenzene 12 detects in 14 samples 1,4 Dichlorobenzene 12 detects in 14 samples Ethylbenzene 12 detects in 14 samples Hexachlorobenzene 0 detects in 14 samples Naphthalene 10 detects in 14 samples 1,2,3 Trichlorobenzenes 0 detects in 14 samples 1,2,4 Trichlorobenzenes 3 detects in 14 samples 1,3,5 Trichlorobenzenes 0 detects in 14 samples 1,2,4 Trimethylbenzene 13 detects in 14 samples Toluene 14 detects in 14 samples Xylenes 14 detects in 14 samples Note: shaded compounds identified as primary odorants

Note: shaded compounds were detected during Phase III mitigation technique tests, but not in untreated tests

Table 6 Phases 1 – III PDI: Additional Compounds Analyzed

Other COIs of Interest Phase 1 Wind Tunnel Phase II/III Odorant Study Phase III Wind Tunnel PAHs Fluorene 1 detect in 15 tests Not Analyzed Not Analyzed Phenanthrene 0 detects in 15 tests Not Analyzed Not Analyzed Pyrene 0 detects in 15 tests Not Analyzed Not Analyzed Other PCBs 0 detects in 2 tests Not Analyzed Not Analyzed Phenol 8 detects in 15 tests Not Analyzed Not Analyzed Mercury 0 detects in 15 tests Not Analyzed Not Analyzed Dimethyl Sulfide Not Analyzed 10 detects in 14 samples Not Analyzed Diemethyl Disulfide Not Analyzed 6 detects in 14 samples 0 detects in 14 samples Hydrogen Sulfide 0 detects in 6 tests 0 detects in 14 samples Not Analyzed Isobutyl Mercaptan 0 detects in 6 tests 2 detects in 14 samples Not Analyzed n-Propyl Mercaptan 0 detects in 6 tests 1 detects in 14 samples Not Analyzed Methyl Mercaptan 0 detects in 6 tests 1 detects in 14 samples Not Analyzed Ammonia 0 detects in 6 tests Not Analyzed Not Analyzed VOCs 1,2-Dichlorobenzene Not Analyzed 12 detects in 14 samples 1 detects in 2 samples 1,3,5-Trimethylbenzene Not Analyzed 12 detects in 14 samples 1 detects in 2 samples 1,1,1-Trichloroethane 10 detects in 15 tests 0 detects in 14 samples Not Analyzed 1,1-Dichloroethane Not Analyzed 0 detects in 14 samples Not Analyzed 1,1-Dichloroethene Not Analyzed 0 detects in 14 samples Not Analyzed 1,2-Dichloroethane Not Analyzed 0 detects in 14 samples Not Analyzed Acetone 12 detects in 15 tests 4 detects in 14 samples Not Analyzed 2-Butanone 15 detects in 15 tests 8 detects in 14 samples Not Analyzed

Table 5 Phase III PDI Wind Tunnel Testing:

Targeted Compounds from Untreated Tests Benzene 14 detects in 14 samples Chlorobenzene 13 detects in 14 samples 1,2 Dichlorobenzenes 11 detects in 14 samples 1,3 Dichlorobenzenes 0 detects in 14 samples 1,4 Dichlorobenzenes 12 detects in 14 samples Ethylbenzene 11 detects in 14 samples Hexachlorobenzene Not Analyzed Naphthalene 11 detects in 14 samples Toluene 14 detects in 14 samples 1,2,3 Trichlorobenzenes 0 detects in 14 samples 1,2,4 Trichlorobenzenes 3 detects in 14 samples 1,3,5 Trichlorobenzenes 0 detects in 14 samples 1,2,4 Trimethylbenzene 11 detects in 14 samples 1,3,5 Trimethylbenzene 10 detects in 14 samples Xylenes (o-,m-,p-) 6 detects in 14 samples

Table 6 Phases 1 – III PDI: Additional Compounds Analyzed (cont’d)

Carbon Tetrachloride Not Analyzed 0 detects in 14 samples Not Analyzed Chloroform 15 detects in 15 tests 0 detects in 14 samples Not Analyzed Cyanide Not Analyzed Not Analyzed Not Analyzed Methyl Tert-Butyl Ether Not Analyzed 0 detects in 14 samples Not Analyzed Methylene chloride 13 detects in 15 tests 1 detects in 14 samples Not Analyzed Tetrachloroethene 12 detects in 15 tests 0 detects in 14 samples Not Analyzed Trans-1,2-Dichloroethene Not Analyzed 0 detects in 14 samples Not Analyzed Trichloroethene 15 detects in 15 tests 0 detects in 14 samples Not Analyzed 1,2,4-Trimethylbenzene Not Analyzed 13 detects in 14 samples Not Analyzed Vinyl chloride Not Analyzed 0 detects in 14 samples Not Analyzed

Appendix F

ProUCL Outputs

Number Percent Maximum Distributional Assessments Exposure Point Concentration

Parameter of Results Nondetect Detecta Meanb Normal Gamma Lognormal  Conclusion Conc. Methodc

Aluminum 37 0 13,000 4,000 No Yes Yes Gamma 4,700 95% KM (BCA)

Antimony 37 0.0811 12 2 No No No No Distribution 4.7 97.5% KM (Chebyshev)

Arsenic 37 0.027 27 6.3 No Yes No Gamma 10 95% KM (Chebyshev)

Barium 37 0 2,800 600 No Yes Yes Gamma 1,200 95% KM (Chebyshev)

Cadmium 41 0.0976 21 3.6 No Yes Yes Gamma 7.2 95% KM (Chebyshev)

Chromium 37 0 3,700 170 No No Yes Lognormal 790 97.5% KM (Chebyshev)

Cobalt 37 0 180 13 No No No No Distribution 48 97.5% KM (Chebyshev)

Copper 41 0 560 84 No Yes Yes Gamma 150 95% KM (Chebyshev)

Manganese 37 0 1,800 350 No No No No Distribution 580 95% KM (Chebyshev)

Mercury 318 0.0283 160 9 No No No No Distribution 15 97.5% KM (Chebyshev)

Nickel 41 0 2,100 110 No No No No Distribution 350 95% KM (Chebyshev)

Vanadium 37 0 280 24 No No No No Distribution 78 97.5% KM (Chebyshev)

Dibenzofuran 33 0.2121 16 2 No Yes Yes Gamma 4.8 95% KM (Chebyshev)

Dieldrin 32 0.5313 0.024 0.0045 No Yes Yes Gamma 0.005 95% KM (t)

Dioxins (as TCDD equivalents) 8 0 230 75 Yes Yes No Normal 130 95% Student's‐t

PCBs 272 0.4007 23 0.59 No No No No Distribution 1.3 97.5% KM (Chebyshev)

1,2,4‐Trichlorobenzene 326 0.6411 380 6.8 No No No No Distribution 17 97.5% KM (Chebyshev)

1,2,4‐Trimethylbenzene 6 0 28 9 Yes Yes Yes Normal 19 95% KM (t)

1,2‐Dichlorobenzene 326 0.3865 1,100 15 No No No No Distribution 38 97.5% KM (Chebyshev)

1,4‐Dichlorobenzene 326 0.3129 3,000 25 No No No No Distribution 84 97.5% KM (Chebyshev)

Benzene 325 0.1415 46 2.8 No No No No Distribution 4.7 97.5% KM (Chebyshev)

Chlorobenzene 325 0.2769 1,500 19 No No No No Distribution 52 97.5% KM (Chebyshev)

Ethylbenzene 325 0.2862 380 4.3 No No No No Distribution 12 97.5% KM (Chebyshev)

Hexachlorobenzene 37 0.2162 19 1.2 No No Yes Lognormal 6.4 99% KM (Chebyshev)

Pentachlorobenzene 4 0 6.7 1.9 na na na na 6.7 Maximum detected

Unres. comb. of 1,2,3/4,5 Tetrachlorobenzene 4 0 21 6.9 na na na na 21 Maximum detected

Xylenes, total 325 0.1938 310 28 No No No No Distribution 45 97.5% KM (Chebyshev)

2‐Methylnaphthalene 33 0.1515 37 6.4 No Yes No Gamma 13 95% KM (Chebyshev)

Benz[a]anthracene 305 0.0492 31 2.3 No No No No Distribution 4 97.5% KM (Chebyshev)

Benzo[a]pyrene 305 0.0721 32 1.8 No No Yes Lognormal 3.2 97.5% KM (Chebyshev)

Benzo[b]fluoranthene 305 0.0721 28 1.9 No No Yes Lognormal 3.2 97.5% KM (Chebyshev)

Benzo[k]fluoranthene 305 0.1541 17 0.88 No No Yes Lognormal 1.5 97.5% KM (Chebyshev)

Chrysene 305 0.0426 30 2.3 No No No No Distribution 3.9 97.5% KM (Chebyshev)

Dibenz[a,h]anthracene 305 0.1574 5.6 0.4 No No Yes Lognormal 0.58 97.5% KM (Chebyshev)

Fluoranthene 305 0.0328 53 5 No No No No Distribution 8.2 97.5% KM (Chebyshev)

Fluorene 305 0.3213 340 3.2 No No Yes Lognormal 10 97.5% KM (Chebyshev)

Indeno[1,2,3‐cd]pyrene 305 0.0951 14 0.92 No No Yes Lognormal 1.5 97.5% KM (Chebyshev)

Naphthalene 322 0.1739 820 83 No No No No Distribution 130 97.5% KM (Chebyshev)

Phenanthrene 305 0.0328 100 7.1 No No No No Distribution 12 97.5% KM (Chebyshev)

Notes:  All concentrations except dioxins are reported in mg/kg and rounded to two significant figures.  Dioxins (as TCDD equivalents) is reported in ng/kg.a ‐  Maximum detected length‐weighted average concentration.b ‐  Arithmetic average concentration with non‐detected results included at the full detection limit.c ‐  ProUCL 4.0 was used for the unweighted assessment based on core length‐weighted average concentrations.  The UCL reported represents thec ‐  method recommended by ProUCL.  For most parameters, this was a Kaplan‐Meier non‐parametric method that accounts for multiple detection limits.

na ‐  not applicable

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

ALUMINUM (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 37

Number of Distinct Detected Data 37 Number of Non‐Detect Data 0

Number of Missing Values 291 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 673.3 Minimum Detected 6.512

Maximum Detected 12538 Maximum Detected 9.436

Mean of Detected 3951 Mean of Detected 8.086

SD of Detected 2404 SD of Detected 0.677

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.906 Shapiro Wilk Test Statistic 0.948

5% Shapiro Wilk Critical Value 0.936 5% Shapiro Wilk Critical Value 0.936

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 3951 Mean 8.086

SD 2404 SD 0.677

   95% DL/2 (t) UCL 4618   95%  H‐Stat (DL/2) UCL 5156

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 2.51 Data Follow Appr. Gamma Distribution at 5% Significance Level

Theta Star 1574

nu star 185.8

A‐D Test Statistic 0.495 Nonparametric Statistics

5% A‐D Critical Value 0.756 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.756 Mean 3951

5% K‐S Critical Value 0.146 SD 2371

Data follow Appr. Gamma Distribution at 5% Significance Level SE of Mean 395.2

  95% KM (t) UCL 4618

Assuming Gamma Distribution   95% KM (z) UCL 4601

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 4618

Minimum 673.3   95% KM (bootstrap t) UCL 4717

Maximum 12538   95% KM (BCA) UCL 4650

Mean 3951   95% KM (Percentile Bootstrap) UCL 4629

Median 4262 95% KM (Chebyshev) UCL 5674

SD 2404 97.5% KM (Chebyshev) UCL 6419

k star 2.51 99% KM (Chebyshev) UCL 7883

Theta star 1574

Nu star 185.8 Potential UCLs to Use

AppChi2 155.2   95% KM (BCA) UCL 4650

   95% Gamma Approximate UCL 4728

   95% Adjusted Gamma UCL 4765

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

ANTIMONY (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 34

Number of Distinct Detected Data 34 Number of Non‐Detect Data 3

Number of Missing Values 291 Percent Non‐Detects 8.11%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.252 Minimum Detected ‐1.378

Maximum Detected 11.65 Maximum Detected 2.455

Mean of Detected 2.132 Mean of Detected 0.152

SD of Detected 2.662 SD of Detected 1.074

Minimum Non‐Detect 0.3 Minimum Non‐Detect ‐1.204

Maximum Non‐Detect 0.471 Maximum Non‐Detect ‐0.754

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 10

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 27

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 27.03%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.704 Shapiro Wilk Test Statistic 0.916

5% Shapiro Wilk Critical Value 0.933 5% Shapiro Wilk Critical Value 0.933

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 1.974 Mean 5.65E‐04

SD 2.606 SD 1.153

   95% DL/2 (t) UCL 2.698   95%  H‐Stat (DL/2) UCL 2.614

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 1.422 Mean in Log Scale 0.0172

SD 3.18 SD in Log Scale 1.128

   95% MLE (t) UCL 2.305 Mean in Original Scale 1.978

   95% MLE (Tiku) UCL 2.327 SD in Original Scale 2.603

  95% Percentile Bootstrap UCL 2.716

  95% BCA Bootstrap UCL 2.848

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.894 Data do not follow a Discernable Distribution (0.05)

Theta Star 2.386

nu star 60.76

A‐D Test Statistic 1.97 Nonparametric Statistics

5% A‐D Critical Value 0.778 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.778 Mean 1.983

5% K‐S Critical Value 0.156 SD 2.564

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.428

  95% KM (t) UCL 2.705

Assuming Gamma Distribution   95% KM (z) UCL 2.687

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 2.704

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 2.971

Maximum 11.65   95% KM (BCA) UCL 2.742

Mean 1.959   95% KM (Percentile Bootstrap) UCL 2.715

Median 0.824 95% KM (Chebyshev) UCL 3.848

SD 2.617 97.5% KM (Chebyshev) UCL 4.655

k star 0.303 99% KM (Chebyshev) UCL 6.241

Theta star 6.471

Nu star 22.41 Potential UCLs to Use

AppChi2 12.64 97.5% KM (Chebyshev) UCL 4.655

   95% Gamma Approximate UCL 3.472

   95% Adjusted Gamma UCL 3.562

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

ARSENIC (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 36

Number of Distinct Detected Data 36 Number of Non‐Detect Data 1

Number of Missing Values 291 Percent Non‐Detects 2.70%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.865 Minimum Detected ‐0.145

Maximum Detected 26.99 Maximum Detected 3.296

Mean of Detected 6.486 Mean of Detected 1.545

SD of Detected 5.221 SD of Detected 0.89

Minimum Non‐Detect 0.35 Minimum Non‐Detect ‐1.05

Maximum Non‐Detect 0.35 Maximum Non‐Detect ‐1.05

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.8 Shapiro Wilk Test Statistic 0.897

5% Shapiro Wilk Critical Value 0.935 5% Shapiro Wilk Critical Value 0.935

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 6.316 Mean 1.456

SD 5.252 SD 1.031

   95% DL/2 (t) UCL 7.774   95%  H‐Stat (DL/2) UCL 9.688

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 6.249 Mean in Log Scale 1.486

SD 5.288 SD in Log Scale 0.948

   95% MLE (t) UCL 7.717 Mean in Original Scale 6.326

   95% MLE (Tiku) UCL 7.656 SD in Original Scale 5.241

  95% Percentile Bootstrap UCL 7.8

  95% BCA Bootstrap UCL 8.083

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 1.565 Data Follow Appr. Gamma Distribution at 5% Significance Level

Theta Star 4.145

nu star 112.7

A‐D Test Statistic 1.077 Nonparametric Statistics

5% A‐D Critical Value 0.764 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.764 Mean 6.335

5% K‐S Critical Value 0.149 SD 5.159

Data follow Appr. Gamma Distribution at 5% Significance Level SE of Mean 0.86

  95% KM (t) UCL 7.787

Assuming Gamma Distribution   95% KM (z) UCL 7.75

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 7.786

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 8.271

Maximum 26.99   95% KM (BCA) UCL 7.786

Mean 6.311   95% KM (Percentile Bootstrap) UCL 7.833

Median 6.028 95% KM (Chebyshev) UCL 10.08

SD 5.258 97.5% KM (Chebyshev) UCL 11.71

k star 0.639 99% KM (Chebyshev) UCL 14.89

Theta star 9.881

Nu star 47.27 Potential UCLs to Use

AppChi2 32.49   95% KM (Chebyshev) UCL 10.08

   95% Gamma Approximate UCL 9.182

   95% Adjusted Gamma UCL 9.335

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

BARIUM (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 37

Number of Distinct Detected Data 37 Number of Non‐Detect Data 0

Number of Missing Values 291 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 31.34 Minimum Detected 3.445

Maximum Detected 2817 Maximum Detected 7.943

Mean of Detected 599.1 Mean of Detected 5.665

SD of Detected 773.5 SD of Detected 1.252

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.7 Shapiro Wilk Test Statistic 0.961

5% Shapiro Wilk Critical Value 0.936 5% Shapiro Wilk Critical Value 0.936

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 599.1 Mean 5.665

SD 773.5 SD 1.252

   95% DL/2 (t) UCL 813.8   95%  H‐Stat (DL/2) UCL 1109

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.763 Data Follow Appr. Gamma Distribution at 5% Significance Level

Theta Star 784.9

nu star 56.49

A‐D Test Statistic 1.01 Nonparametric Statistics

5% A‐D Critical Value 0.785 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.785 Mean 599.1

5% K‐S Critical Value 0.15 SD 763

Data follow Appr. Gamma Distribution at 5% Significance Level SE of Mean 127.2

  95% KM (t) UCL 813.8

Assuming Gamma Distribution   95% KM (z) UCL 808.3

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 813.8

Minimum 31.34   95% KM (bootstrap t) UCL 883.6

Maximum 2817   95% KM (BCA) UCL 821.5

Mean 599.1   95% KM (Percentile Bootstrap) UCL 816.9

Median 285.2 95% KM (Chebyshev) UCL 1153

SD 773.5 97.5% KM (Chebyshev) UCL 1393

k star 0.763 99% KM (Chebyshev) UCL 1864

Theta star 784.9

Nu star 56.49 Potential UCLs to Use

AppChi2 40.21   95% KM (Chebyshev) UCL 1153

   95% Gamma Approximate UCL 841.6

   95% Adjusted Gamma UCL 854.2

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

CADMIUM (mg/kg)

General Statistics

Number of Valid Data 41 Number of Detected Data 37

Number of Distinct Detected Data 37 Number of Non‐Detect Data 4

Number of Missing Values 287 Percent Non‐Detects 9.76%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.0468 Minimum Detected ‐3.061

Maximum Detected 20.84 Maximum Detected 3.037

Mean of Detected 3.95 Mean of Detected 0.399

SD of Detected 5.519 SD of Detected 1.551

Minimum Non‐Detect 0.044 Minimum Non‐Detect ‐3.124

Maximum Non‐Detect 0.987 Maximum Non‐Detect ‐0.0131

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 19

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 22

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 46.34%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.696 Shapiro Wilk Test Statistic 0.972

5% Shapiro Wilk Critical Value 0.936 5% Shapiro Wilk Critical Value 0.936

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 3.59 Mean 0.184

SD 5.352 SD 1.656

   95% DL/2 (t) UCL 4.997   95%  H‐Stat (DL/2) UCL 9.377

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 0.857 Mean in Log Scale 0.169

SD 8.065 SD in Log Scale 1.661

   95% MLE (t) UCL 2.978 Mean in Original Scale 3.584

   95% MLE (Tiku) UCL 3.38 SD in Original Scale 5.356

  95% Percentile Bootstrap UCL 5.008

  95% BCA Bootstrap UCL 5.322

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.596 Data Follow Appr. Gamma Distribution at 5% Significance Level

Theta Star 6.623

nu star 44.13

A‐D Test Statistic 0.727 Nonparametric Statistics

5% A‐D Critical Value 0.8 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.8 Mean 3.588

5% K‐S Critical Value 0.152 SD 5.288

Data follow Appr. Gamma Distribution at 5% Significance Level SE of Mean 0.837

  95% KM (t) UCL 4.997

Assuming Gamma Distribution   95% KM (z) UCL 4.965

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 4.994

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 5.561

Maximum 20.84   95% KM (BCA) UCL 5.113

Mean 3.564   95% KM (Percentile Bootstrap) UCL 4.99

Median 1.071 95% KM (Chebyshev) UCL 7.237

SD 5.368 97.5% KM (Chebyshev) UCL 8.816

k star 0.242 99% KM (Chebyshev) UCL 11.92

Theta star 14.75

Nu star 19.82 Potential UCLs to Use

AppChi2 10.72   95% KM (Chebyshev) UCL 7.237

   95% Gamma Approximate UCL 6.591

   95% Adjusted Gamma UCL 6.745

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

CHROMIUM (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 37

Number of Distinct Detected Data 37 Number of Non‐Detect Data 0

Number of Missing Values 291 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 3.2 Minimum Detected 1.163

Maximum Detected 3711 Maximum Detected 8.219

Mean of Detected 166 Mean of Detected 3.65

SD of Detected 604.5 SD of Detected 1.455

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.256 Shapiro Wilk Test Statistic 0.959

5% Shapiro Wilk Critical Value 0.936 5% Shapiro Wilk Critical Value 0.936

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 166 Mean 3.65

SD 604.5 SD 1.455

   95% DL/2 (t) UCL 333.8   95%  H‐Stat (DL/2) UCL 226.9

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.425 Data appear Lognormal at 5% Significance Level

Theta Star 390.8

nu star 31.44

A‐D Test Statistic 3.087 Nonparametric Statistics

5% A‐D Critical Value 0.825 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.825 Mean 166

5% K‐S Critical Value 0.154 SD 596.3

Data not Gamma Distributed at 5% Significance Level SE of Mean 99.38

  95% KM (t) UCL 333.8

Assuming Gamma Distribution   95% KM (z) UCL 329.5

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 333.8

Minimum 3.2   95% KM (bootstrap t) UCL 1272

Maximum 3711   95% KM (BCA) UCL 370.2

Mean 166   95% KM (Percentile Bootstrap) UCL 360.8

Median 31.72 95% KM (Chebyshev) UCL 599.2

SD 604.5 97.5% KM (Chebyshev) UCL 786.6

k star 0.425 99% KM (Chebyshev) UCL 1155

Theta star 390.8

Nu star 31.44 Potential UCLs to Use

AppChi2 19.63 97.5% KM (Chebyshev) UCL 786.6

   95% Gamma Approximate UCL 265.9

   95% Adjusted Gamma UCL 271.6

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

COBALT (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 37

Number of Distinct Detected Data 37 Number of Non‐Detect Data 0

Number of Missing Values 291 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.29 Minimum Detected ‐1.238

Maximum Detected 179 Maximum Detected 5.187

Mean of Detected 12.72 Mean of Detected 1.322

SD of Detected 34.07 SD of Detected 1.227

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.372 Shapiro Wilk Test Statistic 0.819

5% Shapiro Wilk Critical Value 0.936 5% Shapiro Wilk Critical Value 0.936

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 12.72 Mean 1.322

SD 34.07 SD 1.227

   95% DL/2 (t) UCL 22.17   95%  H‐Stat (DL/2) UCL 13.73

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.494 Data do not follow a Discernable Distribution (0.05)

Theta Star 25.76

nu star 36.53

A‐D Test Statistic 6.056 Nonparametric Statistics

5% A‐D Critical Value 0.811 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.811 Mean 12.72

5% K‐S Critical Value 0.153 SD 33.6

Data not Gamma Distributed at 5% Significance Level SE of Mean 5.6

  95% KM (t) UCL 22.17

Assuming Gamma Distribution   95% KM (z) UCL 21.93

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 22.17

Minimum 0.29   95% KM (bootstrap t) UCL 49.98

Maximum 179   95% KM (BCA) UCL 23.19

Mean 12.72   95% KM (Percentile Bootstrap) UCL 22.66

Median 3.22 95% KM (Chebyshev) UCL 37.13

SD 34.07 97.5% KM (Chebyshev) UCL 47.69

k star 0.494 99% KM (Chebyshev) UCL 68.44

Theta star 25.76

Nu star 36.53 Potential UCLs to Use

AppChi2 23.69 97.5% KM (Chebyshev) UCL 47.69

   95% Gamma Approximate UCL 19.6

   95% Adjusted Gamma UCL 19.98

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

COPPER (mg/kg)

General Statistics

Number of Valid Data 41 Number of Detected Data 41

Number of Distinct Detected Data 41 Number of Non‐Detect Data 0

Number of Missing Values 287 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 1.8 Minimum Detected 0.588

Maximum Detected 561 Maximum Detected 6.33

Mean of Detected 83.68 Mean of Detected 3.874

SD of Detected 101.8 SD of Detected 1.165

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.661 Shapiro Wilk Test Statistic 0.954

5% Shapiro Wilk Critical Value 0.941 5% Shapiro Wilk Critical Value 0.941

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 83.68 Mean 3.874

SD 101.8 SD 1.165

   95% DL/2 (t) UCL 110.5   95%  H‐Stat (DL/2) UCL 151.3

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.979 Data appear Gamma Distributed at 5% Significance Level

Theta Star 85.48

nu star 80.27

A‐D Test Statistic 0.679 Nonparametric Statistics

5% A‐D Critical Value 0.777 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.777 Mean 83.68

5% K‐S Critical Value 0.142 SD 100.6

Data appear Gamma Distributed at 5% Significance Level SE of Mean 15.9

  95% KM (t) UCL 110.5

Assuming Gamma Distribution   95% KM (z) UCL 109.8

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 110.5

Minimum 1.8   95% KM (bootstrap t) UCL 130.1

Maximum 561   95% KM (BCA) UCL 112.5

Mean 83.68   95% KM (Percentile Bootstrap) UCL 111.4

Median 56.32 95% KM (Chebyshev) UCL 153

SD 101.8 97.5% KM (Chebyshev) UCL 183

k star 0.979 99% KM (Chebyshev) UCL 241.9

Theta star 85.48

Nu star 80.27 Potential UCLs to Use

AppChi2 60.63   95% KM (Chebyshev) UCL 153

   95% Gamma Approximate UCL 110.8

   95% Adjusted Gamma UCL 111.9

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

MANGANESE (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 37

Number of Distinct Detected Data 37 Number of Non‐Detect Data 0

Number of Missing Values 291 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 123.6 Minimum Detected 4.817

Maximum Detected 1790 Maximum Detected 7.49

Mean of Detected 349.3 Mean of Detected 5.636

SD of Detected 322.3 SD of Detected 0.591

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.604 Shapiro Wilk Test Statistic 0.878

5% Shapiro Wilk Critical Value 0.936 5% Shapiro Wilk Critical Value 0.936

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 349.3 Mean 5.636

SD 322.3 SD 0.591

   95% DL/2 (t) UCL 438.8   95%  H‐Stat (DL/2) UCL 405.9

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 2.25 Data do not follow a Discernable Distribution (0.05)

Theta Star 155.2

nu star 166.5

A‐D Test Statistic 2.556 Nonparametric Statistics

5% A‐D Critical Value 0.757 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.757 Mean 349.3

5% K‐S Critical Value 0.146 SD 317.9

Data not Gamma Distributed at 5% Significance Level SE of Mean 52.99

  95% KM (t) UCL 438.8

Assuming Gamma Distribution   95% KM (z) UCL 436.5

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 438.8

Minimum 123.6   95% KM (bootstrap t) UCL 512.8

Maximum 1790   95% KM (BCA) UCL 452.9

Mean 349.3   95% KM (Percentile Bootstrap) UCL 441.6

Median 235.1 95% KM (Chebyshev) UCL 580.3

SD 322.3 97.5% KM (Chebyshev) UCL 680.2

k star 2.25 99% KM (Chebyshev) UCL 876.5

Theta star 155.2

Nu star 166.5 Potential UCLs to Use

AppChi2 137.7   95% KM (Chebyshev) UCL 580.3

   95% Gamma Approximate UCL 422.5

   95% Adjusted Gamma UCL 426

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

MERCURY (mg/kg)

General Statistics

Number of Valid Data 318 Number of Detected Data 309

Number of Distinct Detected Data 305 Number of Non‐Detect Data 9

Number of Missing Values 10 Percent Non‐Detects 2.83%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.008 Minimum Detected ‐4.828

Maximum Detected 156 Maximum Detected 5.05

Mean of Detected 9.291 Mean of Detected 0.507

SD of Detected 16.38 SD of Detected 2.244

Minimum Non‐Detect 0.0067 Minimum Non‐Detect ‐5.006

Maximum Non‐Detect 0.361 Maximum Non‐Detect ‐1.018

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 81

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 237

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 25.47%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.294 Lilliefors Test Statistic 0.0662

5% Lilliefors Critical Value 0.0504 5% Lilliefors Critical Value 0.0504

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 9.029 Mean 0.371

SD 16.22 SD 2.362

   95% DL/2 (t) UCL 10.53   95%  H‐Stat (DL/2) UCL 32.87

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 5.513 Mean in Log Scale 0.376

SD 19.8 SD in Log Scale 2.35

   95% MLE (t) UCL 7.344 Mean in Original Scale 9.028

   95% MLE (Tiku) UCL 7.367 SD in Original Scale 16.22

  95% Percentile Bootstrap UCL 10.57

  95% BCA Bootstrap UCL 10.81

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.383 Data do not follow a Discernable Distribution (0.05)

Theta Star 24.26

nu star 236.7

A‐D Test Statistic 6.045 Nonparametric Statistics

5% A‐D Critical Value 0.85 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.85 Mean 9.028

5% K‐S Critical Value 0.0553 SD 16.19

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.909

  95% KM (t) UCL 10.53

Assuming Gamma Distribution   95% KM (z) UCL 10.52

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 10.53

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 10.79

Maximum 156   95% KM (BCA) UCL 10.62

Mean 9.028   95% KM (Percentile Bootstrap) UCL 10.58

Median 1.663 95% KM (Chebyshev) UCL 12.99

SD 16.22 97.5% KM (Chebyshev) UCL 14.71

k star 0.3 99% KM (Chebyshev) UCL 18.08

Theta star 30.11

Nu star 190.7 Potential UCLs to Use

AppChi2 159.7 97.5% KM (Chebyshev) UCL 14.71

   95% Gamma Approximate UCL 10.78

   95% Adjusted Gamma UCL 10.79

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

NICKEL (mg/kg)

General Statistics

Number of Valid Data 41 Number of Detected Data 41

Number of Distinct Detected Data 41 Number of Non‐Detect Data 0

Number of Missing Values 287 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 3 Minimum Detected 1.099

Maximum Detected 2090 Maximum Detected 7.645

Mean of Detected 111.1 Mean of Detected 3.279

SD of Detected 348.3 SD of Detected 1.362

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.325 Shapiro Wilk Test Statistic 0.904

5% Shapiro Wilk Critical Value 0.941 5% Shapiro Wilk Critical Value 0.941

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 111.1 Mean 3.279

SD 348.3 SD 1.362

   95% DL/2 (t) UCL 202.6   95%  H‐Stat (DL/2) UCL 122.2

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.434 Data do not follow a Discernable Distribution (0.05)

Theta Star 255.7

nu star 35.61

A‐D Test Statistic 5.215 Nonparametric Statistics

5% A‐D Critical Value 0.824 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.824 Mean 111.1

5% K‐S Critical Value 0.147 SD 344

Data not Gamma Distributed at 5% Significance Level SE of Mean 54.39

  95% KM (t) UCL 202.6

Assuming Gamma Distribution   95% KM (z) UCL 200.5

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 202.6

Minimum 3   95% KM (bootstrap t) UCL 389.5

Maximum 2090   95% KM (BCA) UCL 215.8

Mean 111.1   95% KM (Percentile Bootstrap) UCL 210.5

Median 27.32 95% KM (Chebyshev) UCL 348.1

SD 348.3 97.5% KM (Chebyshev) UCL 450.7

k star 0.434 99% KM (Chebyshev) UCL 652.2

Theta star 255.7

Nu star 35.61 Potential UCLs to Use

AppChi2 22.95   95% KM (Chebyshev) UCL 348.1

   95% Gamma Approximate UCL 172.3

   95% Adjusted Gamma UCL 175.1

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

VANADIUM (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 37

Number of Distinct Detected Data 37 Number of Non‐Detect Data 0

Number of Missing Values 291 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 1.753 Minimum Detected 0.562

Maximum Detected 279 Maximum Detected 5.631

Mean of Detected 23.65 Mean of Detected 2.248

SD of Detected 53.14 SD of Detected 1.11

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.416 Shapiro Wilk Test Statistic 0.878

5% Shapiro Wilk Critical Value 0.936 5% Shapiro Wilk Critical Value 0.936

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 23.65 Mean 2.248

SD 53.14 SD 1.11

   95% DL/2 (t) UCL 38.4   95%  H‐Stat (DL/2) UCL 27.93

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.629 Data do not follow a Discernable Distribution (0.05)

Theta Star 37.59

nu star 46.56

A‐D Test Statistic 4.282 Nonparametric Statistics

5% A‐D Critical Value 0.797 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.797 Mean 23.65

5% K‐S Critical Value 0.152 SD 52.41

Data not Gamma Distributed at 5% Significance Level SE of Mean 8.736

  95% KM (t) UCL 38.4

Assuming Gamma Distribution   95% KM (z) UCL 38.02

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 38.4

Minimum 1.753   95% KM (bootstrap t) UCL 57.4

Maximum 279   95% KM (BCA) UCL 38.9

Mean 23.65   95% KM (Percentile Bootstrap) UCL 39.84

Median 9.179 95% KM (Chebyshev) UCL 61.73

SD 53.14 97.5% KM (Chebyshev) UCL 78.21

k star 0.629 99% KM (Chebyshev) UCL 110.6

Theta star 37.59

Nu star 46.56 Potential UCLs to Use

AppChi2 31.91 97.5% KM (Chebyshev) UCL 78.21

   95% Gamma Approximate UCL 34.52

   95% Adjusted Gamma UCL 35.1

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

DIBENZOFURAN (mg/kg)

General Statistics

Number of Valid Data 33 Number of Detected Data 26

Number of Distinct Detected Data 26 Number of Non‐Detect Data 7

Number of Missing Values 295 Percent Non‐Detects 21.21%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.0476 Minimum Detected ‐3.044

Maximum Detected 16.35 Maximum Detected 2.794

Mean of Detected 2.417 Mean of Detected ‐0.0461

SD of Detected 4.12 SD of Detected 1.453

Minimum Non‐Detect 0.0537 Minimum Non‐Detect ‐2.925

Maximum Non‐Detect 1.087 Maximum Non‐Detect 0.0837

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 19

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 14

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 57.58%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.532 Shapiro Wilk Test Statistic 0.969

5% Shapiro Wilk Critical Value 0.92 5% Shapiro Wilk Critical Value 0.92

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 1.937 Mean ‐0.597

SD 3.763 SD 1.763

   95% DL/2 (t) UCL 3.046   95%  H‐Stat (DL/2) UCL 6.157

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐0.579

SD in Log Scale 1.674

Mean in Original Scale 1.924

SD in Original Scale 3.768

  95% Percentile Bootstrap UCL 3.088

  95% BCA Bootstrap UCL 3.463

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.606 Data appear Gamma Distributed at 5% Significance Level

Theta Star 3.986

nu star 31.54

A‐D Test Statistic 0.781 Nonparametric Statistics

5% A‐D Critical Value 0.793 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.793 Mean 1.927

5% K‐S Critical Value 0.179 SD 3.709

Data appear Gamma Distributed at 5% Significance Level SE of Mean 0.659

  95% KM (t) UCL 3.043

Assuming Gamma Distribution   95% KM (z) UCL 3.01

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 3.036

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 5.439

Maximum 16.35   95% KM (BCA) UCL 3.102

Mean 1.933   95% KM (Percentile Bootstrap) UCL 3.096

Median 0.624 95% KM (Chebyshev) UCL 4.798

SD 3.765 97.5% KM (Chebyshev) UCL 6.04

k star 0.194 99% KM (Chebyshev) UCL 8.48

Theta star 9.981

Nu star 12.78 Potential UCLs to Use

AppChi2 5.745   95% KM (Chebyshev) UCL 4.798

   95% Gamma Approximate UCL 4.299

   95% Adjusted Gamma UCL 4.489

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

DIELDRIN (mg/kg)

General Statistics

Number of Valid Data 32 Number of Detected Data 15

Number of Distinct Detected Data 15 Number of Non‐Detect Data 17

Number of Missing Values 296 Percent Non‐Detects 53.13%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00104 Minimum Detected ‐6.873

Maximum Detected 0.0243 Maximum Detected ‐3.717

Mean of Detected 0.00619 Mean of Detected ‐5.602

SD of Detected 0.00681 SD of Detected 1.044

Minimum Non‐Detect 0.001 Minimum Non‐Detect ‐6.908

Maximum Non‐Detect 0.00901 Maximum Non‐Detect ‐4.709

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 28

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 4

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 87.50%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.77 Shapiro Wilk Test Statistic 0.924

5% Shapiro Wilk Critical Value 0.881 5% Shapiro Wilk Critical Value 0.881

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 0.00372 Mean ‐6.288

SD 0.00526 SD 1.164

   95% DL/2 (t) UCL 0.0053   95%  H‐Stat (DL/2) UCL 0.00473

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐6.703

SD in Log Scale 1.31

Mean in Original Scale 0.00318

SD in Original Scale 0.00541

  95% Percentile Bootstrap UCL 0.00484

  95% BCA Bootstrap UCL 0.00526

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.928 Data appear Gamma Distributed at 5% Significance Level

Theta Star 0.00667

nu star 27.83

A‐D Test Statistic 0.648 Nonparametric Statistics

5% A‐D Critical Value 0.761 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.761 Mean 0.00357

5% K‐S Critical Value 0.227 SD 0.00516

Data appear Gamma Distributed at 5% Significance Level SE of Mean 9.50E‐04

  95% KM (t) UCL 0.00518

Assuming Gamma Distribution   95% KM (z) UCL 0.00514

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 0.00514

Minimum 0.00104   95% KM (bootstrap t) UCL 0.00622

Maximum 0.0243   95% KM (BCA) UCL 0.00536

Mean 0.00609   95% KM (Percentile Bootstrap) UCL 0.00523

Median 0.00462 95% KM (Chebyshev) UCL 0.00771

SD 0.00519 97.5% KM (Chebyshev) UCL 0.00951

k star 1.511 99% KM (Chebyshev) UCL 0.013

Theta star 0.00403

Nu star 96.7 Potential UCLs to Use

AppChi2 75.02   95% KM (t) UCL 0.00518

   95% Gamma Approximate UCL 0.00785

   95% Adjusted Gamma UCL 0.00796

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

DIOXINS (as TCDD equivalents)

General Statistics

Number of Valid Observations 8 Number of Distinct Observations 8

Raw Statistics Log‐transformed Statistics

Minimum 0.6 Minimum of Log Data ‐0.51083

Maximum 229.25 Maximum of Log Data 5.434813

Mean 74.90375 Mean of log Data 2.891502

Median 64.715 SD of log Data 2.474948

SD 80.30804

Coefficient of Variation 1.07215

Skewness 0.97689

Warning:  There are only 8 Values in this data

Note:  It should be noted that even though bootstrap methods may be performed on this data set,

the resulting calculations may not be reliable enough to draw conclusions

The literature suggests to use bootstrap methods on data sets having more than 10‐15 observations.

Relevant UCL Statistics

Normal Distribution Test Lognormal Distribution Test

Shapiro Wilk Test Statistic 0.883714 Shapiro Wilk Test Statistic 0.809705

Shapiro Wilk Critical Value 0.818 Shapiro Wilk Critical Value 0.818

Data appear Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

   95% Student's‐t UCL 128.6969   95% H‐UCL 425641.8

   95% UCLs (Adjusted for Skewness)   95% Chebyshev (MVUE) UCL 701.9146

   95% Adjusted‐CLT UCL 132.0848 97.5% Chebyshev (MVUE) UCL 936.9635

   95% Modified‐t UCL 130.3313   99% Chebyshev (MVUE) UCL 1398.672

Gamma Distribution Test Data Distribution

k star (bias corrected) 0.366343 Data appear Normal at 5% Significance Level

Theta Star 204.4636

MLE of Mean 74.90375

MLE of Standard Deviation 123.7542

nu star 5.861484

Approximate Chi Square Value (.05) 1.569493 Nonparametric Statistics

Adjusted Level of Significance 0.01946   95% CLT UCL 121.6064

Adjusted Chi Square Value 1.075581   95% Jackknife UCL 128.6969

  95% Standard Bootstrap UCL 118.4452

Anderson‐Darling Test Statistic 0.617797   95% Bootstrap‐t UCL 149.8426

Anderson‐Darling 5% Critical Value 0.770601   95% Hall's Bootstrap UCL 141.6336

Kolmogorov‐Smirnov Test Statistic 0.251247   95% Percentile Bootstrap UCL 119.74

Kolmogorov‐Smirnov 5% Critical Value 0.310868   95% BCA Bootstrap UCL 129.14

Data appear Gamma Distributed at 5% Significance Level 95% Chebyshev(Mean, Sd) UCL 198.6667

97.5% Chebyshev(Mean, Sd) UCL 252.2191

Assuming Gamma Distribution 99% Chebyshev(Mean, Sd) UCL 357.4123

   95% Approximate Gamma UCL 279.7381

   95% Adjusted Gamma UCL 408.1951

Potential UCL to Use Use 95% Student's‐t UCL 128.6969

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

PCBs (mg/kg)

General Statistics

Number of Valid Data 272 Number of Detected Data 163

Number of Distinct Detected Data 163 Number of Non‐Detect Data 109

Number of Missing Values 19 Percent Non‐Detects 40.07%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00448 Minimum Detected ‐5.408

Maximum Detected 23 Maximum Detected 3.135

Mean of Detected 0.96 Mean of Detected ‐1.856

SD of Detected 2.5 SD of Detected 1.917

Minimum Non‐Detect 0.00485 Minimum Non‐Detect ‐5.329

Maximum Non‐Detect 0.278 Maximum Non‐Detect ‐1.282

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 217

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 55

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 79.78%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.351 Lilliefors Test Statistic 0.114

5% Lilliefors Critical Value 0.0694 5% Lilliefors Critical Value 0.0694

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 0.583 Mean ‐2.832

SD 1.987 SD 1.984

   95% DL/2 (t) UCL 0.782   95%  H‐Stat (DL/2) UCL 0.51

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐3.234

SD in Log Scale 2.291

Mean in Original Scale 0.578

SD in Original Scale 1.989

  95% Percentile Bootstrap UCL 0.79

  95% BCA Bootstrap UCL 0.861

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.365 Data do not follow a Discernable Distribution (0.05)

Theta Star 2.629

nu star 119

A‐D Test Statistic 9.933 Nonparametric Statistics

5% A‐D Critical Value 0.852 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.852 Mean 0.58

5% K‐S Critical Value 0.0788 SD 1.984

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.121

  95% KM (t) UCL 0.779

Assuming Gamma Distribution   95% KM (z) UCL 0.778

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 0.779

Minimum 0.00448   95% KM (bootstrap t) UCL 0.882

Maximum 23   95% KM (BCA) UCL 0.794

Mean 0.781   95% KM (Percentile Bootstrap) UCL 0.788

Median 0.342 95% KM (Chebyshev) UCL 1.106

SD 1.949 97.5% KM (Chebyshev) UCL 1.334

k star 0.539 99% KM (Chebyshev) UCL 1.781

Theta star 1.449

Nu star 293.1 Potential UCLs to Use

AppChi2 254.4 97.5% KM (Chebyshev) UCL 1.334

   95% Gamma Approximate UCL 0.899

   95% Adjusted Gamma UCL 0.9

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

1,2,4‐TRICHLOROBENZENE (mg/kg)

General Statistics

Number of Valid Data 326 Number of Detected Data 117

Number of Distinct Detected Data 117 Number of Non‐Detect Data 209

Number of Missing Values 2 Percent Non‐Detects 64.11%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00385 Minimum Detected ‐5.559

Maximum Detected 382.8 Maximum Detected 5.947

Mean of Detected 15.92 Mean of Detected 0.931

SD of Detected 51.51 SD of Detected 2.482

Minimum Non‐Detect 0.00651 Minimum Non‐Detect ‐5.035

Maximum Non‐Detect 32.03 Maximum Non‐Detect 3.467

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 319

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 7

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 97.85%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.379 Lilliefors Test Statistic 0.205

5% Lilliefors Critical Value 0.0819 5% Lilliefors Critical Value 0.0819

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 6.257 Mean ‐1.753

SD 31.67 SD 3.277

   95% DL/2 (t) UCL 9.15   95%  H‐Stat (DL/2) UCL 36.59

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐2.087

SD in Log Scale 2.736

Mean in Original Scale 5.732

SD in Original Scale 31.71

  95% Percentile Bootstrap UCL 8.889

  95% BCA Bootstrap UCL 10.07

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.36 Data do not follow a Discernable Distribution (0.05)

Theta Star 44.22

nu star 84.26

A‐D Test Statistic 4.008 Nonparametric Statistics

5% A‐D Critical Value 0.852 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.852 Mean 5.811

5% K‐S Critical Value 0.0916 SD 31.66

Data not Gamma Distributed at 5% Significance Level SE of Mean 1.761

  95% KM (t) UCL 8.717

Assuming Gamma Distribution   95% KM (z) UCL 8.708

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 8.709

Minimum 0.00385   95% KM (bootstrap t) UCL 12.81

Maximum 382.8   95% KM (BCA) UCL 9.322

Mean 16.38   95% KM (Percentile Bootstrap) UCL 8.979

Median 12.48 95% KM (Chebyshev) UCL 13.49

SD 31.36 97.5% KM (Chebyshev) UCL 16.81

k star 0.822 99% KM (Chebyshev) UCL 23.34

Theta star 19.92

Nu star 536.2 Potential UCLs to Use

AppChi2 483.5 97.5% KM (Chebyshev) UCL 16.81

   95% Gamma Approximate UCL 18.16

   95% Adjusted Gamma UCL 18.17

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

1,2,4‐TRIMETHYLBENZENE (mg/kg)

General Statistics

Number of Valid Data 6 Number of Detected Data 6

Number of Distinct Detected Data 6 Number of Non‐Detect Data 0

Number of Missing Values 311 Percent Non‐Detects 0.00%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.017 Minimum Detected ‐4.075

Maximum Detected 28 Maximum Detected 3.332

Mean of Detected 9.036 Mean of Detected 0.124

SD of Detected 11.97 SD of Detected 3.238

Minimum Non‐Detect     N/A     Minimum Non‐Detect     N/A    

Maximum Non‐Detect     N/A     Maximum Non‐Detect     N/A    

Warning:  There are only 6 Detected Values in this data

Note:  It should be noted that even though bootstrap may be performed on this data set

the resulting calculations may not be reliable enough to draw conclusions

It is recommended to have 10‐15 or more distinct observations for accurate and meaningful results.

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.789 Shapiro Wilk Test Statistic 0.852

5% Shapiro Wilk Critical Value 0.788 5% Shapiro Wilk Critical Value 0.788

Data appear Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 9.036 Mean 0.124

SD 11.97 SD 3.238

   95% DL/2 (t) UCL 18.88   95%  H‐Stat (DL/2) UCL 1.14E+10

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale     N/A    

SD in Log Scale     N/A    

Mean in Original Scale     N/A    

SD in Original Scale     N/A    

  95% Percentile Bootstrap UCL     N/A    

  95% BCA Bootstrap UCL     N/A    

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.275 Data appear Normal at 5% Significance Level

Theta Star 32.89

nu star 3.297

A‐D Test Statistic 0.356 Nonparametric Statistics

5% A‐D Critical Value 0.762 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.762 Mean 9.036

5% K‐S Critical Value 0.356 SD 10.93

Data appear Gamma Distributed at 5% Significance Level SE of Mean 4.887

  95% KM (t) UCL 18.88

Assuming Gamma Distribution   95% KM (z) UCL 17.07

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 18.88

Minimum 0.017   95% KM (bootstrap t) UCL 56.12

Maximum 28   95% KM (BCA) UCL 16.69

Mean 9.036   95% KM (Percentile Bootstrap) UCL 17.04

Median 3.062 95% KM (Chebyshev) UCL 30.34

SD 11.97 97.5% KM (Chebyshev) UCL 39.56

k star 0.275 99% KM (Chebyshev) UCL 57.66

Theta star 32.89

Nu star 3.297 Potential UCLs to Use

AppChi2 0.466   95% KM (t) UCL 18.88

   95% Gamma Approximate UCL 63.98   95% KM (Percentile Bootstrap) UCL 17.04

   95% Adjusted Gamma UCL 140.4

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

1,2‐DICHLOROBENZENE (mg/kg)

General Statistics

Number of Valid Data 326 Number of Detected Data 200

Number of Distinct Detected Data 200 Number of Non‐Detect Data 126

Number of Missing Values 2 Percent Non‐Detects 38.65%

Raw Statistics Log‐transformed Statistics

Minimum Detected 4.20E‐04 Minimum Detected ‐7.775

Maximum Detected 1078 Maximum Detected 6.982

Mean of Detected 23.88 Mean of Detected 0.489

SD of Detected 83.65 SD of Detected 3.325

Minimum Non‐Detect 0.00658 Minimum Non‐Detect ‐5.023

Maximum Non‐Detect 25.32 Maximum Non‐Detect 3.232

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 282

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 44

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 86.50%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.388 Lilliefors Test Statistic 0.165

5% Lilliefors Critical Value 0.0626 5% Lilliefors Critical Value 0.0626

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 14.74 Mean ‐1.343

SD 66.47 SD 3.674

   95% DL/2 (t) UCL 20.81   95%  H‐Stat (DL/2) UCL 226.9

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.37

SD in Log Scale 3.517

Mean in Original Scale 14.66

SD in Original Scale 66.49

  95% Percentile Bootstrap UCL 21.47

  95% BCA Bootstrap UCL 25.41

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.262 Data do not follow a Discernable Distribution (0.05)

Theta Star 91.17

nu star 104.8

A‐D Test Statistic 1.916 Nonparametric Statistics

5% A‐D Critical Value 0.886 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.886 Mean 14.67

5% K‐S Critical Value 0.0696 SD 66.38

Data not Gamma Distributed at 5% Significance Level SE of Mean 3.686

  95% KM (t) UCL 20.75

Assuming Gamma Distribution   95% KM (z) UCL 20.73

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 20.74

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 28.29

Maximum 1078   95% KM (BCA) UCL 21.79

Mean 15.66   95% KM (Percentile Bootstrap) UCL 21.5

Median 2.394 95% KM (Chebyshev) UCL 30.73

SD 66.32 97.5% KM (Chebyshev) UCL 37.69

k star 0.255 99% KM (Chebyshev) UCL 51.34

Theta star 61.39

Nu star 166.3 Potential UCLs to Use

AppChi2 137.5 97.5% KM (Chebyshev) UCL 37.69

   95% Gamma Approximate UCL 18.94

   95% Adjusted Gamma UCL 18.95

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

1,4‐DICHLOROBENZENE (mg/kg)

General Statistics

Number of Valid Data 326 Number of Detected Data 224

Number of Distinct Detected Data 224 Number of Non‐Detect Data 102

Number of Missing Values 2 Percent Non‐Detects 31.29%

Raw Statistics Log‐transformed Statistics

Minimum Detected 8.78E‐04 Minimum Detected ‐7.038

Maximum Detected 3026 Maximum Detected 8.015

Mean of Detected 35.92 Mean of Detected 0.45

SD of Detected 205.8 SD of Detected 3.456

Minimum Non‐Detect 0.00658 Minimum Non‐Detect ‐5.023

Maximum Non‐Detect 25.32 Maximum Non‐Detect 3.232

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 272

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 54

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 83.44%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.431 Lilliefors Test Statistic 0.156

5% Lilliefors Critical Value 0.0592 5% Lilliefors Critical Value 0.0592

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 24.77 Mean ‐1.03

SD 171.3 SD 3.777

   95% DL/2 (t) UCL 40.42   95%  H‐Stat (DL/2) UCL 520.8

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.105

SD in Log Scale 3.698

Mean in Original Scale 24.69

SD in Original Scale 171.3

  95% Percentile Bootstrap UCL 43.12

  95% BCA Bootstrap UCL 58.06

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.23 Data do not follow a Discernable Distribution (0.05)

Theta Star 156.3

nu star 103

A‐D Test Statistic 3.978 Nonparametric Statistics

5% A‐D Critical Value 0.901 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.901 Mean 24.7

5% K‐S Critical Value 0.067 SD 171

Data not Gamma Distributed at 5% Significance Level SE of Mean 9.495

  95% KM (t) UCL 40.36

Assuming Gamma Distribution   95% KM (z) UCL 40.32

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 40.35

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 84.61

Maximum 3026   95% KM (BCA) UCL 45.13

Mean 24.73   95% KM (Percentile Bootstrap) UCL 43.26

Median 0.109 95% KM (Chebyshev) UCL 66.09

SD 171.3 97.5% KM (Chebyshev) UCL 83.99

k star 0.0898 99% KM (Chebyshev) UCL 119.2

Theta star 275.5

Nu star 58.53 Potential UCLs to Use

AppChi2 41.94 97.5% KM (Chebyshev) UCL 83.99

   95% Gamma Approximate UCL 34.52

   95% Adjusted Gamma UCL 34.57

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

BENZENE (mg/kg)

General Statistics

Number of Valid Data 325 Number of Detected Data 279

Number of Distinct Detected Data 275 Number of Non‐Detect Data 46

Number of Missing Values 2 Percent Non‐Detects 14.15%

Raw Statistics Log‐transformed Statistics

Minimum Detected 4.94E‐04 Minimum Detected ‐7.613

Maximum Detected 45.86 Maximum Detected 3.826

Mean of Detected 3.248 Mean of Detected ‐1.101

SD of Detected 5.867 SD of Detected 3.069

Minimum Non‐Detect 0.00139 Minimum Non‐Detect ‐6.578

Maximum Non‐Detect 9.48 Maximum Non‐Detect 2.249

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 300

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 25

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 92.31%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.29 Lilliefors Test Statistic 0.162

5% Lilliefors Critical Value 0.053 5% Lilliefors Critical Value 0.053

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 2.811 Mean ‐1.679

SD 5.548 SD 3.245

   95% DL/2 (t) UCL 3.319   95%  H‐Stat (DL/2) UCL 53.58

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.686

SD in Log Scale 3.214

Mean in Original Scale 2.79

SD in Original Scale 5.551

  95% Percentile Bootstrap UCL 3.311

  95% BCA Bootstrap UCL 3.408

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.301 Data do not follow a Discernable Distribution (0.05)

Theta Star 10.78

nu star 168.2

A‐D Test Statistic 4.065 Nonparametric Statistics

5% A‐D Critical Value 0.87 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.87 Mean 2.796

5% K‐S Critical Value 0.0592 SD 5.542

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.308

  95% KM (t) UCL 3.304

Assuming Gamma Distribution   95% KM (z) UCL 3.303

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 3.304

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 3.401

Maximum 45.86   95% KM (BCA) UCL 3.337

Mean 2.795   95% KM (Percentile Bootstrap) UCL 3.308

Median 0.44 95% KM (Chebyshev) UCL 4.139

SD 5.549 97.5% KM (Chebyshev) UCL 4.72

k star 0.16 99% KM (Chebyshev) UCL 5.861

Theta star 17.47

Nu star 104 Potential UCLs to Use

AppChi2 81.47 97.5% KM (Chebyshev) UCL 4.72

   95% Gamma Approximate UCL 3.568

   95% Adjusted Gamma UCL 3.571

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

CHLOROBENZENE (mg/kg)

General Statistics

Number of Valid Data 325 Number of Detected Data 235

Number of Distinct Detected Data 235 Number of Non‐Detect Data 90

Number of Missing Values 2 Percent Non‐Detects 27.69%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.0012 Minimum Detected ‐6.725

Maximum Detected 1504 Maximum Detected 7.316

Mean of Detected 26.33 Mean of Detected 0.308

SD of Detected 109.5 SD of Detected 3.28

Minimum Non‐Detect 0.00658 Minimum Non‐Detect ‐5.023

Maximum Non‐Detect 25.32 Maximum Non‐Detect 3.232

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 279

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 46

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 85.85%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.405 Lilliefors Test Statistic 0.14

5% Lilliefors Critical Value 0.0578 5% Lilliefors Critical Value 0.0578

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 19.1 Mean ‐1.044

SD 93.77 SD 3.656

   95% DL/2 (t) UCL 27.68   95%  H‐Stat (DL/2) UCL 323.6

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.075

SD in Log Scale 3.594

Mean in Original Scale 19.04

SD in Original Scale 93.78

  95% Percentile Bootstrap UCL 28.77

  95% BCA Bootstrap UCL 34.54

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.241 Data do not follow a Discernable Distribution (0.05)

Theta Star 109.3

nu star 113.2

A‐D Test Statistic 4.032 Nonparametric Statistics

5% A‐D Critical Value 0.896 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.896 Mean 19.05

5% K‐S Critical Value 0.0655 SD 93.64

Data not Gamma Distributed at 5% Significance Level SE of Mean 5.205

  95% KM (t) UCL 27.64

Assuming Gamma Distribution   95% KM (z) UCL 27.61

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 27.63

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 39.2

Maximum 1504   95% KM (BCA) UCL 29.32

Mean 19.07   95% KM (Percentile Bootstrap) UCL 28.68

Median 0.199 95% KM (Chebyshev) UCL 41.74

SD 93.78 97.5% KM (Chebyshev) UCL 51.56

k star 0.0983 99% KM (Chebyshev) UCL 70.84

Theta star 194

Nu star 63.87 Potential UCLs to Use

AppChi2 46.49 97.5% KM (Chebyshev) UCL 51.56

   95% Gamma Approximate UCL 26.2

   95% Adjusted Gamma UCL 26.24

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

ETHYLBENZENE (mg/kg)

General Statistics

Number of Valid Data 325 Number of Detected Data 232

Number of Distinct Detected Data 230 Number of Non‐Detect Data 93

Number of Missing Values 2 Percent Non‐Detects 28.62%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00103 Minimum Detected ‐6.881

Maximum Detected 376.1 Maximum Detected 5.93

Mean of Detected 5.487 Mean of Detected ‐0.9

SD of Detected 27.81 SD of Detected 2.959

Minimum Non‐Detect 0.00139 Minimum Non‐Detect ‐6.578

Maximum Non‐Detect 26.11 Maximum Non‐Detect 3.262

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 322

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 3

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 99.08%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.422 Lilliefors Test Statistic 0.17

5% Lilliefors Critical Value 0.0582 5% Lilliefors Critical Value 0.0582

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 4.117 Mean ‐1.984

SD 23.61 SD 3.325

   95% DL/2 (t) UCL 6.278   95%  H‐Stat (DL/2) UCL 42.51

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐2.201

SD in Log Scale 3.316

Mean in Original Scale 3.92

SD in Original Scale 23.61

  95% Percentile Bootstrap UCL 6.32

  95% BCA Bootstrap UCL 7.705

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.269 Data do not follow a Discernable Distribution (0.05)

Theta Star 20.4

nu star 124.8

A‐D Test Statistic 6.099 Nonparametric Statistics

5% A‐D Critical Value 0.883 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.883 Mean 3.959

5% K‐S Critical Value 0.0655 SD 23.57

Data not Gamma Distributed at 5% Significance Level SE of Mean 1.311

  95% KM (t) UCL 6.12

Assuming Gamma Distribution   95% KM (z) UCL 6.114

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 6.119

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 13.33

Maximum 376.1   95% KM (BCA) UCL 6.595

Mean 4.051   95% KM (Percentile Bootstrap) UCL 6.365

Median 0.165 95% KM (Chebyshev) UCL 9.671

SD 23.6 97.5% KM (Chebyshev) UCL 12.14

k star 0.109 99% KM (Chebyshev) UCL 17

Theta star 37.04

Nu star 71.09 Potential UCLs to Use

AppChi2 52.68 97.5% KM (Chebyshev) UCL 12.14

   95% Gamma Approximate UCL 5.466

   95% Adjusted Gamma UCL 5.474

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

HEXACHLOROBENZENE (mg/kg)

General Statistics

Number of Valid Data 37 Number of Detected Data 29

Number of Distinct Detected Data 29 Number of Non‐Detect Data 8

Number of Missing Values 291 Percent Non‐Detects 21.62%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.0106 Minimum Detected ‐4.551

Maximum Detected 18.51 Maximum Detected 2.918

Mean of Detected 1.342 Mean of Detected ‐1.257

SD of Detected 3.629 SD of Detected 1.699

Minimum Non‐Detect 0.00106 Minimum Non‐Detect ‐6.849

Maximum Non‐Detect 2.5 Maximum Non‐Detect 0.916

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 34

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 3

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 91.89%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.386 Shapiro Wilk Test Statistic 0.972

5% Shapiro Wilk Critical Value 0.926 5% Shapiro Wilk Critical Value 0.926

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 1.12 Mean ‐1.691

SD 3.236 SD 2.115

   95% DL/2 (t) UCL 2.018   95%  H‐Stat (DL/2) UCL 6.058

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.781

SD in Log Scale 1.908

Mean in Original Scale 1.063

SD in Original Scale 3.246

  95% Percentile Bootstrap UCL 2.021

  95% BCA Bootstrap UCL 2.633

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.4 Data appear Lognormal at 5% Significance Level

Theta Star 3.353

nu star 23.22

A‐D Test Statistic 2.07 Nonparametric Statistics

5% A‐D Critical Value 0.829 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.829 Mean 1.075

5% K‐S Critical Value 0.174 SD 3.199

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.535

  95% KM (t) UCL 1.979

Assuming Gamma Distribution   95% KM (z) UCL 1.955

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 1.974

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 5.673

Maximum 18.51   95% KM (BCA) UCL 2.079

Mean 1.097   95% KM (Percentile Bootstrap) UCL 2.044

Median 0.286 95% KM (Chebyshev) UCL 3.409

SD 3.238 97.5% KM (Chebyshev) UCL 4.419

k star 0.192 99% KM (Chebyshev) UCL 6.403

Theta star 5.722

Nu star 14.19 Potential UCLs to Use

AppChi2 6.7   99% KM (Chebyshev) UCL 6.403

   95% Gamma Approximate UCL 2.323

   95% Adjusted Gamma UCL 2.403

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

PENTACHLOROBENZENE (mg/kg)

General Statistics

Number of Valid Data 4 Number of Detected Data 4

Number of Distinct Detected Data 4 Number of Non‐Detect Data 0

Number of Missing Values 293 Percent Non‐Detects 0.00%

Warning: This data set only has 4 observations!

Data set is too small to compute reliable and meaningful statistics and estimates!

The data set for variable PENTACHLOROBENZENE (mg/kg) was not processed!

It is suggested to collect at least 8 to 10 observations before using these statistical methods!

If possible, compute and collect Data Quality Objectives (DQO) based sample size and analytical results.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

UNRES. COMB. OF 1,2,3/4,5 TETRACHLOROBENZENE (mg/kg)

General Statistics

Number of Valid Data 4 Number of Detected Data 4

Number of Distinct Detected Data 4 Number of Non‐Detect Data 0

Number of Missing Values 293 Percent Non‐Detects 0.00%

Warning: This data set only has 4 observations!

Data set is too small to compute reliable and meaningful statistics and estimates!

The data set for variable UNRES. COMB. OF 1,2,3/4,5 TETRACHLOROBENZENE (mg/kg) was not processed!

It is suggested to collect at least 8 to 10 observations before using these statistical methods!

If possible, compute and collect Data Quality Objectives (DQO) based sample size and analytical results.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

XYLENES, TOTAL (mg/kg)

General Statistics

Number of Valid Data 325 Number of Detected Data 262

Number of Distinct Detected Data 262 Number of Non‐Detect Data 63

Number of Missing Values 2 Percent Non‐Detects 19.38%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00245 Minimum Detected ‐6.01

Maximum Detected 314.3 Maximum Detected 5.75

Mean of Detected 34.32 Mean of Detected 0.802

SD of Detected 55.01 SD of Detected 3.604

Minimum Non‐Detect 0.00308 Minimum Non‐Detect ‐5.782

Maximum Non‐Detect 0.971 Maximum Non‐Detect ‐0.0299

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 159

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 166

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 48.92%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.266 Lilliefors Test Statistic 0.174

5% Lilliefors Critical Value 0.0547 5% Lilliefors Critical Value 0.0547

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 27.68 Mean ‐0.281

SD 51.21 SD 3.95

   95% DL/2 (t) UCL 32.36   95%  H‐Stat (DL/2) UCL 2622

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐0.293

SD in Log Scale 3.963

Mean in Original Scale 27.67

SD in Original Scale 51.21

  95% Percentile Bootstrap UCL 32.43

  95% BCA Bootstrap UCL 33.06

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.258 Data do not follow a Discernable Distribution (0.05)

Theta Star 133

nu star 135.2

A‐D Test Statistic 5.503 Nonparametric Statistics

5% A‐D Critical Value 0.889 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.889 Mean 27.67

5% K‐S Critical Value 0.0619 SD 51.13

Data not Gamma Distributed at 5% Significance Level SE of Mean 2.842

  95% KM (t) UCL 32.36

Assuming Gamma Distribution   95% KM (z) UCL 32.35

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 32.36

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 32.89

Maximum 314.3   95% KM (BCA) UCL 32.37

Mean 27.67   95% KM (Percentile Bootstrap) UCL 32.37

Median 1.28 95% KM (Chebyshev) UCL 40.06

SD 51.21 97.5% KM (Chebyshev) UCL 45.42

k star 0.12 99% KM (Chebyshev) UCL 55.95

Theta star 231.3

Nu star 77.77 Potential UCLs to Use

AppChi2 58.46 97.5% KM (Chebyshev) UCL 45.42

   95% Gamma Approximate UCL 36.81

   95% Adjusted Gamma UCL 36.86

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

2‐METHYLNAPHTHALENE (mg/kg)

General Statistics

Number of Valid Data 33 Number of Detected Data 28

Number of Distinct Detected Data 28 Number of Non‐Detect Data 5

Number of Missing Values 295 Percent Non‐Detects 15.15%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.0436 Minimum Detected ‐3.133

Maximum Detected 37.37 Maximum Detected 3.621

Mean of Detected 7.503 Mean of Detected 1.059

SD of Detected 8.905 SD of Detected 1.834

Minimum Non‐Detect 0.0537 Minimum Non‐Detect ‐2.925

Maximum Non‐Detect 0.122 Maximum Non‐Detect ‐2.107

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 7

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 26

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 21.21%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Shapiro Wilk Test Statistic 0.759 Shapiro Wilk Test Statistic 0.893

5% Shapiro Wilk Critical Value 0.924 5% Shapiro Wilk Critical Value 0.924

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 6.372 Mean 0.391

SD 8.62 SD 2.332

   95% DL/2 (t) UCL 8.913   95%  H‐Stat (DL/2) UCL 93.2

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 5.005 Mean in Log Scale 0.577

SD 10.11 SD in Log Scale 2.045

   95% MLE (t) UCL 7.985 Mean in Original Scale 6.384

   95% MLE (Tiku) UCL 7.99 SD in Original Scale 8.61

  95% Percentile Bootstrap UCL 8.92

  95% BCA Bootstrap UCL 9.552

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.595 Data appear Gamma Distributed at 5% Significance Level

Theta Star 12.6

nu star 33.34

A‐D Test Statistic 0.367 Nonparametric Statistics

5% A‐D Critical Value 0.796 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.796 Mean 6.374

5% K‐S Critical Value 0.173 SD 8.487

Data appear Gamma Distributed at 5% Significance Level SE of Mean 1.504

  95% KM (t) UCL 8.922

Assuming Gamma Distribution   95% KM (z) UCL 8.848

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 8.915

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 10.2

Maximum 37.37   95% KM (BCA) UCL 9.04

Mean 6.366   95% KM (Percentile Bootstrap) UCL 8.957

Median 4 95% KM (Chebyshev) UCL 12.93

SD 8.624 97.5% KM (Chebyshev) UCL 15.77

k star 0.186 99% KM (Chebyshev) UCL 21.34

Theta star 34.22

Nu star 12.28 Potential UCLs to Use

AppChi2 5.411   95% KM (Chebyshev) UCL 12.93

   95% Gamma Approximate UCL 14.45

   95% Adjusted Gamma UCL 15.1

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

BENZO(A)ANTHRACENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 290

Number of Distinct Detected Data 287 Number of Non‐Detect Data 15

Number of Missing Values 23 Percent Non‐Detects 4.92%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00251 Minimum Detected ‐5.987

Maximum Detected 31.46 Maximum Detected 3.449

Mean of Detected 2.442 Mean of Detected ‐0.579

SD of Detected 4.901 SD of Detected 1.96

Minimum Non‐Detect 0.00472 Minimum Non‐Detect ‐5.356

Maximum Non‐Detect 1.172 Maximum Non‐Detect 0.159

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 196

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 109

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 64.26%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.309 Lilliefors Test Statistic 0.073

5% Lilliefors Critical Value 0.052 5% Lilliefors Critical Value 0.052

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 2.324 Mean ‐0.765

SD 4.806 SD 2.108

   95% DL/2 (t) UCL 2.778   95%  H‐Stat (DL/2) UCL 5.301

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐0.755

SD in Log Scale 2.073

Mean in Original Scale 2.323

SD in Original Scale 4.807

  95% Percentile Bootstrap UCL 2.792

  95% BCA Bootstrap UCL 2.854

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.438 Data do not follow a Discernable Distribution (0.05)

Theta Star 5.575

nu star 254

A‐D Test Statistic 5.126 Nonparametric Statistics

5% A‐D Critical Value 0.836 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.836 Mean 2.323

5% K‐S Critical Value 0.0568 SD 4.799

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.275

  95% KM (t) UCL 2.777

Assuming Gamma Distribution   95% KM (z) UCL 2.776

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 2.777

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 2.858

Maximum 31.46   95% KM (BCA) UCL 2.792

Mean 2.322   95% KM (Percentile Bootstrap) UCL 2.781

Median 0.519 95% KM (Chebyshev) UCL 3.523

SD 4.807 97.5% KM (Chebyshev) UCL 4.042

k star 0.287 99% KM (Chebyshev) UCL 5.062

Theta star 8.085

Nu star 175.2 Potential UCLs to Use

AppChi2 145.6 97.5% KM (Chebyshev) UCL 4.042

   95% Gamma Approximate UCL 2.794

   95% Adjusted Gamma UCL 2.796

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

BENZO(A)PYRENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 283

Number of Distinct Detected Data 279 Number of Non‐Detect Data 22

Number of Missing Values 23 Percent Non‐Detects 7.21%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00358 Minimum Detected ‐5.632

Maximum Detected 32.12 Maximum Detected 3.469

Mean of Detected 1.886 Mean of Detected ‐0.88

SD of Detected 4.081 SD of Detected 1.908

Minimum Non‐Detect 0.00454 Minimum Non‐Detect ‐5.395

Maximum Non‐Detect 2.429 Maximum Non‐Detect 0.887

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 253

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 52

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 82.95%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.322 Lilliefors Test Statistic 0.0502

5% Lilliefors Critical Value 0.0527 5% Lilliefors Critical Value 0.0527

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 1.758 Mean ‐1.127

SD 3.958 SD 2.102

   95% DL/2 (t) UCL 2.132   95%  H‐Stat (DL/2) UCL 3.427

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.124

SD in Log Scale 2.061

Mean in Original Scale 1.752

SD in Original Scale 3.959

  95% Percentile Bootstrap UCL 2.138

  95% BCA Bootstrap UCL 2.196

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.427 Data appear Lognormal at 5% Significance Level

Theta Star 4.414

nu star 241.9

A‐D Test Statistic 6.681 Nonparametric Statistics

5% A‐D Critical Value 0.839 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.839 Mean 1.754

5% K‐S Critical Value 0.0577 SD 3.953

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.227

  95% KM (t) UCL 2.128

Assuming Gamma Distribution   95% KM (z) UCL 2.127

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 2.128

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 2.213

Maximum 32.12   95% KM (BCA) UCL 2.148

Mean 1.751   95% KM (Percentile Bootstrap) UCL 2.136

Median 0.32 95% KM (Chebyshev) UCL 2.742

SD 3.96 97.5% KM (Chebyshev) UCL 3.17

k star 0.252 99% KM (Chebyshev) UCL 4.01

Theta star 6.94

Nu star 153.9 Potential UCLs to Use

AppChi2 126.2 97.5% KM (Chebyshev) UCL 3.17

   95% Gamma Approximate UCL 2.135

   95% Adjusted Gamma UCL 2.137

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

BENZO(B)FLUORANTHENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 283

Number of Distinct Detected Data 283 Number of Non‐Detect Data 22

Number of Missing Values 23 Percent Non‐Detects 7.21%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00453 Minimum Detected ‐5.398

Maximum Detected 27.8 Maximum Detected 3.325

Mean of Detected 2 Mean of Detected ‐0.691

SD of Detected 3.911 SD of Detected 1.811

Minimum Non‐Detect 0.00454 Minimum Non‐Detect ‐5.395

Maximum Non‐Detect 2.429 Maximum Non‐Detect 0.887

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 247

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 58

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 80.98%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.305 Lilliefors Test Statistic 0.0428

5% Lilliefors Critical Value 0.0527 5% Lilliefors Critical Value 0.0527

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 1.863 Mean ‐0.956

SD 3.8 SD 2.049

   95% DL/2 (t) UCL 2.222   95%  H‐Stat (DL/2) UCL 3.566

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐0.935

SD in Log Scale 1.974

Mean in Original Scale 1.858

SD in Original Scale 3.801

  95% Percentile Bootstrap UCL 2.233

  95% BCA Bootstrap UCL 2.285

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.462 Data appear Lognormal at 5% Significance Level

Theta Star 4.331

nu star 261.3

A‐D Test Statistic 6.587 Nonparametric Statistics

5% A‐D Critical Value 0.83 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.83 Mean 1.859

5% K‐S Critical Value 0.0574 SD 3.795

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.218

  95% KM (t) UCL 2.218

Assuming Gamma Distribution   95% KM (z) UCL 2.217

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 2.218

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 2.291

Maximum 27.8   95% KM (BCA) UCL 2.226

Mean 1.857   95% KM (Percentile Bootstrap) UCL 2.235

Median 0.41 95% KM (Chebyshev) UCL 2.808

SD 3.802 97.5% KM (Chebyshev) UCL 3.219

k star 0.262 99% KM (Chebyshev) UCL 4.025

Theta star 7.096

Nu star 159.6 Potential UCLs to Use

AppChi2 131.4 97.5% KM (Chebyshev) UCL 3.219

   95% Gamma Approximate UCL 2.256

   95% Adjusted Gamma UCL 2.258

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

BENZO(K)FLUORANTHENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 258

Number of Distinct Detected Data 256 Number of Non‐Detect Data 47

Number of Missing Values 23 Percent Non‐Detects 15.41%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00224 Minimum Detected ‐6.1

Maximum Detected 17.25 Maximum Detected 2.848

Mean of Detected 0.962 Mean of Detected ‐1.364

SD of Detected 1.986 SD of Detected 1.777

Minimum Non‐Detect 0.00454 Minimum Non‐Detect ‐5.395

Maximum Non‐Detect 9.3 Maximum Non‐Detect 2.23

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 302

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 3

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 99.02%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.314 Lilliefors Test Statistic 0.0452

5% Lilliefors Critical Value 0.0552 5% Lilliefors Critical Value 0.0552

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 0.847 Mean ‐1.72

SD 1.867 SD 1.969

   95% DL/2 (t) UCL 1.023   95%  H‐Stat (DL/2) UCL 1.296

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale ‐1.782

SD in Log Scale 1.946

Mean in Original Scale 0.819

SD in Original Scale 1.857

  95% Percentile Bootstrap UCL 1.002

  95% BCA Bootstrap UCL 1.029

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.479 Data appear Lognormal at 5% Significance Level

Theta Star 2.009

nu star 247.1

A‐D Test Statistic 5.349 Nonparametric Statistics

5% A‐D Critical Value 0.826 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.826 Mean 0.824

5% K‐S Critical Value 0.0604 SD 1.855

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.107

  95% KM (t) UCL 0.999

Assuming Gamma Distribution   95% KM (z) UCL 0.999

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 0.999

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 1.043

Maximum 17.25   95% KM (BCA) UCL 1.011

Mean 0.828   95% KM (Percentile Bootstrap) UCL 1.008

Median 0.183 95% KM (Chebyshev) UCL 1.288

SD 1.855 97.5% KM (Chebyshev) UCL 1.489

k star 0.199 99% KM (Chebyshev) UCL 1.884

Theta star 4.151

Nu star 121.7 Potential UCLs to Use

AppChi2 97.2 97.5% KM (Chebyshev) UCL 1.489

   95% Gamma Approximate UCL 1.036

   95% Adjusted Gamma UCL 1.037

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

CHRYSENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 292

Number of Distinct Detected Data 288 Number of Non‐Detect Data 13

Number of Missing Values 23 Percent Non‐Detects 4.26%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00453 Minimum Detected ‐5.398

Maximum Detected 29.97 Maximum Detected 3.4

Mean of Detected 2.42 Mean of Detected ‐0.463

SD of Detected 4.383 SD of Detected 1.966

Minimum Non‐Detect 0.00472 Minimum Non‐Detect ‐5.356

Maximum Non‐Detect 0.16 Maximum Non‐Detect ‐1.833

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 74

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 231

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 24.26%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.291 Lilliefors Test Statistic 0.1

5% Lilliefors Critical Value 0.0518 5% Lilliefors Critical Value 0.0518

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 2.318 Mean ‐0.647

SD 4.316 SD 2.125

   95% DL/2 (t) UCL 2.726   95%  H‐Stat (DL/2) UCL 6.313

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 1.448 Mean in Log Scale ‐0.623

SD 5.19 SD in Log Scale 2.075

   95% MLE (t) UCL 1.938 Mean in Original Scale 2.318

   95% MLE (Tiku) UCL 1.94 SD in Original Scale 4.315

  95% Percentile Bootstrap UCL 2.745

  95% BCA Bootstrap UCL 2.787

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.473 Data do not follow a Discernable Distribution (0.05)

Theta Star 5.12

nu star 276.1

A‐D Test Statistic 2.924 Nonparametric Statistics

5% A‐D Critical Value 0.828 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.828 Mean 2.318

5% K‐S Critical Value 0.0563 SD 4.309

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.247

  95% KM (t) UCL 2.726

Assuming Gamma Distribution   95% KM (z) UCL 2.724

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 2.725

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 2.795

Maximum 29.97   95% KM (BCA) UCL 2.725

Mean 2.317   95% KM (Percentile Bootstrap) UCL 2.741

Median 0.693 95% KM (Chebyshev) UCL 3.395

SD 4.316 97.5% KM (Chebyshev) UCL 3.861

k star 0.315 99% KM (Chebyshev) UCL 4.777

Theta star 7.363

Nu star 192 Potential UCLs to Use

AppChi2 160.9 97.5% KM (Chebyshev) UCL 3.861

   95% Gamma Approximate UCL 2.764

   95% Adjusted Gamma UCL 2.767

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

DIBENZO(A,H)ANTHRACENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 257

Number of Distinct Detected Data 254 Number of Non‐Detect Data 48

Number of Missing Values 23 Percent Non‐Detects 15.74%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00153 Minimum Detected ‐6.483

Maximum Detected 5.6 Maximum Detected 1.723

Mean of Detected 0.392 Mean of Detected ‐2.042

SD of Detected 0.723 SD of Detected 1.577

Minimum Non‐Detect 0.00454 Minimum Non‐Detect ‐5.395

Maximum Non‐Detect 9.3 Maximum Non‐Detect 2.23

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 305

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 0

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 100.00%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.295 Lilliefors Test Statistic 0.0466

5% Lilliefors Critical Value 0.0553 5% Lilliefors Critical Value 0.0553

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 0.364 Mean ‐2.309

SD 0.721 SD 1.769

   95% DL/2 (t) UCL 0.432   95%  H‐Stat (DL/2) UCL 0.495

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE method failed to converge properly Mean in Log Scale ‐2.382

SD in Log Scale 1.7

Mean in Original Scale 0.334

SD in Original Scale 0.677

  95% Percentile Bootstrap UCL 0.402

  95% BCA Bootstrap UCL 0.409

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.56 Data appear Lognormal at 5% Significance Level

Theta Star 0.699

nu star 288

A‐D Test Statistic 5.662 Nonparametric Statistics

5% A‐D Critical Value 0.815 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.815 Mean 0.338

5% K‐S Critical Value 0.0601 SD 0.677

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.039

  95% KM (t) UCL 0.402

Assuming Gamma Distribution   95% KM (z) UCL 0.402

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 0.402

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 0.415

Maximum 5.6   95% KM (BCA) UCL 0.406

Mean 0.342   95% KM (Percentile Bootstrap) UCL 0.405

Median 0.109 95% KM (Chebyshev) UCL 0.508

SD 0.676 97.5% KM (Chebyshev) UCL 0.581

k star 0.236 99% KM (Chebyshev) UCL 0.726

Theta star 1.445

Nu star 144.2 Potential UCLs to Use

AppChi2 117.5 97.5% KM (Chebyshev) UCL 0.581

   95% Gamma Approximate UCL 0.42

   95% Adjusted Gamma UCL 0.42

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

FLUORANTHENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 295

Number of Distinct Detected Data 293 Number of Non‐Detect Data 10

Number of Missing Values 23 Percent Non‐Detects 3.28%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00632 Minimum Detected ‐5.065

Maximum Detected 53.18 Maximum Detected 3.974

Mean of Detected 5.152 Mean of Detected 0.224

SD of Detected 9.215 SD of Detected 2.054

Minimum Non‐Detect 0.00489 Minimum Non‐Detect ‐5.32

Maximum Non‐Detect 0.0656 Maximum Non‐Detect ‐2.725

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 44

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 261

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 14.43%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.288 Lilliefors Test Statistic 0.123

5% Lilliefors Critical Value 0.0516 5% Lilliefors Critical Value 0.0516

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 4.983 Mean 0.0528

SD 9.109 SD 2.23

   95% DL/2 (t) UCL 5.844   95%  H‐Stat (DL/2) UCL 16.78

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 4.01 Mean in Log Scale 0.0797

SD 10.11 SD in Log Scale 2.171

   95% MLE (t) UCL 4.965 Mean in Original Scale 4.984

   95% MLE (Tiku) UCL 4.921 SD in Original Scale 9.109

  95% Percentile Bootstrap UCL 5.871

  95% BCA Bootstrap UCL 5.979

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.453 Data do not follow a Discernable Distribution (0.05)

Theta Star 11.37

nu star 267.3

A‐D Test Statistic 3.445 Nonparametric Statistics

5% A‐D Critical Value 0.833 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.833 Mean 4.983

5% K‐S Critical Value 0.056 SD 9.094

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.522

  95% KM (t) UCL 5.844

Assuming Gamma Distribution   95% KM (z) UCL 5.841

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 5.844

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 5.963

Maximum 53.18   95% KM (BCA) UCL 5.885

Mean 4.983   95% KM (Percentile Bootstrap) UCL 5.868

Median 1.56 95% KM (Chebyshev) UCL 7.257

SD 9.109 97.5% KM (Chebyshev) UCL 8.241

k star 0.327 99% KM (Chebyshev) UCL 10.17

Theta star 15.22

Nu star 199.7 Potential UCLs to Use

AppChi2 168 97.5% KM (Chebyshev) UCL 8.241

   95% Gamma Approximate UCL 5.923

   95% Adjusted Gamma UCL 5.928

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

FLUORENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 207

Number of Distinct Detected Data 207 Number of Non‐Detect Data 98

Number of Missing Values 23 Percent Non‐Detects 32.13%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00453 Minimum Detected ‐5.398

Maximum Detected 335.8 Maximum Detected 5.817

Mean of Detected 4.731 Mean of Detected ‐0.657

SD of Detected 24.2 SD of Detected 2.212

Minimum Non‐Detect 0.00296 Minimum Non‐Detect ‐5.824

Maximum Non‐Detect 0.866 Maximum Non‐Detect ‐0.144

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 221

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 84

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 72.46%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.423 Lilliefors Test Statistic 0.0402

5% Lilliefors Critical Value 0.0616 5% Lilliefors Critical Value 0.0616

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 3.229 Mean ‐1.555

SD 20.04 SD 2.358

   95% DL/2 (t) UCL 5.123   95%  H‐Stat (DL/2) UCL 4.588

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.873

SD in Log Scale 2.589

Mean in Original Scale 3.216

SD in Original Scale 20.04

  95% Percentile Bootstrap UCL 5.382

  95% BCA Bootstrap UCL 7.195

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.309 Data appear Lognormal at 5% Significance Level

Theta Star 15.32

nu star 127.9

A‐D Test Statistic 8.606 Nonparametric Statistics

5% A‐D Critical Value 0.867 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.867 Mean 3.219

5% K‐S Critical Value 0.0682 SD 20.01

Data not Gamma Distributed at 5% Significance Level SE of Mean 1.148

  95% KM (t) UCL 5.114

Assuming Gamma Distribution   95% KM (z) UCL 5.108

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 5.112

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 9.329

Maximum 335.8   95% KM (BCA) UCL 5.541

Mean 3.229   95% KM (Percentile Bootstrap) UCL 5.388

Median 0.174 95% KM (Chebyshev) UCL 8.225

SD 20.04 97.5% KM (Chebyshev) UCL 10.39

k star 0.105 99% KM (Chebyshev) UCL 14.65

Theta star 30.84

Nu star 63.87 Potential UCLs to Use

AppChi2 46.48 97.5% KM (Chebyshev) UCL 10.39

   95% Gamma Approximate UCL 4.437

   95% Adjusted Gamma UCL 4.444

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

INDENO(1,2,3‐CD)PYRENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 276

Number of Distinct Detected Data 274 Number of Non‐Detect Data 29

Number of Missing Values 23 Percent Non‐Detects 9.51%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00264 Minimum Detected ‐5.937

Maximum Detected 13.8 Maximum Detected 2.625

Mean of Detected 0.964 Mean of Detected ‐1.381

SD of Detected 1.915 SD of Detected 1.765

Minimum Non‐Detect 0.00454 Minimum Non‐Detect ‐5.395

Maximum Non‐Detect 9.3 Maximum Non‐Detect 2.23

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 302

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 3

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 99.02%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.308 Lilliefors Test Statistic 0.0386

5% Lilliefors Critical Value 0.0533 5% Lilliefors Critical Value 0.0533

Data not Normal at 5% Significance Level Data appear Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 0.898 Mean ‐1.624

SD 1.852 SD 1.96

   95% DL/2 (t) UCL 1.073   95%  H‐Stat (DL/2) UCL 1.479

Maximum Likelihood Estimate(MLE) Method N/A Log ROS Method

MLE yields a negative mean Mean in Log Scale ‐1.644

SD in Log Scale 1.905

Mean in Original Scale 0.876

SD in Original Scale 1.842

  95% Percentile Bootstrap UCL 1.058

  95% BCA Bootstrap UCL 1.078

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.474 Data appear Lognormal at 5% Significance Level

Theta Star 2.035

nu star 261.4

A‐D Test Statistic 6.427 Nonparametric Statistics

5% A‐D Critical Value 0.827 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.827 Mean 0.88

5% K‐S Critical Value 0.0582 SD 1.84

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.106

  95% KM (t) UCL 1.054

Assuming Gamma Distribution   95% KM (z) UCL 1.053

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 1.054

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 1.09

Maximum 13.8   95% KM (BCA) UCL 1.063

Mean 0.88   95% KM (Percentile Bootstrap) UCL 1.059

Median 0.202 95% KM (Chebyshev) UCL 1.34

SD 1.841 97.5% KM (Chebyshev) UCL 1.54

k star 0.26 99% KM (Chebyshev) UCL 1.931

Theta star 3.379

Nu star 158.9 Potential UCLs to Use

AppChi2 130.7 97.5% KM (Chebyshev) UCL 1.54

   95% Gamma Approximate UCL 1.069

   95% Adjusted Gamma UCL 1.07

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

NAPHTHALENE (mg/kg)

General Statistics

Number of Valid Data 322 Number of Detected Data 266

Number of Distinct Detected Data 263 Number of Non‐Detect Data 56

Number of Missing Values 6 Percent Non‐Detects 17.39%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.003 Minimum Detected ‐5.809

Maximum Detected 815 Maximum Detected 6.703

Mean of Detected 100.7 Mean of Detected 1.612

SD of Detected 144.9 SD of Detected 4.025

Minimum Non‐Detect 0.00658 Minimum Non‐Detect ‐5.023

Maximum Non‐Detect 0.341 Maximum Non‐Detect ‐1.076

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 138

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 184

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 42.86%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.243 Lilliefors Test Statistic 0.213

5% Lilliefors Critical Value 0.0543 5% Lilliefors Critical Value 0.0543

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 83.2 Mean 0.431

SD 137.1 SD 4.486

   95% DL/2 (t) UCL 95.8   95%  H‐Stat (DL/2) UCL 57451

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 14.91 Mean in Log Scale 0.574

SD 206.6 SD in Log Scale 4.307

   95% MLE (t) UCL 33.9 Mean in Original Scale 83.2

   95% MLE (Tiku) UCL 36.76 SD in Original Scale 137.1

  95% Percentile Bootstrap UCL 96.19

  95% BCA Bootstrap UCL 97.15

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.238 Data do not follow a Discernable Distribution (0.05)

Theta Star 422.6

nu star 126.8

A‐D Test Statistic 10.39 Nonparametric Statistics

5% A‐D Critical Value 0.898 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.898 Mean 83.2

5% K‐S Critical Value 0.0616 SD 136.9

Data not Gamma Distributed at 5% Significance Level SE of Mean 7.641

  95% KM (t) UCL 95.8

Assuming Gamma Distribution   95% KM (z) UCL 95.77

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 95.8

Minimum 0.003   95% KM (bootstrap t) UCL 97.58

Maximum 815   95% KM (BCA) UCL 96.16

Mean 83.35   95% KM (Percentile Bootstrap) UCL 95.93

Median 2.032 95% KM (Chebyshev) UCL 116.5

SD 137 97.5% KM (Chebyshev) UCL 130.9

k star 0.231 99% KM (Chebyshev) UCL 159.2

Theta star 361.3

Nu star 148.6 Potential UCLs to Use

AppChi2 121.4 97.5% KM (Chebyshev) UCL 130.9

   95% Gamma Approximate UCL 102

   95% Adjusted Gamma UCL 102.1

Note: DL/2 is not a recommended method.

General UCL Statistics for Data Sets with Non‐Detects

User Selected Options

Full Precision    OFF

Confidence Coefficient    95%

Number of Bootstrap Operations    10000

PHENANTHRENE (mg/kg)

General Statistics

Number of Valid Data 305 Number of Detected Data 295

Number of Distinct Detected Data 291 Number of Non‐Detect Data 10

Number of Missing Values 23 Percent Non‐Detects 3.28%

Raw Statistics Log‐transformed Statistics

Minimum Detected 0.00348 Minimum Detected ‐5.66

Maximum Detected 104.2 Maximum Detected 4.646

Mean of Detected 7.366 Mean of Detected 0.443

SD of Detected 13.64 SD of Detected 2.283

Minimum Non‐Detect 0.00454 Minimum Non‐Detect ‐5.395

Maximum Non‐Detect 0.0656 Maximum Non‐Detect ‐2.725

Note: Data have multiple DLs ‐ Use of KM Method is recommended Number treated as Non‐Detect 50

For all methods (except KM, DL/2, and ROS Methods), Number treated as Detected 255

Observations < Largest ND are treated as NDs Single DL Non‐Detect Percentage 16.39%

UCL Statistics

Normal Distribution Test with Detected Values Only Lognormal Distribution Test with Detected Values Only

Lilliefors Test Statistic 0.295 Lilliefors Test Statistic 0.17

5% Lilliefors Critical Value 0.0516 5% Lilliefors Critical Value 0.0516

Data not Normal at 5% Significance Level Data not Lognormal at 5% Significance Level

Assuming Normal Distribution Assuming Lognormal Distribution

DL/2 Substitution Method DL/2 Substitution Method

Mean 7.125 Mean 0.263

SD 13.48 SD 2.458

   95% DL/2 (t) UCL 8.398   95%  H‐Stat (DL/2) UCL 37.85

Maximum Likelihood Estimate(MLE) Method Log ROS Method

Mean 5.432 Mean in Log Scale 0.295

SD 15.19 SD in Log Scale 2.389

   95% MLE (t) UCL 6.868 Mean in Original Scale 7.125

   95% MLE (Tiku) UCL 6.813 SD in Original Scale 13.48

  95% Percentile Bootstrap UCL 8.45

  95% BCA Bootstrap UCL 8.601

Gamma Distribution Test with Detected Values Only Data Distribution Test with Detected Values Only

k star (bias corrected) 0.418 Data do not follow a Discernable Distribution (0.05)

Theta Star 17.62

nu star 246.7

A‐D Test Statistic 3.362 Nonparametric Statistics

5% A‐D Critical Value 0.841 Kaplan‐Meier (KM) Method

K‐S Test Statistic 0.841 Mean 7.125

5% K‐S Critical Value 0.0563 SD 13.46

Data not Gamma Distributed at 5% Significance Level SE of Mean 0.772

  95% KM (t) UCL 8.398

Assuming Gamma Distribution   95% KM (z) UCL 8.395

Gamma ROS Statistics using Extrapolated Data   95% KM (jackknife) UCL 8.398

Minimum 1.00E‐09   95% KM (bootstrap t) UCL 8.614

Maximum 104.2   95% KM (BCA) UCL 8.427

Mean 7.125   95% KM (Percentile Bootstrap) UCL 8.446

Median 2.476 95% KM (Chebyshev) UCL 10.49

SD 13.48 97.5% KM (Chebyshev) UCL 11.95

k star 0.309 99% KM (Chebyshev) UCL 14.81

Theta star 23.06

Nu star 188.4 Potential UCLs to Use

AppChi2 157.7 97.5% KM (Chebyshev) UCL 11.95

   95% Gamma Approximate UCL 8.514

   95% Adjusted Gamma UCL 8.521

Note: DL/2 is not a recommended method.