quantitative passive diffusivequantitative passive...
TRANSCRIPT
Quantitative Passive DiffusiveQuantitative Passive Diffusive--Adsorptive Sampling Techniques for Adsorptive Sampling Techniques for
Vapor Intrusion AssessmentVapor Intrusion AssessmentVapor Intrusion AssessmentVapor Intrusion Assessment
ESTCP Project 08 30ESTCP Project 08-30
T M Al H G lt T G ki S S th thT. McAlary, H. Groenevelt, T. Gorecki, S. Seethapathy, D. Crump, P.Sacco, H. Hayes, M. Tuday, B.
Schumacher, J. Nocerino, P. Johnson
AWMA Specialty Conference – Vapor Intrusion 2010
September 30, 2010Chi IL
11
Chicago, IL
Outline• Challenges with conventional sampling
• Introduction to quantitative passive sampling
• Comparative Studies• Comparative Studies• ESTCP Project 08-30• NAVY SPAWAR Projectj
• Summary of Findings to date
22
Challenges withgConventional Sampling
33
Summa Canister and TO-15
Complex procedure, requires special training ($)
Must be cleaned and certified ($)
Bulky ($) to ship and handleBulky ($) to ship and handle
Maximum ~24 hour samples
Costly: $200 to $300/sample, depending on reporting limit
Not useful for analytes heavier than naphthalene (poor recovery)
4444
55
Automatic Thermal Desorption Tubes / TO-17
Typically customized for each application –high level of training required
Allows longer than 24-hour samples, but the pump must run reliably throughout the sampling periodsampling period
Capable of a larger list of analytes
Typically <$200/sample, depending on analyte list
Potential for breakthrough and competition in high concentration zones
66
Challenging to get sufficient sample volume in low permeability soils
Introduction to Quantitative QPassive Sampling
77
Quantitative Passive Sampling• Expose a sampler for a known time• Ship to laboratory for analysis of adsorbed mass• Calculate concentration
M C0 = concentration of analyte in air (µg/m3)
tkMC 10 M = mass of analyte collected by the sorbent (pg)
k-1 = uptake rate (mL/min)
• Uptake rate is unique to each sampler/analyte pair
tk t = sampling time (min)
• Uptake rate is unique to each sampler/analyte pair• Usually requires expensive controlled experiments• Some passive samplers don’t control uptake – qualitative!
88
Benefits of Passive Sampling• Simple (minimal training)
L i• Less expensive• Low reporting limits with no premium cost
Ti i ht d t ti• Time-weighted average concentration(up to a week or a month if needed)
S ll t hi di t t d l• Smaller – easy to ship, discrete to deploy• Long history of use in Industrial Hygiene
Oth b fit i t h l• Other benefits unique to each sampler
99
ESTCP Study TeamOrganization Name RoleGeosyntec Consultants, Inc.Guelph (Canada)
Todd McAlaryHester Groenevelt
Overall project direction &reporting
US EPA Labs, Las Vegas (NV) Brian SchumacherJohn Nocerino (D)
Experimental Design &Statistics
Arizona State University (AZ) Paul Johnson Practicality for Vapor IntrusionSitesSites
University of Waterloo(Canada)
Tadeusz GoreckiSuresh Seethapathy
PDMS Membrane Sampler
Cranfield University (UK) Derrick Crump ATD Passive and Active Samplers
Fondazione Salvatore Maugeri(Italy)
Paolo Sacco Radiello Samplers
Columbia Analytical Services(CA)
Michael TudayKu-Jih Chen
High Conc. Laboratory TestingUltra II™ samplers & canisters(CA) Ku-Jih Chen Ultra II samplers & canisters
Air Toxics Limited (CA) Heidi HayesStephen Disher
Low Conc. Laboratory TestingATD Passive and Active Samplers
International Team incl ding the labs most familiar ith each of the samplers
1010
International Team, including the labs most familiar with each of the samplers
Four Different Passive Samplers
Waterloo Membrane S l TM
Automated Thermal Desorption (ATD) T bSampler TM
(1-D permeation across a
b CS
Tubes
(1-D diffusion through air, thermal desorption)
membrane, CS2extraction or thermal desorption)
Radiello™
SKC Ultra II
(Badge sampler,
(Radial uptake through porous cylinder, thermal desorption or( g p
1-D uptake through porous plate, thermal desorption)
desorption or solvent extraction)
1111
p )
Desorption Methods
Thermal desorptionThermal desorptionThermal desorptionThermal desorption
Solvent
Sorbent
Solvent desorptionSolvent desorption
1212
Thermal desorption
Sorbent Media SelectionSorbent Description Sampling Properties
Anasorb 747 Beaded active carbon
High capacity for VOCs, solvent desorption
Tenax TA Porous polymer Low capacity, medium to high boiling compounds, thermal
60/80 mesh
g pdesorption
Non porous high capacity forCarbopack B
60/80 mesh graphitized
carbon black
Non-porous, high capacity for VOCs, thermal desorption
1313
There are many, many more sorbents available
Laboratory Analyte List
43
Selected to span a range of compounds of interest for vapor intrusion studies
1414
Selected to span a range of compounds of interest for vapor intrusion studiesChlorinated ethenes, ethanes, methanes, and VOCs of differing solubility and sorptive properties
Low Concentration Tests:To Mimic Indoor and Outdoor Air ConditionsTo Mimic Indoor and Outdoor Air Conditions
Concentration : 1, 50 and 100 ppbvTemperature: 22, 25, 28 °CTemperature: 22, 25, 28 CHumidity: 30, 60 and 90% RHFace velocity: 0.01, 0.2 and 0.4 m/sExposure time: 1 4 and 7 days
High Concentration Tests:
Exposure time: 1, 4 and 7 days
High Concentration Tests:To Mimic Soil Gas Conditions
C t ti 1 10 d 100Concentration : 1, 10 and 100 ppmvTemperature: ambientHumidity: 90-100%Face velocity: very low (5x10-5 m/s)
1515
Face velocity: very low (5x10 5 m/s)Exposure time: 30 minutes
Exposure Chamber• Conditioned gas enters at
bottom and flows past lsamplers
• Samplers rotate on a carousel to control facecarousel to control face velocity
• Baffles help to minimize t b lturbulence
• All surfaces passivated to minimize adsorptionminimize adsorption
1616
Exposure Chamber (cont’d)Off th h lf All fOff the shelf glassware, modified by glass blower
All surfaces passivated
glass-blower
Surrounded by tube of circulating fluid and insulated for temperature controlcontrol
1717
1818
Inter-Laboratory TestingSecondary # of Samplers to
Sampler Type Home Laboratory Laboratories Each Laboratory
W t l M b Ai T i LtdWaterloo Membrane Sampler University of Waterloo
Air Toxics Ltd2
Airzone One
ATD Tubes with Tenax TA Air Toxics Ltd
Columbia Analytical Services 2ATD Tubes with Tenax TA Air Toxics Ltd Services 2
University of Waterloo
ATD Tubes with b k Air Toxics Ltd
Columbia Analytical Services 2CarboPack B Air Toxics Ltd 2
University of Waterloo
SKC Ultra Columbia Analytical Services
Air Toxics Ltd2Services
Airzone One
Radiello Fondazione Salvatore Maugeri
Columbia Analytical Services 2
Air Toxics Ltd
1919
Air Toxics Ltd
Interlab Test – Youden Plot
(blank contamination)
2020
Most results show good agreement between labsMEK was problematic
ANOVA TestingSet the five factors at their midpoints, and run test multiple times
to determine variance from experimental procedures
Factor Units Center Value
Concentration ppb 50
Temperature °C 20Temperature C 20
Gas Flow Velocity m/s 0.2
Sampling Duration days 4
Relative Humidity % 60
2121
e at ve u d ty % 60
Fractional Factorial TestingA series of experiments strategically changing the 5 key factors
Concentration
Temperature
Face Velocity
Sample Time
HumidityHumidity
2222
Chamber Test Results to DateChamber Test Results to Date
2323
Chamber Test Results to DateChamber Test Results to Date
(l bi )(low bias)
2424
Chamber Test Results to DateChamber Test Results to Date
(blank)(blank)
(1-day)
(Calculated Uptake Rate)
2525
Chamber Test Results to DateChamber Test Results to Date
(high bias)
?
?
2626
Chamber Test Results to DateChamber Test Results to Date
2727
Field Testing (SPAWAR/NESDI)Comparative sampling of indoor air, outdoor air, and soil vapor
Indoor Air:
ATD Tubes3M OVM 3500RadielloSKC Ult IISKC Ultra IIWaterloo Membrane
VersusVersus
6-day Summas
With
3 Replicates
2828
Acknowledgements to Ignacio Rivera and Bart Chadwick of SPAWAR for Support
p
Indoor Air (3 Locations in Triplicate)Indoor Air (3 Locations in Triplicate)Indoor Air AverageTrichloroethene
8
10Indoor Air - AverageTrichloroethene
SummaWMS3M OVM
(Anasorb 747)(Activated carbon wafer)
6
8
m3)
ATDRadielloSKC Ultra
Each bar is an average of the three
(Chromosorb 106)(Activated charcoal)
(Chromosorb 106)
4
atio
n (µ
g/m Each bar is an average of the three
samples collected at that location
2
Con
cent
ra
0
IA-1 IA-2 IA-3 Average
C
Sample Location
2929
The low bias for SKC is believed to be due to sorbent selection (Chromosorb 106 worked well for ATD, but SKC has 20X higher uptake rate)TCE was the dominant compound
SubSub--Slab SamplingSlab SamplingTwo types:
1) Fully passive (drop and cork for 24 hours)
2) Semi-passive(drop and purge
ith PID f 10with PID for 10 minutes)
3030
SPAWAR Subslab DataSPAWAR Subslab DataFully passive samples (left)
Fully Passive With Purging
showed “Starvation Effect” in proportion to
t k tuptake rate
Much less “starvation” when PID used to purgePID used to purge gas from hole during 10 min sampling interval
3131
p g
Starvation Factor vs Uptake RateStarvation Factor vs Uptake Rate1000.0
WMS cis-1,2-DichloroetheneWMS Trichloroethene
Starvation Factor = Summa conc’nPassive conc’n
100.0
TrichloroetheneSKC cis-1,2-DichloroetheneSKC Trichloroethene
on F
acto
r
Radiello TrichloroetheneATD cis-1,2-
10.0
Star
vatio
ATD cis 1,2DichloroetheneATD Trichloroethene
3M OVM i 1 2
1.00.01 0.1 1 10 100
Uptake Rate (ml/min)3M OVM cis-1,2-Dichloroethene3M OVM Trichloroethene
3232
Consider a custom “low-uptake” sampler
Modified Uptake RatesModified Uptake Rates
WMS Sampler
ATD Tube & Pinhole Cap
SKC Ultra II and 12-hole Cap
Reduce the uptake rate by reducing the area
3333WMS Pinhole Sampler
p y gLower uptake = less starvation
Passive Soil Gas Passive Soil Gas –– Utah HouseUtah House
3434
6 Passive Samplers6 Passive Samplers(Tenax) (Carbopack) (Cap)
Uptake rates: 0.01 to 75 mL/min
Void space was larger than SPAWAR sub-slab sampling
3535
6 Passive Probes + 1 Conventional6 Passive Probes + 1 Conventional
Modified Latin Square TestAdded one Duplicate
3636
Added one DuplicateTested 1 to 12 day exposuresDrop the sampler/Purge 3 PV
Active Sampler DataActive Sampler Data120
140
TCE in Summa and Hapsite GC/MSPSG-1PSG-2PSG-3PSG-4PSG-5
60
80
100
trat
ion
(ppb
v)
PSG-6Summa - PSG-1Summa - PSG-2Summa - PSG-3Summa - PSG-4Summa - PSG-5
20
40
60
Con
cent Summa - PSG-6
Linear (PSG-1)Linear (PSG-2)Linear (PSG-3)Linear (PSG-4)Linear (PSG-5)
007/16/10 07/21/10 07/26/10 07/31/10 08/05/10 08/10/10 08/15/10 08/20/10 08/25/10 08/30/10 09/04/10 09/09/10
Sampling Date
( )Linear (PSG-6)
160 1,1-DCE in Summa and Hapsite GC/MSPSG-1PSG-2PSG-3PSG 4
100
120
140
on (p
pbv)
PSG-4PSG-5PSG-6Summa - PSG-1Summa - PSG-2Summa - PSG-3
40
60
80
Con
cent
ratio Summa - PSG-4
Summa - PSG-5Summa - PSG-6Linear (PSG-1)Linear (PSG-2)Linear (PSG-3)
3737
0
20
07/16/10 07/21/10 07/26/10 07/31/10 08/05/10 08/10/10 08/15/10 08/20/10 08/25/10 08/30/10 09/04/10 09/09/10
Sampling Date
( )Linear (PSG-4)Linear (PSG-5)Linear (PSG-6)
Fully Passive Fully Passive –– Deep 2Deep 2--inch Probesinch Probes
C/Co = Passive data divided by Mean of all Hapsite data
3838
All but about 3 of these measurements are within a half order of magnitudeLonger samples with weaker sorbent show losses (Tenax)
Fully Passive Fully Passive –– Deep 2Deep 2--inch Probesinch Probes
Longer samples show losses with both Carbopack and Tenax for ATD Tubes
3939
Consistent with theory for incomplete retention (11DCE is 3X less sorptive than TCE)
Starvation Factor Versus Uptake RateStarvation Factor Versus Uptake Rate1000
100
1000
Losses from Tenax
10
or (C
o/C
)
1
vatio
n Fa
cto
0.1
Star
v
0.010.01 0.1 1 10 100 1000
4040
Uptake Rate (mL/min)
After removing artifacts of weak sorbent, trend is encouraging
FlowFlow--Through CellsThrough Cells
Why worry about starvation at all?J st p rge soil gas past the sampler at a modest rate
4141
Just purge soil gas past the sampler at a modest rateComparable to groundwater “Low Flow Sampling Protocol”
FlowFlow--Through Cell ResultsThrough Cell Results
4242
Flow-through cell neutralizes the starvation effectVariability about 1 order of magnitude, but so did Summa cans
Interim Conclusions Encouraging results to dateEncouraging results to date
No samplers eliminatedNo samplers eliminated Challenges with highly sorptive or highly soluble compoundsChallenges with highly sorptive or highly soluble compounds Challenges with very high humidity or long exposures for Challenges with very high humidity or long exposures for
some sampler/sorbent combinationssome sampler/sorbent combinationssome sampler/sorbent combinationssome sampler/sorbent combinations Much better understanding of how to avoid starvation for soil Much better understanding of how to avoid starvation for soil
gas sampling gas sampling –– quantitative results are likely feasiblequantitative results are likely feasibleg p gg p g q yq y Modifications can overcome some of the limitationsModifications can overcome some of the limitations
Custom sorbents and uptake rates Custom sorbents and uptake rates Based on theory & dataBased on theory & data
This is roughly the “halfThis is roughly the “half--way” point way” point –– stay tuned!stay tuned!
4343
Questions/Comments?Questions/Comments?Questions/Comments?Questions/Comments?