Ames Electric Department

Electric Department
502 Carroll, Ames, IA 50010

Customer Service
515 Clark Ave., Ames, IA 50010

Phone:
(515) 239-5120
* * * * *

Project Share Committee

February 5 Minutes (pdf)

Dioxin Report

A Report Prepared for the City of Ames, Iowa
by: Robert C. Brown
Director and Professor
Center for Sustainable Environmental Technologies
Iowa State University
June 26, 2001
EXECUTIVE SUMMARY

Introduction

In September 2000, the North American Commission for Environmental Cooperation (NACEC) released a report entitled "Long-range Air Transport of Dioxin from North American Sources to Ecologically Vulnerable Receptors in Nunavut, Arctic Canada" prepared for the NACEC by Dr. Barry Commoner and his colleagues at the Center for the Biology of Natural Systems (CBNS). This report, henceforth referred to as the "Commoner study," details a computer modeling study of the transport of airborne dioxin (that is, poly-chlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans, also referred to by the acronym PCDD/PCDF) from widely distributed sources on the North America continent to the Canadian polar territory of Nunavut. This computer study suggests that dioxin formed during burning of high chlorine-content fuels, such as municipal solid waste (MSW), can be transported thousands of miles through the atmosphere before being deposited in the Canadian arctic.

The report received wide publicity in Iowa because of its claim that the Ames Municipal Power Plant is among the top one or two sources of dioxin polluting several locations in Nunavut, Arctic Canada (bordering Hudson Bay). This claim, based on computer modeling, is contrary to measurements made almost two decades ago on the Ames Municipal Power Plant, which revealed undetectable levels of dioxin in the flue. Although Ames mixes MSW with the coal it burns, it was concluded that good combustion practice at the facility destroyed organic compounds thought to be precursors to dioxin formation.

When confronted with this information, Commoner’s response was two-fold. First, analytical instrumentation to detect dioxin has vastly improved in the two decades since testing was last performed at the Ames Municipal Power Plant. He implied that the original measurements were incapable of measuring dangerous levels of dioxin emitted from the plant. Second, he noted that recent research demonstrates that dioxin can form at much lower temperatures than was previously suspected. Just because the Ames furnace operates at very high combustion temperatures does not assure that dioxin will not form in gas cooling sections of the boiler. Commoner suggested that, if dioxin is not detected within the boiler, it is very likely forming somewhere in the exhaust plume outside the power plant. Commoner called for comprehensive and periodic testing for dioxin from the Ames Municipal Power Plant.

The present study was commissioned by the city of Ames with the following objectives:

Review the Commoner study to understand its claims as well as its underlying assumptions
Determine regulatory limits and monitoring requirements for dioxin as promulgated by the U.S. Environmental Protection Agency (EPA)
Identify factors that influence dioxin emissions and their relevance to the Ames Municipal Power plant
Estimate dioxin emissions from the Ames Municipal Power plant using available scientific literature and compare these estimates to those made by Commoner

Commoner Study

The computer model used in the Commoner study employs two major components: an air transport model and a dioxin emission inventory. Air transport models have been widely tested and validated by numerous research groups and can be expected to give reasonable semi-quantitative results for well-characterized sources and sinks of transported chemical compounds. The dioxin emission inventory employed by Commoner to arrive at this emission rate for Ames Municipal Power plant is more suspect than the air transport model used to calculate the ATC for Ames and is the focus of the present report.

This number assumes that the amount of dioxin emitted per unit of fuel burned at the Ames facility is equal to the average dioxin emission factor for municipal waste combustors (MWCs) in the United States (31.57 nanograms per kilogram). It also assumes that the Ames facility burns 210 metric tons of fuel per hour, presumably all municipal solid waste (MSW). In fact, the Ames facility only burns 27 metric tons per hour, which immediately reduces the expected dioxin emission to 7.5 g TEQ/yr. Furthermore, as subsequently described, application of an average dioxin emission factor greatly overestimates the emissions from Ames.

Regulation of Dioxin Emissions

For this reason, dioxin emission factors or dioxin concentrations are reported in terms of grams or nanograms toxicity equivalency quotient (g TEQ or ng TEQ) to better reflect the potential impact on human health. The specific emission standards are a function of the size, air pollution control device (APCD) configuration, and age of the facility (note that emission standards are in terms of dioxin concentration in flue gas: ng/dscm where ng is nanograms or billionths of a gram and dscm is "dry standard cubic meters" of flue gas).

Although PCDD/PCDF emissions cannot be continuously monitored at MWCs, the U.S. EPA requires operating and emission parameters that correlate with PCDD/PCDF emissions to be monitored: CO, boiler steam load, particulate matter (PM) control device inlet temperature; opacity and SO2 are monitored to guarantee proper operation of the flue gas cleaning equipment. Each MWC is also subject to annual PCDD/PCDF compliance testing. The Ames facility, although not subject to these limits, is meeting those associated with good combustion practice.

PCDD/PCDF emission standards established by the U.S. EPA (Federal Register, 1999)

Emission standard (ng total/dscm)	Emission standard (ng TEQ/dscm)	Facility age, size, and APCD*
60	1.0	Existing; >225 metric tpd; ESP-based APCD
30	0.5	Existing; >225 metric tpd; non-ESP-based APCD
125	2.1	Existing; >35 to 225 metric tpd
13	0.2	New; >35 metric tpd

* Air pollution control device

Factors Influencing Dioxin Emissions

A number of factors influence the actual dioxin emissions leaving a combustion system. These are discussed in the following paragraphs.

Composition and properties of fuel

It is well known that combustion of fossil fuels like coal generates much less PCDD/PCDF than combustion of MSW. The levels reported for individual dioxin or furan compounds from coal-fired power plants is very low (i.e., 0.033 ng/dscm) or not detected. The reason for the very low emission rate of dioxin from coal-fired boilers is thought to be the relatively high sulfur to chlorine (S/Cl) ratio of coal compared to MSW. Sulfur short-circuits the synthesis of dioxin by converting chlorine gas released from burning plastics to hydrogen chloride, which is not effective in producing dioxin. Thus, power plants that burn a blend of coal and MSW can be expected to produce very little PCDD/PCDF.

Type of Combustion and Air Pollution Control Equipment

Good combustion practice (GCP) is important to reducing dioxin formation by completely oxidizing the organic compounds and carbonaceous solids that are precursors to dioxin formation. Good combustion practice consists of "time, temperature, and turbulence" in the combustion chamber. Combustion temperatures approaching 1800° F at residence times of 1 – 2 s will destroy most gas-phase compounds, conditions achieved at the Ames facility.

Flue gas composition and process conditions downstream of the furnace/boiler section of MWC determine the extent of dioxin synthesis, which occur in the temperature range of 250° – 600° C. At higher temperatures chloro-organics are rapidly destroyed while at low temperatures the reaction rate is minimal.

MWC facilities that employ good combustion practice and perform continuous emission monitoring can achieve dioxin emissions averaging 0.13 ng TEQ/dscm, which is equivalent to a DEF at the Ames facility of only 0.24 ng TEQ/kg.

Estimate of Dioxin Emissions from the Ames Municipal Power Plant

The City of Ames produces a fuel blend that is 80-90%-wt coal and 10-20%-wt MSW. The "worst case scenario" of 20%-wt MSW, which gives the highest chlorine concentrations in the flue gas, is assumed for this analysis. The blended fuel has a minimum sulfur-to-chlorine (S/Cl) ratio of 1.75, well above the value of 1.0 reported in the scientific literature as critical for suppression of dioxin.

Four different dioxin emission scenarios were compared to EPA limits for MWCs as well as to actual measurements made on the Ames facility in 1981. The first scenario uses data employed by Commoner. The second scenario corrects for Commoner’s overestimate of fuel throughput at the Ames facility. The third scenario also corrects for the overestimate of dioxin emission factor (DEF) assumed by Commoner. The fourth scenario recognizes the mitigating effect of sulfur in the coal burned in the Ames facility. The fifth scenario is the actual test data obtained from the Ames facility in 1981, which is consistent with the low dioxin emission scenarios.

This analysis suggests that dioxin emission from the Ames facility most likely falls in the range of 0.14 to 0.25 g TEQ/yr. This range is well within the EPA limit for MWC (1.8 g TEQ/yr for large plants or 3.9 g TEQ/yr for small plants). Furthermore, any one of the three alternative scenarios evaluated removes the Ames facility from the list of "worst polluters" that the Commoner study targeted for remedial action.

Conclusions

Commoner’s study overestimates both the amount of fuel burned at the Ames facility as well as the dioxin emission factor appropriate to the kind of combustion equipment and fuel blends used. Accordingly, the Ames plant emits about four hundred times less dioxin than assumed by Commoner and meets the dioxin emission standards set by the U.S. EPA. Based on this analysis, dioxin testing is not justified at the Ames plant.

Expected dioxin emissions from Ames Municipal Power Plant for various scenarios

Scenario	Dioxin concentration (ng TEQ/dscm) or DEF (ng TEQ/g)	Annual Emission (g TEQ/yr)	Comment
Commoner estimate	31.57 ng TEQ /kg	58	Overestimates fuel use by factor of 10
Commoner estimate with corrected fuel throughput	31.57 ng TEQ /kg	7.5	Uses average DEF for U.S. MWC capacity
Kilgroe (1996) estimate for MWC meeting EPA monitoring requirements*	0.13 ng TEQ /dscm	0.24	Does not account for mitigating effect of sulfur in fuel
Based on DEF for industrial-scale coal-fired combustor (CRE, 1994)	0.6 ng TEQ /kg	0.14	Assumes Ames facility operates with S/Cl > 1.0
Measurements at Ames facility in 1981	< 0.5 ng TEQ/dscm	< 0.9	Measurement was of 2,3,7,8-tetra-CDD with a TEF of 1.0
EPA limit on large MWC, ESP-based APCD	1.0 ng TEQ /dscm	1.8	Assumes 100% fuel burned in Ames is MSW
EPA limit on small MWC	2.1 ng TEQ /dscm	3.9	Assumes 10% of fuel burned in Ames is MSW

1. Introduction

In September 2000, the North American Commission for Environmental Cooperation

(NACEC) released a report entitled "Long-range Air Transport of Dioxin from North American

Sources to Ecologically Vulnerable Receptors in Nunavut, Arctic Canada" prepared for the NACEC by Dr. Barry Commoner and his colleagues at the Center for the Biology of Natural Systems (CBNS) (Commoner, 2000). This report, henceforth referred to as the "Commoner study," details a computer modeling study of the transport of airborne dioxin (that is, poly-chlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans, also referred to by the acronym PCDD/PCDF) from widely distributed sources on the North America continent to the Canadian polar territory of Nunavut. This computer study suggests that dioxin formed during burning of high chlorine-content fuels, such as municipal solid waste (MSW), can be transported thousands of miles through the atmosphere before being deposited in the Canadian arctic.

The report received wide publicity in Iowa because of its claim that the Ames Municipal Power Plant is among the top one or two sources of dioxin polluting several locations in Nunavut, Arctic Canada (bordering Hudson Bay) (Ames Tribune, 2000). This claim, based on computer modeling, is contrary to measurements made almost two decades ago on the Ames Municipal Power Plant, which revealed undetectable levels of dioxin in the flue gas (Junk and Richard, 1981). Although Ames mixes MSW with the coal it burns, it was concluded that good combustion practice at the facility destroyed organic compounds thought to be precursors to dioxin formation.

When confronted with this information, Commoner’s response was two-fold (Ames Tribune, 2000). First, analytical instrumentation to detect dioxin has vastly improved in the two decades since testing was last performed at the Ames Municipal Power Plant. He implied that the original measurements were incapable of measuring dangerous levels of dioxin emitted from the plant. Second, he noted that recent research demonstrates that dioxin can form at much lower temperatures than was previously suspected. Just because the Ames furnace operates at very high combustion temperatures does not assure that dioxin will not form in gas cooling sections of the boiler. Commoner suggested that, if dioxin is not detected within the boiler, it is very likely forming somewhere in the exhaust plume outside the power plant. Commoner called for comprehensive and periodic testing for dioxin from the Ames Municipal Power Plant.

The present study was commissioned by the city of Ames with the following objectives:

Review the Commoner study to understand its claims as well as its underlying assumptions
Determine regulatory limits and monitoring requirements for dioxin as promulgated by the U.S. Environmental Protection Agency (EPA)
Identify factors that influence dioxin emissions and their relevance to the Ames Municipal Power plant
Estimate dioxin emissions from the Ames Municipal Power plant using available scientific literature and compare these estimates to those made by Commoner

The results of these evaluations are detailed in the following sections.

2. The Commoner Study

The computer model used in the Commoner study employs two major components: an air transport model and a dioxin emission inventory. The air transport model, HYSPLIT (Hybrid Single-particle Lagrangian Integrated Trajectory), was originally developed by the National Oceanic and Atmospheric Administration (NOAA) to track the movement of inorganic radionuclides. Air transport models have been widely tested and validated by numerous research groups and can be expected to give reasonable semi-quantitative results for well-characterized sources and sinks of transported chemical compounds. Commoner and his colleagues at CBNS adapted the model to the transport of dioxin in the atmosphere (CBNS, 1995). This adaptation required the addition of a mechanism for photolytic destruction of organic pollutant in the atmosphere by sunlight, but this does not bring into question the basic soundness of the air transport model.

Language in the executive summary of the Commoner study released by NACEC has led to a misperception among some critics of the study that the air transport model assumes point sources, such as Ames Municipal Power plant, each emit one gram Toxic Equivalent Quotient per hour (1 g TEQ/h) of dioxin into the atmosphere. A careful reading of the full report reveals that, to make the computational problem tractable, the 44,091 sources of dioxin accounted for in North America are modeled as "105 standard (hypothetical) source points" each of which is assumed to emit 1 g TEQ/h "puffs" of dioxin in a four-hour interval (p. 3 of Commoner study). The model then determines the amount of this dioxin deposited at downwind locations (receptors) for each of the standard sources: the fraction of the assumed source emission deposited at a given receptor is called the Air Transfer Coefficient (ATC). Interpolation of computational results for the 105 standard source points allows ATC values to be estimated for the 44,091 actual source points in North America. Once an ATC has been determined for a source point, the amount of dioxin deposited at a particular receptor can be calculated from:

Amount deposited at receptor = ATC x amount emitted by source (1)

Whether emission of dioxin from 44,091 actual source points scattered over 9.1 million square miles of the North America continent can be accurately modeled by only 105 standard source points is unclear and can only be resolved by evaluating various discretizations of the computational domain and/or explicitly validating the model by comparing its predictions against experimental data, a point further discussed below. However, what is clear is that the model does not assume 1 g TEQ/h emission of dioxin from the Ames Municipal Power plant. In fact, Table 5.3 on page 83 of the Commoner study assumes dioxin emission from Ames to be fifty-eight grams TEQ per year (58 g TEQ/y or 6.6 x 10-3 g TEQ/h), which is a much smaller but still very large emission rate.

The dioxin emission inventory employed by Commoner to arrive at this emission rate for Ames Municipal Power plant is more suspect than the air transport model used to calculate the ATC for Ames. Separate inventories were developed for Canada, Mexico, and the United States. Commoner reports that the United States inventory was largely based on an inventory developed by the U.S. EPA (1998) although CBNS added several classes absent in the U.S. EPA inventory: iron sintering plants, electric arc furnaces, coal-burning power plants and backyard trash burning. This inventory is not a listing of individual dioxin emission sources, since very few of the thousands of potential emission sources have actually been tested for dioxin. Instead, the inventory is a characterization of twenty-eight different categories of combustion equipment and the range of dioxin emissions that have been measured for similar equipment. The kinds of combustion equipment range from municipal waste combustors (MWC) to metal smelters. Each of these is characterized by an average dioxin emission factor (DEF), which was used to estimate dioxin emissions from a specific facility according to the equation:

Amount emitted by source = DEF x Facility Throughput (2)

For example, the Commoner study assigns MWCs a DEF of 31.57 nanograms TEQ per kilogram throughput (ng TEQ/kg), as reported in Table 2.1 of the Commoner study. A MWC burning 210 metric tons of waste per hour (tph) would emit:

Amount emitted by source = (31.57 ng TEQ /g) x (210 x 103 kg/h)

x (24 h/day) x (365 day/yr) x (10-9 g/ng)

= 58 g TEQ/yr

which is the amount of dioxin Commoner assumed is emitted from the Ames Municipal Power Plant. In fact, the Ames facility is designed to burn only 45 tph of fuel (coal and MSW), which is nearly one-fifth the amount assumed in the Commoner study. Furthermore, the plant typically operates at only a fraction of its design capacity. In 2000, the plant operated at 60% of capacity with a throughput of only 27 tph (Titus, 2001). This difference immediately reduces the upper bound on average dioxin emission from Ames to 7.5 g TEQ/yr.

Furthermore, on page 13 of his report, Commoner reveals that "emission factors for conventional sources, such as incinerators, are based on fuel throughput, not the amount combusted." In other words, Commoner assumes that everything entering the Ames plant is combustible, highly chlorinated MSW, which is far from correct. Less than 20% of the total tonnage burned in Ames is MSW, the balance being coal, which is thought to be an insignificant source of dioxin, as subsequently described. In 2000, the plant burned only 10%-wt MSW in the fuel blend (Titus, 2000). Assuming that dioxin emission is proportional to the amount of highly chlorinated fuel burned, the upper bound on average dioxin emissions from Ames is further reduced to only 0.75 g TEQ/yr. However, dioxin emissions are not simply proportional to the chlorine throughput for a plant. As subsequently explained, even sharper reductions are expected for high sulfur fuels.

This eight to eighty fold reduction in the estimated dioxin emissions from Ames arises solely from errors Commoner made in estimating the throughput of highly chlorinated MSW at Ames. As will be demonstrated in Section 4 and 5, additional factors overlooked by Commoner overestimates the dioxin emission factor in Eq. 2, which further reduces the amount of dioxin emitted from the Ames Municipal Power Plant.

Finally, the study has a shortcoming in research philosophy that diminishes its value as a decision making tool for the community of Ames. The stated objective of the Commoner study is to "test the efficacy of the HYSPLIT air transport model as a means of ranking the North American sources of airborne dioxin with respect to their contribution to the amount of dioxin deposited on ecologically vulnerable receptors in the Inuit territory, Nunavut" (p. 2 of the Commoner study). Thus, one expects that the study will attempt to validate the accuracy of model predictions. In computer simulations, this is customarily accomplished by two complementary techniques.

The first involves exercising the computer model for different consolidations of point sources, different spatial mesh sizes, and different time steps to see if model results converge to the same solution. Page 6 of the Commoner study acknowledges the sensitivity of the model to these various parameters but only the effect of different time steps appears to have been explicitly evaluated. The omission of other parametric evaluations is particularly distressing considering that only 105 standard source points were used to represent 44,091 actual source points spread over 9.1 million miles.

The second method for validating computer models is to compare model predictions to a set of experimental data. Granted, this task can be daunting for a model approaching the global scope of Commoner’s dioxin study of the North American continent. Nevertheless, a computer model, like a scientific theory, is of little value until it has been tested against experiment. The most logical data to collect for validating the model is total deposition rates at selected receptors in the Canadian Arctic. Commoner recognizes the importance of validating models, but he offers very little in the way of experimental data to support his predictions. In the absence of data on receptor deposition rates, he turns to a mere three measurements of dioxin concentration is the tissue of Caribou herds as a proxy. He can make no quantitative comparisons and only the barest of qualitative judgments: both predicted dioxin deposition rates and dioxin concentrations in tissue of Caribou herds increase in moving geographically from west to east.

3. Regulation of Dioxin Emissions

Dioxin refers to 210 compounds – 75 dioxins and 135 furans – with similar structures and properties. Dioxins are collectively known as polychlorinated dibenzo-p-dioxins (PCDD) while furans are polychlorinated dibenzofurans (PCDF). Only 17 of the 210 compounds are toxic, and these differ considerably in their potency. Toxicity, with respect to carcinogenicity, is commonly expressed in terms of the Toxicity Equivalency Factor (TEF), i.e., relative to the most toxic of the dioxin compounds: 2,3,7,8-tetrachloro-dibenzo-p-dioxin (2,3,7,8-TCDD). Thus the concentration of dioxin in flue gas can be expressed in either of two ways: total dioxin concentration or toxic equivalent quotient concentration. Total dioxin concentration, Ctotal, in billionths of a gram per dry standard cubic meter (ng/dscm) is given by:

(3)

where Ci is the concentration of individual dioxin or furan compounds with respect to dry flue gas at standard conditions. The summations are performed over all dioxin and furan compounds present in the flue gas. New Source Performance Standards promulgated by the federal government have typically been expressed in terms of total dioxin concentration.

On the other hand, a better indicator of the health threat of dioxin is the toxic equivalent quotient (TEQ) concentration, CTEQ, in billionths of a gram TEQ per dry standard cubic meter (ng TEQ/dscm):

(4)

Note that the TEQ concentration may be significantly lower than the total concentration because the TEFs for many dioxin and furan compounds are less than one. Since both the Commoner study and most of the scientific literature is reported in terms of TEQ, it shall be the basis of analysis in the present report.

In late 1989, the U.S. EPA proposed new source performance standards (NSPS) on dioxin emission from all new MWCs as well as emission guidelines (EG) for all existing MWCs (EPA, 1989). The following year Congress passed amendments to the Clean Air Act (Clean Air Amendments, 1990) that directed the U.S. EPA to establish MWC emission limits for several pollutants, including PCDD/PCDF. Limits were to be based on maximum achievable control technology (MACT). In 1994 the U.S. EPA proposed revised NSPS and EGs for MWCs. These revised rules required use of good combustion practice (GCP) and MACT flue gas cleaning techniques to continuously limit emission of PCDD/PCDF as well as carbon monoxide (CO), particulate matter (PM), cadmium (Cd), mercury (Hg), lead (Pb), hydrogen chloride (HCl), sulfur dioxide (SO2), and nitrogen oxides (NOx) (EPA, 1994).

On December 19, 1995, EPA promulgated PCDD/PCDF emission standards for all existing and new MWC units with aggregate capacities to combust greater than 35 metric tons per day (Federal Register, 1995). In response to a court remand, the regulations were subsequently amended to remove small MWC units (i.e., units with capacities ranging from 35 to 225 kkg/day) (Federal Register, 1997). EPA reproposed emission standards for small MWCs (defined as units with capacities of between 32 and 224 tons/day) on August 30, 1999 (Federal

Table 1. PCDD/PCDF emission standards established by the U.S. EPA (Federal Register, 1999)

Emission standard (ng total/dscm)	Emission standard (ng TEQ/dscm)	MWC facility age, size, and APCD*
60	1.0	Existing; >225 metric tpd; ESP-based APCD
30	0.5	Existing; >225 metric tpd; non-ESP-based APCD
125	2.1	Existing; >35 to 225 metric tpd
13	0.2	New; >35 metric tpd

* Air pollution control device

Register, 1999). The specific emission standards are a function of the size, air pollution control device (APCD) configuration, and age of the facility as listed in Table 1.

The Ames Municipal Power plant is not classified as an MWC: it is a pulverized coal combustor supplemented with 10 - 20% wt. RDF. However, for comparisons with the EPA regulations on MWC, the Ames facility must be classified as either a small power plant or a large power plant. For total fuel throughput of 386 metric tons per day, the Ames Municipal Power plant might be classified as a large plant with an emission limit of 0.5 ng TEQ/dscm. However, in terms of MSW throughput, which is only 65 metric tons per day, the Ames facility resembles a small plant with an emission limit of 2.1 ng TEQ/dscm.

Although PCDD/PCDF emissions cannot be continuously monitored at MWC, the U.S. EPA requires operating and emission parameters that correlate with PCDD/PCDF emissions to be monitored: CO, boiler steam load, PM control device inlet temperature; opacity and SO2 are monitored to guarantee proper operation of the flue gas cleaning equipment. Each MWC is also subject to annual PCDD/PCDF compliance testing. The operating and emissions limits for PCDD/PCDF compliance as well as the corresponding parameter for the Ames facility are listed in Table 2. The Ames facility, although it is not subject to these limits, is meeting those associated with good combustion practice: both CO and opacity are well below permitted levels. However, SO2 runs higher than permitted for MWC because no sulfur control equipment is used at the Ames facility. Also, the temperature entering the APCD, which is a hot ESP, is about 100º C hotter than is considered desirable by the U.S. EPA (cold ESP or, better yet, fabric filters generally give lower dioxin emissions that hot ESP). The European Union has mandated more stringent dioxin regulations, requiring incinerators to reduce dioxin emissions to 0.1 ng TEQ/dscm since 1996 (European Union, 1994).

Table 2. EPA limits for PCDD/PCDF compliance from existing MWCs

Emission limit	Small plant (35 –225 tpd)	Large plant (> 225 tpd)	Ames Facility
CO* (continuous)	150 ppm	150 ppm	< 100 ppm
Steam load (continuous)	Normal	Normal	60% Design
PM control device inlet temperature (continuous)	250° C	250° C	~360° C
Opacity (continuous)	10%	10%	1.5%
SO2 (continuous)	80 ppm (or 50% reduction)	35 ppm (or 75% reduction)	144 ppm

* Coal/RDF mixed fuel-fired

As a final note, clear progress has been made in the United States in improving combustion equipment to limit dioxin emissions once the problem was recognized. The U.S. EPA estimates that total dioxin emissions in the United States during 1995 was 1,100 g TEQ /yr. In contrast, municipal waste combustors in the late 1980’s are estimated to have emitted nearly 8,000 g TEQ/yr of dioxin (U.S. EPA, 2000).

4. Factors Influencing Dioxin Emissions

Combustion-generated dioxin can arise from three mechanisms (Altwicker, 1992): gas phase reactions involving chlorinated precursors such as chlorobenzenes (CB), chlorophenols (CP) and polychlorinated biphenyls (PCB); condensation reactions involving gas-phase precursors and fly ash; and solid-phase reactions involving metal chlorides and fly ash carbon, a process known as de Novo synthesis. The first mechanism is not thought important in MWC flue gases (Shaub and Tsang, 1983). The second mechanism can be largely avoided by complete combustion of gas-phase organic compounds. The third mechanism, de Novo synthesis, since it involves hard-to-oxidize carbonaceous material in fly ash, can be the predominant source of dioxin in combustion equipment. A number of factors influence the actual dioxin emissions leaving a combustion system. These are discussed in the following paragraphs.

Composition and properties of fuel

The amount of dioxin formed is believed to increase substantially during combustion upsets associated with improper feed conditions. It is important to blend or mix waste prior to combustion to reduce variations in heating content, volatility, and moisture content (Kilgroe et al., 1990).

It is well know that combustion of fossil fuels like coal generates much less PCDD/PCDF than combustion of MSW. The U.S. Department of Energy sponsored a project in 1993 to assess emissions of hazardous air pollutants at coal-fired power plants. As part of this project, CDD/CDF stack emissions were measured at seven U.S. coal-fired power plants (Riggs et al., 1995). The levels reported for individual dioxin or furan compounds were very low (i.e., 0.033 ng/dscm) or not detected. Riggs could not attribute variation in emissions between plants to any specific fuel or operational characteristic.

Bremmer et al. (1994) reported the results of emission measurements at two coal-fired facilities in The Netherlands. The emission factor reported for a pulverized coal electric power plant equipped with an ESP and a wet scrubber for sulfur removal was 0.35 ng TEQ /kg of coal fired (equivalent to 0.02 ng TEQ /dscm at 11% O2). The emission factor for a chain grate stoker equipped with a cyclone APCD was 1.6 ng TEQ /kg of coal fired (or 0.16 ng TEQ /dscm at 11% O2).

No testing of the PCDD/PCDF content of air emissions from commercial/industrial (as opposed to electric utility) coal-fired combustion units in the United States have been identified. However, CRE (1994) reported results of testing at 13 commercial/ industrial coal-fired boilers in the United Kingdom. The emission factors ranged from 0.04 to 4.8 ng TEQ /kg coal combusted (mean value of 0.6 ng TEQ /kg).

The reason for the very low emission rate of dioxin from coal-fired boilers is thought to be the relatively high sulfur to chlorine (S/Cl) ratio of coal compared to MSW. Griffin (1986) noted that S/Cl ratios in coal are 5:1 while the S/Cl ratios in MSW are only 0.33:1. Griffin hypothesized that the higher chlorine content of MSW allows molecular chlorine (Cl2) to form by the Deacon reaction catalyzed by metals; e.g., copper:

CuCl2 + ½ O2 ® CuO + Cl2 (5)

CuO + 2HCl ® CuCl2 + H2O (6)

The net reaction of this catalytic process is described by:

2HCl + ½ O2 ® H2O + Cl2 (7)

Molecular chlorine readily chlorinates aromatic compounds both in the gas phase and in the presence of fly ash (Gullet, 1990). Without fly ash Cl2 was 4 times more efficient than HCl in chlorinating these compounds (Gullet et al. 1994).

However, this chlorination path can be short-circuited by sulfur in the form of sulfur dioxide (SO2) (Griffen, 1986):

Cl2 + SO2 + H2O ® SO3 + 2HCl (8)

Raghunathan and Gullet (1996) showed that addition of HCl produced PCDD/PCDF while addition of SO2 suppressed the formation of PCDD/PCDF. The authors determined a critical S/Cl ratio of about 1.0 in coal-flames above which dioxin formation was suppressed.

Thus, power plants that burn a blend of coal and MSW can be expected to produce very little PCDD/PCDF if sulfur in the coal is in sufficient quantities to react with chlorine in the MSW, a fact confirmed in several studies. Kimble and Gross (1980) detected no appreciable tetrachlorinated dibenzodioxin, the most toxic of the dioxin and furan compounds, in the exhaust of a combined coal/MSW power plant. More recently, pilot-scale tests by Xie and coworkers (2000) found undetectable levels of PCDD/PCDF when coal and PVC, a chlorinated plastic, were burned together. Ohlsson et al. (1990) found that co-firing of coal with RDF produced undetectable levels of PCDD/PCDF. Tests in a large-scale MSW incinerator found appreciably lower PCDD/PCDF concentrations when high sulfur coal was added to the fuel (Lindbauer et al., 1992). Even modest blending of coal, as little as 5%-wt, with MSW can dramatically reduce PCDD/PCDF emissions, as demonstrated by Gullet et al. (1998).

Type of Combustion and Air Pollution Control Equipment

Fuel composition and process conditions in the furnace/boiler sections of MWC determine the specific products of incomplete combustion (PIC) that are precursors for PCDD/PCDF formation. Implicated PICs include: chlorobenzenes (CBs), chlorophenols (CPs), polychlorinated biphenyls (PCBs), and the carbon in fly ash (Kilgroe, 1995). Furnace destruction of organics must include both gas- and condensed-phase organics. Field tests have shown that dioxin formation correlates with CO and total hydrocarbon (THC) concentrations in flue gas and with the amounts of PM carried out of the combustor with the flue gas (Kilgroe, 1990).

Thus, good combustion practice (GCP) is important to reducing dioxin formation by completely oxidizing the organic compounds and carbonaceous solids that are precursors to dioxin formation. Good combustion practice consists of "time, temperature, and turbulence" in the combustion chamber. Time scales at elevated temperatures required to destroy PICs are milliseconds for gaseous components and second to minutes for small solid particles. Combustion temperatures approaching 980° C (1800° F) at residence times of 1 – 2 s will destroy most gas-phase compounds (Kilgroe, 1990). Excess air and good mixing of fuel and air (turbulence) are also important to complete combustion. However, GCP may not be satisfactory for oxidizing all unburned carbon in fly ash, which can participate in heterogeneous reactions that form PCDD/PCDF by the third of the three dioxin formation mechanisms described above.

Flue gas composition and process conditions downstream of the furnace/boiler section of MWC determine the extent of de Novo reactions, which occur in the temperature range of 250° – 600° C (Vogg and Stieglitz, 1986). At higher temperatures chloro-organics are rapidly destroyed while at low temperatures the reaction rate is minimal. Maximum formation rates for tetra- through octa-CDD/PCDF (the most important compounds from a risk perspective) occurs at 300° C. Accordingly, the preferred location to generate dioxin is the economizer and equipment for dedusting, especially electrostatic precipitators (Vogg, 1995), which operate near this temperature. Studies in Germany suggest that dust deposits and longer residence times of particles in the cooling zone at temperatures of 250° C - 450° C may favor de Novo synthesis of dioxin (LAI, 1995).

Clearly, the amount of dioxin formed by the de Novo mechanism depends upon the length of time flue gas spends in the temperature range of 250° C - 450° C, a fact confirmed by laboratory studies (Fangmark et al., 1994). This temperature range is most typically encountered in economizers, air heaters, and PM control devices. PM control device, because they concentrate fly ash and operate at temperatures favorable for de Novo synthesis, can function as chemical reactors that generate and emit PCDD/PCDF (Kilgroe, 1990). Limiting the temperature at which PM control devices are operated is important in controlling the formation and emission of PCDD/PCDF 914-16). Lowering PM control device operating temperatures to less than 250° C results in a major reduction in PCDD/PCDF formation rates and alters the partitioning of vapor- and solid-phase PCDD/PCDF (EPA, 1989). In general, APCDs based on fabric filters is preferred to cold ESP, which is preferred to hot ESP.

The U.S. EPA has tabulated dioxin emission factors (DEF) for several categories of MWC, including three varieties of mass burn units, two varieties of modular units, dedicated RDF units, and fluidized beds, in combination with various pollution control devices (EPA, 2000). A representative sample of this information is summarized in Table 3. None of these units resemble the Ames facility, which is a pulverized coal boiler with hot electrostatic precipitators (H-ESP). Commoner employed a DEF of 31.57 ng TEQ/kg, which appears to be an average value for total MWC capacity in the United States derived from U.S. EPA data (EPA, 2000). However, MWC facilities that employ good combustion practice and perform continuous emission monitoring of the process variables listed in Table 2 can achieve dioxin emissions

Table 3. Dioxin Emission Factors (DEF) for Selected Combinations of MWC and Air Pollution Control Devices (EPA, 2000)

MWC Designation	Air Pollution Control Device	Average DEF (ng TEQ/kg)
Mass Burn Water Wall	C-ESP	6.1
Mass Burn Water Wall	H-ESP	473
Mass Burn Refractory	DS-FF	0.65
Mass Burn Rotary Kiln	C-ESP	47.0
RDF Dedicated	DS-FF	0.24
Modular – Starved Air	C-ESP	16.2
Modular – Starved Air	H-ESP	79.0
Modular – Excess Air	DS-FF	16.2
Fluidized Bed RDF	DS-FF	0.63

C-ESP: Cold electrostatic precipitator

DS-FF: Dry scrubber combined with fabric filter

H-ESP: Hot electrostatic precipitator

averaging 0.13 ng TEQ/dscm, which is equivalent to a DEF at the Ames facility of only 1.0 ng TEQ/kg (Kilgroe, 1996). Clearly, the DEF value employed by Commoner overestimates dioxin emissions from the Ames facility even if it is classified as an MWC.

5. Estimate of Dioxin Emissions from the Ames Municipal Power Plant

The first step in this analysis is to determine the elemental analysis for the coal/MSW fuel blend used at the Ames Municipal Power Plant. Table 4 lists the most recent elemental analyses for the Rochelle North Antelope Western coal and the MSW burned at the Ames facility. Chlorine content of MSW is highly variable; the value in Table 4 is the maximum concentration measured, which is thirty times higher than the chlorine content of the coal. The City of Ames produces a fuel blend that is 80-90%-wt coal and 10-20%-wt MSW. The "worst case scenario" of 20%-wt MSW, which gives the highest chlorine concentrations in the flue gas, is assumed for this analysis and is presented in Table 4. Note that the blended fuel has a minimum sulfur-to-chlorine (S/Cl) ratio of 1.75, well above the value of 1.0 reported in the scientific literature as critical for suppression of dioxin (Raghunathan and Gullet, 1996).

The blended fuel requires 6.2 kg dry air per kg fuel for stoichiometric combustion, yielding 6.5 kg dry flue gas per kg fuel burned. In practice, anywhere from 20% to 50% air in excess of this amount is employed to assure complete combustion. For estimating annual emissions of dioxin (g/yr) based on dioxin concentrations, the combustion air requirement must be adjusted to 7% oxygen, which is equivalent to 49% excess air for this particular fuel blend. This adjustment increases the amount of flue gas produced to 9.5 kg/kg fuel. Furthermore, the amount of flue gas must be presented on a volumetric basis at standard temperature and pressure (20° C and 1 atmosphere), which is 7.67 scm/kg for the blended fuel described in Table 4. The annual fuel consumption at the Ames Municipal Power plant is reportedly 16 tph. Thus, estimates of annual dioxin emissions based on dioxin concentrations (ng TEQ/dscm) employ the formula:

Dioxin emission (g TEQ/yr) = CTEQ x (7.67 dscm/kg) x (27 x 103 kg/hr) x 8760 hr/yr

x (10-9 g/ng) (9)

Estimates of annual dioxin emissions based on dioxin emission factors (DEF), which are based on ng dioxin emitted per kg of fuel burned, employ the formula:

Dioxin emission (g TEQ/yr) = DEFTEQ x (27 x 103 kg/hr) x 8760 hr/yr x 10-9 g/ng (10)

Table 5 compares expected emissions for different scenarios to EPA limits for MWCs as well as to actual measurements made on the Ames facility in 1981. The first scenario uses data

Table 4. Fuel analysis for coal, MSW, and blended fuel burned at Ames Municipal Power Plant

Constituent	Coal (%-wt)	MSW (%-wt)	*Blend (%-wt)**
C	51.7	35.5	48.5
H	3.5	4.8	3.8
O	12.6	25.1	15.1
N	0.73	0.40	0.66
S	0.22	0.15	0.21
Cl	0.02	0.59 (max)	0.12 (max)
Moisture	27.0	26.5	26.9
Ash	4.38	7.6	5.0
S/Cl ratio	11.0	0.25 (min)	1.75 (min)

*Fuel blend is assumed to be 80%-wt coal + 20%-wt MSW

Table 5. Expected dioxin emissions from Ames Municipal Power Plant for various scenarios

Scenario	Dioxin concentration (ng TEQ/scm) or DEF (ng TEQ/g)	Annual Emission (g TEQ/yr)	Comment
Commoner estimate	31.57 ng TEQ /kg	58	Overestimates fuel throughput by factor of 10
Commoner estimate with corrected fuel throughput	31.57 ng TEQ /kg	7.5	Uses average DEF for total MWC capacity in U.S.
Kilgroe (1996) estimate for MWC meeting EPA monitoring requirements*	0.13 ng TEQ /dscm	0.24	Does not account for mitigating effect of sulfur in fuel
Based on DEF for industrial-scale coal-fired combustor (CRE, 1994)	0.6 ng TEQ /kg	0.14	Assumes Ames facility operates with S/Cl > 1.0
Measurements at Ames facility in 1981	< 0.5 ng TEQ/dscm	< 0.9	Measurement was of 2,3,7,8-tetra-CDD, which has a TEF of 1.0
EPA limit on large MWC, ESP-based APCD	1.0 ng TEQ /dscm	1.8	Assumes 100% fuel burned in Ames is MSW
EPA limit on small MWC	2.1 ng TEQ /dscm	3.9	Assumes 10% of fuel burned in Ames is MSW

* See Table 2

employed by Commoner. As already explained, Commoner overestimated throughput of fuel at the Ames facility by over a factor of almost ten. Correcting for this error reduces annual emission estimates from 58 g TEQ/yr to 7.5 g TEQ/yr.

The Commoner study also overestimates the DEF for the Ames facility. Since the study included over 44,000 emission sources, it was not practical for his research team to individually classify each of the MWC included in the study. Instead, they assumed an average DEF of 31.57 ng TEQ/kg for all MWC. In fact, dioxin emissions vary several orders of magnitude for different kinds of MWC. Although the U.S. EPA does not classify the pulverized coal-fired boiler in Ames as a MWC, its achievement of good combustion practice, including combustion temperatures exceeding 980 C and operating conditions specified in Table 2, suggests that its DEF is closer to the average DEF of 0.13 ng TEQ /dscm reported by Kilgroe (1996) for well-operated MWC than is the industry-averaged value employed by Commoner. Accordingly, if the Ames facility were classified as an MWC, likely emissions of dioxin would be around 0.24 g TEQ/yr (the fact that one of the two Ames boilers employs a hot ESP rather than a cold ESP or fabric filter adds some uncertainty to this conclusion).

However, even this evaluation ultimately overestimates the likely dioxin emissions from Ames because it assumes all the fuel burned is highly chlorinated and ignores the mitigating effect of sulfur in the fuel. Since the coal/MSW fuel blend used by Ames has sulfur to chlorine ratio exceeding 1.0 (and very little copper and zinc to catalyze dioxin forming reactions), dioxin emissions are expected to more closely resemble those of a coal-fired plant than those of a MWC. As previously noted, many coal-fired utility boilers have undetectable or nearly undetectable dioxin emissions. Adopting a conservative approach, the CRE average value of 0.6 ng TEQ /kg for industrial and commercial coal-fired boilers is employed in Table 5 to estimate the effect of combined coal/MSW firing (CRE, 1994). In this case, dioxin emissions are very likely on the order of 0.14 g TEQ/yr. This value is consistent with dioxin measurements made on the Ames facility in 1981. Junk and Richard (1981) of the Ames National Laboratory reported no measurable levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin, which is by far the most toxic of the dioxin and furan compounds (TEF = 1.0). The detection limit in these tests was 0.5 ng TEQ/dscm, which translates to less than 0.9 g TEQ/yr of dioxin emitted from the Ames plant.

This analysis suggests that dioxin emission from the Ames facility most likely falls in the range of 0.14 to 0.25 g TEQ/yr. This range is well within the EPA limit for MWC (1.8 g TEQ/yr for large plants or 3.9 g TEQ/yr for small plants). Furthermore, any one of the three alternative scenarios presented in Table 5 removes the Ames facility from the list of "worst polluters" that the Commoner study targeted for remedial action.

As a final point of concern, Commoner raised the specter of dioxin forming in the gas plume leaving the exhaust stack, downwind of the plant. This possibility was presented as a counterargument to the contention that no dioxin was detected in the Ames facility twenty years ago. Commoner’s suggestion is based on the fact that de Novo synthesis of dioxin can occur at much lower temperatures than was expected several years ago.

In a refractory walled incinerator with no heat recovery, post-stack synthesis of dioxin is a credible possibility since the temperature of exhaust gases can be several hundred degrees Centigrade. However, in a power plant, flue gas is cooled to less than 250° C before it is emitted from the plant, which is below the range for significant de Novo synthesis of dioxin.

Although Commoner’s argument for post-stack synthesis of dioxin is flawed, it illustrates the difficulties of attempting to respond to the charges he has made against the Ames facility. If Ames undertook in-plant testing demanded by Commoner and dioxin emissions were found to be negligible, the controversy would not be at an end. Instead, Commoner could turn to his post-stack theory of dioxin synthesis and suggest plume sampling as the only way to resolve definitively the controversy. It is unlikely that such difficult measurements have ever been undertaken. However, the cost would undoubtedly be several fold more expensive than the $80,000 in-plant testing currently being contemplated by the community of Ames.

6. Conclusions

7. References

Altwicker, E. Fundamental aspects of dioxins (PCDD) from combustion, Report 92-3, New York State Energy Research and Development Authority.

Ames Tribune (2000) November 15.

Center for the Biology of Natural Systems (1995) Quantitative estimation of the entry of dioxins, furans, and hexachlorobenzene into the Great Lakes from airborne and waterborne sources. Final report to the Joyce Foundation.

Clean Air Amendments (1990) P. L. 101-549, U.S. Congress, Washington, DC (November 15).

CRE (1994) Emissions of environmental concern from coal utilisation. United Kingdom: Coal

Research Establishment. ECSC 7220-EC/011. June 1994.

Commoner, B., Woods Bartlett, P.W., Eisel, H., Couchot, K., Long-range Air Transport of Dioxin from North American Sources to Ecologically Vulnerable Receptors in Nunavut, Arctic Canada, Final Report, North American Commission for Environmental Cooperation, September 2000.

European Union (1994) Directive 94/67/CE, Appendix I.

Fangmark, I., et al. (1994) Environ. Sic. Technol. 28, 624.

Federal Register (1995) Standards of performance for new stationary sources and emission guidelines for existing sources: municipal waste combustors, final rule.

F.R. (December 19) 60:65387-65436.

Federal Register (1997) Standards of performance for new stationary sources and emission guidelines for existing sources; final rule. F.R. (September 15) 62:48347-48391.

Federal Register (1999) Emission guidelines for existing stationary sources: small municipal waste combustion units; proposed rule. New source performance standards for new small municipal waste combustion units. F.R. (August 30) 64:47233-47307.

Griffin, R. D. (1986) A new theory of dioxin formation in municipal solid waste combustion. Chemosphere 15, 1987-1990.

Gullett, B. K., Bruce, K. R., and Beach, L. O. (1990) Formation of chlorinated organics during solid waste combustion. Waste Manage. Res. 8, 203- 214.

Gullett, B. K., Lemieux, P. M., and Dunn, J. E. (1994) Role of combustion and sorbent parameters in prevention of polychlorinated dibenzo-p-dioxin and polychlorinated dibenzofuran formation during waste combustion. Environ. Sci. Technol. 28, 107-118.

Gullett, B. K., Raghunathan, K, and Dunn, J. E. (1998) "The effect of cofiring high-sulfur coal with municipal waste on formation of polychlorinated dibenzodioxin and polychlorinated dibenzofuran, Environ. Eng. Sci. 15, 59-70.

Junk, G. A. and Richard, J. J. (1981) Dioxin not detected in effluents from coal/refuse combustion, Chemosphere 10, 1237-1241.

Kilgroe, J. D., et al. (1990) Combust. Sci. Tech. 74, 223.

Kilgroe, J. D. (1996) Control of dioxin, furan, and mercury emissions from municipal waste combustors, J. Hazardous Materials 47, 163-194.

Kimble, B. J. and Gross, M. I. (1980) Tetrachlorodibenzo-p-dioxin quantification in stack-collected coal fly ash, Science 207, 59.

LAI (LanderausschluB fur Immissionsshutz) (1995) Determination of requirements to limit emissions of dioxins and furans. UBA Texte 58/95. Environmental Agency, Berlin, Germany.

Lindbauer, R. L., Wurst, F., and Prey, T. (1992) Combustion dioxin suppression in municipal solid waste incineration with sulfur additives. Chemosphere 25, 1409 – 1414.

Ohlsson, O. O., Livengood, C. D., and Daugherty, K. E. (1990) Results of emissions and ash testing gin full-scale co-combustion tests of binder-enhanced dRDF pellets and high-sulfur coals, Air and Waste Management Association Forum 90, Pittsburgh, PA, June 24-29.

Raghunathan, K. and Gullelt, B. K. (1996) Role of sulfur in reducing PCDD and PCDF formation. Environ. Sci. Technol. 30, 1827 – 1834.

Riggs, K.B., Brown, T.D., Schrock, M.E. (1995) PCDD/PCDF emissions from coal-fired power plants. Organohalogen Compounds 24, 51-54.

Shaub, W. M. and Tsang (1983) W. Environ. Sci. Technol. 17, 721.

Titus, G. (2000) Personal communication, May 18.

U. S. Environmental Protection Agency (1989) Background information for proposed standards, post-combustion technology performance, Vol. 3, EPA-450/3-89-27c (NTIS PB90-154865) Research Triangle Park, NC, August.

U. S. Environmental Protection Agency (1989) Air pollution standards of performance for new stationary sources and emission guidelines; Final rules, 40 CFR Parts 60, 51, and 52 (December 20).

U. S. Environmental Protection Agency (1994) Standards of performance for new stationary sources: Municipal waste combustors; Emission guidelines: Municipal waste combustors; proposed rules, 40 CFR Part 60 (September 20).

U. S. Environmental Protection Agency (1998) The inventory of sources of dioxin in the United States, EPA/600/P-98/002Aa.

U.S. Environmental Protection Agency (2000) E-mail from John Bachmann (EPA/OAQPS) to Dwain Winters (EPA/OPPT). Revised MWI, MWC dioxin estimates. May 1, 2000.

U.S. Environmental Protection Agency (2000) Draft Dioxin Reassessment Report, Volume 2: Sources of Dioxin-Like Compounds in the United States, Chpt. 3. Combustion sources of CDD/CDF: Waste Incineration (http://www.epa.gov/ncea/pdfs/dioxin/part1/volume2/chap3.pdf)

Vogg, H. and Stieglitz, L. (1986) Thermal behavior of PCDD/PCDF in fly ash from municipal incinerators. Chemosphere 15, 1373-1378.

Vogg, H. (1995) PCDD/PCDF und Abfallverbrennung. Organohalogen Compd. 22, 31-48.

Xie, Y., Xie, W., Kunlei, L. Dicen, L., Pan, W. P., and Riley, J. T. (2000) "The effect of sulfur dioxide on the formation of molecular chlorine during co-combustion of fuels," Energy & Fuels 14, 597-602

Back to Electric Services

Hit Counter views since January 27, 2004