...... t ............................................................. ............................... / ;,2 /

How to Evaluate a W/Oxidation System

Ultraviolet oxidation is a proven technology for ground and process water treatment. Learn to identify when to consider it and how much it will cost. ................................. by Robert Notarfonzo and Wayne McPhee .................................

No single technology offers the solution for every water treatment problem. Although the conventional technologies of air strip- ping and activated carbon have proved robust and usually cost effective, continued

advances in ultraviolet (UV)/oxidation have made it the U.S. Environmental Protection Agency (EPA) proven tech- nology of choice in an ever increasing number of ground- water and process water applications. The number of full- scale UV/oxidation installations has increased in the last five years from a handful to more than 150.

Principles of UV/oxidation In the UV/oxidation process, a high-powered lamp emits W radiation through a quartz sleeve into the contaminated water. An oxidizing agent, typically hydrogen peroxide, is added, which is activated by the UV light to form oxidiz- ing hydroxyl radicals:

H,O, + UV + 2.OH These radicals indiscriminately destroy the toxic organic

compounds in the water. Depending on the nature of the organic species, two types of initial attack are possible: it can abstract a hydrogen atom to form water, as with alkanes or alcohols, or it can add to the contaminant, as is the case for olefins or aromatic compounds. The following equation represents the simplified general oxidation process:

Chlorinated 0, 0, organic ++ Oxygenated ++ CO,+H,O+Cl-

molecule .OH intermediates .OH The attack by hydroxyl radicals, in the presence of oxy-

gen, initiates a cascade of reactions leading to mineral- ization, such as CO, and H,O. In certain applications, catalysts, which are photo active and non-toxic, are added to significantly enhance the system’s performance. A UV/oxidation system can be designed to treat to any dis- charge requirement.

W/oxidation’s key advantage is its inherent destructive nature; contaminated water is detoxified with no require- ment for secondary disposal. There is no transfer of con- taminants from one medium to another. Furthermore, UV systems in combination with hydrogen peroxide have no vapor emissions, hence no air permit is required. The equipment is quiet, compact and unobtrusive, and preven- tive maintenance and operating requirements are low in a carefully designed system.

A typical system In a typical UV/oxidation system, reagents are injected and mixed using metering pumps and an in-line static mixer. The contaminated water then flows sequentially through one or more UV reactors, where treatment occurs.

Pretreatment, such as solids removal, pH adjustment and oil and grease removal, sometimes is required. In practice, if the W system is designed carefully with provisions for auto- mated cleaning of the quartz sleeve which surrounds the UV lamp, pretreatment often can be avoided, reducing both the capital investment and the ongoing maintenance costs.

The UV lamp inside the reactor is operated at high volt- age, typically between 1000 and 3000 volts. Safety inter- locks are fitted to protect personnel from both the UV radiation and the high voltage supply. These interlocks usually are linked to a programmable logic controller (PLC), which can be used to control the whole installa- tion, including feed pumps, the UV lamps and the reagent delivery systems. A PLC can be accessed via a modem to facilitate diagnostics for easier servicing, and can be reprogrammed to accommodate changes in operation throughout the remediation cycle.

In most groundwater applications, the material specified for the UV reactor is 316L stainless steel, which protects against the oxidants and the UV light while providing excellent resistance to corrosion.

UV/oxidation checklist The competitive universe for UV/oxidation systems is not between the various vendors who manufacture these sys- tems, but with the conventional media transfer technologies

74 POLLUTION ENGINEERING OCTOBER 1994

SPILSTOPPER' 'x

PROTECTIVE COVER FOR DRAINS, GRATES & MANHOLES AGAINST ACCIDENTAL SPILLS

US. PATENT NO. 4838732

SPILSTOPPER an effective precautionary measure against accidental flooding of:

Manholes Storm Drains

Vents Meters Cleanouts

0

0

el

0

E?

SAVES MONEY - SPILSTOPPER reduces loss ed fluids; is much less expensive than landscaping, and other precautionary means; eliminates fines and penalties polluting drainage systems; reduces liabilities; and significantly reduces

FLEXIBLE - Extreme flexibility of allows it to effectively seal off mounted manholes and grates,

has been proven to be a in the following fields:

sand, grass, asphalt or concrete pave Ynt . - B ~ L K OIL TERMINALS

LIGHT WEIGHT - SPILSTOPPER is heivy enough to seal off spills, but is light enough to beleasily manag- ed by one man.

DURABLE - SPILSTOPPER is cov red with a tough elastomer which can be stretched to nearly double its length and recover completely.

INERT - SPILSTOPPER does not react chemically with water, petroleum products or most caustics.

EAS!LY STORED - SPILSTOPPER can be stored in an easily handled roll on tank trucks or on site. (SPILSTOPPER storage bin can be purchased separately.)

E/

GAS'OLINE STATIONS PETROLEUM TANK TRUCKS HAZARDOUS MATERIALS HANDLING TANK TRUCKS CHEMICAL PLANTS ANY BUSINESS HANDLING E.P.A. REGULATED MATERIALS MILITARY INSTALLATIONS PUBLIC WORKS DEPARTMENTS FIRE AND EMERGENCY DEPARTMENTS SP!LL CONTROL RESPONSE TEAMS

DISTRIBUTED BY: MANUFACTURED BY: PRODUCTS rLARK CO.. INC

916 W. 25th Street Norfolk, Virginia 23517 (804) 625-5917

For INFO circle 55 on Reader Service Card OCTOBER 1994 POLLUTION ENGINEERING 73

9 .....................................................................................................................................................................................................................

of activated carbon and air stripping. Before conducting a preliminary cost estimate, potential applications can be quickly screened with yes or no answers to the following checklist questions:

Is a destruction technology preferred? Are there any restrictions on air discharge? Does the principle contaminant air strip poorly -

Henry’s Law Constant 100 atm/mole fraction? Do any of the principle contaminants load poorly on acti-

vated carbon in the liquid phase - 50 mg/g carbon at 1 ppm? Does the background water chemistry consume large

amounts of liquid phase activated carbon or interfere with carbon operation, such as high iron or high chemical oxy- gen demand (COD)?

Are there handling or disposal concerns associated with loading contaminants on to activated carbon, such as load- ing and concentrating explosive or carcinogenic com- pounds on to the carbon bed?

If the result is three or more yes answers, UV/oxidation should be considered and a preliminary cost estimate is in order using the design parameters provided as shown in Calculation Example 1. If the preliminary cost estimate looks favorable, arrange a design test to confirm and guar- antee costs.

Electrical energy per order and UV dose The key design variables are the exposure to UV radiation and the number of orders of contaminant concentration removed. These two variables are combined into a single function, the Electrical Energy per Order (EE/O). The EE/O is a powerful scale-up parameter and is a measure of the treatment obtained in a fixed volume of water as a function of exposure to UV light. It is defined as the kilo- watt hours of electricity required to reduce the concen- tration of a compound in 1000 gallons by one order of magnitude, or 90 percent. The units for EE/O are kWh/lOOO gallodorder.

For example, if it takes 10 kWh of electricity to reduce the concentration of a target compound from 10 ppm to 1 ppm - 1 order of magnitude or 90 percent - in 1000 gal- lons of groundwater, then the EE/O is 10 kWh/1000 gaVorder for this compound. It will then take another 10 kWh to reduce the compound from 1 ppm to 0.1 ppm, and so on.

The EWO measured in a design test is specific to the water tested and to the compound of interest, and it will

Figure 1. Esfimafed capifal costs are given in ranges.

vary for different applications. Typical EE/Os for a range of organic contaminants are provided in Table 1.

With the EE/O determined, either through design tests or by using Table 1, the UV dose, or the amount of elec- trical energy required to treat 1000 gallons, needed to treat a specific case is simply calculated using the follow- ing equation: (1) UV dose (kWW1000 gallons) = EE/O X log (initial/

final), where initial is the starting concentration (any units), and final is the anticipated or required discharge standard (same units as initial).

For streams with several contaminants the required energy is not additive but determined by the contami- nant requiring the greatest UV dosage. See Calculation Example 2.

Operating costs Once the required UV dose is known, the electrical oper- ating cost associated with supplying the UV energy can be calculated using the following equation: ( 2 ) Electrical cost ($/lo00 gal) =

UV dose (kWh/1000gal) X Power cost ($/kWh) The second key parameter from the design test is the con-

centration (ppm) of chemical reagents used, specifically hydrogen peroxide and any catalyst added to improve per- formance. The peroxide dose is based on the UV absorbance and COD of the water, and is typically in the range of 50 to 200 ppm (mg/L). For the purpose of a pre- liminary cost estimate, the simplest rule of thumb for esti- mating the amount of peroxide necessary is the greater of 25 ppm or twice the COD concentration. Hydrogen perox- ide cost varies from $0.005 to $0.008 per ppm concentra- tiordl000 gal. If a catalyst is required, its selection and con- centration will vary with the target compound and must be based on design test results. Lamp replacement costs typi- cally range between 40 percent to 50 percent of the electri- cal cost. Therefore, ( 3 ) Total operating cost ($/lo00 gal) = 1.45 X Electrical

cost + peroxide cost, where peroxide cost = (H202 concentration in ppm) X ($O.O05/ppm/1000 gal).

OCTOBER 1994 POLLUTION ENGINEERING 15

....................................................................... r, .......................................................................

UVpower is influenced more by a change infow rate than a change in concentration.

Capital costs Capital cost is a function of system size, which is a func- tion of the UV power required to destroy the selected con- taminants. Using the EE/Os provided in Table 1, the fol- lowing equation is used to determine the total UV power (kW) required: (4) UV power (kW)

= EE/O X 60 X flow (gpm) X log (initial/final)

= UV dose X 60 X flow (gpm) 1000

1000 Once the required UV power is known, Figure 1 can be

used to look up the associated capital cost in U.S. dollars. The capital costs are given as ranges to allow for the actu- al number of discrete reactors which will be required along with any additional system options required.

The total UV power varies proportionally with flow rate and orders of concentration of contaminant removed. For example, doubling the flow rate, or treating from 10 ppm to 0.1 ppm instead of down to 1 ppm, will double the UV power required. The equation theoretically can be used to obtain total UV power for any combination of flow ratc or concentration, but its accuracy depends on the EE/O, which

76 POLLUTION ENGINEERING OCTOBER 1994

for a single contaminant can vary significantly from those listed depending on the water matrix and concentration.

It is important to note that the total UV power, and hence the capital and operating cost, is influenced much more by a change in flow rate than by a similar change in concen- tration because of its logarithmic dependence on concen- tration through the log (initialhnal) term.

These are the most accurate cost figures possible without performing actual design tests. The design tests consist of batch treatment runs of sampled water while varying UV dosage and reagent concentrations. Capital and operating costs are optimized over the lifetime of the remediation project by selecting the combination of total UV power and reagent concentration where the most economical treat- ment is obtained.

Cost comparisons Compared with other treatment technologies, such as acti- vated carbon, unit operating costs for UV/oxidation increase much more slowly with increasing influent concentrations. Typical operating costs as a function of influent concentra- tion are shown in Table 2. UV/oxidation treatment costs are almost always less for those contaminants which load poor- ly on carbon, regardless of Concentration. UV/oxidation treatment costs are competitive for most of the average load- ing compounds but a definitive answer can only come from

Since UV/oxidation capital costs typically are higher than that of carbon for the same size flow, longer term pro- jects may favor UV/oxidation where the cumulative savings in operating costs offset the higher capital expense.

the results of 2 design t m .

Robert Notarfonzo is a market analyst and Wayne McPhee is a process engineer with Solarchem Environmental Systems, Toronto, Ontario, 905-477-9242.

Please circle the appropriate number on the Reader Service Card to indicate the level of interest in the article. High 412 Medium 413 Low 414