Foreward

4.3. Site Visit

During a site visit health assessors could see, smell, or feel blowing fugitives. Observing the site characteristics and proximity of the community will aid them in evaluating the potential for the public to be affected by reasonable worst-case fugitive emissions.

ATSDR recommends health assessors conduct a site visit of the thermal treatment facility to observe waste and residuals handling practices, facility operation, and environmental sampling. A site visit is necessary to understand the configuration and operation of the facility, its relationship to the community, susceptible populations, and any topographic features in the area that could affect the dispersion of facility emissions. Health assessors will find the site visit most productive if they review in advance the facility design and operating conditions, the environmental monitoring plan, and the health and safety plan.

During site visits to thermal treatment facilities, health assessors should also look for such things as:

• How close are residences to the facility or site?
  
  
• Are residential areas likely to be contaminated from site runoff?
  
  
• Are residences in the predominant wind direction from the site?
  
  
• Are day care facilities, schools, hospitals, nursing homes, or other institutions that are occupied 24 hours a day, or other sensitive populations, etc. near the site?
  
• Are large or tall buildings nearby that could affect down-wash or could be fumigated by stack emissions?
  
  
• Are hills, mountains, valleys, canyons, large lakes, oceans, or other topographic features in the area affecting the wind currents that must be considered when modeling?
  
  
• Does the area experience frequent inversions or other weather conditions that should be considered when evaluating the facility emissions?
  
  
• Are food or water supplies such as home gardens, diaries, cattle ranches, farms, etc. likely to be affected by fallout from the facility?
  
  
• Does a potential for worker exposure exist, and does the site have an adequate worker protection program? Health assessors should also note any worker safety issues they observe while on site.

If health assessors identify worker safety problems or are unsure whether the facility's program is adequate, they should discuss their findings with their supervisor before contacting the National Institute for Occupational Safety and Health (NIOSH) or the Occupational Safety and Health Administration (OSHA).

When visiting RCRA or CERCLA sites, ATSDR Health assessors must comply with OSHA training and medical monitoring regulations and wear the appropriate personal protective equipment (PPE) for the site conditions (ATSDR 1991).

To evaluate the potential public health effects of a thermal treatment technology, health assessors must be familiar with the design of incinerators and desorbers. This chapter discusses the common subsystems, explains the significant features of various types of equipment that could be used, and describes the emissions that could be of public health concern.

5.1. Thermal Treatment Facility Designs

An incinerator removes and destroys organic constituents in the waste. A desorber removes and captures or treats organic constituents.

Thermal treatment units should be designed to handle the specific type(s) of waste to be treated. Properly designed thermal desorption units can effectively remove volatile organic compounds (VOCs), semi-volatile organic compounds (SVOCs), PCBs, pesticides, and petrochemicals from solid wastes such as contaminated soils. Some desorber designs can also decontaminate small amounts of sediment or liquid waste in conjunction with solid waste. Incinerators can be designed to treat all types of wastes simultaneously or a single type. Table 2 lists desorbers, the major incinerator types, and the various physical wastes they can treat.

The four major subsystems in a thermal treatment facility are (1) waste preparation and feed systems, (also known as pretreatment), (2) combustion or desorption chambers, (3) air pollution control equipment, (also known as gas post treatment), and (4) liquids and ash handling, (also known as solids post treatment), and residuals management systems. Figure 1, in Appendix C, shows the general orientation of the subsystems and typical process component options for incinerators. Figure 2, in Appendix C, shows the general orientation of thermal desorption systems. The desorber unit could be a rotary kiln/dryer, thermal screw, fluidized bed, distillation chamber, or belt conveyor system (EPA 1993). A desorption system and an incinerator have some design differences. A desorption system has a chamber operating at temperatures just high enough to effectively vaporize, but not combust, the organic compounds from the contaminated material. It also has condensers, carbon adsorption units or both in addition to or instead of any of the other air pollution control equipment (APCE) shown in Figure 1. An incinerator has one or two refractory lined combustion chambers operating at high enough temperatures to vaporize any organic compounds and destroy them (i.e., convert them to carbon dioxide, water, and acid gases such as hydrochloric acid vapors).

The so-called three T's - time, temperature, and turbulence - are important for complete incineration of organic chemicals. For complete combustion to occur, the volatilized or partially destructed organic chemicals need sufficient time at a high enough temperature and turbulence to mix thoroughly the off-gases with excess oxygen.

5.1.1. Pretreatment - Waste Preparation and Feed Systems

The type of waste preparation and feed system is determined by the physical form of the wastes, the type of contaminants, and the type of primary combustion chamber (PCC) or desorption chamber (DC). The physical forms of the wastes (gas, liquid, sludge, or solid) and the type of contaminants (flammable, VOCs, etc.) to be treated can also affect the design of the other three subsystems.

Gaseous wastes are piped into the combustion or desorption chambers. If the gases are flammable or ignitable they can be fed to the DC, PCC, or SCC. But non-flammable gases are usually fed only to the DC or PCC.

Liquid wastes are usually blended agitated or both, then pumped through nozzles or atomizing burners into either one or both of the incinerator combustion chambers. Liquid wastes should be screened to avoid clogging the small nozzle or atomizer openings. Small quantities of liquids can be treated in desorbers by spraying the liquids on solid wastes during pretreatment or injecting them into the chamber while feeding solid wastes.

Sludges are usually fed into the DC or the incinerator PCC through water-cooled lances. In desorbers, the percentage of liquids in waste feeds must often be limited because some desorption chamber designs cannot accommodate sludges or slurries. In those cases in which limitation is required, sludges or slurries should be dewatered or mixed in small amounts with the solid wastes, then fed to the DC.

Solid wastes are fed into the DC or PCC. Containerized wastes can only be incinerated, and are typically fed through an airlock ram feeder into the PCC. But they can also be gravity fed through a chute into the PCC. Special solid wastes, such as debris or PCB capacitors, are usually shredded before they are fed into the PCC; they are not normally treated in desorbers. Bulk solid wastes, such as contaminated soil, are typically fed to the desorber or incinerator via vibratory or screw feeders or conveyor belts. If the solid wastes are fed to a fluidized bed incinerator or desorber, all solids generally must be screened, crushed, or shredded to less than 2 to 2.5 inches in diameter. Some desorber designs (e.g., 4-inch screw conveyors) can accept waste only 0.5 inches in diameter or even smaller. If the screw conveyor is 12 inches in diameter or larger, the materials can be 2 inches in diameter or greater (Kulweic 1985). In this regard, it is important to note that screening, crushing, pulverizing, or shredding wastes can increase the potential for fugitive emissions.

Flammable or ignitable wastes require special nonsparking equipment to prepare and feed the waste. If the waste feed or preparation areas are enclosed, fire codes could require an automatic fire suppression system in addition to good ventilation to prevent vapors from accumulating and approaching the lower explosive limit (LEL). Good ventilation is also necessary to prevent oxygen displacement and worker asphyxiation hazard-especially if the pretreatment area is enclosed and the wastes contain VOCs. The air from enclosed pretreatment areas should be vented to the thermal treatment chamber or a carbon filter system. To prevent volatilization of the VOCs, waste piles containing VOCs should be covered.

Waste analysis is another important part of waste preparation. Waste analysis ensures that waste does not contain any chemicals e.g., PCBs, dioxins, certain metals, etc., that should not be treated in a particular treatment facility. Analysis also provides information, e.g., pH, viscosity, water content, concentrations of various metals and certain organics, solids, compatibility with other wastes stored on site, etc., necessary to protect the equipment and to properly manage the waste. The parameters for analysis at each facility is determined on a site-specific basis, depending on the type of equipment in use and permit conditions.

5.1.2. Combustion and Desorption Chambers

Incinerators operating with PCB and hazardous waste permits typically have two combustion chambers. This is to meet the EPA regulatory requirement to destroy or remove 99.9999% of the PCBs and dioxins present in the wastes, and 99.99% of the other designated hazardous organic constituents. If sludges or solids are to be incinerated, the PCC might be a rotary kiln, fixed hearth, infrared, or fluidized bed, which can handle solids and sludges as well as liquid and gaseous wastes. If only liquid wastes are to be treated, a liquid injection incinerator is typically used. Most PCB and hazardous waste incinerators also have a fired secondary combustion chamber (SCC) or afterburner to provide the temperatures (1800F - 2200°F) and residence time (2 seconds) necessary to destroy effectively the wastes. Because of the high temperatures, incinerator combustion chambers are refractory lined. The lining can be firebrick, cast-in-place refractory, or a sprayed-on insulation. Under the CERCLA program, the chambers' design can depend on whether TSCA or RCRA regulations are ARARs for the site.

Desorbers are used to clean up contaminated soils, including PCBs, petroleum, hazardous waste, and other Superfund wastes. EPA has not specified performance standards for desorbers. If they treat wastes classified as hazardous, they are regulated under 40 CFR Part 264 Subpart X, Miscellaneous Units. Site-specific operating conditions are set for each desorption facility based on the contaminants present in the waste and the design of the unit (see Chapter 2 for further discussion). Currently, no fixed commercial desorber facilities treating hazardous wastes are operating under a RCRA permit. The most commonly used desorption chambers (DCs) are rotary dryers (rotary kilns), thermal screws, indirect calciners, and belt desorbers.

Typical desorber operating temperatures are 200°F - 1000°F (EPA 1993). They could, however, operate as high as 1500°F, depending on the contaminant, waste matrix, and desorber construction materials. The air pollution control equipment train immediately follows the DC. DCs are not usually refractory lined because the flame is usually not in contact with the metals or wastes in the chamber. If, however, the DC must be run at higher operating temperatures to decontaminate effectively the soil being treated, it could require refractory lining.

5.1.3. Air Pollution Control Equipment

The types of wastes to be treated will also dictate the type of APCE needed to protect the public from exposure to harmful stack emission concentrations. All incinerators treating wastes containing halogens, such as chlorinated solvents, PCBs, or metal halides, need APCE designed to remove or neutralize acid gases. Acid gases, such as hydrogen chloride (HCl) could be generated when these wastes are combusted. Also, desorbers sometimes operate at such high temperatures that acid gases are produced from decomposition of halogenated compounds.

Acid gases are removed by either wet or dry scrubbing systems.

A typical wet scrubbing system consists of a quench (to reduce rapidly the gas temperature, which reduces recombinant reactions that could otherwise form dioxins, furans, etc., and to remove some particulate and acid gases), a venturi scrubber (to control particulate and to remove acid gases), a packed bed or tray tower (to
remove acid gas and additional particulate), and a demister (to reduce the visibility of the vapor plume). Generally, for acid absorption and neutralization, caustic soda or lime is added to the water circulating through the venturi scrubber and packed bed or tray tower.

A typical dry scrubbing system uses a waste heat boiler after SCC or thermal oxidization to cool the off-gas to 400°F - 600°F before it flows into an absorber. Lime, in the form of calcium hydroxide slurry, is injected into the absorber as a finely atomized spray, in the 5% - 50% slurry to water range. The temperature of the gases exiting the absorber are normally maintained in the 250°F - 300°F range. The flue gases next flow into a particulate collection device, such as an electrostatic precipitator (ESP) or a baghouse (also called a fabric filter).

To detect organic breakthrough as quickly as possible, it is very important to have a hydrocarbon monitor after a carbon adsorption unit.

To remove the organic contaminants volatilized from the wastes, desorbers typically have a wet or dry scrubbing system for particulate removal, followed by a series of condensers, then activated carbon adsorption units. If the condensers are not operated at temperatures well below the lowest-boiling organic expected in the flue gas, the carbon adsorption units could quickly become saturated and ineffective. Also, carbon adsorption units do not efficiently remove very low-boiling VOCs. Another problem with these units is that channeling can occur, which allows some of the flue gas to pass straight through the unit without filtration.

In order to detect organic breakthrough as quickly as possible, a CEM to detect total hydrocarbons (THC) should be installed after a carbon adsorption unit. The facility should have procedures in place to change out the carbon units when breakthrough occurs. The CEM flue gas monitors, such as for carbon monoxide (CO), total hydrocarbon (THC), radioactivity, or monitors for specific extremely toxic chemicals, are typically installed in the stack after the APCE.

High efficiency particulate air (HEPA) filters are usually only used on TT systems treating radioactive or extremely toxic wastes. To avoid some of the problems with condensers, some of the newer desorption facilities are using catalytic oxidizers to destroy the organic compounds in the off-gases. Many TT facilities could be adding catalytic oxidizers, carbon injection, carbon adsorption beds-or a combination of these-to achieve, as the regulatory standards require, lower dioxin and furan emissions.

The final piece of equipment for all TT systems is a stack or exhaust pipe, which should be designed to meet or exceed the EPA Air Program's Good Engineering Practice (GEP) standards to disperse adequately the stack emissions and to minimize public exposure. Minimum GEP is defined as follows:

GEP = H + 1.5L H = height of a nearby structure, and L = the lesser dimension of the height or projected width of the nearby structure.

The maximum GEP stack height that EPA will credit when it conducts stack emissions modeling is defined as the greater of 65 meters or H + 1.5L. If the stack is taller than the maximum GEP stack height, the facility will have better dispersion of its plume; thus ATSDR should not discourage tall stacks or exhaust pipes. That said, however, this discussion must NOT be taken to imply that ATSDR supports "dilution is the solution to pollution!" ATSDR supports proper design and operation of all technologies to minimize all emissions. Nevertheless, because zero emissions cannot be achieved, good public health practice recommends the use of distance, such as a buffer zone or a sufficiently tall stack, to prevent harmful exposures.

5.1.4. Residuals Management

The types of residuals generated by a thermal treatment facility depend on the types of waste feeds and APCE. The quality of the residuals depends on the treatment temperature, moisture content, residence time, and whether additional chemicals are added to the treatment chamber or APCE. Under RCRA, any process residual from treating a listed hazardous waste is still a listed hazardous waste and must be handled accordingly.

5.1.4.1. Liquids

Employees at thermal treatment facilities generate liquid waste streams when they wash the equipment and the waste processing and unloading areas. If it is included in the APCE design, a wet scrubber generates a liquid waste. The facility can dispose of the waste waters by

• Injecting them into the incinerator combustion chamber; injecting them into a desorber is, however, generally not economical.
  
  
• Treating and discharging them through the sewer system to the publicly owned treatment works (POTW).
  
  
• Treating and discharging them to a nearby body of water, such as a stream, lake, or surface impoundment.
  
  
• Spraying them on the hot soil or ash to cool it as it is discharged from the PCC or DC.

A permit is required to discharge waste waters. Depending on the facility location and the constituents present in the waste waters, the local POTW might require the facility to pretreat the waste waters. When evaluating a facility that discharges to a POTW, health assessors should also evaluate the impact of the thermal treatment facility on the POTW's sludge-especially if the facility's waste water stream has significant concentrations of metals or organics. Contaminants of concern can concentrate in POTW sludge. Sludges from POTWs are often used as fertilizers by farmers and homeowners, and could therefore be a completed exposure pathway requiring evaluation.

EPA regulations require all except CERCLA facilities (see section 3.2.) to obtain a National Pollution Discharge Elimination System (NPDES) permit for all discharges into certain bodies of water. The permit specifies parameters such as pH, biochemical oxygen demand (BOD), and chemical concentrations that can be discharged, and, possibly, treatment prior to waste water discharge. Nevertheless, because not all toxic chemicals in hazardous waste are regulated under the NPDES program, the discharge could still contaminate the local surface water, sediments, and food chain pathways.

The waste water generated from facility and equipment washing could contain small concentrations of any of the hazardous substances present in the wastes being processed. Although the waste waters from the wet scrubbing system could contain higher concentrations of some of the metals than were in the original wastes, they usually contain low concentrations of only a few organic chemicals. Waste waters (also known as process waters) need to be analyzed before they are sprayed onto hot soil or ash. This will ensure no recontamination of the treated materials. In that regard, TSCA requires the process water to contain less than 3 parts per billion (ppb) PCBs.

Desorbers having condensers as a part of their APCE generate liquid waste containing the same organic chemicals desorbed from the waste and liquified in the condensers. Because organic contaminants will be concentrated in this waste stream, employees need training in how to manage this waste safely.

5.1.4.2. Ash

High-ash-content liquid wastes, sludges, or solid wastes fed to a thermal treatment facility generate bottom ash and fly ash. Bottom ash is generated in and discharged from the PCC or DC. If the unit is used to decontaminate soil, the bottom ash is usually referred to as the decontaminated soil rather than the ash. Fly ash is the solid particles entrained in the flue gas and metals, or the organics that are volatilized and later, when the flue gases are cooled, condense in the air pollution control equipment. When the flue gas in the unit is fast and turbulent, dry friable soils and small granular materials can cause an increased particulate loading on the APCE. The bulk of the fly ash is captured and removed by the APCE to keep the facility in compliance with its particulate emission limitation.

5.1.4.2.1. Bottom Ash - Decontaminated Soil

Bottom ash or decontaminated soil (if soil is being treated) discharged from the PCC or DC should contain only very low concentrations of organic contaminants. This ash or soil could be a fine material with metals concentrated in it that is easily wind-dispersed. Some facilities discharge the ash or soil into a water quench for cooling and to prevent it from blowing. Some facilities have a water spray directed on the ash or soil conveyor, while others have a shroud around the conveyor where water sprays onto the hot ash or soil. Sometimes, however, the bottom ash from a high-temperature incinerator is a molten slag material that does not contribute to fugitive emissions.

5.1.4.2.2. Fly Ash - Particulate Matter

Particles vary in diameter, but those 2.5 microns or less have the greatest respiratory effect on humans. But even particles up to 10 microns are respirable and thus can cause respiratory problems.

Thermal treatment facilities using a dry scrubber system, or that treat contaminated soil or high-solids waste, generate fly ash-also known as particulate matter (PM)-that is entrained in the flue gas. The fly ash can be removed from the flue gas by cyclones, scrubbers, bag houses, or ESPs. Some facilities discharge the fly ash removed by the APCE directly into drums or other containers for disposal. This is usually effective in preventing fugitive dust emissions. Some facilities move the fly ash and the bottom ash to a residuals management area where they are analyzed before being mixed or discarded. To prevent fugitive dust emissions, these conveyors should be enclosed or shrouded and kept moist or under negative pressure.

5.1.5. Other Design Features - Thermal Relief Ventsa

The activation of the TRV must be tied into the AWFCO system to prevent additional wastes from being fed to the unit while the TRV is open.

Some desorbers-and most incinerators that treat solid wastes, sludges, or thick liquids with a high solids content-have a thermal relief vent (TRV) immediately after the combustion or desorption chamber(s). The TRV is also known as an emergency relief vent or dump stack. The hot gases can be diverted through the TRV if downstream equipment malfunctions. TRVs are necessary on certain facility designs to prevent downstream equipment fires and to prevent hot combustion gases from venting at ground level-which would be a greater hazard to facility workers and nearby residents. A TRV is not necessary if the desorber operates at temperatures low enough for the APCE to withstand in an emergency situation. But because the emergency vent allows the flue gases to bypasses the APCE when the stack is opened, the public could be exposed to metals, particulates, acid gases, and possibly organic chemicals if the emergency relief vent is opened while waste is still in the combustion or desorption chamber(s).

Today, TRV openings are infrequent; however, prior to RCRA permitting, TRV openings occurred frequently. According to a recent EPA survey of RCRA permit writers, a facility might not have a TRV opening for several months, then could have an equipment malfunction and experience several openings in a short time. Thirty TRV openings a year were once common. The bottom line is that facilities should work toward minimizing TRV openings, or eliminating them altogether.

5.2. Emissions of Public Health Concern

The two categories of emissions of potential public health concern are stack emissions and fugitive emissions.

5.2.1. Stack Emissions

5.2.1.1. Organics

Although thermal treatment units can efficiently destroy or remove organic chemicals in wastes, they can also emit low concentrations of some organics. The organics in incinerator stack emissions are known as products of incomplete combustion (PICs). Various PIC definitions exist, but this document will use the term generally to refer to any organic compound found in incinerator stack emissions. This section discusses PICs found in most waste incinerator emissions. EPA has sampled and analyzed the stack emissions of a number of hazardous waste incinerators for PICs. Table 3 shows the emission rates of some chemicals EPA found in stack emissions of hazardous waste incinerators (EPA 1990). Table 4 is a compilation of chemicals that have been detected in full-scale incinerators and research reactors (Costner and Thornton 1990).

Thermal desorption stacks can also emit small concentrations of organic chemicals and volatile metals present in the wastes being treated. Partial degradation or pyrolytic breakdown byproducts (carbon monoxide and hydrocarbons) can also be emitted. It should be noted, however, that thermal desorber stack emissions have not been characterized as extensively as incinerator emissions.

5.2.1.2. Dioxins and Furans

Test data from some TT facilities show low concentrations of polychlorinated dibenzo dioxins (dioxins or PCDDs) and polychlorinated dibenzo furans (furans or PCDFs) in the stack emissions of PCB and RCRA incinerators and thermal desorption facilities. The test data suggest that incomplete destruction of organic material in the combustion zone and adsorption of this material on entrained fly ash increases the possibility of PCDD and PCDF formation in the APCE. This is particularly true if the gas temperature or downstream surfaces are in the 400°F-650°F (200°C-340°C) range (EPA 1989). PCDDs and PCDFs are formed where particulates are held up while in the 400°F-650°F temperature range, such as when thermal treatment units treating soils have (1) a bag house after the primary or desorption chamber; or (2) when heat transfer surfaces (such as boilers or heat exchangers) are present, thus allowing the deposition of particulates. PCDD and PCDF emissions can be reduced by a rapid quench and by reducing the dioxin or furan precursors. A rapid quench stops recombinant reactions which could generate dioxins and furans. The most common method of cooling the gases quickly is water quenching; injecting air is another common method. Research has shown that good operating practices can also minimize the formation of chlorinated dioxins and furans.

Good operating practices to minimize dioxin and furan formation include: Maintaining low carbon monoxide (less than 100 ppm), Maintaining low total hydrocarbon (less than 10 ppm) levels in the stack gas, and Quickly cooling post combustion/desorption gases to below 400°F (200°C).

The data EPA used to establish the maximum achievable control technology (MACT) standards for hazardous waste incinerators establish that when the particulate matter control device is operating at or below 400°F, dioxin and furan emissions are below 0.4 nanograms toxicity equivalency quotient per dry standard cubic meter (ng TEQ/dscm), unless the incinerator is equipped with a waste heat recovery boiler.

Chlorinated dioxins and furans can be controlled through using APCE for the removal of the dioxins and furans from flue gases. Facilities can use activated carbon injection, catalytic oxidizers (Maaskant 2001), catalytic membrane filter systems (Pranghofer 2001), or a wet carbon adsorption/oxidation scrubber system (Siret and Bessy 2001) to lower the dioxin and furan emissions below levels expected to cause adverse health effects.

Research on the formation of dioxins and furans has primarily been conducted on incinerators. Still, using the same good operating practices at desorbers helped reduce dioxin and furan formation at some desorption facilities.

5.2.1.3. Metals and Halogens

Concentrations of metals in stack gases can be affected by: Solids temperatures, Chlorine Volatility of metals, and Type of APCE.

If activated carbon is used for effective mercury, dioxin, and furan control, the flue gas temperature must be below 400°F. The temperature should also be maintained below 400°F to avoid carbon fires. Arsenic, beryllium, cadmium, and chromium are metals sometimes found in wastes and stack emissions that because of their carcinogenicity could be of health concern. Other metals possibly present are antimony, barium, lead, mercury, nickel, selenium, silver, and thallium. Because of its volatility, mercury is a particularly difficult metal to capture in the APCE. It is important to control mercury emissions because mercury bioaccumulates in several animal species. Injecting carbon into the APCE is effective in adsorbing mercury, dioxin, and furan.

Because metals are elements, they cannot be destroyed by incineration or any other treatment technology. They could therefore remain in the bottom ash, be carried into the APCE, and removed as fly ash or in the scrubber liquor, or be emitted in the stack gases. The concentration of metals present in stack gases can be affected by the following:

• Solids temperatures-as the temperature of solids increases in the primary or desorption chamber, the concentration of some metals (e.g., barium and silver) in the flue gases also increases (see Table 6).
  
  
• Chlorine-with a few exceptions, chlorine generally increases the volatility of metals (see Table 7).
  
  
• Volatility of metals-Table 7 contains the volatility temperatures of 12 metals with and without chlorine. These metals volatilize at normal incineration temperatures; however, a few might not volatilize in desorbers, depending on their operating temperature.
  
  
• Type of APCE-Table 8 shows conservative metals removal efficiencies of different types of APCE. Performance burns almost always demonstrate that a facility's APCE has better removal efficiencies than those shown in the table, so the values in Table 8 could be used to estimate potential health impacts when health assessors do not have facility stack emission data.

Little data are available for desorber stack emissions. Stack sampling and monitoring requirements for desorbers have not been as stringent as those for incinerators, except under the PCB program. EPA requires identical monitoring for PCB desorber and incinerators during performance tests.

5.2.2. Fugitive Emissions

The two primary sources of fugitive emissions at any thermal treatment facility are the waste processing/feed area and the residuals management area.

Fugitive emissions from the waste processing/feed area can be volatilized organics or contaminated particulate. Particulates contain organics and metals blown off waste piles or waste transfer equipment or emitted through cracks in conveyor systems and storage and waste processing buildings. Fugitive emissions from the waste processing/feed area can be effectively prevented by a number of methods or combinations of methods, such as (1) covering the waste with tarps or plastic, (2) spraying it with water, (3) spraying it with foam or other coating materials, or (4) unloading and processing the waste inside a building maintained under negative pressure. The air drawn from the building should be piped directly to the incinerator or desorber for disposal, or vented through a carbon filter system to remove the organics before the air is released to the atmosphere. While water spray can be a very common and cost effective way to control metals and particulate matter, it is not very effective for VOC control. Other disadvantages include the need to manage contaminated water runoff and process problems, particularly if the waste is too wet or there is a broad variation in the moisture content of the waste feed (EPA 1992). If bulk liquids are received and stored in tanks, the tanks should be vented through pipes to a carbon filter system or piped directly to the incinerator combustion chamber.

To ensure effective fugitive emissions management, ambient air monitoring is recommended at TT facilities. If the facility does not have fixed monitoring stations, hand-held monitors should be used to screen during spills or other releases and to periodically screen the area and potential release points, e.g., flanges, valves, etc.

Fugitive emissions from the solid residuals (fly ash, bottom ash, or decontaminated soil) handling area are generally fine particulate matter that can, if not managed properly, easily become airborne. At most facilities the fly ash and bottom ash are stored separately, at least until they are analyzed, but some CERCLA facilities combine them within the treatment system.

5.3. Design and Operating Considerations Important to Public Health

The following thermal treatment facility design considerations are important in minimizing or preventing public exposure (see also Table 9):

• Controlling fugitive emissions can be done by physical or mechanical means, operating conditions, or enclosures or buildings. A facility could also be designed with a large buffer zone so fugitive emissions do not migrate off site. In this latter case, health assessors should determine whether access to the site is restricted to prevent exposure of people and pets, or animals that could become a part of the food chain.
  
  
• Venting waste feed tanks to the combustion chamber(s) when the incinerator is operating or through a carbon filter system.
  
  
• Installing an AWFCO system connected to the key desorption, combustion, and APCE operating conditions to prevent continued operation under poor operating conditions that could cause excess stack emissions.
  
  
• Triggering the opening of a TRV-if one exists-but only under extremely critical operating conditions. If health assessors are not sufficiently experienced with the design of incinerators to evaluate what parameters trigger the opening of the TRV, they should contact ATSDR combustion specialists. They should also ask how often the TRV is used, how long it stays open, and if any stack or ambient monitoring has occurred during its use.
  
  
• If the facility generates solid residuals, designing and operating the ash handling equipment to prevent blowing fugitive particulates.
  
  
• Installing CEMs that monitor stack emissions. ATSDR recommends that all TT facilities have a total hydrocarbon (THC) monitor. But if a carbon monoxide (CO) monitor at a combustion facility has historically run well below 100 ppm on volume basis (ppmv), then an THC monitor might not be needed. Desorbers including condensers, carbon adsorption units or both need an THC monitor to monitor VOC emissions.
  
  
• Locating the facility and stack high enough in relation to the surrounding terrain and buildings to comply with GEP will provide good dispersion of the plume. This will also minimize the likelihood of a plume down draft or fumigation, and minimize exposure of the public and workers from the stack emissions.

The following operating limits are key to preventing poor combustion, desorption, or APCE operation (see section 6.1.1.4. for more on recommended operating conditions). As appropriate on a site-specific basis, these operating limits should trigger the facility's AWFCO system (see Table 10).

• Minimum temperature in the desorption and each combustion chamber, and in a catalytic oxidizer.
  
  
• Maximum pressure in the desorption or primary combustion chamber.
  
  
• Maximum feed rate of each waste type, i.e., liquids, sludges, solids, etc., to the desorption chamber and each combustion chamber or maximum feed rate for each feed location.
  
  
• Maximum chlorine and metals feed rates.
  
  
• Maximum size of batches or containerized wastes.
  
  
• Maximum carbon monoxide and/or hydrocarbon stack emissions.
  
  
• APCE inlet gas temperature.
  
  
• Critical APCE operating parameters specific for each type of equipment at the facility.
  
  
• Maximum flue gas flowrate or velocity.
  
  
• Maximum temperature in the desorber or each combustion chamber if metals or NOx are a health concern at the site.

To thoroughly evaluate a thermal treatment facility's potential impact on the community, a team approach is recommended. Team members familiar with the design and operation of thermal treatment technologies should evaluate the following:

• Design and operating conditions,
  
  
• Trial burn or performance test plan,
  
  
• Test report,
  
  
• Inspection schedules,
  
  
• Operator training, and
  
  
• Contingency plans.

A site visit is necessary to understand the configuration and operation of the facility, its relationship to the community, susceptible populations, and any topographic features in the area that may affect the dispersion of the facility emissions.

The stack and fugitive emissions should be modeled using an EPA-approved dispersion model and reviewed by a team member experienced in air modeling. A toxicologist, physician, industrial hygienist, or other health specialist should evaluate ambient air and stack emissions for potential for health effects. An industrial hygienist or environmental scientist should evaluate the ambient air sampling and monitoring plans for stack and fugitive emissions, action levels and associated response actions, health, safety, and contingency plans, and community demographics. Experienced health assessors could do several or all of these tasks. After the thermal treatment facility is constructed, the person reviewing the thermal treatment design-if not all the members of the site team-should tour the facility, the community, and the locations of any on- and off-site ambient air monitoring and sampling stations. See section 4.3 for further discussion on site visits. As discussed in Chapter 4, referencing anywhere in this document items such as modeling, ambient monitoring stations, risk assessments, etc., does not imply that every facility will or should have all of these items. This chapter discusses how to evaluate different items or information if they are available for the site being evaluated.

If ATSDR is involved early in the planning stage of the project, the following 3-phase review can be conducted: the pre-operational phase, the testing phase, and the operational phase. This chapter discusses the types of information normally available in each of these phases and how to evaluate the information. This chapter also explains terms and acronyms normally used in the thermal treatment industry. If ATSDR becomes involved late in the process, such as after the facility is operational, staff should still obtain and review the information discussed in each phase below.

6.1. Pre-operational Phase - Information to Review for Health Implications

The pre-operational phase covers design through the construction of the thermal treatment facility.

Although several stages of design drawings could be available, ATSDR might not have the resources to review and comment on each level of design before plans are finalized. This could be the case despite the fact that EPA and the facility might want the agency's input early in the process. Yet, if the agency becomes involved only after the facility is already constructed, public health officials should still at least review the final design or as built drawings and descriptions. Also, if EPA and the facility staff understand ATSDR's public health concerns as outlined in this document, they should be able to address all of the agency's technical concerns. Thus ATSDR's involvement even late in the process should not be too disruptive.

6.1.1. Design and Operating Considerations Pertinent to Protecting Public Health

6.1.1.1. Effectiveness of the Technology

If the technology does not effectively destroy or decontaminate the waste on the first pass through the unit, public exposure to contaminants could increase.

If the technology does not effectively decontaminate the solid waste on the first pass through the unit, worker exposure to contaminants could be increased. This is especially the case if workers handle the partially treated waste as if it is clean prior to receiving the treated waste analysis. Because of the greater potential for fugitive emissions, reprocessing and additional handling of partially treated solid wastes can also increase off-site exposure. Although the partially treated waste might not be an acute hazard, the cumulative dose should be considered. While the cost of waste treatment is not a major consideration for ATSDR, it could be more economical to run the unit a few degrees hotter, increase the solids retention time, or enhance the agitation or tumbling of the solid waste. Any or all of these measures could ensure that wastes do not have to be reprocessed, rather than attempting to run the unit at a lower temperature or shorter solids retention time, and then having to reprocess a substantial portion ( i.e., more than 10%) of the batches. The key to effective decontamination is sufficient solids time at the required temperature.

At a pre-operational facility, data from treatability tests and waste analysis, or data from previous sites where the unit was operated, will help address these issues. For example, has the unit (or a similarly designed unit) treated a similar waste? If so, what were the concentrations of the contaminants, the operating conditions, and performance test and environmental monitoring results? Do current plans indicate the design will remain the same and that the facility will operate under the same conditions? If the only data available are from treatability tests, do the plans indicate that the proposed operating conditions are at least as conservative as, if not more conservative than, the conditions demonstrated in a successful treatability test? Have problems identified during the treatability test been addressed?

To ensure effective treatment of the waste, a minimum solids retention time should be set in conjunction with a minimum flue gas exit temperature in the desorption or primary combustion chamber. Or, in the alternative, a minimum bottom ash discharge temperature should be specified. These parameters should be a part of the AWFCO system.

To ensure effective treatment of the flue gas, the state or EPA should establish operating conditions for the APCE and for monitoring of the stack gas. The APCE operating conditions should be specified for each air pollution control device (APCD) based on the operating conditions during a successfully completed performance test. See section 6.1.1.4. for recommendations on APCE operating conditions.

6.1.1.2. Fate of Contaminants of Health Concern

ATSDR staff should evaluate the potential for public exposure to both process residuals and effluents. Therefore, health assessors need to know the fate of any contaminants of concern that are either present in the waste or created by the treatment process. For example, VOCs in desorber stack gas could include vinyl chloride or trichloroethylene, depending on temperatures or whether base-catalyzed dechlorination is conducted simultaneously in the desorber.

The design estimate of the fate of the contaminants should be verified by analyzing all residuals during the performance test.

If metals are present in the wastes, where will they ultimately end up? Tables 6 and 7 show the boiling points of some of the metals found in hazardous wastes, as well as the affect chlorine (if present in the waste streams) could have on the metals. Table 8 shows the contaminant removal efficiency of different air pollution control systems. Using these tables and the proposed operating conditions, health assessors should be able to estimate whether the metals will be left in the treated waste, captured in the air pollution control system effluent(s), or exit through the stack. Mercury, for example, is a highly volatile metal that is difficult to capture in APCE, so the amount of mercury in the total waste feed to thermal treatment units should be limited, if it is present at all in wastes at the site.

Any halogens present in the wastes could be converted to acid gases (e.g., hydrogen chloride, hydrogen bromide) in incinerators or high temperature desorbers, but might or might not be decomposed in low-temperature thermal desorbers (LTTDs). Acid gases can be removed, neutralized, or both in the APCE (see section 5.1.3.). Acid gases in the concentrations typically found in uncontrolled flue gases before the APCE of incinerators burning halogenated compounds can cause acute respiratory problems and initiate asthma attacks (see sections 8.1.1. and 8.1.2.). Nevertheless, a well-designed and operated thermal treatment facility can easily remove acid gases and particulates.

The fate of organic chemicals differs for incinerators and for desorbers.

Incinerators can destroy or remove > 99.99% to > 99.9999% of the organic chemicals present in waste streams. The fate of organic chemicals in desorbers depends on the type of air pollution control equipment employed. Most desorbers are designed to vaporize (volatilize)--but not destroy--the organic chemicals present in the waste(s). The primary chamber is usually operated at low temperatures (i.e., relative to incinerators). Most desorbers use condensers to liquefy the organics volatilized in the primary chamber and remove them from the flue gas. The flue gas then passes through carbon filters to remove any remaining organics not removed by the condensers/chillers. Catalytic oxidizers can also be used to destroy the organics in the flue gases. Desorber stack emissions should be analyzed to determine whether decomposition products such as vinyl chloride are blowing through (not being adsorbed by) the carbon filter. Depending on the concentration of organic solvents present in the original waste(s), a sufficient quantity of organics can be removed in the condensers, making it possible to recycle the condensed solvent(s).

The operating conditions for the APCE for desorbers are extremely important for protecting the public. If not captured in the air pollution control system, the organic contaminants present in the waste or soil are transferred to the flue gases and emitted from the stack. Most desorbers have carbon filters as the last air pollution control device to remove any organics in the flue gas after the condenser. To ensure the organics are removed from the flue gases, ATSDR strongly recommends that all the stack effluent of all desorbers be continuously monitored for THCs. And if the THC reading exceeds 10 ppm (see the hazardous waste incinerator and industrial furnace standard [64 Federal Register [FR] 52869-52870]) an interlock system should shut off the waste feed to the desorber. On a site-specific basis, other CEMs could be used to monitor for hydrocarbon breakthrough. Health assessors should evaluate carefully the CEM's ability to detect small changes in hydrocarbon emissions.

An increase of THC in the stack gases indicates (1) organics are breaking through the desorber's carbon filter and it must be changed, (2) the catalytic oxidizer's catalyst has been sintered or deactivated (blinded or poisoned), or (3) poor combustion is occurring in an incinerator. The recommended 10 ppm THC limit is based on what EPA deems technically feasible for incinerators; it is not a health-based standard per se. According to the EPA preamble to the Final Standards for Hazardous Air Pollutants for Hazardous Waste Combustors (64 FR 52869), "More than 85 percent of test conditions in our data base have hydrocarbon levels below 10 ppmv, and nearly 75 percent have levels below 5 ppmv." Although desorbers are not likely to have been included in the EPA combustors data base, ATSDR recommends that any alternative thermal technology used in lieu of an incinerator be at least as protective of public health as would be an incinerator.

One of ATSDR's missions is to prevent or minimize harmful exposures. If, based on a site-specific toxicological and modeling review, a THC limit lower than 10 ppm is deemed necessary, ATSDR staff should discuss with the facility operators and EPA or the state regulatory agency the need for tighter operating controls. These tighter controls should result in reduced emissions, the setting of a lower THC emission rate, or the monitoring of stack gas for the specific organic contaminant to be limited. But monitoring the stack for specific organic chemicals will rarely be appropriate, unless an extremely toxic chemical is being treated, such as nerve agents. The toxicological review should be based on the organics and metals detected in fugitives and stack gas samples during the performance test, and any other compounds on the approved waste list that were not in the performance test waste feed or were not analyzed for during the performance test. Nondetect results should be included on a compound and site-specific basis (see ATSDR 1992 for further guidance on this issue). If a site-specific evaluation determines that the facility needs a higher than 10 ppm THC limit to be able to operate continuously, then a toxicological evaluation of the hydrocarbons being emitted during the performance test should be conducted.

6.1.1.3. Engineering Design Considerations Affecting Stack Emissions

The only way to prevent the release of stack emissions from any thermal treatment technology at levels that would be a public health hazard is to design properly the unit for the waste to be treated and to set operating controls. The control conditions should ensure that the facility is operated in the same manner or more conservatively than during the performance or risk burns or both that were passed (see section 6.1.1.5. for a discussion on performance test burns). Some facility operators might argue that operating controls are not necessary, and that their staff would not operate the facility improperly because staff members are concerned about their own safety. Other operators argue that if public health officials feel they need assurances, they should rely on the CEMs to ensure that the emissions are not a health hazard; CEMs measure what is actually being released out of the stack. ATSDR agrees that CEMs are an important part of monitoring and controlling the system. Nonetheless, problems arise when relying only on stack CEMs.

CEMs usually only measure indicators of stack emission quality. For example, opacity or PM monitors are indicators of respirable particulates and metal emissions, CO monitors are indicators of good combustion in incinerators, THC monitors are indicators of good combustion and low PICs in incinerators and effective operation of the air pollution control or treatment system in desorbers, and oxygen (O2) and stack gas flow meters are indicators of favorable combustion conditions. But none of these monitors measure specific chemicals or metals of health concern. A flame ionization detector (FID) THC monitor does not report the full mass of chlorinated compounds (e.g., a FID might report the mass of a highly chlorinated compound as a negative value) and thus could be a poor indicator of organic emissions at a particular site. Stack gas monitors might be available in the near future to measure specific metals. Some facilities can measure on a fairly frequent, but not continuous, basis a few organic chemicals in stack gases using a gas chromatograph with a mass spectrometer detector (GC/MS). High moisture, particulate loadings or both could render chemical-specific measurements in the stack gases impossible at some facilities. The point is, controlling the waste treatment facility using only CEMs is not likely to prevent harmful exposure from occurring--operating controls are also needed.

EPA regulations do not set specific standards for desorbers as they do for incinerators. Instead, EPA relies on the permit writer to set whatever conditions are necessary to protect human health and the environment. Even though EPA guidance recommends that incinerator standards be considered when permitting a desorber, many facility operators argue against having to analyze their stack emissions and also having CEMs.

Most health assessors lack the engineering expertise to review design calculations and detailed engineering specifications to ensure that a thermal treatment system can achieve the projected operating conditions and performance standards. Still, when reviewing the information available during the pre-operational phase, health assessors should look for several key design and operational features. The following sections discuss the major equipment subsystems and suggest specific items affecting emissions and, ultimately, the level of public exposure to site contaminants and byproducts.

6.1.1.3.1. Waste Feed Handling

Fugitive emissions from the waste feed handling area could be the source of the highest public exposure to contaminants.

The more homogeneous the waste feed material is, the easier it is to ensure consistent operation of the treatment unit as a whole and consistent levels of stack emissions. Consequently, facility operators often have a waste feed preparation area where front-end loaders or other equipment mix contaminated soils or other wastes with high organic, moisture, or metals content with less-contaminated wastes. The wastes could also be crushed or sorted by size; other debris might have to be removed for further pretreatment or treatment by another method. The solid waste materials to be thermally treated are then placed on conveyor(s) or in augers or ram feeders and fed into the PCC or DC. If the waste material is relatively dry and blows easily or if some of the contaminants are fairly volatile, fugitive emissions from the waste feed handling area could be the source of the highest public exposure to contaminants from the site.

Some sites conduct air monitoring with hand-held monitors to ensure their controls are effective and to alert staff to take corrective action if fugitive emissions exceed a predetermined level.

Liquid wastes should be stored in tanks equipped with agitators, mixers, or recirculating pumps to keep the liquid waste feed homogeneous. Open top tanks should not be used if wastes contain VOCs. Feed lines should have screens to remove objects that might clog injection nozzle(s). Leaks of hazardous wastes from pipes, pumps, and valves can also be sources of fugitive emissions.

If the community is close to the waste handling area or excavation areas at a CERCLA site, the operator could erect a waste feed handling building, tent, or wind screen to control airborne emissions. The operator could install water runoff control devices such as berms, fabric fencing, hay bales or all of these around the excavation area. In cold or wet climates, waste feed handling enclosures might also be needed to ensure a more consistent waste feed and economical operation. Feed conveyors should be enclosed in a building or by shrouds or other enclosures to minimize worker and community exposure to fugitive emissions. During site visits, health assessors should evaluate whether fugitive emissions from the waste excavation, handling, and storage areas are under control.

6.1.1.3.2. Combustion/Desorber Chambers

Numerous designs of thermal treatment chambers (e.g., rotary kilns, fixed or moving hearths, hollow augers, conveyor belts, cylinders) exist. Incinerators normally have two combustion chambers, referred to as primary and secondary combustion chambers (PCC and SCC). Desorbers usually have only one chamber, the desorption chamber. The function of the PCC and the desorption chamber is essentially to volatilize the contaminants in the waste. Agitation or tumbling of solid wastes can increase contaminant volatilization by exposing the wastes to the hot flue gasses or heat source. In an incinerator, the combustion or decomposition of the organic chemicals present in the waste also begins in the PCC. The amount of time a waste needs to remain in the desorber or PCC depends on the size of waste particles, the volatility and concentration of the contaminants, the operating temperature, and the effectiveness of the heat transfer. Typically, solid wastes remain in this chamber between 15 and 45 minutes. The treatment time and minimum bottom ash temperature actually needed to ensure the waste consistently meets the site-specific treatment standards are determined during the system testing phase (see discussion in sections 6.1.1.1. and 6.2.).

Desorbers--The externally heated carrier gas, air or an inert gas such as nitrogen, flows through the desorption system, to convey through the post-desorption treatment system the contaminants removed from the waste. The volume of carrier gas used in a desorption system is generally much less than the volume used in an incinerator. If air is the carrier gas in a desorber, to prevent an explosion from occurring, the operator must ensure sufficient dilution air is present to, in turn, ensure that 25% of the lower explosion limit (LEL) is not exceeded. The American Academy of Environmental Engineers also warns that, "In gas post treatment, a potential fire hazard exists in the baghouse if hydrocarbons or other combustible materials are allowed to collect on the filters. This presents a potential problem especially in the countercurrent rotary desorber configuration when used to treat material contaminated with heavier organics."(AAEE 1993). The carrier gas flow can be co-current (i.e., flow in the same direction as the waste), or counter-current (i.e., flow in the opposite direction). The carrier gas is also known as flue gas.

Incinerators--To supply oxygen for the flame in the burners and to support combustion of the VOCs, an incinerator uses air as the carrier gas. Some incinerators also add oxygen gas through the burner block to improve the oxidation of the volatilized organics while maintaining a lower air flow through the system.

In most thermal treatment systems, the carrier gas is drawn through the entire system by an induced draft fan between the APCE and the stack. The flow should be sufficient to create a negative pressure at the so-called face of the PCC/DC where the waste is fed into the system, or the system should be sealed--both to prevent the escape of the contaminants being volatilized in the chamber, and to carry them through the air pollution control equipment.

6.1.1.3.3. Treated Waste Handling

Because all the moisture and organic chemicals have been volatilized from the waste, the treated waste (bottom ash or decontaminated soil) exiting the PCC/DC is hot (typically 400°F-1000°F), and usually dusty. The bottom ash can be cooled by (1) dropping it into a water chamber from which it is usually removed via an ash drag type conveyor or (2) spraying it with water through nozzles in the shroud/conveyor cover as it is conveyed to the treated waste storage area. Some facilities recycle water by using the water from their wet scrubber to cool the bottom ash. If VOCs have condensed in the wet scrubber they can be re-volatilized when the scrubber water is sprayed on the hot treated waste. Because the conveyor shrouds/covers are not typically sealed, the VOCs can be emitted as fugitive emissions through cracks and openings in the conveyor system. For example, at one desorber facility this practice caused acute health problems for people off site (ATSDR 1997a). Also, an explosion can occur if the LEL is reached in the conveyor equipment. But this is only a problem with desorbers, not incinerators, because the VOCs are destroyed in the incinerator combustion chambers.

If the waste is not treated sufficiently, i.e., if the temperature is not hot enough in the PCC/DC to volatilize all the organic chemicals, the steam generated by the hot ash falling into the water chamber or by spraying water on the conveyor could strip additional organic chemicals out of the waste. Those fugitive organics will then be emitted with the steam, and could pose a hazard to workers or nearby residents. Because this can be a problem with both incinerators and desorbers, a minimum bottom ash exit temperature should be specified to ensure that the waste is properly decontaminated on the first pass through the treatment system.

6.1.1.3.4. Flue Gas Treatment/Air Pollution Control System

A major difference between desorbers and incinerators is that desorbers' flue gas contain all the organic chemicals volatilized from the waste.

Different types of equipment exist for treating flue gases. Incinerators almost always have a SCC, sometimes called a thermal oxidizer or afterburner, in which additional burners complete the combustion of the organic chemicals volatilized from the waste. The flue gases from the SCC and the desorption chamber contain entrained particulate matter and volatilized metals (if metals are present in the waste).

If halogens are in the waste, desorber flue gases might contain acid gases. Incinerator flue gases contain acid gases if halogens are present in the waste, except when a fluidized-bed incinerator is used. If halogens are present in the waste, caustic material is often used in the fluidized-bed material to neutralize the acid as soon as it is generated.

The flue gas treatment system is known as the air pollution control system. It should contain a series of APCE that will remove the contaminants to below levels of health concern before the gas is vented through the stack. Health assessors should review site documents and the information in Tables 6-8 to determine whether the facility has appropriate APCE to treat the flue gases. See also sections 5.1.3. and 6.1.1.4.

6.1.1.3.5. Process Monitoring Equipment

Process monitoring equipment examples: thermocouples, manometers, pressure drop indicators, flow rate meters, pH meters, continuous emission monitors [CEMs].

EPA or state regulatory staff should conduct an in-depth review of the monitoring equipment specifications to ensure that the equipment is appropriate for the operating conditions it is intended to monitor. The facility's inspection schedule should include calibration of the various monitors, with National Institute of Standards and Technology (NIST) traceable gases on an appropriate schedule. Facility operators also often install duplicate monitors on some of the key operating conditions (e.g. thermocouples).

All the key operating conditions should have monitoring equipment connected to a data logger that continuously logs those operating conditions.

Health assessors might not have the expertise to do an in-depth review; however, if they do, they should review the equipment descriptions for the items below.

• Does the operating condition to be monitored occur in the middle of the monitor's operating range? The monitor's range should not be too large or it will not be very accurate in the range the facility will be operating. Conversely, if the monitor has a narrow operating range, it will not register the true high or low value reached by the facility. This could cause a lower rolling average, and the number to recover or come back into range more quickly.
  
  
• When calculating so-called rolling averages, what number does the equipment use in the calculation when the monitor is pegged out at the bottom or top of the range? Typically, the monitor uses the value where the unit is pegged-out (the maximum or minimum reading available). The rolling average was introduced to not force an AWFCO for instantaneous transient spikes, consequently allowing for greater overall operation approaching a steady state.
  
  
• Does the equipment have a dual operating range that automatically expands if the readings exceed the normal operating range? A narrow range should only be used if the monitor has a dual operating range that automatically expands when the narrow range is exceeded. This is particularly important on CO monitors since CO concentrations are typically less then 10 ppm, but could spike into the thousands (2,000-5,000 ppm).

During the operating phase, health assessors should review the following items:

• A drift test and relative accuracy testing audit (RATA) should be passed before full production begins. These tests should be done during shakedown or the performance test.
  
  
• Monitoring equipment should be calibrated in accordance with manufacturer's instructions and a written calibration protocol. Calibration should occur at a minimum of two points that bracket the normal facility operating range; however, three-point calibrations are preferred (top, middle, and bottom of monitor's range).
  
  
• CEMs should be calibrated daily with National Institute of Standards and Technology traceable gases. Thermocouples, flowrate meters, pressure drop indicators, and other monitoring equipment that are either sealed or cannot readily be calibrated can be calibrated monthly, quarterly, or annually. During the operational phase, health assessors should verify that the equipment is installed, inspected, and calibrated on schedule.
  
  
• If a CEM that is tied to an AWFCO cannot be properly calibrated within 15-20 minutes, the waste feed should be shut off until the CEM calibration is completed.

6.1.1.3.6. Stack Height

The taller the stack, the better the plume dispersion.

Proper stack height lowers the potential for harmful exposure to the public. The stack should be taller than any nearby buildings, so the plume will not directly affect the building's windows or air intakes. Some facility operators prefer a short stack so their facility will not be noticed. But this potentially increases the exposure of on-site workers and nearby residents or businesses. The EPA Air Program requires that stacks meet GEP standards, thus ensuring good plume dispersion and minimizing plume downwash or fumigation of nearby buildings and personnel. See section 5.1.3. for further discussion of GEP and how to calculate GEP stack height. The EPA CERCLA program commented that it does not consider GEP an ARAR. If the stack height is below GEP, or site documents do not address this issue, health assessors should pay particular attention to the stack height, and the distance from and height of, nearby buildings. This data will assist them to evaluate the potential for downwash or fumigation of nearby buildings and personnel.

If the area has frequent inversions lasting several hours and the stack height is below the typical inversion height, the modeling of the stack emissions could indicate that administrative controls, such as limiting the operating hours, are necessary during prolonged inversions. This is not a recommendation that should be suggested unless there is good justification. Starting and stopping the operation can sometimes cause more emissions and equipment problems then steady state operation.

6.1.1.3.7. Handling of Process Residuals

Health assessors should observe the handling of process residuals (e.g., fly ash, condensate, scrubber water, spent carbon, spent filters, etc.) during site visits to determine whether fugitive emissions from these areas could impact the public. They should also investigate whether the ultimate disposal of the residuals and any off-site transportation of process residuals could result in additional public exposure. If the facility operator is required to reprocess a substantial portion of the batches, health assessors should note the potential exposure routes that will be impacted. See sections 5.1.4., 5.2.2., and 6.1.1.1. for further discussion.

6.1.1.4. Operation and Maintenance Plan (O&M Plan)

The O&M plan, or permit application, should specify the operating conditions for each piece of equipment at the facility. It should also state that, based on the operating conditions actually demonstrated during the fully successful test(s), the operating conditions will be modified after the performance test. Health officials need to understand that the facility and regulatory agencies must carefully set the range of operating conditions to assure safe operations consistent with the performance test conditions, and to allow some flexibility in operations so that AWFCOs do not cause frequent upset conditions.

Key conditions to be continuously monitored should be interlocked with the AWFCO system. Waste feed should automatically be cut off when the conditions starred (*) below are exceeded. The O&M plan should specify the following conditions to ensure effective treatment of the waste:

• Minimum PCC flue gas exit temperature for incinerators, and minimum bottom ash temperature for desorbers and incinerators treating contaminated soil.*
  
  
• Minimum SCC flue gas exit temperature for incinerators.*
  
  
• A measure of solids residence time in the PCC or desorption chamber such as kiln rotation speed.
  
  
• Maximum size of solid materials to be fed to the unit (usually 2 inches or less).
  
  
• Maximum waste feed rate for each waste type to each chamber.*
  
  
• Maximum viscosity of pumpable wastes.
  
  
• Maximum burner turndown.
  
  
• Minimum atomization fluid pressure.
  
  
• Maximum size and/or btu content, and frequency of addition of containerized waste.
  
  
• Maximum quantity of waste and minimum treatment cycle for batch operated facilities.

The plan should specify the following conditions for APCE, as applicable to the facility, to ensure the effective flue gas treatment:

• Maximum inlet temperature to APCE.
  
  
• Minimum quench flow rate.
  
  
• Minimum pressure differential across a venturi scrubber.*
  
  
• Minimum/maximum nozzle pressure to scrubber.
  
  
• Minimum pressure differential across a baghouse.*
  
  
• Maximum pressure differential across a baghouse to activate the bag cleaning system.
  
  
• Minimum liquid-to-gas ratio and pH to a wet scrubber.*
  
  
• Minimum caustic feed to dry scrubber.*
  
  
• Minimum kilovolt-amperes (kVA) settings for electrostatic precipitators (wet/dry).*
  
  
• Carbon injection rate* and location.
  
  
• Minimum kilovolt (kV) and minimum liquid flowrate to ionized wet scrubber (IWS).*
  
  
• Maximum inlet temperature to carbon bed or carbon injection system.
  
  
• Maximum exit temperature from condensers.*
  
  
• Maximum temperature to catalytic oxidizer.*

The plan should specify the following operating conditions to ensure that stack and fugitive emissions are maintained below levels of health concern:

• Maximum total halides and ash feed rate to the system.
  
  
• Maximum CO emissions or maximum total hydrocarbon emissions measured at the stack or other appropriate location.*
  
  
• Flue gas temperature to APCE below 400°F to prevent dioxin production.*
  
  
• Maximum flue gas flow rate or velocity measured at the stack or appropriate location.*
  
  
• Maximum pressure in the desorption chamber or PCC.*
  
  
• Maximum metals feed rate of each metal of concern.
  
  
• Maximum of 25% of LEL in desorber gases using a combustible gas monitor in the stack or APCE (if air is the carrier gas).*
  
  
• Maximum gas exit temperature from PCC and desorption chamber, if the wastes contain metals that could be volatilized at temperatures higher than those used during the performance test.*
  
  
• Maximum air monitoring values allowable at facility fence line. More than one number should be set, i.e., one set of numbers for particulate matter and contaminants of concern (or VOCs) that trigger on-site measures to control fugitive emissions, and a higher number for each to trigger shut down of excavation, facility operations, or both. See section 6.1.2.1. for further discussion on ambient air sampling and monitoring.

6.1.1.5. Performance Test Plans

Performance tests could be labeled compliance tests, MACT tests, or trial burns for RCRA facilities. At CERCLA facilities they are also known as trial burns, performance tests, or proof of process (POP) tests. In this document the term "performance test" refers to any of these tests. The operating conditions during the performance test are used to set the facility operating conditions. When setting the operating conditions one should allow maximum operating range consistent with the performance test operating conditions. If the operating range is narrow, the frequency of AWFCOs will increase. This could cause unstable operations and increase the potential for PIC formation.

If a risk assessment is conducted, the thermal treatment facility could also do a separate risk burn. The risk burn measures the stack emissions during normal operation of the facility. By contrast, the performance test emissions should be the worst stack emissions that could occur during routine fluctuations in operating conditions of the thermal treatment system. Rather than doing two separate tests, some facility operators combine the risk burn with the performance test.

Performance tests should be conducted at worst-case operating conditions, i.e., at the minimum temperatures in the desorption and combustion chambers for organic chemicals, but at the maximum temperatures during normal operations for metals.

To provide the facility maximum operating flexibility, performance tests for thermal treatment facilities are usually conducted under worst-case operating conditions. At some sites, however, the operator's contract could require the performance test to be run at the facility's proposed operating conditions (normal conditions rather than worst-case). Because the operating conditions are set from the performance test, the performance test is the de facto worst-case operating condition. The stack emissions during normal operation should not be worse than those measured during the test because the AWFCOs are set at the operating conditions used during the performance test.

Worst-case operating conditions for organic chemical emissions include maximum feed and flue gas flow rate; minimum temperature in the desorption and combustion chambers; maximum halides, metals, ash, and water feed rates, as well as the minimum or maximum (as appropriate) operating conditions for all the starred operating conditions in the previous section. Unless the operator is willing to accept a narrow range of operating temperatures, it is impossible to do worst-case operating conditions for organic chemicals and metals at the same time. As a result, if metals are a health concern at that site, two performance tests are usually conducted. Worst-case operating conditions for metals emissions include maximum feed rate, maximum halides and metals feed rates, maximum PCC or desorber exit gas temperatures, but at the same minimum or maximum APCE operating conditions.

RCRA stack sampling requirements differ for incinerators and for desorbers. EPA does not have specific regulations for desorbers. RCRA permit writers usually use the incinerator regulations as guidelines for desorbers. The desorber performance test plan should include stack emissions sampling and analysis (S&A) for destruction and removal efficiency (DRE), PICs, TICs, and metals (if they are a health concern at the site). Some desorber operators might object to the DRE and PIC requirements, arguing that the standards are not applicable because they are not combusting or destroying the waste. Nevertheless, at the temperatures used in desorbers some decomposition of organics will occur. Because the condensers or carbon units-or both-might not capture all the organics in the flue gas, the identity and concentration of the organics in the stack emissions must be evaluated. The DRE test primarily measures the removal efficiency of the desorption facility, and can thus be called the removal efficiency test. To provide equivalent protection of the public, the removal efficiency for a RCRA desorber should be numerically the same as the RCRA incinerator DRE, i.e., 99.99% for the POHCs in the waste and 99.9999% for wastes classified as dioxin, furan, or PCB wastes (if over 50 ppm PCBs).

At CERCLA sites, site-specific concentrations are established for stack emissions and site clean-up concentrations, and are based on ARARs or a risk assessment (see section 3.2 or Appendix E for additional discussion of ARARs). If the stack emissions and clean-up criteria are protective of public health, health assessors do not need to consider removal efficiencies for CERCLA incinerators or desorbers. The stack emissions, however, do need to be characterized during the performance test to make sure they meet the approved site-specific values.

To ensure normal-to-worst case emissions during desorber testing, if the facility has a carbon adsorption bed, the performance tests should be conducted during the middle-to-end of the carbon adsorption units change-out cycle rather than immediately after changing the carbon bed. For more information on carbon adsorption systems see section 5.1 of EPA 1992. At CERCLA sites, EPA often requires the performance test to be conducted as soon as possible. If the unit has few operating problems during shakedown and the site-specific carbon bed life expectancy is calculated to be many months or years, the performance test should not be delayed solely to accommodate this recommendation. Condensers should operate at the maximum temperature allowed during normal operations.

Incinerators do not normally have carbon adsorption beds, but can have carbon injection systems to control dioxin, furan, and mercury emissions. Carbon injection systems should be operated at the minimum carbon injection feed rate for worst case operating conditions. For further information on carbon injection systems see the Federal Register for September 30, 1999, Part 63, Subpart EEE and the preamble thereto.

To show compliance with RCRA incinerator emissions regulations under their preferred operating conditions, operators often spike the waste feed with organic or metal compounds or both. The organic spikes are selected from lists of difficult-to-destroy compounds and are called POHCs, or target compounds. If POHCs or metals are spiked into the waste feed to demonstrate worst-case feed rates and compliance with the performance standards, the spiked waste feed should be fed to the thermal treatment unit before stack sampling begins. The minimum amount of time in advance of stack sampling that the spiked waste feed should be fed equals the solids retention time. Thus when sampling begins, the thermal treatment unit is operating at steady conditions.

Each performance test should consist of three runs under the same operating conditions. Some facilities do an additional run so they have backup data in the event some of the samples are lost or broken. Performance test plans should cover the following topics:

• Operating conditions for each piece of equipment,
  
  
• Waste feed sampling and analysis (S&A),
  
  
• Stack emissions S&A - POHCs/target compounds, PICs, products of contaminant degradation or reformation (i.e., organics not removed in the APCE), TICs, and metals,
  
  
• CEMs,
  
  
• Residuals (including treated waste) S&A - to characterize them and to identify and quantify the fate of contaminants,
  
  
• Treated waste S&A at CERCLA sites - to verify decontamination, and
  
  
• Quality assurance and quality control (QA/QC) - chain of custody for samples, field and method blanks, and laboratory spikes.

If EPA approved stack sampling, analytical, and QA/QC procedures are followed, the data can be assumed to be adequate for making public health determinations. That is, unless the QC data indicate the method was not sensitive or accurate enough for determining whether public exposure will be potentially harmful (i.e., below levels of health concern). EPA approved methods are in the EPA Publication SW-846, Test Methods for Evaluating Solid Wastes, Physical/Chemical Methods and the EPA Air Program regulations in 40 CFR Part 60 (see "http://www.epa.gov.") Other sampling and analytical methods or modifications of the EPA QA/QC procedures could provide data of adequate quality for making health determinations, but they must be evaluated based on the available QC data (i.e., demonstrated sensitivity and accuracy).

6.1.2. Additional Considerations Important to Public Health

ATSDR staff should also consider the items addressed below.

6.1.2.1. Ambient Air Sampling and Monitoring Plan

Health assessors should recommend that thermal treatment facilities have an ambient air sampling and monitoring plan. This is especially true if the stack data or risk assessment indicates the potential for air releases of contaminants at concentrations that could cause adverse health effects, and if the facility is allowed to operate at the conditions used during the stack testing. Table 11 summarizes the issues the plans should address. In addition, if public is concerned, health assessors should consider recommending that the facility or EPA conduct ambient air monitoring and sampling to provide data that will document the level of public exposure, despite the fact that in theory, the contaminants are not likely to be at concentrations expected to cause adverse health effects. Nevertheless, the health assessor should be aware that the chance of measuring the ambient air concentration predicted by modeling is extremely remote. Therefore, the need for ambient air sampling, monitoring, or both should be determined on a site-specific basis. Usually, EPA or the state regulatory agency will not allow continued operation at conditions likely to generate stack emissions at levels of health concern. But fugitive emissions might not be so easy to control. Health assessors should know that under RCRA, ambient air sampling can only be required on a site-specific basis through the omnibus provision, so the sampling must be deemed necessary to ensure protection of human health or the environment.

Ambient air monitors are direct reading instruments that continuously, or frequently (<15-minutes turnaround time), sample the ambient air and provide real-time data on levels of airborne contaminants (specific compounds or groups of compounds). Monitors should be located at the fence line of the facility, and where air dispersion modeling indicates that people could be exposed to fugitive or stack emissions from the thermal treatment facility at potentially harmful levels. Fence-line monitoring stations should be in the predominant wind direction(s) near where fugitive emissions are likely to be generated and detected, as well as upwind from the facility. Examples of air monitors include photo ionization detectors (PIDs), flame ionization detectors (FIDs), oxygen meters, combustible gas monitors, colorimetric instruments, remote optical sensors (ATSDR/EPA 1997) and gas chromatographs.

Ambient air monitors continuously sample the ambient air and provide real-time data. Ambient air samples are time-integrated samples that are analyzed in a laboratory at the end of the sampling period.

If the site is large, such as a military base, the monitors should be at the borders or fence line of the thermal treatment site, not the fence line of the base. Modeling should be done to determine other locations where ambient air monitors or sampling could be needed, such as on-site housing, schools, day-care centers, and hospitals, as well as off-site locations where susceptible populations and the maximum-exposed receptor are located. For worker protection, personal monitoring and on-site ambient monitoring should also be conducted in the excavation and waste handling areas.

Health assessors are cautioned against making broad statements about public exposure based solely on ambient air monitoring or sampling data. The data from one station is only relevant to residents in that area-not necessarily the whole community. Fugitive emissions are usually carried by ground level wind currents. Thus stations between the fugitive source and the community are more likely to detect fugitive than stack emissions. A number of factors influence where point source emissions, such as stack effluents, are likely to have an impact at any given time. It is also important to remember that ambient air monitoring provides only an indication of potential deposition of the monitored substances that could enter the food chain pathway. If the pathway analysis indicates potential harmful exposures through the food chain pathway, then biota sampling could also be needed.

One to four chemicals should be chosen as indicators of exposure to site contaminants. Following are criteria for the selection of such compounds:

• Method available. Chemicals that can be easily and quickly quantified, preferably on a continuous basis, or at least a frequent basis (not greater than 15 minutes between sampling and report of results).
  
  
• Detectable concentrations. Chemicals present in highest concentrations on site or in sufficient concentration to be detectable if migrating off site as a fugitive or stack emission.
  
  
• Unique. Not likely to also be present in the air from other sources, e.g., traffic, industry, agricultural usage, etc.
  
  
• Representative of the various classes of chemicals present.
  
  
• Toxicity. Present on site in concentrations that could cause acute or chronic health effects.

Action levels should be specified that will trigger specific actions to reduce emissions.

The contingency or site-safety plan(s) should specify "on-site action levels"or "fence-line action levels" or both, as appropriate, depending on the toxicity of the contaminants and the distance to the nearest off-site population. Action levels are the concentration of the monitored chemicals that will trigger specific actions by facility staff to reduce emissions. Usually, several action levels will be set. For example, when the lowest on-site threshold is exceeded in the work or excavation area, hand-held monitors should be used to try to identify the source of the release. Actions can then be initiated to reduce emissions, to signal an increased level of protective equipment being worn by on-site personnel, or both. If concentrations continue to increase and exceed the second on-site action level or exceed a fence-line action level, work can be stopped, waste materials covered or wetted down, or other appropriate actions initiated depending on the source of the release. If the fence-line monitors then exceed a third specified concentration (the second fence-line action level), the thermal treatment unit can be shut down and monitoring and sampling in the community can be initiated.

Action levels should be set based on air monitoring data rather than air sampling data.

If contaminants are present on site in concentrations that could foreseeably cause acute health effects off site, the community first-responders should be notified to stand by in case the evacuation or shelter-in-place action level is reached. Some sites have fewer action levels; they can choose to directly shut down the facility, determine the source of the release(s), and fix the problem. If there are multiple contractors working on the site, health assessors should verify whether everyone is using the same action levels-not different contractors using different criteria.

In addition to ambient air monitoring, time-weighted ambient air samples (8-hour, 24-hour, etc.) should be taken. Air sampling consists of analytical techniques requiring either on- or off-site laboratory analysis. Components commonly used in sampling and analysis include filters, tubes, cartridges, impingers, bubblers, badges, bags, and canisters (ATSDR/EPA 1997). In addition to fence-line ambient air monitoring, ambient air samples should be taken in the community at locations likely to be impacted by fugitive or stack emissions as shown by air dispersion modeling. Air sampling and air monitoring are often used interchangeably, but they are not the same. Air monitors are direct reading instruments providing analytical data within seconds or minutes. Perimeter or fence-line sampling is conducted to verify the monitoring results. Air samples must be analyzed in a laboratory. If the samples are not fed directly into an on-site gas chromatograph (GC), it could be several days or weeks before the analytical data are returned. Therefore, action levels should be set based on air monitoring data rather than air sampling data. Ambient air samplers take samples periodically. The samples might or might not be composited over several days or weeks, depending on the toxicity and reactivity of the chemical, the quantity of sample necessary to achieve a detection limit at or below the level of health concern, and other factors. For example, a sample might be collected and composited every 15 minutes for 24 hours and sampling repeated every 5th day, or a sample could be collected every half-hour for 3 days. Local meteorologic conditions should be recorded during each sampling period.

If continuous air monitoring is conducted at a site, it should occur whenever there is activity on site (e.g., excavation or facility operation).

Air monitoring and air sampling programs should begin prior to initiating site operations to establish baseline chemical levels in the community. These samples can be used to look at the background exposure of the
community and to evaluate the impact of the facility on the overall contaminant level in the air, as well as the community's exposure. Air monitoring and air sampling should occur when the facility is operating to the
maximum extent possible. If excavation or treatment unit operations occur only from 7 am to 5 pm, doing a
24-hour composite sample will low-bias the sample because the samples of presumably clean air taken between 5 pm and 7 am will dilute the presumably contaminated samples taken during operating hours.

ATSDR recommends that perimeter and community air sampling occur daily during shakedown, the performance test, and the first weeks of full operation, i.e., any time the thermal treatment facility is processing hazardous wastes from startup until the data is reviewed by EPA and ATSDR staff. When the contaminant air concentrations are consistently well below levels of health concern, sampling frequency can be reduced to once every 3 to 6 days, or collected for a 3-7 day period.

In summary, the ambient air sampling and monitoring plan should provide:

• Data necessary to access episodic and chronic exposures,
  
  
• Assurance of worker protection,
  
  
• On-site action levels and response actions,
  
  
• Fence-line action levels and response actions,
  
  
• Community action levels and response actions,
  
  
• S&A for contaminants of health concern,
  
  
• Sampling to verify real-time monitoring results, and
  
  
• Sampling during facility operations.

6.1.2.2. Contingency and Site Safety Plans

EPA requires that every RCRA and CERCLA site have a site safety plan containing an on- and off-site contingency plan, which should be a part of the health and safety plan (H&S plan). In addition to the action levels and on- and off-site responses discussed in section 6.1.2.1., RCRA requires the contingency plan to describe the safety equipment and its location. It should describe emergency situations such as chemical spill, fire, explosion, and flooding (if in flood zone), and emergency procedures the facility staff will carry out to prevent or minimize exposure or danger to the community and on-site workers. The plan(s) should specify who will respond and list the names and phone numbers of the site safety officer(s), hospital(s), and emergency personnel that have agreed to provide additional services, such as the fire department, Chemtrec, or Poison Control Center. The plan should contain or give the on-site location of the Material Safety Data Sheets (MSDSs) for all the hazardous chemicals present on site.

Only one overall health and safety plan should exist, even at sites with multiple contractors.

If several contractors are working on site, they might have their own site safety plan. All site safety and contingency plans should be reviewed to ensure they agree on action levels and response actions, and that all important items are included. Sites with multiple contractors should have one overall H&S plan that includes all contractor specific duties, safety items, and monitoring requirements. The contingency plan often addresses only on-site activities. The H&S plan should also address off-site action levels and activities, and it should integrate with the local emergency response plan.

If the ambient air monitoring and sampling plan is not a part of the H&S plan, health assessors should make sure that action levels and response actions specified in each plan are consistent with each other and are logical.

6.1.2.3. Maintaining Good Performance

The key to maintaining good performance is having well-trained, experienced, skilled, dedicated, and conscientious employees. Routine operator and facility staff training is mandatory at RCRA and CERCLA facilities. Because thermal treatment units are only temporarily at CERCLA sites, the contractor typically brings only a few key personnel to the site and locally hires temporary workers. While these temporary employees can be trained on safety and OSHA requirements and can learn to do their specific jobs, they will often lack experience. That means they will need close supervision by technical staff who possess the appropriate educational background, experience, and training.

Even with all the proper design features, skilled operators are essential for a safe, effective treatment program.

Operators should understand the principles of good combustion and desorption and be thoroughly familiar with all major and support systems at their plants. Careful attention to proper waste treatment rates and waste blending, as needed, helps to ensure that the systems are not overloaded and that the AWFCOs are not activated excessively. Routine maintenance, inspection, and instrument calibrations should be conducted and recorded. Safety and emergency response plans that thoroughly address likely failure scenarios (including loss of power and operational failures) should be in place, documented, and shared with local officials. Emergency release drills should be conducted periodically with the knowledge and involvement of local emergency response personnel. Additionally, all employees should be adequately trained in appropriate health and safety procedures for the safe day-to-day operation of the facility.

While ATSDR recognizes that having well-trained staff who are vigilant in their inspection and maintenance of the plant is critical for the proper performance of thermal treatment facilities and the protection of the public, most health assessors are not trained to review critically the adequacy of training plans and inspection schedules. Health assessors should at least discuss the plans with EPA or the regulatory agency to determine whether these program areas are addressed by the site and that knowledgeable staff have reviewed and approved the plans.

Facilities should have detailed checklists to ensure that facility staff are properly inspecting and maintaining the facility.

Thermal treatment facilities are complex industrial plants requiring constant inspection, maintenance, and adjustment. To ensure that employees check all the necessary monitors, gauges, valves, or equipment, RCRA requires approval of inspection schedules and preparation of detailed checklists. Documentation of inspections, including problems found and repaired, must be kept at the RCRA facility. CERCLA facilities should also have detailed and comprehensive checklists to ensure that facility staff are properly inspecting and maintaining the facility.

To ensure that the system operates in a manner consistent with the operating conditions specified in the RCRA permit or CERCLA documents, ATSDR recommends that EPA or state environmental staff conduct frequent, random, unannounced inspections and make the results available to the public. Under some circumstances, a few state environmental programs have assigned permanent on-site inspectors at thermal treatment facilities.

Each time a CERCLA thermal treatment system is relocated, it should be retested.

Another way to ensure continued satisfactory operation is to retest periodically the thermal treatment system. This would be appropriate for all RCRA and CERCLA facilities if they operate at the site for an extended period of time, if other conditions indicate it might not be operating properly, or if the waste feed changes. RCRA regulations allow for a less rigorous performance test if the unit or a similarly designed incinerator successfully passed a performance test on similar wastes at another site.

6.1.2.4. Transportation of Wastes

If the facility treats wastes transported from other sites, the transportation aspects should be carefully considered. Routes of access should be selected to minimize accident (release) potential, and, if possible, avoid narrow or winding roads and residential, school, and play areas. For remediation of Superfund sites for which no over-the-road hauling is required, care is still needed to avoid spills, blowing soil, or other types of releases when transporting the wastes on site.

When waste is hauled off site, trucks should be decontaminated before leaving the hazardous waste site. The Department of Transportation regulations require hazardous wastes to be hauled in containers or tightly covered and properly placarded trucks to prevent blowing during transit. Residuals or treated soils being transported off site should also be transported in a manner to minimize any fugitive emissions during transit. Even though the residuals should not contain contaminants at levels of health concern, excessive fugitive particulate should be controlled during transportation.

6.1.2.5. Location of the Unit

Another consideration relevant to public health is the location of the thermal treatment facility with respect to the community. Community members want to know the possible health effects associated with living or working in the path of stack or fugitive emissions fallout. To address those concerns, when reviewing the location of a treatment facility with stack and potential fugitive emissions, regulatory agencies combine air dispersion models and local meteorologic data to determine conditions necessary to protect human health and the environment. Such modeling results are helpful in identifying prevailing wind transport patterns and deposition.

Ideally, the facility should not be located where modeling indicates that ground-level concentrations of stack or fugitive emissions will potentially cause harmful public exposures. Dispersion models can also help evaluate the need for, and the location of, off-site air monitors used to detect fugitive emissions associated with excavation, waste handling and storage facilities, process equipment, and residuals management areas. If concern arises about the impact of the facility on a specific food source, such as a dairy, farm, fish hatchery, etc. and ATSDR has data regarding the uptake of the contaminants of concern by the particular food chain species, dispersion modeling can be used to estimate the deposition of contaminants that would be available for food chain uptake.

Little flexibility exists in selecting a site for a Superfund treatment facility, except with regard to where it is placed within the boundaries of the actual site. When EPA is conducting the feasibility study, if the site is surrounded by residential areas, modeling should be used to determine whether thermal treatment is a preferable technology for cleanup of that particular site. Health assessors should tour the area to determine the land use and get an idea of the demographic structure of any nearby residential areas and areas that modeling has indicated are likely to be impacted by the facility. A detailed demographic analysis can be obtained from the ATSDR GIS Activity Group in the Program Evaluation, Records, and Information Services Branch.

6.1.2.6. Community Involvement

This section is not intended to prescribe that all the communications styles discussed should be used in every community where a thermal treatment facility is to be located. Different people learn new information in different ways, so a variety of media and communication styles should be used in each community to meet the needs of the various groups within the community.

ATSDR recommends that information and data concerning a thermal treatment facility's design, testing, operation, and monitoring be made public.

Technical staff from either the regulatory agencies, the facility, the public health agencies or all of them should make a special effort to explain facility information in plain language to which a layperson can understand and relate. Explanations of the facility's design and safety features, the roles of different equipment in assuring emissions will be controlled, and how the facility will operate, be tested, etc., should be provided to the public during the planning stage, i.e., before the facility is built. After the facility is built, but before any waste is fed to it (to avoid the risk of public exposure), local officials and the public should be allowed to tour the facility's equipment and control room, and be provided an explanation of the design and operation of the facility. Afterwards, the facility operator might want periodically to hold open houses.

Agency staff and site personnel should make a special effort to have open discussions with the public in small groups, or one-on-one in public availability type meetings so they can get to know each other and possibly build public confidence and trust in the staff and in the safety of the facility. Information about all the data that goes into the air modeling, such as 5 years of meteorological data, topography, and land use should be explained to the public, so they will understand how modeling can be used to project potential public exposure.

It is human nature to fear the unknown.

Several sites have set up advisory committees comprised of representatives from a number of local organizations as another way to disseminate information to the community and to allow the community to bring its concerns to the facility staff and regulators.

6.2. Testing Phase - Operating Conditions to Protect Public Health

The testing phase starts at the completion of construction and continues until the unit is approved to begin full operation.

The testing phase covers, (1) so-called shaking down or testing the equipment to make sure everything is working properly, (2) the performance test, and, (3) the post test, pre-operational period when the data is being reviewed, and the final operating conditions are being set.

6.2.1. Pre-Performance Test Period - Equipment Shakedown Phase

Initially, the treatment system will be heated to ensure that all the mechanical systems, monitoring and control equipment, and automatic waste feed shutoff systems are functioning properly. After the entire thermal treatment system has been checked and determined to be operating properly, the facility operator will then process clean materials or soils to further test all the systems. After clean materials have been successfully processed, EPA or the state will allow contaminated waste to be processed at reduced feed rates and conservative operating conditions to test the system further to ensure everything is working properly and to prepare for the trial burn. For RCRA incinerators, the pre-performance test burn operating period is usually limited to a maximum of 720 hours of operating time actually treating contaminated materials. During the pretest period the operator should be allowed a limited number of hours (perhaps 40 hours) to operate at the maximum feed rate and worst-case operating conditions to ensure that all systems will function properly during the performance test. Otherwise, conservative operating conditions should be set for all desorber and incinerator operating periods prior to the review of the performance test data and setting of the full operating conditions based on that data. The EPA requirements for desorbers are not as extensive as for incinerators, but EPA usually applies the same standards to both types of facilities.

6.2.1.1. Conservative Operating Conditions

The waste feed rates and the operating temperatures should be set conservatively.

During the pretest period, when equipment malfunctions are more likely to occur, the waste feed rates and the operating temperatures should be set conservatively, e.g., 50% -60% feedrate, temperature 100°F - 500°F hotter, etc. to minimize the potential for harmful public exposure. As operating experience is gained and the unit is fully tested, waste feed rates should be gradually increased until feed rates are around 75% - 80% of maximum to be demonstrated during the performance test. These rates should be exceeded only for the limited time period when the facility is allowed to operate at maximum test conditions. But, at no time should the facility be allowed to exceed the manufacturer's specifications. The temperatures in the PCC, SCC, and DC should be maintained 10% -20% higher than proposed for normal operating conditions. Condenser and chiller temperatures on desorbers should be maintained at lower temperatures than proposed full operation temperatures to help ensure that any extra loading of organics which could occur due to process upsets during start-up are condensed.

6.2.1.2. Site-specific Needs Considerations

Health assessors should evaluate the site-specific need for increased stack monitoring during the pre-performance test period. If the thermal treatment unit was approved to have neither CO nor THC CEMs during full operation, ATSDR strongly recommends that all units have one or both CEMs monitoring the stack emissions during the performance and risk burn tests. The CO and THC monitors will help ensure that good combustion or adsorption of desorbed organics occurs. Depending on the type of desorber unit, a CO monitor might not be appropriate. CO monitors are almost always required by EPA on incinerators. A THC monitor or some other CEM should monitor the desorber stack emissions to ensure that the volatilized organics are condensed, destroyed, or adsorbed.

If EPA does not require the unit to have a sulfur dioxide (SO₂) monitor, but the wastes contain sulfur in sufficient concentrations to theoretically generate SO₂ at levels of health concern if APCE malfunctioned, then health assessors should consider recommending a SO₂ monitor, at least during the testing phase. Factors to be considered when evaluating the need for increased stack monitoring during this period include:

• Proximity of the community to the unit,
  
  
• Past performance and operating experience of the thermal treatment unit company,
  
  
• Level of protective equipment routinely worn by on-site personnel, and
  
  
• Presence of and type of ambient air monitors in continuous operation on site.

6.2.1.3. On-site and Community Ambient Air Monitoring and Sampling

If the facility has an ambient air monitoring program, health assessors should determine whether the frequency should be increased during operational changes such as when starting up, increasing the feed rate, processing different wastes, or during the performance test, to determine if worker or community exposure rates are affected by the process change.

Consider the need to sample for additional or different contaminants when different wastes are processed.

More frequent monitoring should continue until the analytical sampling data indicates that levels are below health concern. Health assessors should also consider the need to sample for additional or different contaminants during start-up and when different wastes are processed. The evaluation should include compounds possibly formed during the treatment process. See section 6.1.2.1. for a more extensive discussion of ambient air sampling and monitoring.

6.2.2. Performance Test Period

As soon as the thermal treatment system can operate at a steady state, the performance test should be conducted.

While the facility staff should have some operating time to gain experience on how to operate the equipment in the most efficient and protective manner, stack emission data should also be obtained as quickly as possible to determine what the potential public health impact is from a new facility. This is particularly important at Superfund sites. Some CERCLA facility operators have proposed testing phase feed rates and time periods used at RCRA type sites. If approved, this would allow the operators to process more than half the contaminated soils and wastes on site before the performance test is even conducted, the test report and environmental data reviewed, and before the agency could determine whether the unit performs in a manner that is protective of public health.

6.2.2.1. Worst-case Operating Conditions

ATSDR recommends conducting performance tests for thermal treatment facilities at worst-case operating conditions, so the stack emissions will be the worst that are likely to occur during normal operation. But at facilities where only one waste stream will be burned, such as CERCLA sites, the performance test can be conducted at their proposed operating conditions. This should be a site-specific determination. Worst-case operating conditions include

• Maximum feed rate,
  
  
• Maximum concentration of minimum Btu wastes,
  
  
• Maximum flue gas flow rate,
  
  
• Minimum temperature in DC or PCC and SCC,
  
  
• Maximum halides, metals, ash, and water feed rates, and
  
  
• Minimum or maximum (as appropriate) operating conditions for all the starred operating conditions in section 6.1.1.4.

If metals are a health concern at the site, it is difficult to do worst-case operating conditions for organic chemicals and worst-case operating conditions for metals at the same time; two performance tests are usually needed. The primary difference is that, to be worst-case for organic chemicals, performance tests should be done at the minimum temperatures in the PCC/DC and SCC, but should be at the maximum temperatures allowed during normal operations for metals. See sections 6.1.1.4. and 6.1.1.5. for additional discussion.

6.2.2.2. Stack Testing

Desorbers should have a THC CEM as a standard operating condition.

Thermal treatment stack sampling should be conducted, at a minimum, for POHCs, PICs, VOCs, SVOCs, metals, acid gases, particulate, THC, O₂, dioxins, and furans. If contaminants containing sulphur or nitrogen are present in the wastes to be treated, the stack emissions also should be sampled and analyzed for sulfur oxides (SO_X) or nitrous oxides (NO_X) during the performance test.

If the incinerator operates at temperatures greater than 2200°F, the nitrogen in the combustion air could form NO_X, so stack sampling for NO_X should be required. Incinerator stack emissions should also be analyzed for carbon monoxide using a CEM. If an THC monitor is not required on the incinerator or desorber as an operating condition, it should be required during the performance test.

Each performance test should consist of three stack sampling runs. Some facilities do a fourth run so they have backup data in case some of the samples are lost or broken. Representatives from the EPA or state environmental protection department knowledgeable in stack sampling procedures should be present during most or all the runs to ensure that the performance test is conducted according to the approved plan.

6.2.2.3. Consider Site-specific Need for Additional CEMs

If the thermal treatment facility is not required to monitor continuously the stack for THC, ATSDR strongly recommends that a THC CEM be in place at least during the performance test. Health assessors and the environmental oversight agency staff should discuss the site-specific need to have other CEMs monitoring stack emissions during the performance test if they are not already required for operation, such as carbon monoxide, carbon dioxide (CO₂), SO₂, or opacity. See sections 6.1.1.2., 6.2.1.2., and 6.2.2.2. for further discussion.

6.2.2.4. On-site, Fence Line, and Community Sampling and Monitoring

If the facility has an ambient air monitoring system, health assessors should consider the need to have the frequency increased during the performance test to determine whether worker or community exposure rates or both are affected. The thermal treatment unit could be operating at maximum capacity and worst-case operating conditions, which could cause increased fugitive or stack emissions or both. Health assessors should also consider the need to sample for additional or different contaminants during the performance test, especially if the waste is spiked with additional chemicals or if different chemicals (PICs) could be formed during treatment of the spiked waste. See sections 6.1.2.1. and 6.2.1.3. for further discussion.

6.2.3. Posttest Period to Operational Phase

ATSDR recommends that on a site-specific basis, CERCLA and new RCRA thermal treatment facilities be shut down after the performance test and not be allowed to operate until the test report is reviewed and the final operating conditions are set based on the operating conditions demonstrated during the performance test. If a new facility is the same design and will be processing essentially the same wastes as a facility that has previously, successfully demonstrated safe operations, then it would not be necessary to shut down. A site and facility specific determination might also conclude that sufficient information is included in a pre-test risk assessment to determine that continued operation is unlikely to cause a harmful exposure. Public exposure to stack emissions should be minimized until data demonstrate that the emissions will not pose a health risk to the community.

EPA regulations allow RCRA incinerators (new and existing interim status facilities) to continue operating at conservative operating conditions that the permit writer believes to be protective of public health during the post-test period. RCRA permits have often required one-half the facility's proposed feed rate and higher operating temperatures. RCRA regulations for desorbers (40 CFR 264 Subpart X) are silent on this issue, but permit writers usually apply the RCRA incinerator regulations to desorbers that are relevant to the desorber's design. Nevertheless, CERCLA and new RCRA facilities do not have a long operating history on which to base post-test period operating condition. Thus unless a site-specific determination is made that continued operation is unlikely to cause harmful public exposures, ATSDR recommends that these units be shut down during this period. If existing facilities have an operating history of low CO and THC (CO <100 ppm and THC <10 ppm) and stable operations (few TRV openings and AWFCOs), then stack emissions will not likely be a health threat if operations continue during the posttest period. ATSDR staff should not routinely oppose continued operation at these facilities during this period.

Facility staff usually collect at least one stack sample during shakedown to determine whether the facility is ready to be tested. The facility can submit the analysis of the pretest samples (usually without QA/QC data) as justification for continued operation after the performance test. A minimum of three data points are needed to determine the reliability of a measurement. ATSDR staff should therefore carefully consider on a site-specific basis the advisability of continued post-test operation based on limited pretest data especially if it is a new emerging thermal treatment technology with very little operating experience on waste streams or unusual waste configurations.

ATSDR recommends that CERCLA and new RCRA thermal treatment facilities be shut down after the performance test and not allowed to operate until the test report is reviewed and the final operating conditions are set.

At a minimum, post-test operations should (1) be for a limited time period or maximum amount of waste that can be processed at CERCLA sites-or both, and (2) require that the unit be immediately shut down if the stack data shows any of the performance criteria were not met. The limit on the amount of waste that is allowed to be processed at a CERCLA site before the performance test data is reviewed and final operating conditions set should be set on a site-specific basis, but should generally be limited to less than 25% of the amount of waste to be treated at the site. Agency staff should expedite their review of performance test results to ensure the public is not unduly exposed and the agency does not cause unnecessary delays.

6.2.3.1. Modeling of Stack Emissions Data

The stack emissions should be modeled using EPA approved air dispersion models to determine the maximum ground level concentrations and deposition rates that could occur from stack emissions during full operation. Five years of meteorologic data from a National Weather Station or one year of on-site meteorologic data should be input into the model to ensure that realistic worst-case exposures are considered.

Maximum annual, 24-hour, and hourly concentrations should be calculated to evaluate possible chronic as well as acute exposures.

The models, meteorological data, and assumptions used in the model should be reviewed by an ATSDR meteorologist or other staff or contractor experienced in modeling to ensure that the modeling done is appropriate to use in evaluating public health implications of the thermal treatment facility. Modeling can be used to evaluate future and past exposure pathways. Ambient air monitoring and sampling data, along with modeling data, can be used to evaluate the public's current exposure. See section 6.1.2.1. for a detailed discussion of ambient air monitoring and sampling plans.

When identifying the maximum exposed individual (MEI) or maximum exposed receptor that could be impacted by the site, ATSDR staff should use realistic assumptions about how that individual could be impacted. This is called the reasonable maximum exposure (RME) (EPA 1994, EPA 1998b). The RME is not necessarily a residence; it can be any location or a composite of several locations that are frequently used, such as a business, commercial or industrial area, or schools, recreational areas, parks, or a farm, dairy, ranch, or hatchery. For RCRA facilities, where a potential exists for future residential development around the facility, health assessors should consider in their evaluation any area that might become inhabited. For a Superfund site where the thermal treatment unit will only be temporarily located, only currently habitable areas should be considered in the modeling and exposure evaluation. Unhabitable areas, such as swamps, deserts, cliffs or rugged terrain, lakes, forest preserves, or areas of national or state parks that are only infrequently accessed, should not be considered when evaluating the RME. Areas of national or state parks frequently accessed, such as visitor centers, camping, hiking trails, etc. should be included in the RME evaluation.

In addition to inhalation exposure, health assessors should consider indirect exposure through the food chain if farms, ranches, hatcheries, or other food sources could be impacted by deposition of stack or fugitive emissions. EPA risk assessments have shown that the majority of potential risk from combustion facilities is through the indirect pathways rather than direct inhalation of their emissions (EPA 1998b). Others have pointed out that the indirect exposure pathway importance is an artifact of all the uncertainty assumptions used in risk assessments, and that actual environmental and food-chain sampling does not support the conclusion that the majority of potential risks from thermal treatment facilities is through indirect exposure. Whenever environmental sampling data is available, e.g., ambient air sampling, soil or water samples, food-chain samples, soil or water samples, etc., the health assessor should use that data rather than modeled values in their health evaluation. See Chapter 7 for further discussion of exposure evaluation. The health assessor is reminded that ambient air monitoring and sampling data reflect the total exposure of people near the sampling location to all sources emitting the chemicals being measured. Unless the chemicals analyzed for are unique to a particular facility, health assessors should not attribute the exposure to a specific facility.

6.2.3.2. Evaluating Reports

Health assessors should obtain and review a copy of any risk assessment which has been conducted. If EPA, the state, or the facility conducted a pre-operational risk assessment, they usually revise it after the performance test, based on the stack emissions data. Most risk assessments include an evaluation of the air and indirect exposure pathways along with the emissions modeling data and assumptions. Using the risk assessment as a starting point can save health assessors a lot of time-if they agree with the modeling and exposure assumptions used in the risk assessment.

During the post-test period, prior to the facility resuming operation, in addition to the risk assessment and performance test report, health assessors should review the on- and off-site ambient air monitoring and sampling reports and any worker personal monitoring data that was collected until that time. A toxicologist or health professional should review the concentrations of contaminants in the ambient air relative to appropriate standards and guidelines, and render an opinion on whether the facility operations under those conditions are likely to cause adverse health effects if those operations are allowed to be completed in the proposed operational period. While ambient air data are the best data to use in making health calls, only a limited number of constituents are typically available, so modeled performance and risk burn data should also be considered. Often the toxicity of the contaminant levels measured in the stack gases are sufficiently low that a toxicologist can eliminate all but a few contaminants for which modeling might be needed to evaluate the potential public exposure.

6.2.3.3. Evaluating Operating Conditions

EPA or the state regulatory agency staff are responsible for reviewing all the data collected during the performance test and for specifying how the thermal treatment facility must be operated (if the data supports allowing the facility to operate). If the EPA permit writer, remedial project manager (RPM), or on-scene coordinator (OSC) follows the EPA incineration guidance on how to set operating conditions based on test burn results (see EPA 1989) and sets the operating conditions listed in section 6.1.1.4., the facility's operation should not cause public health impacts. Still, all of the incineration conditions and guidance might not be applicable to some desorbers. Appropriate operating conditions have to be determined on a site-specific basis. ATSDR staff should also review:

• CEM data,
  
  
• S&A data for the feed materials, residuals, and stack emissions,
  
  
• QA/QC report,
  
  
• Operating conditions during the performance test, and
  
  
• Data from any risk burns that are used to set operating conditions.

Health assessors should determine whether the EPA-specified facility operating conditions ensure that normal stack emissions will not exceed those measured during the performance test.

Frequent waste feed cut offs and restarts can create unstable operating conditions and increase stack and/or fugitive emissions.

Generally maximum and minimum operating conditions should be set by first discarding any spike values (outliers) excessively higher or lower than the readings during the balance of the run, and then averaging the lowest (or highest) value in each run of the test. Some facilities have a computerized data logger which records the individual readings and also calculates a rolling average for each operating condition. A more stable operation will be achieved if rolling average operating conditions are used. Rolling averages, however, mask the full range of operating conditions that occur. If instantaneous readings are used to set operating conditions that are tied to the automatic waste feed cutoff system, then the facility should set an alarm point at a slightly higher (or lower) reading so that facility operators are notified of a pending shutoff and can take action to correct quickly that condition and avert an AWFCO, thus providing more stable operations. The method used to record the operating conditions during the performance test (rolling average or instantaneous) should be used to set the final operating conditions.

6.3. Operational Phase - Information to Review to Protect Public Health

The thermal treatment facility can be brought back on line after the performance test data have been reviewed, the decision has been made to allow the thermal treatment unit to begin operating again, and the operating conditions have been specified for full operation. Depending on the availability of agency resources and priorities, ATSDR staff might only occasionally review facility reports during the operating phase or they could be asked to review all CERCLA site reports. It would be too resource intensive for ATSDR staff to review all RCRA reports. This section discusses the types of reports and information possibly available during the operational phase and their relevancy to public health.

6.3.1. Equipment Operation Protective of Public Health

An important key to preventing public exposure to hazardous emissions is to have a well-operated thermal treatment facility. The more stable the facility operator can maintain the operating conditions, the lower the overall emissions.

6.3.1.1. Incident Reports

EPA often requires the operator to submit incident or noncompliance reports whenever there is a TRV opening (see section 5.1.5. for more details), noncompliance with operating conditions, or on- or off-site ambient air action levels are exceeded. ATSDR should look at any available ambient air monitoring and sampling data for the time period covered by the incident to see whether an increase in the potential public exposure to site contaminants occurred.

6.3.1.2. Inspection Reports

Reports prepared by EPA or state environmental department compliance inspectors will list any violations discovered during inspections of RCRA permitted facilities. Most Superfund facilities are not inspected by the RCRA compliance inspectors. Their operations are overseen by the RPM, OSC, or the EPA oversight contractor on a very frequent basis. Inspectors or overseers who have completed the RCRA Incinerator Inspection or Incineration Permit Writers training course or have experience with thermal treatment facilities will provide the best oversight of the facility operations.

Inspection reports, notices of deficiencies (NODs), and notices of violations (NOVs) are good sources of information on the past operation of RCRA facilities. These documents might also contain data on releases that exceed permitted limits if they have occurred, and give ATSDR staff an indication of how well the facility has been operated in the past in order to estimate past exposure pathways. If enforcement actions are pending, the most recent inspection report might not be available for some time after the inspection.

6.3.1.3. Continuous Monitoring Reports

RCRA facilities must maintain on-site records of all continuously monitored operating conditions. The types of conditions usually continuously monitored are listed in section 6.1.1.4., which includes all of the stack continuous emission monitors. These data could be in electronic or hard copy format. Because data are continuously recorded, the volume of data is extensive, making it impractical to look at all the available data to see how well the facility has been operated. If, however, there are questions about a particular time in the past when there could have been releases, this could be a good data source if health assessors are familiar with the permit conditions and know how to interpret monitoring data and operating conditions.

Superfund sites should also keep the records of all their continuously monitored operating conditions on site. Likewise, if questions about the potential for releases during a particular time period arise, ATSDR staff familiar with this type of data should visit the site and examine the data.

ATSDR staff should primarily review the CEM data for signs of fluctuations in the emissions or operating conditions that could indicate the potential for stack or fugitive emissions releases during and around the time period in question. If there is the possibility that emissions did increase, health assessors should also look at any on- and off-site ambient air data available for that time period. Health assessors might lack the expertise to evaluate the facility operating conditions and thus not be unable to judge the potential for past exposure. If that level of review is necessary, and EPA compliance inspection reports that address the issue(s) are not available, the assistance of a DHAC or contractor combustion specialist may be requested.

6.3.1.4. Additional Stack Emissions Testing Reports

EPA requires all RCRA and some CERCLA thermal treatment facilities, depending on the length of time they will be operating, to retest their stack emissions periodically. Health assessors should review all emission test reports to understand the potential for public exposure in the past as well as the current operating status of the facility. New modeling might need to be done if there have been

• Modifications to the thermal treatment unit or air pollution control equipment or operating conditions that would change the stack height, stack gas temperature, or exit velocity,
  
  
• Substantial changes in waste feed that could affect the stack gas velocity or temperature or the deposition rate,
  
  
• Construction or other modifications of nearby buildings that could affect the plume, or
  
  
• Development of the area within a 1-mile radius of the site if the area was previously rural or different receptors need to be considered.

The emissions test report should include the stack gas temperature and velocity which health assessors can compare to the numbers used in the original modeling for the facility. If there are only minor differences in the stack gas temperature or velocity and they are within normal site fluctuations, the dispersion coefficients from the original modeling can be used to estimate maximum potential acute and chronic public exposure.

6.3.2. Overall Facility Conditions Protective of Public Health

6.3.2.1. Ambient Air Reports - Fence Line and Community

Whenever good quality ambient air monitoring and sampling data are available for a facility, ATSDR staff should rely on that rather than modeling as the main source of data by which to characterize the community's potential exposure. Because ambient air monitors and samplers measure the total impact on the residents from the thermal treatment facility as well as all other sources in the area emitting the chemicals being measured in the ambient air, the ambient air data are a better indicator of whether residents could have been exposed to harmful levels of the chemicals. Health assessors should be careful in how they word statements regarding ambient air data because (1) the data only represent the exposure to the chemicals measured, (2) they represent the exposure of residents near the monitor, not necessarily the community at large, and (3) the chemicals measured can rarely be traced to their origin unless they are unique to that thermal treatment facility.

Modeling of the facility's stack emissions and information from the EPA Toxic Release Inventory (TRI) database for that area can also be used to estimate the community's potential total exposure if ambient air data are not available. However, all point sources are not required to report their emissions to EPA for inclusion in the TRI database, and only certain specified chemicals have to be reported. Therefore, TRI data could or could not be a good indicator of the community's total exposure to a particular chemical. See section 6.1.2.1. for more details on ambient air monitoring and sampling.

6.3.2.2. Residuals Analysis Reports

All thermal treatment facilities should routinely analyze their process residuals (bottom ash, fly ash, condensate, scrubber water, spent carbon, spent filters, etc.). Superfund sites are normally required to analyze each batch or day's run of solids or soils processed. Other process residuals are typically analyzed less frequently. See sections 5.1.4., 6.1.1.1., 6.1.1.2., and 6.1.1.3.7. for additional discussion on process residuals. If there is a potential for the public to be exposed to process residuals, the reports on the analysis of those residuals will be an important data source.

The ATSDR Public Health Assessment Guidance Manual, Chapter 7 provides guidance on the public health implications of exposure to contaminants of concern (ATSDR 1992b). Addressed here are only the key issues for evaluating pathways related to thermal treatment facilities. The hierarchy of comparison values and the latest values are found in the ATSDR Comparison Values tables, updated quarterly and available on the ATSDR Web page. Because inhalation studies have not been conducted on all chemicals, toxicologists often look at oral studies and the health effects found in those studies. Then they use their professional judgement to evaluate the potential public health implications of the air exposure pathway.

In addition to inhalation exposure, if gardens, fishing areas, farms, ranches, hatcheries, or other food sources could be impacted by deposition of stack or fugitive emissions, health assessors should consider indirect exposure through the food chain. During the site visit, ATSDR staff should drive through areas within several miles of the facility (particularly areas where modeling predicted the greatest deposition) and note observed food or animal feed sources. In the final analysis, reasonable worse-case scenarios based on site-specific observations and community input should be used for the health evaluation.

EPA risk assessments that assume a farmer and his family are located at the highest deposition location, and that they raise everything they eat, even when there is no farming activity in the area, have shown the majority of potential risk from combustion facilities is through indirect pathways, rather than direct inhalation of emissions (EPA 1998b). Others have pointed out that the assumed importance of the indirect exposure pathway is an artifact of all the uncertainty assumptions used in risk assessments. They also say that actual environmental and food-chain sampling does not support the conclusion that the majority of potential risks from thermal treatment facilities is through indirect exposure. The point here is, however, that in health evaluations whenever environmental sampling data is available, e.g., ambient air sampling, soil or water samples, food samples, etc., the health assessor should use that data rather than modeled values.

Key issues health assessors should consider when evaluating exposure pathways at thermal treatment facilities include:

• In assessing individual exposures, ambient air monitoring and sampling data are preferable to stack data. But detection levels of most chemicals make it advantageous to sample at the stack, where concentrations should be the highest. Also, because fewer compounds can be analyzed in ambient air, health assessors should review both sets of data.
  
  
• If stack emissions and ambient air concentrations are below comparison values, no further analysis of the air pathway is needed. If, however, stack emissions are above ATSDR screening values and ambient air sampling data are not available, modeling of the stack emissions is required to evaluate potential health impacts.
  
  
• In thermal treatment facilities, dioxins and furans are always a public concern. For existing facilities EPA hazardous waste combustion regulations specify a toxicity equivalency quotient (TEQ) stack emission concentration less than or equal to 0.4 nanograms per dry standard cubic meter (ng/dscm). For new facilities the requirement is 0.2 ng/dscm (64 FR 52828). ATSDR toxicologists should evaluate, on a case-by-case basis, the dioxins/furans congener-specific, stack-emission data or ARAR set for the site.
  
  
• On occasion, dioxins, furans, PCBs, polycyclic aromatic hydrocarbons (PAHs), other persistent organic compounds, or toxic metals are detected in the stack gases at levels of health concern. When those substances are detected, health assessors should also evaluate the potential for indirect exposure of the public through any food chain contamination resulting from stack emission deposition. In that regard, if EPA, the state, or the facility has already done a risk assessment, it should include indirect exposure scenarios. The health assessor should by all means obtain and review that document.
  
  
• Reasonable worst-case scenarios should be used to evaluate the indirect or food-chain exposure pathway. The scenarios should be based on the types of food and feed sources in that area, as well as input from the community. For example, health assessors should consider the location of farms, fishing and recreation areas, and home gardens. Assessors should also consider the types of farms (e.g., livestock, poultry, vegetables, feed crops, etc.), the size and flow rates of local bodies of water, and food consumption practices of the local population.
  
  
• When looking at chronic exposures the life of the facility should be considered. RCRA facility life expectancy could be over 30 years. CERCLA sites are usually active less than 2 years. A general rule is that in making public health determinations, health assessors should use a facility's maximum projected operating life.
  
  
• If the unit has a TRV and is at an operating RCRA facility, or a CERCLA unit that has been operated at other sites, information should be obtained on the frequency and number of TRV openings and length of each opening. If no data exists, toxicologists should assume at least 12 TRV openings per year, lasting 30 minutes each (or at least as long as the solids' residence time), and add this concentration to the normal stack emissions when evaluating chronic exposure. Health assessors should assume that during TRV events, all acid gases, particulate matter, and metals are emitted from the TRV. If the TRV is after the SCC, assume all halogens are converted to acid gases, the metals are vaporized (if the SCC temperature is greater than the temperatures specified in Table 7 for each metal). Assume also the DRE initially to be 99.99%, but that it will decrease during a TRV event, based on a site-specific determination.
  
  
• Fence-line or community ambient air data or observations during a site visit could indicate that nearby residents could be exposed to fugitive emissions. If so, health assessors should consider potential exposure (particularly of small children) to chemicals that might be deposited on surface soils and indoor dust, as well as air and indirect exposure pathways.

In September 1993, ATSDR convened a panel of 30 national experts not affiliated with the agency to identify and evaluate information related to the public health implications of human exposure to PCBs. The experts were divided into three panels. One panel discussed the health effects of PCBs, one discussed incineration of PCB-contaminated waste, and the third panel discussed technically feasible, non-incineration remedial technologies. One of the recommendations from the panel was to conduct baseline surveys of exposures and health status of communities and workers at sites where new incinerators were to be located. These surveys were to be conducted at the site before operation, with follow-up testing during or after incineration. The panel also recommended that studies be conducted at existing incineration facilities (ATSDR 1993c).

ATSDR has conducted or funded several health studies and exposure investigations in communities near hazardous waste combustion facilities. Several of these health studies involved collecting biological samples of residents living near contaminated sites. The samples reflected conditions before, and after incinerators were used to remediate the sites. The object was to determine whether residents were exposed to dioxins during the operation of the incinerator.

It is important to note that no other remediation technology has undergone as many stack emission tests, as much ambient air monitoring, or as many health studies as has incineration. It is equally important to note that only one incineration facility, the Caldwell Systems, Inc. hazardous waste incinerator in Caldwell County, N.C., was implicated by the ATSDR-funded studies as the potential cause of adverse health effects in some workers and community members. Summaries of ATSDR-conducted or funded studies are included here, so that health assessors can learn of the research findings, and status of studies still in progress at the time of publication of this document. Table 12 provides a list and brief summary of each of these studies.

The National Institute for Occupational Safety and Health (NIOSH) conducts health hazard evaluations (HHEs) at work sites. This chapter also includes summaries of the evaluations NIOSH staff, contractors, or grantees have conducted at incineration facilities. No HHEs were found for thermal desorption facilities. Summaries of all types of incineration facilities have been included because hazardous wastes might be occasionally mixed with other wastes. If wastes are mixed, health assessors must consider in their evaluation that combined waste stream. Included are all HHEs conducted at facilities incinerating hazardous wastes and several HHEs at municipal solid waste (MSW) and medical waste combustion (MWC) facilities, also known as biological waste incinerators (BWIs).

8.1. ATSDR Funded Health Studies Related to Combustion

This section provides a summary of each of the six health studies related to hazardous waste combustion and funded by ATSDR.

8.1.1. Caldwell Systems, Inc. (NC) Name of Study

Study of Symptom and Disease Prevalence
Caldwell Systems, Inc. Hazardous Waste Incinerator
Caldwell County, North Carolina
September 1993 (ATSDR 1993b)

Type of Waste

Hazardous waste, predominantly lacquer chips and dust from the furniture industry, waste torpedo fuel from the U.S. Navy, and pipeline industry wastes.

Facility Description

From 1976 to September 1987 the Caldwell incinerator consisted of a combustion chamber and short stack. Flames and smoke were frequently seen exiting the stack-an indication of poor combustion. The Caldwell Systems, Inc. incinerator (CSI) is not representative of hazardous waste combustion facilities operating today. Currently, most hazardous waste incinerators have received final RCRA permits and are subject to performance standards, emissions testing requirements, and operating restrictions to ensure attainment of emissions limits. Interestingly, a few months before it was shut down, the incinerator was upgraded to include air pollution control equipment and automated operating controls. The facility was also a waste blending, bulking, and storage facility.

Background

In 1976, Caldwell County built the CSI incinerator adjacent to the county landfill. The semi-rural area was about 7 miles south of Lenoir, NC. The county operated the facility until 1977, when it leased the incinerator to CSI, who until September 1987 operated it without any air pollution control equipment and few waste feed restrictions.

From November 1980 until it closed in 1987, CSI operated the hazardous waste incinerator as an interim status facility under the RCRA regulations. Regulations for interim status facilities at that time were less stringent than those for fully permitted facilities. No restrictions were placed on feed rates, waste characteristics, or stack emissions. According to the Air Program permit granted by North Carolina in 1977, CSI was allowed to burn 4,100 pounds per hour-more than twice as much waste as the design rate of 1,800 pounds per hour. Furthermore, former CSI employees told ATSDR investigators that the licensed rate was frequently exceeded.

On May 31, 1988, Caldwell County ordered the incinerator to cease operations. The plant did, however, continue until December 1989 to operate as a waste blending, bulking, and storage facility.

On September 13, 1989, a fire in a roll-off container generated smoke and irritant fumes, and nearby residents had to evacuate the area. This led the Caldwell County Superior Court on September 28, 1989, to issue an order directing CSI to cease operations completely, and to remove all waste no later than December 1, 1989.

In April 1990, EPA Region IV asked ATSDR to perform a health consultation to evaluate the health complaints of residents living near the incinerator as well as those of former CSI employees and their families. ATSDR concluded that workplace conditions during the years of incinerator operation at CSI presented a threat to human health for some former CSI employees and the members of their households, as well as pulmonary health problems for residents living near the former incinerator. Employees told of engaging in sludge fights (throwing hazardous waste sludges at fellow employees) and dunking employees' heads in drums of hazardous waste. In July 1990, ATSDR issued a public health advisory concluding that a significant threat to human health was associated with past work practices at CSI. The advisory further recommended a study of the population living near the site and of household contacts of the former CSI workers. During the fall of 1991 NIOSH performed a health study of former CSI workers (see section 8.2.1. or NIOSH 1992b). In July 1991, ATSDR conducted a retrospective cross-sectional symptom and disease prevalence study of 713 residents living within a 1.5-mile radius of the CSI site.

Study Objective

To determine whether the prevalence of specific symptoms or diseases in persons living within a 1.5-mile radius of the former CSI site differed from the symptoms and diseases in similar persons not living near the site.

Study Design

Retrospective cross-sectional symptom and disease prevalence study.

Type of Data Collected

Symptom and disease questionnaire data, including a detailed respiratory section adapted from the American Thoracic DLD-78 questionnaire. The questionnaire covered 14 self-reported symptoms and 26 self-reported, physician-diagnosed diseases.

Study Group

To select the study group, investigators sampled all residents living within a 0.9-mile radius of the incinerator and half of all residents (every other household) living from 0.9 to 1.5 miles from the incinerator. The comparison group was drawn from the village of Gamewell in Lenoir Township, approximately 8 miles west of the site. Ninety-six percent (713) of eligible participants from the target area and 91% (588) of eligible participants from the comparison area were included in the study. To be eligible for the study, residents had to (1) be 3 through 79 years of age, (2) have resided in one of the two study areas for at least 6 months prior to May 1988, and (3) not be a former worker at CSI or a family member of a former worker.

Summary

In comparison to the control group, target group participants had an increased prevalence of self-reported irritant and respiratory symptoms, but not respiratory diseases or hospital admissions for these diseases since the onset of incinerator operation (adjusting for age, sex, and cigarette smoking). Residents of the target area were almost nine times more likely than residents of the control area to report symptoms of recurrent wheezing or cough following a respiratory insult (Respiratory Symptom Complex A), adjusting for cigarette smoking, asthma, and environmental concern. Furthermore, residents living within 0.9 miles of the site were almost twice as likely as those living from 0.9 to 1.5 miles from the site to report symptoms consistent with Respiratory Symptom Complex A (adjusting for smoking, asthma, and environmental concern). Residents of the target area were approximately one and a half times more likely to report neurologic symptoms than were residents of the comparison area, and almost two and a half times as likely to report the diagnosis of selected neurologic diseases, after adjusting for smoking, sex, diabetes, alcohol ingestion, and environmental concern.

Within the target area, statistically significant odds ratios (OR) were found for neurologic symptoms (dizziness and poor coordination) in the areas north and south of the incinerator (consistent with the prevailing winds) compared with areas east and west of the incinerator. Among persons living less than 0.9 miles from the incinerator compared with those living farther away in the target area, statistically significant, increased odds ratios appeared for chest pain, poor coordination, dizziness, and irritative symptoms (using stratified analysis). Multiple logistic regression analysis demonstrated that among target area residents, neither distance nor direction played an important role for respiratory symptoms, but that people living within 0.9 miles of the incinerator were more likely to report the occurrence of respiratory symptoms (Complexes A and B) than if they lived farther away.

The prevalence of self-reported cancer was not higher in the target area than in the comparison area. The residents of the target area were not more likely to report adverse reproductive outcomes than target area residents.

Conclusions

The prevalence of self-reported irritant, respiratory, and neurologic symptoms was significantly higher in the target area versus the comparison area.

Neither self-reported, physician-diagnosed respiratory diseases, nor hospital admissions for these diseases, differed in prevalence between the target and comparison areas when all participants were compared. Nevertheless, the prevalence of self-reported physician-diagnosed respiratory diseases (bronchitis and pneumonia) was higher in the comparison area among participants who reported no health problems attributable to environmental causation.

Within the target area, self-reported irritant, respiratory, and neurologic symptoms were more prevalent among participants living 0.9 miles or less from the CSI incinerator versus those living between 0.9 and 1.5 miles away. Neurologic symptoms among the study group participants were also consistent with the prevailing winds (north-south).

The increased prevalence rates of self-reported irritant, respiratory, and neurologic symptoms within the target area remained statistically significant after adjusting for smoking, sex, employment in the furniture industry, and environmental concern.

Although the study was not designed to determine a causal association between environmental exposures and health outcomes, significant differences in self-reported symptoms were found between the target and comparison areas.

Further examination of the residents living near the incinerator is indicated. This conclusion is based on an increased prevalence of self-reported respiratory, neurologic, and irritant symptoms.

Public Health Assessment Implications

The target area participants living closest to the CSI incinerator reported irritant, respiratory, and neurologic symptoms more often than did participants living more than 0.9 miles away. Although the study was not designed to determine a causal association between environmental exposures and health outcomes, significant differences were found between symptoms reported in the target and comparison areas.

No ambient air monitoring data were available during the CSI operating years to verify the exposure of residents near the site. However, the study findings are likely associated with higher concentrations of stack and fugitive emissions due to

1. CSI's short stack height,

2. Incomplete combustion due to poor operating practices, e.g., exceeding design capacity, flames and smoke in the stack emissions, etc.,

3.No APCE to remove the acid gases, volatilized metals, particulate matter, or PICs from the stack emissions, and

4. Sloppy waste handling practices, e.g., frequent spills, open containers, employee horseplay, fires in storage areas, etc.

No differences appeared in self-reported physician-diagnosed respiratory diseases. The implication is that people living closer to the hazardous waste facility were more likely to occasionally have acute symptoms, but at that time (3 years after the incinerator shut down) there were no indications of incipient chronic illness.

The CSI incinerator is not representative of hazardous combustion facilities currently in operation. Today, most hazardous waste incinerators have received final RCRA permits. And to ensure attainment of emissions limits, they are subject to performance standards, emissions testing requirements, and operating restrictions. The CSI incinerator in all likelihood produced substantially more hazardous stack and fugitive emissions than incinerators operating today.

Public health officials should work with state and federal regulatory officials to prevent stack and fugitive emissions from hazardous waste thermal treatment facilities to the maximum extent practical. No thermal treatment unit should operate without air pollution control equipment. To evaluate the residents' exposure, ambient air monitoring should be conducted in communities near facilities with known air releases.

8.1.2. Caldwell Systems, Inc. (NC) Name of Study

Health Outcome Follow-up Study of Residents Living Near the
Caldwell Systems, Inc. Site
Caldwell County, North Carolina
August 1998 (ATSDR 1998)

Type of Waste

Hazardous waste, predominantly lacquer chips and dust from the furniture industry and waste torpedo fuel from the U.S. Navy, and pipeline industry wastes.

Facility Description

From 1976 to September 1987 the incinerator consisted of a combustion chamber and short stack. Flames and smoke were frequently seen exiting the stack-an indication of poor combustion. A few months before it was shut down, the incinerator was upgraded to include air pollution control equipment and automated operating controls. The CSI incinerator is not representative of hazardous waste combustion facilities operating today. Currently, most hazardous waste incinerators have received final RCRA permits and are subject to performance standards, emissions testing requirements, and operating restrictions to ensure attainment of emissions limits. The facility was also a waste blending, bulking, and storage facility.

Background

This is a followup to ATSDR's 1993 study on the prevalence of symptoms and diseases among residents living within 1.5 miles of the CSI incinerator. Because ATSDR's initial study demonstrated higher prevalence of respiratory, neurologic, and irritant symptoms among target area residents, in 1993 ATSDR conducted a follow-up study to compare further the respiratory, neurologic, and immune systems of residents within the target area to residents in the comparison area.

For further background information, see section 8.1.1. or the Study of Symptom and Disease Prevalence, Caldwell Systems Inc. (ATSDR 1993b).

Study Objective

• To compare the prevalence of abnormalities in pulmonary and neurobehavioral functioning and immune biomarkers among three groups selected from participants in the original study: the original symptomatic group, the original asymptomatic group, and the comparison group.
  
  
• To evaluate the extent to which pulmonary and neurologic symptoms reported in the follow-up study reflected actual organ-system effects that could be demonstrated using biomarker panels.
  
  
• To determine the extent to which participants who reported pulmonary symptoms in the first study reported the same symptoms again 2 years later.

Study Design

Cross-sectional follow-up study

Type of Data Collected

• Pulmonary function tests (PFTs) performed on all participants,
  
  
• Forced vital capacity (FVC),
  
  
• Forced expiratory flow (25-75%),
  
  
• Forced expiratory volume in 1 second (FEV1),
  
  
• Forced expiratory volume in 1 second to forced vital capacity (FEV1/FVC),
  
  
• ATSDR basic immune test battery (total lymphocyte count, T-cells count, B-cells count, cluster designation 4+ (CD4+) lymphocytes count, CD8+ lymphocytes count, CD4/CD8 ratio, level of immunoglobins A, G, and M, and C reactive protein),
  
  
• Adult environmental neurobehavioral test battery, and
  
  
• Questionnaire data, using the same questionnaire from the original study.

Study Group

The study group contained 3 subgroups:

• 52 of 77 target area participants who were symptomatic in the original study,
  
  
• 112 of 150 selected target area participants who were asymptomatic in the original study, and
  
  
• 96 of 150 selected comparison area participants from the original study.

The overall participation rate was 69%.

Summary

The prevalence of existing respiratory symptoms was related to study group status. The original symptomatic group reported significantly more respiratory symptoms than the original asymptomatic group: 53.8% versus 28.6% for cough, 51.9% versus 23.2% for phlegm production, 50% versus 31.3% for shortness of breath, and 61.5% versus 32.1% for wheezing. After adjusting for the effect of smoking, only phlegm production and wheezing remained statistically significant.

Pulmonary function test results were worse in the original symptomatic group than in the original asymptomatic group. Abnormal PFTs were uniformly more prevalent in the original symptomatic group than in the other two groups; however, these differences were not significant after controlling for smoking status. Pulmonary function test results were significantly worse among target area participants who had respiratory symptoms at followup than among those who did not.

There were no statistically significant differences among the study groups in the mean values for various immune biomarkers.

Women from the original symptomatic group demonstrated significantly lower sensitivity to vibration (vibrotactile threshold test), worse coordination (Santa Ana test), and lower strength (dynamometer test) compared to women from the original asymptomatic group or the comparison group. The reason for the differences in the neurobehavioral tests and the clinical significance of the finding is unknown.

No significant differences appeared in the neurobehavioral test results among the three groups of men.

Conclusions

Self-reported respiratory symptoms were more prevalent in the original symptomatic group than in the original asymptomatic group. But after adjusting for the effect of smoking, only phlegm production and wheezing remained statistically significant.

After controlling for smoking status, differences in pulmonary function test results were not statistically significant.

The results of the immune test battery were similar for the three groups.

Some differences in neurobehavioral test results were observed among women in the original symptomatic study group in comparison to women in the other two groups. Women from the original symptomatic group demonstrated significantly lower sensitivity to vibration (vibrotactile threshold test), worse coordination (Santa Ana test), and lower strength (dynamometer test) compared to women from the original asymptomatic group or from the comparison group. The clinical significance of this finding is unknown.

Public Health Assessment Implications

The authors of this study concluded that the original symptomatic group had worse pulmonary function test (PFT) results in comparison to the original asymptomatic and comparison groups. Furthermore, they concluded that target area participants who had respiratory symptoms at followup had worse pulmonary function test results in comparison to target area participants without them. Nevertheless, the fact that the differences were not statistically significant after controlling for smoking implies that the abnormal PFTs measured were more likely due to smoking than the CSI incinerator.

After adjusting for the effect of smoking, phlegm production, and wheezing, some neurobehavioral test results in women were statistically significant, so one cannot rule out the possibility that some effects could be related to proximity to the CSI facility. Although this study cannot prove what caused the measured health effects, public health officials should work with state and federal regulatory officials to prevent stack and fugitive emissions to the maximum extent practical from hazardous waste thermal treatment facilities. No thermal treatment unit should operate without air pollution control equipment. Ambient air monitoring should be conducted in communities near facilities with known air releases to evaluate the residents' exposure.

The CSI incinerator is not representative of hazardous waste combustion facilities currently in operation. Currently, most hazardous waste incinerators have received final RCRA permits and are subject to performance standards, emissions testing requirements, and operating restrictions to ensure attainment of emission limits. The CSI incinerator likely produced more hazardous emissions than incinerators operating today.

8.1.3. Calvert City Industrial Complex (KY) Name of Study

Symptom and Illness Prevalence with Biomarkers Health Study
for Calvert City and Southern Livingston County, Kentucky
May 1995 (ATSDR 1995)

Type of Waste

Hazardous waste and chemical manufacturing emissions

Facility Description

Seventeen companies, many involved in manufacturing and handling of chemical compounds, occupy the Calvert City Industrial Complex (CCIC), including LWD, Inc., a commercial hazardous waste incineration and treatment facility that has been operating waste incinerators since 1978. Information was not provided on the design of the two LWD incinerators. At least one of the other large chemical manufacturing companies in the complex also had a hazardous waste incinerator.

Background

The CCIC started in the late 1940s. In addition to the hazardous waste incinerators, the CCIC contains two EPA National Priority List sites.

In response to a petition request filed in May 1987 by a resident of Livingston County, ATSDR initiated a public health assessment at the site. ATSDR completed the health assessment in 1992 and recommended that the agency consider Calvert City and the southern Livingston County area for future health studies. In 1993, ATSDR initiated this health study.

Meetings held between ATSDR staff and the local community elicited three general areas of concern: (1) exposure to toxic substances, (2) increased prevalence of various symptoms and illnesses, and (3) excess cancer deaths. Chemicals of concern included heavy metals, dioxin, various organic chemicals, such as vinyl chloride, neurotoxic chemicals, and others. Some members of the community stated there was a high rate of birth defects in the Calvert City area. Other symptoms and illnesses of concern were infertility, miscarriages, low birth weight, skin rashes, chronic obstructive pulmonary disease, respiratory problems, asthma, eye irritation, neurologic disease, lupus erythematosus, and silicosis.

Study Objective

• To compare the prevalence of certain self-reported diseases and symptoms of participants in the target group with prevalence in the comparison group,
  
  
• To characterize the distribution of selected biomedical, pulmonary function, and volatile organic compound (VOC) exposure test values in the target study group,
  
  
• To compare them with both standard reference ranges for these tests and the distribution of biomedical test values in the comparison study group, and
  
  
• To examine the effect of other factors, such as age, sex, smoking history, and occupation, on biomedical test results and symptom and disease prevalence.

Study Design

Cross-sectional symptom and illness prevalence study

Study Group

The study group was comprised of 357 target area residents and 363 comparison area residents. Blood specimens for the VOC exposure test were collected from 100 of the target area residents and from 107 of the comparison group. The participation rates for the target and comparison areas were 48% and 55%, respectively.

The target area residents were drawn from two areas thought to have the greatest potential for exposure to hazardous substances emanating from the CCIC. Target Area A included all residents living within defined city limits of Calvert City (most of these residents live within 3 miles of CCIC), Target Area B included all residents living in the southern region of Livingston County just across the Tennessee River from the CCIC.

Based on location and socio-demographic factors, ATSDR selected Cadiz, a city 40 to 50 miles southeast of Calvert City, for the comparison community.

ATSDR defined the target and comparison areas, then conducted a census of randomly selected residences in each area to generate a list of eligible residents. The sample size for the study was then selected to allow 80% or greater power to detect realistic differences over a spectrum of background symptoms and disease prevalence.

Type of Data Collected

• Questionnaire data (to assess symptoms, illnesses, and confounding factors),
  
  
• Pulmonary function tests, and
  
  
• Biologic specimens to test for organ function and recent exposure to VOCs.

Summary

The study was conducted in two phases. The first phase was the census of randomly selected residences to generate a list of eligible residents. The census also included a well water survey. The second phase consisted of administering a standardized symptom and disease prevalence questionnaire, conducting pulmonary function tests, and collecting biological specimens.

Target and comparison area study participants had similar age, sex, race, education, income, and length of residency distributions. Target area participants, however, were more likely than comparison area participants to have worked in a position that could have exposed them to chemicals.

The self reporting of two illnesses was significantly higher in the target area: weakness or paralysis of limbs and gallbladder disease (odds ratios of 3.3 and 2.6, respectively). The odds ratios for these two diseases remained statistically significant after adjusting for confounding factors. Breast cancer was the only cancer with enough cases to make a statistical comparison between the two groups, but no significant difference in occurrence appeared between the target and comparison area women.

There were no statistical differences in the reproductive histories between women in the target and comparison areas. Odds ratios for miscarriages, stillbirths, and birth defects were less than 1.0, indicating less occurrence in the target area, but none were statistically significant.

Odds ratios were calculated to determine whether there were differences between target and comparison area groups when comparing biomarker results. Serum total protein results (to assess liver function) of target area study participants were more frequently outside the established reference range than results of comparison area participants (odds ratio 2.3, 95% confidence interval 1.2-4.5). No odds ratios comparing renal or immune system biomarkers were statistically significant.

There were no statistical differences in the pulmonary function test results between target and comparison area participants.

Of the 16 compounds with detectable levels, nine were significantly higher (though not statistically) in comparison area participants and six were significantly higher (though not statistically) in target area participants.

Conclusions

• Target area study participants reported illnesses slightly more often and symptoms less often than comparison area study participants; however, no clear pattern of symptoms or illness was observed.
  
  
• There were no statistically significant differences in reproductive histories between target and comparison area women.
  
  
• Biologic test results of the hepatobiliary, renal, immune, and hematopoietic systems revealed no clear differences between the target population and established reference levels.
  
  
• There were no statistically significant differences in pulmonary function tests between target and comparison area participants.
  
  
• There was no association between area of residence and VOC exposure, nor excessive recent exposure occurring in either study or community.
  
  
• Of the 31 chemicals tested for in the VOC exposure testing, only acetone was found at a statistically significantly higher mean concentration in target area participants than in comparison area participants. Still, both target and comparison levels were below the national reference level.

Public Health Assessment Implications

Even though this study found no association between the area of residence and any measured adverse health effects, the same might not be true for other industrial parks or hazardous waste incinerators. The potential for public exposure at each hazardous waste facility must be evaluated on a case-by-case basis. This study adds to the growing database indicating that adverse health effects are not usually found at the chemical concentrations found in the environment, even in communities near 50 year-old industrial parks.

8.1.4. Times Beach (MO)

Name of Study

Dioxin Incinerator Emissions Exposure Study
Times Beach, Missouri
July 1999 (Roberts 1999)

Type of Waste

Hazardous waste contaminated soil containing 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)

Facility Description

A transportable incinerator was constructed on the Times Beach Superfund site. The incinerator consisted of a rotary kiln operated around 1650°F with oxygen enrichment and a secondary combustion chamber operated around 2100°F. The emergency vent located between the two combustion chambers had two propane burners that fired automatically to provide extra combustion of the emissions if the vent opened and there was a flame out (lose of flame) in one or both of the combustion chambers.

The primary and the secondary combustion chambers (PCC and SCC) were fired with natural gas and operated under negative pressure. The air pollution control chain consisted of a quench, an hydro sonic scrubber, a wet ESP, a carbon adsorption bed, and the stack. All material handling occurred inside structures (portable buildings) that were under negative pressure. Trucks bringing contaminated soil to the facility entered the buildings through an air lock, dumped their load, and went through a decontamination chamber before exiting the building (ATSDR 2000).

Background

The 1988 EPA Record of Decision for 27 dioxin sites in eastern Missouri called for thermal destruction of contaminated soils and other materials. Approximately 265,000 tons of soil and other materials containing TCDD from these sites were burned at the Times Beach, Missouri Superfund site between March 17, 1996 and June 20, 1997. TCDD concentrations in the soil materials ranged from one part per billion (ppb) to approximately 3,000 ppb.

Times Beach, formerly an incorporated city in southwest St. Louis County, is approximately 20 miles from St. Louis. Times Beach and the 26 other eastern Missouri sites were contaminated in the early 1970s when waste oil contaminated with TCDD was sprayed on roadways and other areas for dust suppression.

Study Objective

To determine whether serum concentrations of TCDD, and other related compounds significantly increased in persons living in the vicinity of the Times Beach site during the time frame of the incineration operation compared to persons who lived farther away from the site.

Study Design

Prospective cohort

Study Group

The study group consisted of 76 of 245 randomly selected people from the survey census of the target area. The comparison group consisted of 74 of 245 randomly selected people from the survey census of the comparison area. To be eligible for the study, participants had to be between 18 and 64 years of age, and could not be employed in occupations likely to have exposure to TCDD and other related compounds. Pregnant or nursing mothers were also excluded from the study.

The target area was within a 4-kilometer radius of the incinerator. To determine the most likely neighborhoods for exposure, the location of the incinerator was compared to air-dispersion modeling maps prepared by an EPA contractor. The comparison area was selected using the following criteria: (1) the area needed to be in St. Louis, but approximately 16 kilometers from the incinerator, (2) no industries likely to produce TCDD or related compounds could be present in the area, and (3) no known TCDD site could exist in the area. Furthermore, no comparison area participant could be exposed in the target area vicinity on a daily basis.

After defining the target and comparison areas, ATSDR conducted a census of each household in the two areas, randomly selected 245 names from each area census, and recruited study participants from these lists of names.

Type of Data Collected

• Questionnaire data, and
  
  
• Serum TCDD levels (reported on a lipid-adjusted basis).

Summary

Blood samples were collected in September 1995 (before incineration), July 1996 (shortly after incineration began), and June 1997 (just before the conclusion of the incineration). The blood levels of most analytes decreased from pre-incineration to the end of the incineration. There were no differences in the mean blood levels between the target and comparison areas for any analysis, except log-transformed polychlorinated biphenyl (PCB), which was slightly higher in the comparison area. After controlling for age, gender, participant's weight and height, head of household income and head of household education level, there were still no differences in the levels of analytes between the target and comparison areas.

TCDD levels decreased from a mean of 1.79 parts per trillion (ppt) to 1.23 ppt in the target area, and from 1.46 ppt to 1.23 ppt in the comparison area. The TEQ showed a similar decrease over time.

The maximum and minimum levels of most analytes tested fell within the Centers for Disease Control and Prevention (CDC) reference ranges normally seen for these chemicals.

Most demographic, socio-economic, and behavioral characteristics did not differ significantly between the two groups. However, participants in the comparison group smoked more cigarettes, used more lawn-care services, and weighed more than those in the target group.

TCDD levels for the target and comparison area were combined and stratified on a number of participant characteristics. None of the comparisons resulted in statistically significant differences. ATSDR conducted the same analysis for TEQ and the only significant difference was for living in a home with smokers. The average TEQ for participants living in a home with smokers was 12.77 ppt compared to 9.36 ppt in homes without smokers.

Conclusions

ATSDR concluded that incineration of TCDD-contaminated soil and other material at the Times Beach incinerator did not result in any measurable exposure to the population surrounding the incinerator, as indicated by serum TCDD levels. There was no evidence of any changes in serum levels of analytes between pre- and post-incineration.

Public Health Assessment Implications

The study's findings support the use of incineration for similar materials contaminated with dioxin-like compounds, as long as the excavation, waste processing, and incineration are conducted in a controlled manner. The slight decrease in Serum TCDD levels could be an indication of reduced exposure to dioxins due to the sites being cleaned up.

8.1.5. VERTAC/Hercules Site (AR)

Name of Study

Adverse Reproductive Outcomes in Pulaski County
for Years 1980 Through 1990
March 1998 (Breasted et al 1998)

Type of Waste

Although this study did not involve a hazardous waste incinerator, it did involve exposure to dioxins and other hazardous substances from the production of phenoxy herbicides. On-site incineration of low dioxin 2,4-Dichlorophenoxy acetic acid (2,4-D) wastes and high dioxin 2,4,5-Trichlorophenoxy acetic acid (2,4,5-T) wastes was planned for the future.

Facility Description

A herbicide manufacturing facility surrounded by homes in close proximity to the site. The site was later remediated using a transportable hazardous waste incinerator.

Background

The Jacksonville community in Pulaski County, Arkansas, has documented exposures to dioxin and possible exposures to chlorophenol and chlorophenoxy acid potentially related to multi-year production of phenoxy herbicides at the VERTAC site. The planned incineration of 2,4-D and 2,4,5-T wastes created an additional potential for further environmental exposure.

As reported in the 1986 Report to the Policy Advisory Commission by the Arkansas Reproductive Health Monitoring System (ARHMS), an apparent increase in fetal loss (20%) was noted for Jacksonville births from 1980 to 1982 when compared to the remainder of Pulaski County rate (5.9% compared to 4.5%). This difference was statistically significant for all races combined and for whites. The difference, however, was not statistically significant for African-Americans. The fetal loss ratio studies were for those fetuses known to be at least 8 weeks or older in gestational age, (old enough to minimize the potential for over- or under-counting very early losses.)

In this study, ATSDR and the Arkansas Department of Health sought to extend the time period studied from 1980 through 1990, and to examine a broader range of adverse pregnancy and developmental outcomes, including birth defects, low birth weight, and specific developmental disabilities occurring in early childhood-in addition to fetal loss.

Study Objective

To examine an 11-year data series of reproductive outcomes for evidence of possible association with the VERTAC Superfund site of Jacksonville, Arkansas, and to provide pre-incineration baseline data regarding these health outcomes, which could be compared to future post-incineration data.

Specific objectives included (1) ascertaining all Pulaski County resident's cases of birth defects and fetal loss occurring in the 1980-1990 birth cohort, and all cases of developmental disabilities occurring in the 1985-1990 birth cohort, (2) investigating whether prevalence rates of adverse pregnancy outcomes differed between Jacksonville and the rest of Pulaski County, (3) performing extensive mapping studies and cluster analyses to search for possible local aggregations of cases and their potential relationship with the VERTAC site, and (4) estimating baseline birth prevalence rates of these reproductive outcomes for possible later comparison to post-incineration rates.

Study Design

Retrospective analysis of prevalence rates

Type of Data Collected

• Birth defects and other adverse birth outcome cases collected by the ARHMS. Cases are sought from hospitals, clinics, diagnostic centers, Head Start programs, community programs, vital records and other sources, and
  
  
• Births, deaths, and fetal deaths data collected by the Arkansas Department of Health.

Study Group

The study group included all birth outcomes for Jacksonville residents considered potentially exposed to the VERTAC site. The comparison group included all birth outcomes for non-Jacksonville residents of Pulaski County. Over the 11-year period there were 67,545 (64% white) births in all of Pulaski County and 9,602 (82% white) births in Jacksonville. Because Little Rock Air Force Base abutted Jacksonville, the study group differed significantly from the comparison group with respect to military employment (48.7% of births vs. 3.1%). Other differences between the two groups included maternal race, socioeconomic status, and birth weight. To avoid the expected large bias of military out-migration prior to diagnosis, the authors limited many analyses to nonmilitary births only.

Summary

The excess of fetal loss in Jacksonville in the early 1980s reported earlier by ARHMS was confirmed. Trend analysis indicated a decrease to the lower rates of the balance of Pulaski County, followed by a rapid increase in fetal loss county-wide. The analyses suggest that the excess fetal loss in the early 1980s occurred among military families who were stationed at the Little Rock Air Force Base abutting Jacksonville. No evidence was found to link this reproductive outcome to the VERTAC site.

No evidence was found between occurrence rate of birth defects and the VERTAC site, considering time trends, prevalence rates, and spatial analyses.

Isopleth mapping identified clustering of low birth weight (1985-1990) between 1 and 2 miles southeast of the VERTAC site.

The analyses suggested a possible association between the prevalence of neonatal seizures and seizure disorders and proximity to the VERTAC site (1985-1990), although case numbers were very low. The identified clusters of children diagnosed with seizure disorders, neonatal seizures, mental retardation, and low birth weight around and to the southeast of the VERTAC site suggests the possibility of a common factor which might or might not be associated with the site.

The predicted under-ascertainment of diagnosed cases of birth defects and especially of developmental disabilities among military families was shown. Inability to link study records of 1980-1984 to birth certificates restricted the analysis of nonmilitary events to the 1985-1990 time period.

Because of the conversion of rural route addresses to road addresses during the study decade, 100% geocoding of data was not attained. Geocoding failures among Jacksonville residents primarily occurred for addresses to the north and west of Jacksonville; potential clusters in those areas might have been missed.>

Racial differences in prevalence rates of adverse outcomes and different racial distribution in the study and control group both required race-specific analyses that further reduced the statistical power of the study.

Conclusions

• The increased fetal loss rates observed in the Jacksonville area in the early 1980s were not associated with spatial proximity to the VERTAC site.
  
  
• No indication of an excess in birth defects related to the VERTAC site was observed.
  
  
• Weak associations were found for a few developmental problems, such as seizures and neonatal seizures, and the VERTAC site.
  
  
• Several clusters of low birth weight were noted in Pulaski county, including one to the southeast of the VERTAC site.
  
  
• No indication of an association of fetal loss, birth defects or developmental disabilities with the passage of time was demonstrated.

The localized excess in low birth weight could account for the weak clusters of developmental disabilities. There is no direct evidence relating the low birth weight cluster to the southeast of the VERTAC site to site activities. ATSDR and the Arkansas Department of Health recommended further spatial study of low birth weight and developmental disabilities, including evaluating possible relationships to body burdens and environmental data.

Public Health Assessment Implications

This study did not find any associations between fetal loss, birth defects, or low birth weight and proximity to the production of phenoxy herbicides that contained dioxins at the VERTAC site. Weak associations were found between a few developmental problems and proximity to the site. While this study documents the status of the community's health prior to incineration for potential comparisons to any health effects measured after the site is remediated, it does not document the exposure (if any) of the pregnant mothers to dioxins or the herbicides manufactured at the VERTAC site.

On-going Study - Incinerator Exposure Assessment

Purpose: To help determine if remediation activities associated with the VERTAC site cleanup, particularly handling and incineration of drummed wastes, result in increases in body burdens of site-related chemicals by conducting pre- and post-incineration biomonitoring activities for estimating blood lipid concentrations and urine concentrations of site contaminants in nearby residents.

Biological samples were taken and analyzed before incineration began and after it was completed. Preliminary reports indicate that there was no increase in dioxin body burdens in residents living near the site attributable to the incineration of VERTAC wastes. Two residents' dioxin blood levels were higher during the post-incineration testing, so ATSDR conducted an exposure investigation and found residential sources of exposure which EPA remediated (ATSDR 1997b).

Public Health Assessment Implications

• Incineration of dioxin contaminated wastes can be done safely without causing measurable public exposures.
  
  
• Potential adverse health outcomes related to current or past site exposures should be taken into account when evaluating the safety of proposed on-site incineration of hazardous wastes.
  
  
• If health studies determine that people have been exposed to toxic chemicals, public health officials need to determine the source of the exposure and see that the exposure is eliminated.

8.1.6. Three Waste Incinerators (NC)

Name of Article

Do Waste Incinerators Induce Adverse Respiratory Effects?
An Air Quality and Epidemiological Study of Six Communities (Shy 1995)

Type of Waste

One incinerator burns municipal solid waste (garbage) and one burns medical waste. The industrial furnace burns coal and liquid hazardous wastes.

Facility Description

The biomedical waste facility was a commercial facility with two continuous-feed incinerators. The combined capacity of the two units was 35 metric tons per day. No air pollution controls were in use during the first year of the study. The incinerators burned boxed medical wastes containing microbiological wastes, pathological tissue, needles, discarded instruments and utensils, plastics, paper, pigments, and discarded biologicals and chemicals used in laboratories. No radioactive wastes were incinerated.

The municipal waste facility was a publicly owned facility with two incinerators and a total capacity of 224 metric tons per day. The units operated continuously, burning paper, plastics, and other household wastes to generate steam. The incinerators had an electrostatic precipitator and a 73-meter stack.

The third facility had four rotary kilns which heated crushed slate to produce lightweight aggregate for use in construction materials. The kilns normally burn coal, but during the second and third years of the study one of the kilns also burned liquid hazardous wastes at a maximum rate of 1,220 liters per hour. The maximum heat input to each kiln was 35 million British thermal units per hour.

Background

ATSDR and EPA provided funding and support to the University of North Carolina (UNC) at Chapel Hill for this 3-year study. UNC conducted an epidemiological study of the prevalence and incidence of respiratory effects among residents of communities surrounding three types of waste combustion facilities (a biomedical incinerator, a municipal waste incinerator, and a liquid hazardous waste-burning industrial furnace), and three matched comparison communities.

Twelve-hour ambient air sampling was conducted (day and night samples) for 35 days each year in each of the six study communities for particulate matter (PM_2.5 and PM₁₀), aluminum, iron, sulfur, silicon, zinc, sulfur dioxide (SO₂), hydrogen chloride (HCl), nitrous acid (HNO₂), and nitric acid (HNO₃). Simultaneously, the cohorts in that community recorded twice daily in a diary any respiratory symptoms they experienced, and a subgroup performed baseline spirometry at the beginning of the month, and peak expiratory flow rates twice daily. A second subgroup performed spirometry and provided a sample of nasal washings once each year.

Study Objective

• To compare the prevalence of chronic respiratory symptoms, respiratory hypersensitivity, diminished lung function, upper respiratory tract inflammatory reactions, and upper and lower respiratory tract diseases in exposed and nonexposed communities, adjusting for the distribution of known risk factors for these conditions,
  
  
• To select subcohorts of normal and of hypersensitive adults in these exposed and control communities and to obtain daily measurements of lung function and respiratory symptoms in these persons over a one month period, annually, for 3 years, with simultaneous daily measurements of air quality in each community, and
  
  
• To identify whether subgroups of the population are at higher risk of lung and respiratory disease from exposure to fugitive or stack emissions from incinerators.

Study Design

Longitudinal prospective cohort study of prevalence and incidence of respiratory effects

Study Group

The cohort consisted of 3,479 persons living in three communities within a 2 by 5 kilometer (km) elliptical area around the three facilities. The controls were 3,392 persons matched to the cohort by socioeconomic characteristics who lived in three communities upwind of and no closer than 5 km to the facilities. Each person (or an adult in the household) participated in a 20-25 minute telephone survey. Questions were drawn from the respiratory disease questionnaire of the American Thoracic Society with additional questions on household characteristics and demographics, smoking, chemical exposures at work and home, type of heating and cooking appliances, and perceived quality of the outdoor air in the immediate neighborhood.

Based on these responses, UNC recruited 80 persons from each community for the longitudinal component of the study. Forty subjects in each community were selected because they had asthma or asthma-like symptoms during the past 12 months. The other 40 were selected because they gave negative responses to all questions regarding chronic or acute respiratory symptoms. All subjects were nonsmokers and were not regularly exposed to cigarette smoke in their home. An additional 25 subjects from each community were recruited to participate in a once yearly collection of nasal lavage samples and to perform spirometric lung function tests.

Type of Data Collected

• Questionnaire data,
  
  
• Daily diaries of respiratory symptoms (35 days each year),
  
  
• Peak flow measurements (twice daily same 35 days each year),
  
  
• Ambient air environmental samples (day and night samples same 35 days each year),
  
  
• Spirometry (once a year), and
  
  
• Nasal washings (once a year).

Summary

The data from the first year failed to show any pattern of excess chronic or acute respiratory symptoms in any of the incinerator communities. Among the so-called normals, mean forced expiratory volume in 1 second (FEV₁) and peak expiratory flow rate values were consistently higher in incinerator than in comparison communities. Among the so-called sensitives, mean peak expiratory flow rate values were higher in the municipal and hazardous waste incinerator communities than in their paired comparisons, whereas these values were slightly lower in the biological waste incinerator community than in its comparison community. FEV₁ results did not show a consistent difference between community pairs among the sensitive subgroup. In the nasal lavage analysis neither cell counts nor biochemical indices of inflammation suggested an inflammatory effect of residence in the incinerator versus comparison communities. There were no important differences in average peak flows or in the diurnal change in peak flows between incinerator and comparison communities.

No differences in concentrations of particulate matter were detected among any of the three pairs of study communities. Average fine particulate (PM_2.5) concentrations measured for 35 days varied across study communities from 16 to 32 micrograms per cubic meter (µg/m³). Within the same community, daily concentrations of the fine particulate varied by as much as eightfold, from 10 to 80 µg/m³, and were nearly identical within each pair of communities. Direct measurements of air quality and estimates based on a chemical mass balance receptor model showed that incinerator emissions did not have a major or even a modest impact on routinely monitored air pollutants. During the first year of this 3-year study, the industrial furnace was burning coal, not hazardous wastes.

A one-time baseline descriptive survey (n = 6,963) did not reveal consistent community differences in the prevalence of chronic or acute respiratory symptoms between incinerator and comparison communities, nor were differences seen in baseline lung function tests or in the average peak expiratory flow rate measured over a period of 35 days.

Conclusions

Based on this analysis of the first year of the study, investigators concluded that they had no evidence of acute or chronic respiratory effects or lung function abnormalities associated with residence in any of the three incinerator communities. They also stated, however, that their "failure to reject the null hypothesis does not warrant acceptance of the null as fact...to the degree that other incinerators burn different wastes or operate under different conditions, our conclusions are only applicable to the three specific incinerators in our study communities."

The 3-year study was completed in 1994, but UNC has not completed the study report at this time. Preliminary indications are that findings from the second and third year were the same as for the first year, and that there were no differences in the measured values between the three incinerator communities when compared to the three comparison communities.

Public Health Assessment Implications

This study is important for public health officials who evaluate incineration facilities, because it was the first study to obtain simultaneously direct measurements of both air quality and respiratory function and symptoms in incinerator and comparison communities. The study included cohorts with pre-existing respiratory problems, sensitive population, as well as normal individuals. It combined environmental sampling concurrently with respiratory symptoms daily diaries (to eliminate recall bias) and objective biomedical testing.

This well-designed study shows that incineration of hazardous, municipal, and biomedical wastes can be done in a safe manner.

The differences (if any) between year 1 and years 2 and 3 in the hazardous waste industrial furnace community will provide for comparisons of health effects when an industrial furnace was and was not burning hazardous wastes. Preliminary reports indicate that there were no differences.

8.2. NIOSH Studies

On-site workers involved with the operation or maintenance of a facility often experience the greatest exposure to contaminants from the facility. Review of the literature identified several NIOSH health hazard evaluations (HHEs) evaluating health implications associated with occupational exposures at incineration facilities. Only four HHEs involved hazardous waste facilities. These studies will be presented first. Summaries of several NIOSH studies at municipal waste combustors (MWCs) and medical/biological waste incinerators (BWIs) are included because the types of problems identified could also be found at hazardous waste thermal treatment facilities. The MWCs and BWIs are also included because occasionally health assessors may be asked to evaluate those types of facilities or hazardous wastes may sometimes be co-treated with those wastes. Sometimes more information on the facility design and background has been included than was in the NIOSH report because of ATSDR staff's familiarity with the facility. The Public Health Assessment Implications section will discuss the application of the study to the public health assessment process, which will help the reader apply the study to other sites.

8.2.1. The Caldwell Group (NC)

Name of Report

Health Hazard Evaluation Report
The Caldwell Group
North Carolina
HETA 90-240-2259 (NIOSH 1992b)

Type of Waste

Hazardous waste, mostly liquid and solid wastes from furniture industry and wastes containing Otto Fuel II (torpedo propellant from U.S. Navy)

Facility Description

The Caldwell Group consisted of three companies, Caldwell Systems, Inc. (CSI), Mitchell Systems, Inc. (MSI) and Caldwell Industrial Services, Inc. (CIS). CIS provided hazardous waste (HW) transportation services. CSI in Lenoir, NC, and MSI, in Spruce Pines, NC, were interim status commercial hazardous waste facilities, each with an incinerator, storage tanks, a drum storage area, and blending tanks. The CSI incinerator is the same facility as discussed in the sections 8.1.1. and 8.1.2.

The incinerators had similar designs, primarily for burning liquid wastes; however, solid or slurried wastes were also fed to them. Neither of the incinerators had any air pollution control equipment until 1987, when CSI was remodeled and air pollution control equipment added. Wastes were often fed to the incinerators far in excess of their design capacity. Prior to 1987, flames and smoke (indicating poor combustion conditions) were often seen coming out of their relatively short stacks.

Background

The 2-acre CSI site is adjacent to the Caldwell County landfill. The incinerator was originally built in 1976 by the county, who operated it until March 1977, when it was leased to CSI. In September 1987, CSI upgraded the facility by adding air pollution control equipment. The Caldwell County Health Department ordered the incinerator closed on May 31, 1988. The plant, however, continued to operate as a waste blending, bulking, and storage facility until September 13, 1989, when a fire in a roll-off container emitted smoke and irritant fumes, causing the evacuation of residents in the area. On September 28, 1989, the Caldwell Superior Court issued an order requiring CSI to cease operations and remove all waste from the facility by December 1, 1989.

The MSI facility was built by the Caldwell Group and began operation in 1980. It was closed in 1985, and later dismantled. In 1983, a local waste transportation and clean-up firm was purchased by the Caldwell Group and renamed CIS. CIS was sold after both incinerators were closed.

In 1987, the NC Division of Occupational Safety and Health (NC-DOSH) investigated the CSI facility and found no excessive exposures to HW during the time of their investigation. In 1989, NIOSH received reports of neurologic problems in former Caldwell workers. In October 1989, NC-DOSH investigated CIS in response to reports of work-related illnesses; however, they found no evidence of hazardous chemical exposures. In August 1990, NIOSH investigators made an unannounced visit to CIS. At the time of the visit, activities with potential for exposure were not being performed, and no HW were seen on site. NIOSH investigators later collected personnel records, and the HW manifests for waste Otto Fuel II.

Summary

In September 1990, NIOSH investigators medically evaluated 14 former Caldwell Group employees to independently assess the workers' neurologic conditions, and to develop a case definition for an epidemiologic study to determine whether the reported neurologic disorders could be associated with work at Caldwell. NIOSH staff confirmed the finding of disabling movement disorders (myoclonus and tremor) in two of the 14 former employees evaluated. Because there was not a high prevalence of an objectively quantifiable finding that could be used as part of a specific case definition, a valid epidemiologic study was not feasible. Still , screening examinations were offered to the 313 other current and former Caldwell employees to address concerns that other employees might have undetected or unreported neurologic disorders.

In November 1991, 54 current and former Caldwell Group employees participated in the screening exams conducted by NIOSH. No additional cases of disabling movement disorders characterized by myoclonus and tremor were found. The most frequent neurologic finding was a mild postural tremor in eight participants.

Some former employees' descriptions of environmental conditions and work practices at the Caldwell facilities suggest that substantial exposures, especially before 1987, might have occurred, although NC-DOSH investigations in 1987 and 1989 found no evidence of hazardous chemical exposures during their visits. The disabled and other former employees reported that their acute symptoms were worst when they handled wastes containing the Otto Fuel II waste torpedo propellant. The disabled employees reported heavy direct skin contact and inhalation exposures. The principal component in Otto Fuel II is propylene glycol dinitrate.

Conclusions

NIOSH concluded that these results did not represent the entire Caldwell workforce or HW workers in general because only 17% of 313 eligible employees participated. NIOSH also concluded that the question of whether the symptoms or movement disorders are related to the HW exposures during work at the Caldwell Group facilities remains unanswered; and that the absence of an indisputable medical explanation for the symptoms could increase the employees' anxiety, which in turn could exacerbate their symptoms.

Public Health Assessment Implications

Because of the inherent potential for adverse health effects related to HW exposures, health assessors should evaluate the potential for exposures to occur during waste unloading and handling. If HW unloading and handling frequently occurs in open areas, i.e., not in buildings under negative pressure, activities that cause worker exposure will likely result in fugitive emissions that could also affect nearby residents. Facilities should use engineering controls, require employees to wear personal protective equipment and clothing, and provide worker training to reduce the employees' exposure to organic solvents and other toxic chemicals.

When an indisputable medical explanation for a worker's or community's symptoms does not exist, supportive counseling may be needed, not on the basis that the medical problem is "all in the patient's head," but rather, as a way to increase coping mechanisms for dealing with something that cannot be understood.

8.2.2. ENSCO (AR)

Name of Report

Health Hazard Evaluation Report
ENSCO
El Dorado, Arkansas
HETA 86-519-1874 (NIOSH 1988)

Type of Waste

PCBs and hazardous wastes (HWs)

Facility Description

The ENSCO incineration facility is a commercial facility that disposed of PCBs and HWs in El Dorado, Arkansas at the time of this NIOSH evaluation. In 1991, ENSCO ceased combusting PCBs. As a part of their RCRA permit ENSCO has been required to fund an Arkansas Department of Environmental Quality (ADEQ) on-site inspection program since 1990. ENSCO employed about 275 workers, including 50 truck drivers. NIOSH described the incinerator as a rotary kiln (PCC) followed by a liquid injection thermal oxidation unit (SCC). The flue gas goes through a scrubber before being discarded through a tall stack. The solids and ash from the kiln discharge into 55-gallon drums.

Wastes are transported to ENSCO by truck or rail, tested by the on-site laboratory, and stored in warehouses prior to incineration. The old PCB warehouse stores liquid PCBs. Drums are opened and the liquids pumped into a holding tank. Solids and sludges are packaged in 10-gallon plastic drums for subsequent shredding and incineration.

Some capacitors and PCB-contaminated solids were stored in the north warehouse, located several miles from the main site, prior to being transported to the main site for incineration. ENSCO was phasing out this warehouse at the time of the NIOSH visits. The new PCB warehouse is on site and is designed for storage of PCB-contaminated solids and capacitors prior to incineration.

Capacitors, other solids, and drummed wastes are hauled by truck from the warehouses to the kiln dock, where they are unloaded onto the dock and then fed by lift truck into a shredder which feeds directly into the rotary kiln. A ram feeder injects plastic drums of flammable liquid organic wastes into the kiln. Liquid PCB and HW are pumped from storage tanks and injected through nozzles in the SCC in a closed system.

Background

In December 1986, the Arkansas Department of Health asked NIOSH to evaluate worker exposure to PCBs at ENSCO. NIOSH personnel conducted an initial inspection on January 27, 1987, and then an in-depth environmental and medical survey March 23-26, 1987. A review of the ENSCO environmental sampling and blood PCB data from May 1985 through December 1986 indicated that the kiln dock area was the highest exposure area and that the old PCB warehouse and North warehouse were intermediate exposure areas.

The wipe sampling at ENSCO documented surface contamination in the warehouses and kiln dock areas, as expected. NIOSH also found significant surface contamination in the break rooms, lunch room, and shower area.

NIOSH's environmental sampling consisted of area air samples, personal air samples, and surface wipe samples. The medical survey included, for each participant, (1) an interview regarding work history and potential exposure to PCB, (2) an examination of the skin of the head and neck for signs of chloracne, and (3) measurement of serum PCB concentration.

Summary

Forty of the 41 air samples contained PCB concentrations greater than the NIOSH recommended exposure limit of 1 µg/m³. The air concentrations ranged from 0.85 to 40 µg/m³ with the highest levels in the kiln dock area. None of the air samples exceeded the OSHA permissible exposure limit (PEL) for PCBs.

Twenty-five of the 28 low-contact surface wipe samples (floors) exceeded the 100 micrograms per square meter (µg/m²) guideline for PCB levels. That guideline is based on the background level in non-industrial buildings. Only 12 of the 20 samples from high-contact surfaces, such as control panels and desk and table tops were greater than 100 µg/m².

All five surface wipe samples analyzed for PCDDs and PCDFs exceeded the 1 nanogram per square meter guideline for total 2,3,7,8-TCDD TEQ. The concentrations ranged from 4-40 ng TCDD TEQ/m²; however, 2,3,7,8-TCDD was not found in any of the samples.

Forty of the 81 serum PCB levels were greater than the general population background level of 20 ppb. Ten PCB blood samples were in the 20-50 ppb range, 14 in the 51-100 ppb range, 10 in the 101-200 ppb range, 1 in the 201-300 ppb range, and 5 were greater than 300 ppb. The serum PCB levels ranged from 2 to 385 ppb. Employees in the production department (which included the kiln dock) had the highest PCB levels with a median concentration of 98 ppb. Warehouse and maintenance workers also had elevated PCB levels with medians of 52 and 46 ppb, respectively. Nine workers had skin findings suggestive of chloracne, but only four of the nine had PCB serum levels greater than 20 ppb, and none of the workers in the 201-385 ppb serum PCB level had signs of chloracne.

Conclusions

NIOSH concluded that the lack of a consistent association between skin findings suggestive of chloracne and serum PCB levels suggests that either the skin findings were not due to chloracne or the cause was something other than PCBs, perhaps PCDD or PCDF.

The environmental and medical data documented excessive exposure to PCBs. The environmental data also documented the presence of PCDD and PCDF, though not 2,3,7,8-TCDD, on environmental surfaces at ENSCO.

NIOSH concluded that additional engineering and administrative controls, work practices, personal protective equipment (PPE), and exposure monitoring were needed at ENSCO to reduce employee exposures.

Public Health Assessment Implications

When ATSDR health assessors are conducting site visits they should (1) wear appropriate PPE for the chemicals present at the site, (2) minimize contact with surfaces in a facility, (3) follow site decontamination procedures, and (4) always wash their hands before eating and before and after using the restroom to prevent exposure and carrying contamination off site.

Waste handling and feed areas are usually the areas with the highest air concentrations of contaminants. The potential for fugitive emissions from these areas reaching the community should be carefully evaluated during site visits.

8.2.3. Allied Chemical (LA)

Name of Report

Health Hazard Evaluation Report
Allied Chemical
Baton Rouge, Louisiana
HETA 80-232-1055 (NIOSH 1982a)

Type of Waste

Hazardous waste

Facility Description

The Allied Chemical plant (Allied) is a polyolefin manufacturing plant located next door to Rollins Environmental Services, Inc. (Rollins), a commercial HW treatment, storage, and disposal facility in Baton Rouge Parish, LA. See section 8.2.4. for a description of the Rollins facility.

Background

In October 1980, NIOSH investigated complaints by 10 Allied employees of health problems they attributed to chemical wastes migrating from the Rollins facility, which was located immediately north of the Allied plant. NIOSH staff inspected the Allied plant for potential workplace exposure due to Allied's manufacturing activities and administered questionnaires to 108 of Allied's 233 employees. The questionnaire collected information on work history, medical history, and symptoms.

Summary

No industrial hygiene samples were collected based on outdoor workers at Allied reporting a significantly higher frequency of symptoms and the complaints being plant-wide, rather than associated with any particular process or operation at Allied. Outdoor workers reported a higher incidence of watery burning eyes, dry mouth, sore throat, cough, chest tightness, chest pain, suffocating feeling, headache, weakness, and nausea than indoor workers. But in this study, no chronic medical problems were significantly elevated.

Fence-line ambient air sampling data collected by the state environmental agency and EPA throughout 1980 were used to evaluate the potential exposures of Allied employees.

Conclusions

Because the measured concentrations of volatile organic vapors did not approach an occupational health standard either singly or in combination, NIOSH concluded that "airborne chemical waste which may migrate to the Allied plant from Rollins does not constitute a serious occupational health hazard."

Public Health Assessment Implications

• Low level concentrations of VOCs migrating from a HW facility could result in complaints of eye and respiratory symptoms.
  
  
• Ambient air sampling data may be available from other agencies for nearby industries that can be used to evaluate off-site exposures.
  
  
• Even though the concentrations of chemicals in the air are not at levels expected to cause health effects, some people could experience symptoms that can cause alarm in the community.
  
  
• Ambient air sampling at the fence line and/or in the community should be conducted for an extended period of time to effectively evaluate public exposure.

8.2.4. Rollins Environmental Services (LA)

Name of Report

Health Hazard Evaluation
Rollins Environmental Services
Baton Rouge, Louisiana
HETA 81-037-1055 (NIOSH 1982b)

Type of Waste

Hazardous wastes

Facility Description

Rollins Environmental Services (Rollins) is a commercial hazardous waste treatment, storage, and disposal facility located on approximately 160 acres in an industrial area of East Baton Rouge Parish. The site has an incinerator, biological stabilization and treatment, landfill, land farm, surface impoundments, drum storage area and crusher, and chemistry laboratory. The plant had about 50 employees at that time on three shifts, who handled approximately 193,000 tons of HW per year. Rollins was later purchased by Safety-Kleen who closed the former Rollins Baton Rouge incineration facility in 1997.

Background

In response to a request from 34 Rollins employees, NIOSH medical and industrial hygiene staff inspected the plant on November 5-6, 1980, conducting 45 medical interviews and collecting 49 screening air samples which were analyzed for VOCs, metals, pesticides, polynuclear aromatics (PNAs), PCBs, hydrogen sulfide, and acid anions. Samples were collected in the unloading pump pad, landfill, land farm, biosystem, oil/water separator ponds, and the drum crusher. Grab samples were taken from the liquid in the land farm holding pit and the cement kiln dust.

NIOSH conducted a follow-up visit March 17-20, 1981, and gathered personal air sampling to evaluate the magnitude of employee exposure to the ubiquitous chemicals (benzene, toluene, and xylene) found in the samples taken in November 1980. NIOSH staff took two samples in the breathing zone of 19 workers. Silica, respirable and total dust, and noise monitoring were also conducted because of the earth moving operations.

Summary

All job categories were exposed to cyclohexane solubles, benzene, toluene, and xylene; however, the exposures did not exceed recognized health standards either singly or when identical target organs and additivity of effect were assumed.

Operations personnel were exposed to naphthalene, indan, indene, and anthracene. Maintenance workers were exposed only to anthracene. Transportation workers had naphthalene and indan exposure. Laboratory personnel were not exposed to any of the specific PNAs measured. Because of the sampling technique used, NIOSH staff felt that the PNAs and arenes detected were in a vapor state.

Eighteen samples for total and respirable dust were collected on the terrigator and grader and one personal respirable dust sample was obtained. Two samples exceeded the OSHA total dust standard and two samples exceeded the 10-hour NIOSH recommended time-weighted average (TWA) of 50 µg/m³ for respirable free silica. All respirable samples exceeded 50 µg/m³ before allowing for nonexposure during nonworking hours.

The eight noise dosimeter measurements taken indicated that high noise exposure occurred during grading operations at the land farm. The incinerator was not operating during the NIOSH visits.

Conclusions

NIOSH concluded that Rollins personnel are exposed to solvents, PNAs, and arenes which were very similar to those compounds measured in the bulk land farm holding pit samples. Excess respiratory and eye irritation among Rollins employees, complaints of eye and lung irritation among Allied workers who worked outdoors, and excess eye irritation complaints from citizens of Alsen may be attributed to dust-bearing PNA materials blowing throughout the Rollins site and off site. The cyclohexane-soluble fraction of airborne particulate concentrations measured at Rollins are known to produce eye irritation.

Land farm workers are overexposed to noise and silica. The landfill personnel could have similar exposures, but were not included in this study.

NIOSH recommended:

• Basic hygiene facilities and storage facilities for protective equipment must be improved.
  
  
• Gloves and aprons should be selected to protect against carcinogenic skin hazards.
  
  
• If crushing drums with residual contents is necessary, the process should be done remotely to prevent operator exposure.
  
  
• The use of pit water for dust control should be discontinued.
  
  
• Wastes should be labeled in a manner understandable to workers, including its hazards and the protective equipment needed when handling it.
  
  
• Supervisors should have first aid training.
  
  
• Landfill personnel should be monitored for silica and noise exposure.
  
  
• Medical surveillance should be established, including pre-employment and periodic follow-up examinations.

Public Health Assessment Implications

Measuring worker exposures to chemicals on site where concentrations are highest will give health assessors data to determine what could be causing community exposures, complaints, or both. All areas of a HW thermal treatment facility (not just the thermal unit) should be evaluated for fugitive emissions and potential for worker and community exposures.

Even though the measured exposures to chemicals in the breathing zone are not at concentrations believed to cause adverse health effects, they could still cause eye or respiratory irritation in some people and be responsible for complaints by workers and community members.

8.2.5. Grosse Pointes-Clinton Refuse Disposal Authority (MI)

Name of Report

Health Hazard Evaluation Report
Grosse Pointes-Clinton Refuse Disposal Authority
Mount Clemens, Michigan
HETA 90-348-2135 (NIOSH 1991a)

Type of Waste

Municipal solid wastes

Facility Design

The Refuse Authority was formed in the early 1960s. The report did not state when the incinerator was constructed.

The incinerator is a continuous feed/mass burner design with two parallel PCCs. The floor of each PCC is an inclined reciprocating grate approximately 12 feet wide by 40 feet long. Refuse is dumped from the garbage trucks into a storage pit, an overhead crane then transfers the refuse to a continuous feed hopper which drops the refuse into the PCC at a controlled rate. The reciprocating grate slowly pushes refuse from the top of the chamber toward a water-filled ash-collection system. When refuse first enters the PCC the heat drives off the water and could pyrolyze some of the more volatile materials. The solids and the pyrolyzed gases ignite as the refuse moves down the grate.

After about 20 minutes the unburned materials and bottom ash reach the end of the PCC grate and fall into a water-filled channel. The 1500°F gases and fly ash are carried into a SCC and then into a conditioning and cooling tower. The cooling tower has water sprays to cool the exhaust gases and capture some of the particulate matter. An induced draft fan downstream of the electrostatic precipitator next draws the gases through the electrostatic precipitator which traps most of the remaining fly ash before exiting the stack. The induced draft fan maintains the system under negative pressure to minimize the escape of gases and fly ash. After the refuse is lit, it burns continuously all week without needing auxiliary fuel burners.

The bottom ash, unburned refuse, and fly ash are all collected in the water-filled conveyor channel located under the incinerator and air pollution control equipment. The fly ash collected in the electrostatic precipitator is carried to the water channel in an enclosed mechanical conveyor. Eventually, all the materials collected in the water channel are removed by a continuous drag conveyor which carries it up a ramp and dumps it into a truck for disposal.

Friday evenings, the furnaces are shut down to allow them to cool overnight. On Saturdays, the PCCs are opened and two to four workers enter them to clean the grates with push brooms. Any built-up slag is broken off with a pick axe. After the PCCs are cleaned, the access doors to the under-fire air chamber below the reciprocal grates are opened to remove, if necessary, any built-up ash or debris. One worker climbs in to perform the cleanup.

Background

In July 1990, NIOSH received a request from employees to evaluate possible employee exposures to silica, lead, cadmium, and mercury from inhalation of ash, particularly during the once-a-week furnace cleanout. The only personal protection equipment employees wear while cleaning the furnace are disposable dust masks. NIOSH investigators visited the incinerator facility in November 1990 and March 1991, to observe the operations and to collect air and ash samples.

Summary

Because the highest exposures were expected to occur during the furnace cleanout, 25 air samples were collected at that time. Some samples were collected in a worker's breathing zone using a battery-powered pump attached to the employee's belt, and some area samplers were suspended from pipes on the inside wall of the PCC at a height of about 5 feet. Ten samples were analyzed for respirable dust (<10 microns in diameter) and crystalline silica, ten samples for total dust and elements, and five samples for mercury.

Four bulk samples of ash were analyzed for arsenic, beryllium, cadmium, chromium, lead, nickel, and zinc. No beryllium or arsenic was detected. While the number of bulk ash samples is too limited to allow for conclusions, the data seem to suggest that fly ash collected in the electrostatic precipitator could have higher metal content than the bottom ash from the PCC. Employees are exposed to the bottom ash when cleaning out the PCC, a task that takes 1 to 2 hours every week.

Two of the four area samplers and one of the six personal samplers indicated average dust concentrations that were well above the limits specified by OSHA and American Conference of Governmental Industrial Hygienists (ACGIH), if one assumes that those exposures continued for 8 hours each day. Because the furnace cleanout lasts less than 2 hours, the 8-hour TWA must be used. Only two of the samples, the area sample taken under the burner grate and one of the personal samples worn by an employee involved in the furnace cleanout, indicated an OSHA violation for total dust. Respirable dust levels were close to, but did not exceed, the OSHA standard. The concentrations of metals in the dust samples were highly variable, but none of them exceeded OSHA or NIOSH recommended limits. One personal sample slightly exceeded the NIOSH standard for crystalline silica.

Conclusions

A dense cloud of ash and other particles became airborne during the incinerator cleanout. Because of the constantly changing composition of municipal refuse from week to week it is likely that the elemental content of the ash will also vary. NIOSH recommended that exposures to dust be minimized by

• Spraying water on the ash before employees enter the furnace to sweep the grates,
  
  
• Controlling the direction of air movement inside the furnace, if possible, to carry the dust away from the employees,
  
  
• Requiring workers to wear full-face piece respirators in a pressure-demand or positive-pressure mode,
  
  
• Not allowing workers to enter the under-fire area until the upper chamber is completely cleaned,
  
  
• Providing workers with goggles and disposable coveralls and requiring them to shower and change before leaving the work site, and
  
  
• Requiring workers entering the under-fire area to wear a safety line and not allowing them to work alone.

Public Health Assessment Implications

This study concludes that workers who enter incinerators to clean them out could be exposed to dust and other particulates (metals, silica, etc.) at levels of health concern if they do not wear appropriate protective equipment. Because the content of the wastes being burned is highly variable at municipal incinerators, as well as at hazardous waste incinerators and desorbers, workers entering thermal treatment units should wear full-face piece respirators and coveralls, and change and shower before leaving the work site.

Confined-space entry procedures should be used when entering incinerator or desorber chambers or air pollution control equipment.

Health assessors should evaluate whether cleaning out incinerators is likely to cause off-site exposure to the dust that is generated.

8.2.6. Delaware source Recovery Facility (PA) Name of Report

Delaware County Resource Recovery Facility
Chester, Pennsylvania
HETA 91-0366-2453 (NIOSH 1994)

Type of Waste

Municipal

Facility Design

The Delaware County Resource Recovery Facility (DCRRF) is a waste-to-energy incinerator that began burning waste on March 6, 1991. The facility incinerates MSW and refuse derived fuel (RDF), a shredded form of MSW, to produce energy. MSW contains primarily paper products, yard and food wastes, metals, rubber, and glass.

The facility has six combustor-boiler trains that use proprietary design water-walled rotary combustors engineered to burn approximately 448 tons of MSW per day. Each combustor consists of a 13.3-foot diameter rotating inclined cylinder and induced draft fans to provide combustion air to the burners. After employees sort the MSW at the plant, the MSW is moved by conveyor to an inclined feed chute leading into the combustor.

Air pollution is controlled using spray dryers and baghouses for each unit. Pulverized lime slurry is injected into a reaction vessel where acid gases (mainly sulfur dioxide and hydrochloric acid) are absorbed. The system design incorporates flue gas evaporation in an atomized lime slurry to produce a dry calcium salt. A baghouse is used downstream of the spray dryer to collect the spray dryer reactant products, unreacted sorbent, and fly ash.

Background

On August 8, 1991, NIOSH received a request for a health hazard evaluation of the DCRRF. The request was submitted on behalf of the contractors regarding conditions during construction and initial operation of the facility. It specifically identified complaints of eye irritation: ear, nose, and throat problems, skin rash, heat stress, and concern about exposure to the lead-containing dust of incinerator ash.

In December 1991, NIOSH conducted a site visit and interviewed employees, collected medical monitoring records and bulk samples of incinerator ash and settled dust samples. Based on the results of this initial walk-through, NIOSH returned in June 1992, to conduct full-shift personal monitoring for lead and other metals, silica, and respirable dust.

Summary

NIOSH collected 34 personal breathing zone and general area air samples. Employees wore two sampling trains: one to sample for respirable dust and silica, and the other to sample for metals. Because NIOSH suspected that workers were transporting ash on their boots to a carpeted administrative area at the DCRRF, investigators collected dust samples from the carpet and from a chair. Wipe samples were also collected from workers' hands and from the tops of tables in the contractor's break/lunch trailer.

No metals were detected at elevated levels in the air samples. None of the breathing zone samples for lead exceeded the OSHA PEL of 50 µg/m³ for an 8-hour TWA. Chromium and cadmium samples did not exceed their respective OSHA PELs, ACGIH Threshold Limit values (TLVs), or NIOSH Recommended Exposure Limits (RELs). Nickel was not detected.

Although Breathing zone and area samples for respirable dust were all below the OSHA PEL of 5 mg/m³, NIOSH concluded that this criteria might not be appropriate because exposure to these dusts also involves a potential exposure to toxic metals. Neither respirable quartz nor cristobalite were detected in the respirable dust samples.

Lead, chromium, cadmium, and nickel were present on the wipe samples from workers' hands and from the tops of lunch tables. Lead was found in greatest abundance. The presence of metals in the wipe samples indicated an increased risk of ingestion of toxic metals for workers in contact with incinerator ash.

Although NIOSH did not cite the results of its bulk analysis, it reported that, according to the DCRRF Municipal Incinerator Ash Residue Monitoring Report dated March 20, 1991, the lead content of the ash was 6.44% lead on a dry-weight basis.

Conclusions

NIOSH concluded that inhalation exposure to lead was not an occupational hazard at the DCRRF. Although lead was found in bulk and wipe samples, the breathing zone samples ranged from nondetectable to 4.6 µg/m³. NIOSH concluded, however, that a potential occupational exposure to toxic metals via ingestion existed. Lead, chromium, cadmium, and nickel were detected in wipe samples collected from lunch tables and the hands of employees. Although breathing zone samples for respirable dust did not exceed the OSHA PEL, NIOSH concluded that this PEL is probably insufficiently protective because of the presence of toxic elements in the incinerator ash. Respirable silica was not detected.

NIOSH recommended improvements in workplace hygienic practices-specifically, hand washing.

NIOSH also recommended characterizing the fugitive ash emissions from the residue building. During the second NIOSH visit, investigators noted that airborne fly ash was escaping from the residue building and blowing towards neighborhoods in nearby Chester, Pennsylvania. The situation appeared to be compounded by the wind coming off the Delaware River. NIOSH noted that the layout and construction of the residue building, with large doors opened at either end, favored wind funneling through the building as the scrap trucks were loaded.

Public Health Assessment Implications

NIOSH investigators did not find any workplace inhalation exposures to metals, respirable dust, or silica in excess of occupational standards. But NIOSH did note that fugitive ash emissions from the residue building were blowing towards neighborhoods. Residents of these neighborhoods could have a small risk of inhalation or dermal exposure to the toxic metals in these dusts. During site visits health assessors should look for potential sources of fugitive emissions that could be migrating off site.

8.2.7. Monroe County Incinerator (FL)

Name of Report

Health Hazard Evaluation Report
Monroe County Incinerator, Key Largo, Florida
HETA 82-056-1186 (NIOSH 1982c)

Type of Waste

Municipal solid wastes and medical wastes

Facility Design

The facility has three reciprocating grate furnaces with combined capacity of 300,000 pounds per 24-hour period. Each incinerator is batch-charged every 8 minutes with 560 pounds of mixed municipal and hospital wastes, or municipal wastes alone. Each batch typically includes about 180 to 200 pounds of hospital waste.

The grate is designed to retain the ash and cinders; however, part of the ash falls through the grate into an ash hopper. An employee must go beneath the incinerator periodically and remove an iron cover allowing ash to fall on the floor, then scrape the balance out by hand or by water spray. Most of the ash is conveyed from the incinerator in a quenched ash conveyor channel. The conveyor channel lacks guard rails at several points.

Rain water from the incinerator building roof drains into a cistern. If the temperature goes above 1800°F, the collected rain water is used to cool the combustion chamber. Auxiliary water is fed to the cistern from the county drinking water system. The type of connection used creates the potential for contaminated cistern water to backflow into the drinking water system. News reports in 1982 indicated a continual low-pressure problem in the county water system.

Background

On November 24, 1981, NIOSH received a request to evaluate the Monroe County Incinerator, North Key Largo Plant #1 in Key Largo, FL. The employees were concerned about biohazards and exposure to fly ash from the incinerated hospital waste. The facility was evaluated on January 6-7, 1982, by the UNC's Occupational Health Studies Group.

The facility began operation in September 1981, and initially burned only household wastes. Private haulers pick up the hospital wastes and bring them to the incinerator. The driver typically backs his truck into the incinerator building and unloads the bagged hospital wastes by hand from the truck into the front-end loader bucket.

Summary

Investigators observed the transferring of hospital waste from trucks to the charging hopper, one employee was observed while he was removing the ash from the ash hopper under the grate, and samples of the grate siftings (ash) were analyzed.

Several employees were interviewed, but none complained of adverse health effects. Employees reported that occasionally the plastic bags of hospital waste break during handling or transport in the truck. One employee reported seeing human body parts and blood being spilled from the truck onto the incinerator building floor. Other employees had not seen body parts but had seen blood leaking from bags onto truck beds and the incinerator floor. All the hospital wastes appeared to be double-bagged and none were broken or leaking during the site visit.

The employee cleaning out the grate siftings wore monogoggles but no dust respirator. The procedure took about 30 minutes. Samples of the ash contained 0.08 to 0.12% volatile material when ignited to 1000°F. The percent of volatile material was consistent with or less than the concentration expected. There was no indication that hospital wastes are leaving the incinerator partly unburned.

Conclusions

• Guard rails are needed along the ash conveyor channel to prevent workers from falling into the channel and being mangled or drowned.
  
  
• The backflow connection between the cistern and the county water line needs to be inspected and replaced if needed to prevent contamination of the county drinking water system.
  
  
• Equipment and procedures to remove and convey hopper ash with minimal worker exposure should be provided. In the meantime, disposable NIOSH approved dust masks and eye protection should be worn by employees while cleaning ash hoppers.
  
  
• No immediate biohazard was identified during the site visit; however, worker exposure to infectious wastes due to breakage of the bags is possible.

Public Health Assessment Implications

Workers at facilities handling infectious or medical waste should wear PPE to prevent exposure to infectious wastes or particulates. The public's off-site exposure to infectious wastes is unlikely. The only potential public exposure identified by this study is the potential for backflow of water from the cistern into the public drinking water supply. At industrial sites, all tanks and water pipes that are connected to a public water system should have a valve to prevent backflow.

8.2.8. Northwest Incinerator (PA)

Name of Report

Northwest Incinerator
Philadelphia, Pennsylvania
HETA 88-207-2195 (NIOSH 1992a)

Type of Waste

Municipal

Facility Description

Built in 1959, the Northwest Incinerator consists of two furnaces. The furnaces operate at 1650°F to 1900°F, and can each burn approximately 700 tons of garbage per day. Crane operators in enclosed cabs dump the garbage into hoppers leading to the inclined stoker and horizontal conveyor belt inside of each furnace. The emission control system for each furnace consists of a cooling tower, drying tower, and an electrostatic precipitator which are fed by an induced draft. Fly ash from the emission system and the incinerator are quenched with water and then cooled in a residue tank. The ash is then dumped into a truck and transported to a landfill.

Background

In March 1988, NIOSH received a joint request from Philadelphia and the American Federation of State, County, and Municipal Employees, District 33, Local 427 to evaluate potential employee exposures at the city's Northwest Incinerator. The request was submitted in response to a March 1988 report issued by EPA and ATSDR, evaluating potential community exposures from the incinerator. Although the report concluded that no significant health hazard existed for the community, it indicated that incinerator workers were the potentially exposed population of most concern, and recommended that the city request NIOSH to evaluate potential worker exposures.

NIOSH conducted an initial site visit on June 9, 1988, and returned for environmental sampling on June 22-26, 1988.

Summary

NIOSH focused its investigation on evaluating potential employee exposure to PCDDs, PCDFs, metals, silica, total dust, and respirable dust. NIOSH collected 12 general air area and 15 personal breathing zone samples (PBZs) for dust, metals and silica. For the PBZs, employees were asked to wear two personal sampling pumps: one measured exposures to respirable dust and silica, the other measured exposure to total dust and metals. NIOSH collected six general air samples for PCDDs and PCDFs; however, PBZs for PCDDs and PCDFs were not collected.

The PCDDs and PCDFs samples were expressed as 2,3,7,8-TCDD equivalents, using both EPA 1987 toxicity equivalency factors (TEFs) and the 1989 International TEFs. Using the 1987 criteria, the six air samples ranged from 0.01 to 12.8 picograms per cubic meter (pg/m³), while using the 1989 criteria the samples ranged from less than 0.001 to 24.2 pg/m³. The only sample exceeding the National Research Council (NRC) guideline of 10 pg/m³ was collected during furnace cleaning.

Five wipe samples were collected for PCDDs and PCDFs at the main office, lunchroom, change room, incinerator floor and a hotel room (for a background sample). Only the sample collected from the incinerator floor was at the NRC acceptable limit for dioxin. The results, however, showed that PCDDs and PCDFs were being transported to the office, lunchroom, and change room via air or on the clothes and shoes of employees.

Airborne concentrations of respirable nuisance dust (27 samples) were all well below the OSHA Permissible Exposure Limit (PEL) of 5 mg/m³. Two of the 27 samples contained trace amounts of silica. Of the 27 total dust samples collected, one exceeded the OSHA PEL of 15 mg/m³ and two exceeded the lower ACGIH TLV of 10 mg/m³. One breathing zone sample exceeded the OSHA PEL for lead and the ACGIH TLV (but not the OSHA PEL) for cadmium. One breathing zone and one area air sample exceeded the NIOSH REL for nickel (0.026 mg/m³).

Seven surface wipe samples collected for metals indicated that the major constituents of the surface dust were metals with relatively low toxicity. But one sample did contain a significant amount of lead. The results suggested that incinerator ash was being transported on the clothes and shoes of employees from the work areas to other areas of the facility.

Conclusions

The environmental sampling data indicated that the potential existed for employee overexposure to PCDDs, PCDFs, and metals via inhalation and surface contact from the incinerator ash. Because the Northwest incinerator closed immediately following the NIOSH survey, the NIOSH recommendations focused on remediation or reopening of the site. NIOSH recommended that remediation or reopening of the site should follow all requirements of RCRA and CERCLA to ensure that personnel entering the site are adequately protected. NIOSH also recommended that a competent industrial hygiene and safety professional be used to help establish an adequate health and safety program if the facility reopens.

Public Health Assessment Implications

Although the environmental sampling data indicated the potential for employee overexposure to both PCDDs and PCDFs via inhalation and dermal contact with the ash, it is unlikely that the public would have been impacted by fugitive emissions from the plant. Employees, however, tracked dioxins and metals to other parts of the plant so they could have also tracked contamination to their vehicles and residences. If employees do not shower and change clothes before leaving work, health assessors should consider potential exposure of employees' families to contaminants at the site.

8.2.9. Lutheran Medical Center (NY)

Name of Report

Lutheran Medical center
Brooklyn, New York
HETA 88-314-2152 (NIOSH 1991b)

Type of Waste

Medical, infectious, pathological, biological

Facility Description

The pathologic waste incinerator was located in a partial 6th floor that covered a portion of the roof. Licensed to burn 175 pounds of infectious waste per hour, the incinerator was manually loaded and employed two stage combustion followed by a scrubber for final control of emissions.

Incinerator emissions passed through the scrubber located in the penthouse above the incinerator room before being released to the outside atmosphere. The scrubber stack extended beyond the roof of the penthouse by approximately 8 feet and the top of the stack was estimated to be approximately 35 feet above the roof of the medical center building.

Air handling units were contained within fan enclosures on the roof of the 5th floor at varying distances from the incinerator/scrubber stack.

Background

On July 12, 1988, NIOSH received a request from employees at the Lutheran Medical Center concerning a variety of noxious odors in the hospital that were potentially originating from the medical center's pathologic waste incinerator. They were concerned that exposures to incineration products could be affecting their health.

On November 29, 1988, NIOSH investigators conducted a preliminary evaluation of conditions at the hospital and interviewed workers to determine the extent of their work-related health complaints. Employees reported headaches, nausea, hair loss, and dermatitis. Several workers indicated that the Neonatal Intensive Care Unit was affected by intermittent episodes of noxious odors, and that patients had been evacuated due to the odors on one occasion. Based on this initial investigation, NIOSH returned in April 1989, to conduct an industrial hygiene survey and tracer gas study to evaluate the potential for reentry of incinerator exhaust emissions into the hospital's ventilation systems.

Summary

NIOSH focused its investigation on assessing the potential for the hospital's ventilation system to entrain emissions from the hospital's pathologic waste incinerator. In addition to obtaining and reviewing drawings of the hospital's ventilation system, NIOSH investigators conducted a tracer gas study and collected environmental samples. Collecting both general-area air and breathing zone samples, NIOSH sampled for total and respirable particulates, metals, and VOCs. NIOSH also monitored carbon dioxide levels throughout the hospital.

The highest detectable airborne concentrations of total and respirable particulates were less than 5% of the NIOSH REL of 10 mg/m³ and 5 mg/m³, respectively.

General-area air samples collected for qualitative screening for VOCs showed that all samples contained toluene, xylenes, isopropanol, some alkanes, and a series of various aliphatic hydrocarbons. Based on the results of the qualitative screenings for VOCs, 23 quantitative samples were analyzed for the compounds detected qualitatively. Toluene, xylene, isopropanol, trichloroethylene, and 1,1,1-trichloroethane were detected in some samples but all detected levels were less than 1% of NIOSH RELs.

Carbon dioxide levels inside the hospital ranged from 575 ppm to 850 ppm. As the indoor concentration approaches and exceeds 1,000 ppm there is an indication that inadequate amounts of fresh air are being delivered to those areas.

Tracer-gas sampling in rooms served by air units #1 and #4 showed that entry of incinerator stack emissions occurred under certain meteorological conditions. In the afternoon, prevailing winds blew the incinerator emissions towards the air units fresh-air intakes. But the calculated dilution factors were large, and emissions were greatly diluted.

Conclusions

The tracer-gas evaluation showed that entrainment of incinerator stack emissions was possible under certain meteorological conditions; however, there were no documented over-exposures to any of the chemical substances evaluated. NIOSH recommended evaluating the effect of stack height, modifying the air handling unit fresh air intakes-or both-in the event the pathologic waste incinerator was restarted.

Public Health Assessment Implications

When evaluating thermal treatment facilities, public health officials should determine whether stack emissions are likely to enter the fresh air intake units for on-site and nearby buildings.

The books listed here do not exhaust all the resources available on the topic of incineration and desorption technologies. Nevertheless, we believe they are the resources health assessors will find most useful in addressing health concerns and technical issues at hazardous waste thermal treatment technologies.

9.1. ATSDR Resource Books

The following ATSDR documents are available through ATSDR or the National Technical Information Services (NTIS). Many of them are also on ATSDR's Web site at http://www.atsdr.cdc.gov

• Public Health Overview of Incineration as a Means to Destroy Hazardous Waste - Guidance to ATSDR Health Assessors
  
  
• Environmental Data Needed for Public Health Assessments - A Guidance Manual
  
  
• Public Health Assessment Guidance Manual
  
  
• Toxicological Profiles (for different chemicals)

9.2. EPA Resource Books

Copies of the following EPA documents can be obtained from NTIS. Copies of the EPA hazardous waste regulations, including the preamble and background documents, as well as a number of memos, and letters regarding incineration, industrial furnaces, and thermal desorption can be found on EPA's Web site at http://www.epa.gov/oswer

EPA also has two combustion training modules on their web site under the heading "RCRA Hotline Training Modules"; one is called Introduction to Hazardous Waste Incineration and the other one is Introduction to Strategy for Hazardous Waste Minimization and Combustion.

• Guidance Manual for Hazardous Waste Incinerator Permits. Volume I of the Hazardous Waste Incineration Guidance Series. Office of Solid Waste and Emergency Response. SW966. July 1983.
  
  
• Handbook: Guidance on Setting Permit Conditions and Reporting Trial Burn Results. Volume II of the Hazardous Waste Incineration Guidance Series. Office of Research and Development. EPA/625/6-89/019. January 1989.
  
  
• Handbook: Hazardous Waste Incineration Measurement Guidance Manual. Volume III of the Hazardous Waste Incineration Guidance Series. Office of Solid Waste and Emergency Response. EPA/625/6-89/021. June 1989.
  
  
• Handbook: Quality Assurance/Quality Control (QA/QC) Procedures for Hazardous Waste Incineration. Office of Research and Development. EPA/625/6-89/023. January 1990.
  
  
• Technical Implementation Document for EPA's Boiler and Industrial Furnace Regulations. Office of Solid Waste and Emergency Response. EPA-530-R-92-011. March 1992.
  
  
• Final Technical Support Document for Hazardous Waste Combustor MACT Standards, Volume IV: Compliance with the MACT Standards. Office of Solid Waste and Emergency Response. July 1999.
  
  
• Engineering Handbook for Hazardous Waste Incineration. Office of Solid Waste. SW-889. September 1981.
  
  
• Background Information Document for the Development of Regulations for PIC Emissions from Hazardous Waste Incinerators. Office of Solid Waste. October 1989.
  
  
• Human Health Risk Assessment Protocol (HHRAP) for Hazardous Waste Combustion Facilities. EPA 530-D-98-00l A. B, and C. July 1998

9.3. Other Technical Resource Books

Numerous books and journal articles are available on the various aspects of incineration as a technology and on the permitting of hazardous waste incinerators. However, references regarding thermal desorption are limited. Following is list of books which readers may find useful.

• Hazardous Waste Incineration - a Resource Document. The American Society of Mechanical Engineers. New York: ASME. January 1988.
  
  
• Innovative Site Remediation Technology - Thermal Desorption. American Academy of Environmental Engineers. Anderson WC, editor. Annapolis, MD: AAEE copyright; 1993.
  
  
• Innovative Site Remediation Technology - Thermal Destruction. American Academy of Environmental Engineers. Anderson WC, editor. Annapolis, MD: AAEE copyright; 1994.
  
  
• Hazardous Waste Incineration: Evaluating the Human Health and Environmental Risks. Roberts, Teaf, Bean, ed. Lewis Publishers. Baco Raton, FL:1999.
  
  
• Hazardous Waste Incineration and Human Health. Travis CC, Cook SC. CRC Press, Inc. Boca Raton, FL: 1989.
  
  
• Health Effects of Municipal Waste Incineration. Hattemer-Frey HA, Travis CC, ed. CRC Press, Inc. Boca Raton, FL: 1991.
  
  
• Waste Incineration and Public Health. National Research Council. National Academy Press. Washington, DC: Sep 1999.
  
  
• Proceedings - International Conference on Incineration and Thermal Treatment Technologies for years: 1991, 1993, 1995, 1998, 1999, 2000. Copyright: The Regents of the University of California, Irvine.

References

AAEE 1993               American Academy of Environmental Engineers. 1993. Innovative site remediation technology - thermal desorption. Anderson WC, editor. Annapolis: AAEE.

ATSDR 1991            Agency for Toxic Substances and Disease Registry. 1991. Health and safety program for hazardous substance field activities: policy and procedures. Atlanta: US Department of Health and Human Services.

ATSDR 1992a           Agency for Toxic Substances and Disease Registry. 1992a. Public health overview of incineration as a means to destroy hazardous wastes - guidance to ATSDR health assessors. Atlanta: US Department of Health and Human Services.

ATSDR 1992b           Agency for Toxic Substances and Disease Registry. 1992b. Public health assessment guidance manual. Atlanta: US Department of Health and Human Services.

ATSDR 1993a           Agency for Toxic Substances and Disease Registry. 1993a. Public health advisory procedures. Atlanta: US Department of Health and Human Services.

ATSDR 1993b           Agency for Toxic Substances and Disease Registry. 1993b. Study of symptom and disease prevalence Caldwell Systems, Inc. hazardous waste incinerator, Caldwell County, North Carolina. Atlanta: US Department of Health and Human Services.

ATSDR 1993c           Agency for Toxic Substances and Disease Registry. 1993c. Proceedings of the expert panel workshop to evaluate the public health implications for the treatment and disposal of polychlorinated biphenyls-contaminated waste. Lichtveld MY,Susten AS, editors. Atlanta: US Department of Health and Human Services.

ATSDR 1995             Agency for Toxic Substances and Disease Registry. 1995. Symptom and illness prevalence with biomarkers health study for Calvert City and Southern Livingston County, Kentucky. Atlanta: US Department of Health and Human Services.

ATSDR 1997a           Agency for Toxic Substances and Disease Registry. 1997a. ATSDR record of activity for telephone communication with Bob Montione of New York State Department of Health. Atlanta: US Department of Health and Human Services.

ATSDR 1997b           Agency for Toxic Substances and Disease Registry. 1997b. Exposure investigation - VERTAC, Incorporated. Atlanta: US Department of Health and Human Services.

ATSDR 1998             Agency for Toxic Substances and Disease Registry. 1998. Health outcome follow-up study of residents living near the Caldwell Systems, Inc. site, Caldwell County, North Carolina. Atlanta: US Department of Health and Human Services.

ATSDR 2000             Agency for Toxic Substances and Disease Registry. 2000. ATSDR record of telephone communication with Robert Field of EPA Region VII. Atlanta: US Department of Health and Human Services.

ATSDR/EPA 1997                     EPA designs for air impact assessments at hazardous waste sites.1997. Manual containing lectures by ATSDR and EPA/ERT for course at ATSDR. Atlanta: US Department of Health and Human Services.

Breasted et al 1998                     Breasted MA, Kirby RS and Canino C. 1998. Adverse reproductive outcomes in Pulaski County for years 1980 through 1990. By Arkansas Reproductive Health Monitoring System. Atlanta: US Department of Health and Human Services.

Costner and Thornton 1990         Costner P and Thornton J. 1990. Playing with fire-hazardous waste incineration. Washington, DC: Greenpeace USA.

EPA 1989                US Environmental Protection Agency. 1989. Guidance on setting permit conditions and reporting trial burn results, Vol II of the hazardous waste incineration guidance series. EPA/625/6-89/019. Washington, DC: EPA.

EPA 1990                US Environmental Protection Agency. 1990. Minimization and control of hazardous combustion byproducts. EPA/600/2-90/039. Washington, DC: EPA.

EPA 1992                US Environmental Protection Agency. 1992. Control of air emissions from superfund sites. EPA/625/R-92/012. Washington, DC: EPA.

EPA 1993                US Environmental Protection Agency. 1993. Engineering bulletin - thermal desorption treatment. Vol. 2. EPA/540/0-00/000. OERR, Washington, DC and ORD, Cincinnati, OH.

EPA 1994                US Environmental Protection Agency. 1994. Exposure assessment guidance for RCRA hazardous waste combustion facilities. EPA530-R-94-021. Washington, DC: EPA.

EPA 1998a               US Environmental Protection Agency. 1998a. RCRA, Superfund & EPCRA hotline training module - introduction to: applicable or relevant and appropriate requirements. EPA540-R-98-020 OSWER9205.5-10A. Washington, DC: EPA.

EPA 1998b               US Environmental Protection Agency.1998b. Human health risk assessment protocol for hazardous waste combustion facilities. EPA530-D-98-001A, B & C, peer review draft. Washington, DC: EPA.

EPA 2001                 US Environmental Protection Agency. 2001. Electronic mail to Betty C. Willis at ATSDR from Andrew Opalko, subject: revised thermal desorber definition. Washington, DC: EPA.

64 FR 52869-70       NESHAPS: final standards for hazardous air pollutants for hazardous waste combustors. Federal Register 1999 Sep 30; 64:52828-53077.

Kulwiec 1985             Kulwiec RA, editor. 1985. Materials handling handbook. New York: John Wiley & Sons.

Maaskant 2001          Maaskant OL. 2001. The Shell system for NOx removal and dioxin destruction from incineration flue gas. 3rd International Symposium on Incineration and flue gas treatment technologies; Brussels, Belgium.

NIOSH 1982a           National Institute of Occupational Safety and Health. 1982a. Health hazard evaluation report HETA 80-232-1055 Allied Chemical, Baton Rouge, Louisiana. Cincinnati: US Department of Health and Human Services.

NIOSH 1982b           National Institute of Occupational Safety and Health. 1982b. Health hazard evaluation report HETA 81-037-1055 Rollins Environmental Services, Baton Rouge, Louisiana. Cincinnati: US Department of Health and Human Services.

NIOSH 1982c           National Institute of Occupational Safety and Health. 1982c. Health hazard evaluation report HETA 82-056-1186 Monroe County incinerator, Key Largo, Florida. Cincinnati: US Department of Health and Human Services.

NIOSH 1988             National Institute of Occupational Safety and Health. 1988. Health hazard evaluation report HETA 86-519-1874 ENSCO, El Dorado, Arkansas. Cincinnati: US Department of Health and Human Services.

NIOSH 1991a           National Institute of Occupational Safety and Health. 1991a. Health hazard evaluation report HETA 90-348-2135 Grosse Pointes-Clinton Refuse Disposal Authority, Mount Clemens, Michigan. Cincinnati: US Department of Health and Human Services.

NIOSH 1991b           National Institute of Occupational Safety and Health. 1991b. Health hazard evaluation report HETA 88-314-2151 Lutheran Medical Center, Brooklyn, New York. Cincinnati: US Department of Health and Human Services.

NIOSH 1992a           National Institute of Occupational Safety and Health. 1992a. Health hazard evaluation report HETA 88-207-2195 Northwest incinerator, Philadelphia, Pennsylvania. Cincinnati: US Department of Health and Human Services.

NIOSH 1992b           National Institute of Occupational Safety and Health. 1992b. Health hazard evaluation report HETA 90-240-2259 the Caldwell Group, North Carolina. Cincinnati: US Department of Health and Human Services.

NIOSH 1994             National Institute of Occupational Safety and Health. 1994. Health hazard evaluation report HETA 91-0366-2453 Delaware County Resource Recovery Facility, Chester, Pennsylvania. Cincinnati: US Department of Health andHuman Services.

NIOSH 1995             National Institute of Occupational Safety and Health. 1995. Health hazard evaluation report HETA 90-0329-2482 New York City Department of Sanitation, New York, New York. Cincinnati: US Department of Health and Human Services.

Pitts 1998                   Pitts D. 1998. Metals: control, compliance and monitoring-advanced tutorial. Presented at: International Conference on Incineration and Thermal Treatment Technologies Meeting; 1998 May 11-15, Salt Lake City.

Pranghofer and Fritsky 2001             Pranghofer GG and Fritsky KJ. 2001. Destruction of polychlorinated dibenzo-p-dioxins and dibenzo furans in fabric filters: recent experiences with a catalytic filter system. 3rd International Symposium on Incineration on Flue Gas Treatment Technologies. Brussels, Belgium.

Roberts 1999              Roberts DW. 1999. Dioxin incinerator emissions exposure study, Times Beach, Missouri. By Missouri Department of Health. Atlanta: US Department of Health and Human Services.

Shy et al. 1995            Shy CM et al. 1995. Do waste incinerators induce adverse respiratory effects? An air quality and epidemiology study of six communities. Env Health Perspectives 103(7-8):714-24.

Siret and Bessy 2001   Siret B. and Bessy C. 2001. DEDIOXLAB: an efficient process for removing and destroying dioxins from flue gas. 3rd International Symposium on Incineration on Flue Gas Treatment Technologies. Brussels, Belgium.

---

This page last updated on May 10, 2002

Lateefah A. Wooten / lpw1@cdc.gov

---

ATSDR Home  |  Search  |  Index  |  Glossary  |  Contact Us
About ATSDR  |  News Archive  |  ToxFAQs  |  Public Health Assessments
Privacy Policy  |  External Links Disclaimer  |  Accessibility
U.S. Department of Health and Human Services