Journal of the Air & Waste Management Association

ISSN: 1096-2247 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uawm20

An Assessment of Costs and Benefits Associated with Mercury Emission Reductions from Major Anthropogenic Sources Jozef M. Pacyna , Kyrre Sundseth , Elisabeth G. Pacyna , Wojciech Jozewicz , John Munthe , Mohammed Belhaj & Stefan Aström To cite this article: Jozef M. Pacyna , Kyrre Sundseth , Elisabeth G. Pacyna , Wojciech Jozewicz , John Munthe , Mohammed Belhaj & Stefan Aström (2010) An Assessment of Costs and Benefits Associated with Mercury Emission Reductions from Major Anthropogenic Sources, Journal of the Air & Waste Management Association, 60:3, 302-315, DOI: 10.3155/1047-3289.60.3.302 To link to this article: http://dx.doi.org/10.3155/1047-3289.60.3.302

Published online: 24 Jan 2012.

Submit your article to this journal

Article views: 377

View related articles

Citing articles: 27 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uawm20 Download by: [37.44.207.62]

Date: 22 January 2017, At: 20:26

ISSN:1047-3289 J. Air & Waste Manage. Assoc. 60:302–315 DOI:10.3155/1047-3289.60.3.302

TECHNICAL PAPER

Copyright 2010 Air & Waste Management Association

An Assessment of Costs and Benefits Associated with Mercury Emission Reductions from Major Anthropogenic Sources Jozef M. Pacyna NILU-Norwegian Institute for Air Research, Kjeller, Norway; and Faculty of Chemistry, Gdansk University of Technology, Gdansk, Poland Kyrre Sundseth and Elisabeth G. Pacyna NILU-Norwegian Institute for Air Research, Kjeller, Norway Wojciech Jozewicz ARCADIS, Research Triangle Park, NC John Munthe, Mohammed Belhaj, and Stefan Astro¨m IVL Swedish Environmental Research Institute, Gothenburg, Sweden

ABSTRACT Several measures are available for reducing mercury emissions; however, these measures differ with regard to emission control efficiency, cost, and environmental benefits obtained through their implementation. Measures that include the application of technology, such as technology to remove mercury from flue gases in electric power plants, waste incinerators, and smelters, are rather expensive compared with nontechnological measures. In general, dedicated mercury removal is considerably more expensive than a co-benefit strategy, using air pollution control equipment originally designed to limit emissions of criterion pollutants, such as particulate matter, sulfur dioxide, or oxides of nitrogen. Substantial benefits can be achieved globally by introducing mercury emission reduction measures because they reduce human and wildlife exposure to methyl mercury. Although the reduction potential is greatest with the technological measures, technological and nontechnological solutions for mercury emissions and exposure reductions can be carried out in parallel.

IMPLICATIONS The Governing Council of the U.N. Environmental Programme (UNEP), in its decision 24/3 IV in 2007 on chemicals management, established the Ad Hoc Open-Ended Working Group on Mercury of governments, regional economic integration organizations, and stakeholder representatives to review and assess options for enhanced voluntary measures and new or existing international legal instruments for addressing the global challenges presented by mercury. In that context, a general qualitative assessment of potential costs and benefits were prepared for strategic objectives required to reduce the environmental and health risks from mercury. This paper reports on the outcome of this assessment.

302 Journal of the Air & Waste Management Association

INTRODUCTION Mercury (Hg) emissions from anthropogenic sources are contributing significantly to human and wildlife exposure to methyl mercury (MeHg). Hg is toxic, persistent, and a pollutant of global concern because of a residence time in the atmosphere of some of its chemical forms of approximately 2 yr. Hg can be transported with air masses over the globe, and its atmospheric deposition may occur far away from its emissions regions. Thus, environmental consequences of Hg fate in the atmosphere can be observed in the regions where emissions of this pollutant are insignificant (e.g., in polar regions).1 From an economic perspective, Hg pollution may cause quite significant costs, not only because of the damage costs from negative impacts on human health and the environment, but also because of abatement costs (investment and operational) and risk communication costs. Societal benefits occur when there is a reduction in societal, ecological, or human health damage costs. In this paper, benefits are regarded as large if their estimated value exceeds the measure cost by a factor of 2, but are regarded as low if they are equal or lower than the measure costs. Medium benefits are between the large and small benefits. From Hg exposure effects on human health, a benefit that has been given particular attention is reduction of neurological damage that otherwise would lead to impaired development of the brain and the subsequent loss of intelligence quotient (IQ) from exposure in the prenatal phase.2 Deposited Hg can be taken up by biota and bioaccumulate and biomagnify in the form of MeHg in food webs, particularly aquatic food webs, to levels that can be harmful to organisms, including humans. The most cited studies on neurotoxic impacts due to Hg have followed groups of children among three populations in New Zealand,3 the Seychelles,4 and the Faroe Islands5 whose diet contains a particularly large Volume 60 March 2010

Pacyna et al. Table 1. A qualitative assessment of the costs and benefits from different Hg byproduct reduction options. Option 1 2 3 4

Reduction

Costs

Benefits

Reduction from coal usage Reduction from industrial processes Reduction from waste Reduction from chlor-alkali industry

Medium 3 Large Medium 3 Large Small 3 Large Small 3 Large

Large Medium 3 Large Large Medium 3 Large

portion of seafood. On the basis of the findings in these studies, Trasande et al.6 considered several possible forms of the dose-response function in estimating the societal cost of the IQ decrement in the United States. This function is also used in the Spadaro and Rabl study.7 Spadaro and Rabl concluded on the basis of a literature review on damage costs based on IQ decrement (all studies conducted in the United States) that it is proper to use $18,000 (all costs US$) per IQ point. They also estimated that the worldwide average damage costs per kilogram of Hg emitted are $1,500 because of ingestion. This estimate was based on damage costs in the United States and was then applied for the rest of the world by adjusting for purchase power parity per capita in each country. The costs of reducing Hg emissions are generally linked to the economic costs of introducing the necessary equipment or other necessary actions to obtain the reduction. These costs consist of investment costs and operation and maintenance (O&M) costs, and they are defined in this paper as small, medium, or large in relation to the highest cost of abatement for a given strategy or category. In general, the effect of external factors such as pollution controls on production costs may be absorbed by the producer or passed on to the consumer depending on the elasticity of the market. It is necessary to examine efficiencies and costs of available options to reduce emissions and exposure to Hg to develop cost-efficient strategies to reduce environmental and human health impacts of Hg. A full societal costbenefit analysis would require detailed information on alternative strategies and associated costs of reducing emissions as well as quantitative source-receptor and dose-response descriptions and quantified economic benefits for reducing the impacts on human health and ecosystems. As a first step toward a sound basis for decision support, the focus in this paper is to qualitatively and quantitatively assess various measures and their associated costs and benefits. Quantitative information in this paper is regarded as a supplement to the qualitative evaluations. Artisanal gold mining, which is one of the largest emitters of Hg, is not discussed in this paper because reduction options for this category are still under discussion. Table 1 summarizes the qualitative costs and benefits assessment of the major byproduct emission categories.8 To examine the effects of available options to reduce emissions and exposure to Hg, current and future emission levels need to be assessed. A global inventory of anthropogenic emissions to the atmosphere in 2005 was prepared by the U.N. Environmental Programme (UNEP) Chemicals and Arctic Monitoring Assessment Programme (AMAP)9 as a contribution to the UNEP report “Global Atmospheric Mercury Assessment: Sources, Emissions and Transport.” Hg emissions and emission trends by country, Volume 60 March 2010

region, and sector were assessed, and the anthropogenic emissions estimates were geospatially distributed. Also, scenario emissions inventories for 2020 were assessed on a global scale to investigate the implications of actions to reduce Hg emissions. ATMOSPHERIC EMISSIONS OF MERCURY Anthropogenic emissions of Hg are generated during the production of industrial goods, the consumption of raw materials (e.g., fossil fuels and ores), and intentional use of Hg in various consumer products and industrial processes, including the chlor-alkali industry. Emissions from the former group of sources are often called “byproduct emissions,” whereas emissions from the other group of emission sources are known as “intentional use emissions.” The largest single source of anthropogenic emissions to the atmosphere is estimated to be the combustion of fossil fuels, mainly coal in utility, industrial, and residential boilers.5 Although Hg concentrations found in coals and other fossil fuels are low, their large quantity use in production leads to significant byproduct emissions. The most recent assessment of Hg emissions from various anthropogenic sources in various parts of the world was prepared by AMAP/UNEP in 2008.9 It was estimated that the total anthropogenic emissions of Hg in 2005 were approximately 1930 t, distributed among various emission source categories. As much as 45% of the total emissions of Hg from all anthropogenic sources worldwide in 2005 originated from combustion of fossil fuels. Emissions during the artisanal small-scale gold production contributed approximately 18%, followed by nonferrous metal manufacturing, including gold (⬃13%), and cement production (⬃10%). Waste incineration was responsible for 6.5% of the emissions, whereas the chloralkali sector represented 2.4%. As much as 60% of the total global Hg emissions were estimated as coming from Asia, which is the largest contributing continent. An overview of 2005 Hg emissions to the air from various anthropogenic sources in different regions is presented in Figure 1. The major source categories discussed in this paper cover approximately 80% of the total global anthropogenic emissions in 2005. Emission projections for Hg in 2020 were estimated within the AMAP/UNEP 2008 project9 and another project prepared for the Nordic Council of Ministers.10 Three scenarios were developed: the status quo (SQ) scenario, the extended emission control (EXEC) scenario, and the maximum feasible technological reduction (MFTR) scenario. The SQ scenario assumes that current patterns, practices, and Hg uses that result in Hg emissions to air will continue in the future. Economic activity in various industrial sectors is assumed to increase, including in those Journal of the Air & Waste Management Association 303

Pacyna et al.

Figure 1. Emissions of Hg to air in 2005 from various anthropogenic sectors in different regions.9

sectors that produce Hg emissions, but emission control practices remain unchanged. The EXEC scenario assumes economic progress at a rate dependent on the future development of industrial technologies and emission control technologies (i.e., Hgreducing technology currently used throughout Europe and North America would be implemented elsewhere). It further assumes that emission control measures currently implemented or committed to in Europe to reduce Hg emission to air or water would be implemented around the world. These include measures adopted under the Long-Range Transboundary Air Pollution (LRTAP) Convention, European Union (EU) directives, and agreements to meet the Intergovernmental Panel on Climate Change (IPCC) Kyoto targets on reduction of greenhouse gases (GHGs), causing climate change (which will also result in reductions of Hg emissions). The MFTR scenario assumes implementation of all solutions/measures leading to the maximum degree of reduction of Hg emissions and its loads discharged to any component of the environment; cost is taken into account but only as a secondary consideration. An overview of the estimated emissions of Hg to air in 2005 and within the scenarios is presented in Figure 2.9 It can be concluded from the scenario estimates that a significant increase of about one-third of the 2005 Hg emissions is expected by 2020 if no major changes in the efficiency of emission control are introduced (the SQ scenario). A decrease by one-third of the total emissions of Hg in 2005 can be expected in 2020 if the assumptions of the EXEC scenario are met. As much as a half of the 2005 total emission can be reduced by 2020 if the assumptions of the MFTR scenario are met. Details on methodology of emission scenario development and results of its application are available from AMAP/UNEP.9 304 Journal of the Air & Waste Management Association

REDUCTION OF MERCURY EMISSIONS FROM COAL USAGE Hg Abatement Efficiency and Costs Coal combustion is the largest emission source of Hg to the atmosphere, as concluded in AMAP/UNEP.9 The options for Hg abatement include primary measures (improved energy efficiency, coal washing, coal blending, and coal additives) and secondary (or technical) measures of abatement such as co-benefit Hg removal by the existing criterion pollutant controls or dedicated Hg control (activated carbon injection). For any given coal combustion facility, including its air pollution control equipment configuration, the amount of Hg emissions from the plant is directly related to the amount of coal burned to generate a unit of electricity. It follows that if the amount of coal burned to generate a unit of electricity could be reduced, then the overall Hg emission from a given power plant would also decrease. This could be accomplished by measures undertaken to improve the efficiency of a power plant. Therefore, the first step in the review of strategies for controlling Hg emissions should be to consider any viable means to improve the efficiency of the plant. Improvement of plant efficiency also provides for reduction of GHG emissions in addition to reduction of Hg emissions. Improvement of plant efficiency may involve several measures designed to conserve fuel (coal) and, as a result, reduce the amount of Hg emissions. Some of the most commonly applicable measures at a coal-burning facility include new burners, improved air preheater, improved economizer, improved combustion measures, minimization of short cycling, minimization of gas-side heat transfer surface deposits, and minimization of air infiltration. In addition, O&M practices have a significant impact on plant performance, including its efficiency, reliability, and operating cost. Two examples of good O&M practices include steam line maintenance and water treatment. A well operated and maintained plant will experience less

Figure 2. Trends in estimated emissions of Hg to air in 2005 from byproduct sectors and the chlor-alkali industry in different regions compared with three different scenarios for 2020.9 Volume 60 March 2010

Pacyna et al. rapid deterioration of heat rate; hence, O&M practices themselves influence coal use and Hg emissions. Good O&M practices should be an ongoing concern in daily plant operation. Depending on the coal used, a certain decrease of Hg emissions may be obtained by the deployment of coal treatment technologies before combustion. Coal treatment technologies considered in the context of plant efficiency and Hg removal include conventional coal washing, coal beneficiation for Hg content, coal blending, and coal additives. Fuel substitution schemes are also capable of delivering decreased Hg emissions (e.g., substitution of coal with natural gas or renewable energy sources). However, fuel substitution schemes are beyond the scope of this paper and are not discussed here. The next step in Hg emission control strategy, beyond efficiency improvement and coal treatment, is the set of approaches designed to maximize the so-called cobenefit removal, or the amount of Hg removal that is realized as the effect of operation of air pollution control equipment originally designed to limit emissions of criterion pollutants (e.g., particulate matter [PM], sulfur dioxide [SO2], or oxides of nitrogen[NOx]) and already in place at the power plant. Depending on the available air pollution control equipment, these approaches could include modernization of electrostatic precipitators (ESPs), modification of wet desulfurization scrubber chemistry, alteration of selective catalytic reduction (SCR) operation, or a combination of these. Should a higher level of Hg emission control be desired beyond what can be achieved through the co-benefit removal, a power plant would have to deploy a dedicated Hg removal technology. Because dedicated Hg removal requires the addition of hardware and usage of resources during its operation (capital and O&M expenses), it is generally considerably more expensive than the co-benefit strategy. This section describes coal treatment methods that may be applicable to reduce Hg emissions. Next, the concept and advantages of the co-benefit removal of Hg are presented, followed by dedicated Hg removal via activated carbon injection (ACI). Energy efficiency measures are beyond the scope of this paper and are not discussed here. Coal Treatment Methods. Coal treatment methods that can be used to decrease Hg emissions from coal-fired power plants include coal washing, beneficiation, blending, and coal additives. Conventional coal cleaning facilities use physical cleaning techniques to reduce the mineral matter and pyritic sulfur content. Physical coal cleaning typically involves a series of process steps including size reduction and screening, gravity separation of coal from sulfur-bearing mineral impurities, and dewatering followed by drying. As a result, the coal product has a higher heating value and less variability (compared with untreated coal) so that power plant efficiency and reliability are improved. An added benefit to these processes is that emissions of SO2 and other pollutants, including Hg, can be reduced. The efficiency of this removal depends on the cleaning process used, type of coal, and the contaminant content of coal. Although conventional coal cleaning methods will remove some of the Hg associated with the Volume 60 March 2010

incombustible mineral materials, they will typically not remove Hg associated with the organic carbon structure of the coal. Kraus et al.11 indicate that 10 –50% of the Hg in coal can typically be removed in the coal cleaning process. There are reports of conventional coal cleaning reducing the Hg content of bituminous coal from 12 to 78% (average 37%), depending on the coal source.12 The variation in Hg reductions quoted above might be a function of the type of process used to clean a given coal and the nature of the Hg in the coal matrix. Coal beneficiation is capable of improving coal properties beyond what can be achieved with conventional coal washing alone. An example of coal beneficiation is the K-Fuel process, which includes coal washing as a primary step, but then utilizes additional treatment to reduce the Hg content of coal. The process results in reduction of Hg emissions in addition to lower PM, SO2, and NOx emissions. K-Fuel is a beneficiated coal that is derived from subbituminous or lignite coal.13 The resulting fuel is lower in ash, higher in heating value, and produces lower pollutant emissions than untreated coal. K-Fuel technology may also be applicable to bituminous coal. However, the supplier has focused exclusively on subbituminous and lignite applications because the K-Fuel process is a moisture and ash reduction process, and these coals are characteristically high in moisture and ash content. Presently, there is no adequate information to estimate the price of the processed coal. In the future, the price for K-Fuel may be based on the price of competing coal on a heating value basis plus credits for environmental benefits (SO2, NOx, and Hg emission reductions). Coal blending/switching at power plants is commonly used as power plants attempt to cost-effectively meet SO2 emission limits. As a side effect of this SO2 emission control strategy, Hg speciation may be shifted toward oxidized Hg because of increased content of chlorine in the flue gas; therefore, given a favorable configuration of air pollution control equipment, Hg removal may be augmented. Oxidized Hg is generally easier to remove than elemental Hg. For example, a field study at a large utility plant firing a 60% subbituminous and 40% bituminous blend at two identical boilers (one with and the other without SCR) demonstrated an increase in the oxidized Hg fraction from 63% without SCR to 97% with SCR. The unblended subbituminous coal would have achieved between 0 and 40% oxidized Hg.13a Another comprehensive study of the effects of coal blending on Hg speciation in the presence of SCR examined the oxidation of Hg using blends ranging from 10 to 40% subbituminous with bituminous coal.14 The amount of the acrossthe-SCR Hg oxidation increase was found to be larger for a corresponding increase in the percent of bituminous coal. As described above, the extent of Hg oxidation increases with the amount of chlorine in the coal. However, the concentration of chlorine in the coal often may not be sufficient to achieve a high level of Hg oxidation. To overcome this issue, approaches have been designed to add halogen compounds such as bromine or chlorine salts to coal before combustion. Alternatively, hydrochloric acid (HCl) or ammonium chloride (NH4Cl) may be added. Journal of the Air & Waste Management Association 305

Pacyna et al. Table 2. Average Hg capture by existing postcombustion control configurations installed on coal-fired units.a Average Hg Capture by Control Configuration (%) Coal Burned in PC-Fired Boiler Unit Postcombustion Control Strategy PM control only

PM control and SDA

PM control and wet FGD systemb

Postcombustion Emission Control Device Configuration

Bituminous Coal

Subbituminous Coal

Lignite

ESPc ESPh FF PS SDA ⫹ ESP SDA ⫹ FF SDA ⫹ FF ⫹ SCR PS ⫹ FGD ESPc ⫹ FGD ESPh ⫹ FGD FF ⫹ FGD

36 14 90 Not tested Not tested 98 98 12 81 46 98

9 7 72 9 43 25 Not tested 10 29 20 Not tested

1 tested tested tested tested 2 not tested Not tested 48 Not tested Not tested Not Not Not Not

Notes: aTable is based on data from ref 38; bEstimated capture across both control devices. PC ⫽ pulverized coal, ESPc ⫽ cold-side ESP, ESPh ⫽ hot-side ESP, PS ⫽ particle scrubber, SDA ⫽ spray dryer absorber system.

Halogen additives oxidize elemental Hg and make it available for capture devices downstream. They may be particularly useful for improving Hg removal for units firing low-chlorine subbituminous coals. They may be sprayed on coal, injected into the boiler, or added as solids upstream of a coal pulverizer. Bromine is thought to have an advantage over chlorine because its Deacon-type reactions are more favorable and it is consumed by SO2 to a lower degree than chlorine.15 Secondary Measures to Reduce Hg Emissions from Coal Combustion. Secondary measures include co-benefit Hg removal by the existing criterion pollutant controls and dedicated Hg control (i.e., ACI). To define the co-benefit removal of Hg from coal-fired flue gas, one needs to consider the forms of Hg typically present and their amenability for control. One of the most complete efforts to date to understand the potential for co-benefit removal of Hg by the existing air pollution equipment is the information collection request (ICR) that started in late 1998 in the United States and was mandated by the U.S. Environmental Protection Agency (EPA). On the basis of ICR data, Table 2 summarizes the average reduction in total Hg emissions from units burning different coals and equipped with different postcombustion equipment configurations. It should be noted that the data presented in Table 2 pertain to air pollution equipment operating on flue gas generated by combustion of U.S. coals. Combustion of non-U.S. coals may result in different Hg captures for the same type of coal because of differences in composition. As can be seen from Table 2, the capture of Hg across the existing equipment can significantly vary based on coal type and equipment configuration. The level of recorded co-benefit removal ranges from negligible to more than 90%. The strongest factor is the type of coal used, which in turn affects Hg speciation at the inlet to the control device(s). In general, units burning subbituminous and lignite coals demonstrate significantly lower Hg capture than similarly equipped bituminous-fired units. Units that burn bituminous coals typically have relatively 306 Journal of the Air & Waste Management Association

high concentrations of oxidized Hg at the inlet to the control device(s). Gaseous compounds of oxidized Hg are water-soluble and can be absorbed in a wet flue gas desulfurization (FGD) system; however, elemental Hg vapor is insoluble in water and therefore does not absorb in such slurries. The effectiveness of a wet FGD in controlling Hg may be augmented by an upstream SCR, which has the potential to alter Hg speciation in a way that increases the amount of oxidized Hg in the flue gas (as explained earlier for coal blending effects). Particulate-bound Hg can be captured by ESPs and fabric filters (FFs). Therefore it is important to maintain an ESP or FF in good operation to maximize its Hg removal potential. For plants where the amount of Hg capture desired is beyond what can be achieved through co-benefit removal, deployment of dedicated Hg control technologies may be required. To date, use of sorbent injection has shown the most promise as a near-term Hg control technology. In the basic scenario envisioned for sorbent injection, powdered activated carbon (PAC) is injected between the air heater and the particulate control device, as shown schematically in Figure 3. As shown in Figure 3, PAC is typically injected into the duct upstream of the particulate control device (most often an ESP). After adsorbing the Hg from the combustion flue gas, PAC is captured together with the fly ash in the PM control device. There are concerns involving mixing fly ash with the PAC following its injection for Hg control because of potentially decreased saleability of

Figure 3. Schematic of ACI.37 Volume 60 March 2010

Pacyna et al. Table 3. Annual investment and operating costs for various types of emission control equipment used in hard and brown coal combustion as calculated within the EU ESPREME project.b Annual Costs (US$ 2008/MWh)

Emission Control Technology

Estimated Hg Reduction (%)

Investment Cost

Operating Cost

Total Costa

24 90 32 98 98

0.5 0.5 0.9 2.7 2.7

0.9 1.5 0.5 3.0 2.4

1.4 1.9 1.4 5.7 5.1

Dry ESP FF Dry ESP—retrofitted from medium to high control efficiency FF ⫹ wet or dry scrubber ⫹ sorbent injection Dry ESP ⫹ wet or dry scrubber ⫹ sorbent injection

Notes: aAnnual operating costs of about US$ 20/kWh could be expected for emerging technologies such as electrocatalytic oxidation or integrated gasification combined cycle; bThe accuracy of cost estimates in the table is within ⫾50%.

the fly ash. These concerns have spurred the development of the toxic emission control process (TOXECON). TOXECON is a process developed by the Electric Power Research Institute (EPRI) in which PAC is injected downstream of the existing particulate control device and a pulsejet fabric filter (PJFF) is added downstream. The TOXECON process makes the separation of fly ash from the spent sorbent possible and thus preserves the quality of the fly ash. A drawback of this approach is the increased capital cost due to the addition of the PJFF. To mitigate this capital expenditure, the TOXECON II process was developed. TOXECON II injects sorbent directly into the initial collecting fields of an ESP. The process takes advantage of the fact that most fly ash is collected in the upstream collecting fields (initial fields that flue gas passes through). Thus, injection of PAC into the downstream collecting fields results in only a small portion of fly ash being contaminated by the sorbent (carbon). This contaminated fly ash comes from the final fields of the ESP where only a small mass fraction of fly ash is collected. In addition to PAC injection, several dedicated multipollutant technologies are being developed that are capable of simultaneously removing Hg, SO2, PM, and NOx from flue gas. These technologies are at varying levels of development and commercial availability and will not be discussed in this paper. The incremental cost of Hg reduction (i.e., the cost, in US$/kg Hg removed, to achieve a specific reduction) is influenced largely by the level of baseline Hg capture exhibited by the existing air pollution control device (APCD) configuration and coal Hg content. For example, the incremental cost of Hg control will increase when (1) baseline Hg capture by existing APCD is high, or (2) the coal Hg content is low, because a smaller quantity of Hg is removed from the flue gas for a given level of control. In terms of raw monetary cost, reducing Hg from coal combustion can be quite expensive. The incremental cost of Hg emission reduction varies substantially depending on factors such as type of coal used, type of combustion unit, type of APCD already in place to control other pollutants, facility configuration, and percent reductions expected from the existing APCD. For example, wet scrubbers installed primarily for Hg have been estimated to cost between $76,000 and $174,000/lb Hg removed (or between $168,000 and $384,000/kg Hg removed). This result is Volume 60 March 2010

very close to the cost of $234,000/kg Hg removed, estimated and used in a study of the effectiveness of the U.N. Economic Commission for Europe (ECE) heavy metals (HM) protocol and cost of additional measures.16 Early estimates of Hg control costs indicated that it would cost between $67,700 and $70,000/lb (or between $149,300 and $154,000/kg) to achieve a 90% control level using sorbent injection.17 However, following the Research, Development, and Demonstration (RD&D) activities sponsored by the U.S. Department of Energy (DOE), the costs of sorbent injection for Hg removal have shown significant advances along with the potential for reductions in overall installation and operational costs. A DOE economic analysis released in 2007 indicated that the cost of Hg control could be drastically lowered compared with original estimates because of a reduction in the injection rate of a sorbent. The analysis indicated that a levelized incremental cost of 90% Hg emission control by means of PAC injection ranged from approximately $30,000 to less than $10,000/lb Hg removed for DOE field testing sites.18 These DOE test sites used a chemically treated (brominated) activated carbon. Although the capital cost of a Hg control system (i.e., ACI) is relatively low, the major expenditure comes from the cost of carbon itself. Brominated carbon generally affords much lower injection rates (mass sorbent/flue gas flow) than the untreated carbon to accomplish the same level of Hg removal. Thus, although chemically treated carbons are more expensive than untreated ones, the use of chemically treated carbons allows for significantly lower costs of Hg removal. A summary of information on annual investment and operating costs for various types of emission control equipment that may be deployed at plants burning hard or brown coal is presented in Table 3. As can be seen from Table 3, a considerable amount of Hg removal could be achieved with PM control equipment alone at annual operating costs below $2.00/kWh. Significantly higher annual operating costs were noted for PM control equipment combined with wet or dry scrubber and sorbent injection. Benefits of Hg Emissions Abatement A qualitative benefit assessment shows that globally and locally there is a large potential for emission reductions of Journal of the Air & Waste Management Association 307

Pacyna et al. Hg as well as reduction of other air pollutants with consequent health benefits from Hg abatement measures.8 Information on the benefits and costs of reduction of Hg emissions from coal combustion was reviewed by the Northeast States for Coordinated Air Use Management (NESCAUM).19 The NESCAUM study describes the results from a comprehensive assessment of the health benefits of reducing Hg emissions from coal-fired power stations in the United States. It has been anticipated that reductions in Hg emissions from coal-fired power plants will decrease MeHg concentrations in fish. The model developed in the study assumes that equilibrium currently exists between deposited Hg and fish MeHg concentrations and between fish MeHg concentrations and MeHg exposures to individuals who consume these fish. Changes in the quantity of Hg deposited are assumed to lead to leaner and proportional changes in fish MeHg concentrations, assuming that no other factors change. Two potential health effects were considered: cognitive abilities and cardiovascular events. The results from epidemiological studies were used to develop association between MeHg exposures in males and increased risks of myocardial infarction and premature mortality. Using a contingent valuation approach, it has been estimated that the value of premature fatality is approximately $6 million (in 2000 US$), but it was indicated that this value should be taken with caution. The 2005 NESCAUM study also described the possible benefits of the U.S. power plant Hg emission controls in terms of IQ increases in the annual birth population. The predicted annual benefit associated with IQ increases in the annual birth population ranged from $75 million to $288 million, estimated within two scenarios related to different emissions projections in the U.S. power plants where the decrease in deposition levels were up to 7 and 8% in scenarios 1 and 2, respectively, in a certain geographical area.19 Societal benefits related to global Hg emissions were estimated in Pacyna et al.20 The societal benefits were estimated as a difference between the societal damage costs related to the emission reductions calculated for the scenario assuming the status quo of environmental pollution between the years 2005 and 2020 and the emission reductions projected in the scenario in which application of modern emission control devices is used. The marginal damage cost per metric ton of Hg emitted was separately estimated for inhalation of Hg-polluted air and ingestion of Hg-contaminated food. The damage cost of $1,500/kg Hg emitted was accepted for the ingestion pathway. However, the damage cost from inhalation of Hg was estimated to be relatively insignificant compared with the damage costs from ingestion. Societal damage costs were related to IQ loss through loss of earning, education, and opportunity cost while at school. The two-thirds of the societal costs due to Hg pollution are associated with the societal costs due to Hg pollution of the environment by emissions from coal combustion. On the basis of results from the study, the annual social benefits associated with the IQ change due to Hg emission reductions from coal combustion worldwide can be estimated to be more than $0.9 billion. 308 Journal of the Air & Waste Management Association

REDUCTION OF MERCURY EMISSIONS FROM INDUSTRIAL PROCESSES Industrial processes contribute approximately 25% to the total emissions of anthropogenic Hg to the atmosphere.9 Emissions from the nonferrous and ferrous metal industries are estimated to contribute approximately 10% each to the total emissions. With regard to the Hg emissions from nonferrous metal production, their amounts depend mainly on (1) the content of Hg in nonferrous metal ores used mostly in primary processes or scrap used in secondary nonferrous production, (2) the type of industrial technology used in the production of nonferrous metals, and (3) the type and efficiency of emission control installations. The content of Hg in ores varies substantially from one ore field to another,21,22 as does the Hg content in scrap. Depending on the country, Hg emissions from primary production using ores in nonferrous smelters are between 1 and 2 orders of magnitude higher that the Hg emissions from secondary smelters with scrap as the main raw material. Pyrometallurgical processes in the primary production of nonferrous metals, which use high-temperature roasting and thermal smelting, emit Hg and other raw material impurities, mostly to the atmosphere. Nonferrous metal production with electrolytic extraction is responsible more for risks of water contamination. The primary sources of Hg emissions from Portland cement manufacturing contribute approximately 10% to the total anthropogenic emissions of this element. Hg emissions are mostly generated during the processing of raw materials in the kiln. Kiln operations consist of the pyroprocessing (thermal treatment) of raw materials that are transformed into clinkers. Raw material processing differs somewhat for the wet- and dry-kiln processes. Regardless of kiln type, Hg is introduced into the kiln with raw material (limestone) and with fuels (e.g., coal), which are used to provide heat for calcination and sintering of raw materials. Other fuels, such as shredded municipal garbage, chipped rubber, petroleum coke, and waste solvents, are also being used frequently and may contribute to Hg emission from cement production. In addition, fly ash from coal combustion may be added to clinker for concrete production. This added fly ash may contain Hg as a result of the condensation of gaseous Hg on fine fly ash particles in the flue gas before the collection of fly ash in particulate control devices such as ESPs or FFs.23 EPA reported that Hg is retained by fly ash with the use of sorbents for enhanced Hg capture and is thus unlikely to be leached to the environment.24 The same results came from the EPA study on wet FGD scrubbers.25 Hg Abatement Efficiency and Costs Large nonferrous smelters use high-efficiency APCDs to control PM and SO2 emissions from roasters, smelting furnaces, and convertors.26,27 ESPs are the most commonly used devices for the control of PM. Control of SO2 emissions is achieved by absorption to sulfuric acid in sulfuric acid plants, which are a common part of smelting plants. Hg is mostly emitted in a gaseous elemental form from large nonferrous smelters, and the ESPs are therefore not very effective in its removal. The elemental Hg does not end up in sulfuric acid plants and is instead emitted to the atmosphere from the smelter stacks. The amount of Volume 60 March 2010

Pacyna et al. these emissions depends on the content of Hg in the ore. This content varies substantially from one ore field to another. Only limited information has been gathered on Hg emission rates from nonferrous smelters.28 Because Hg emissions are mostly in the elemental form, there may be a need to deploy ACI or a wet scrubber to control them. Other options to control Hg emissions from industrial sources include those discussed previously for the coal combustion sector. Hg can be emitted to the atmosphere during the production of metallurgical coke, which is used in the iron and steel industry. ESPs or FFs (and less frequently wet scrubbers) are used in the coke production plants to control emissions, particularly those generated during quenching. Quenching is performed to cool down coke and to prevent complete combustion of coke upon exposure to air. Although no data are available for the performance of the ESPs or FFs in coke production plants, it is expected that Hg removal is limited.28 Thus, ACI would need to be deployed to control Hg emissions from metallurgical coke production facilities. A major assessment of costs and benefits of reducing Hg emissions from various industrial sources has been prepared for the UNEP activities mentioned earlier in this paper by the U.S. Department of State. One such industry is secondary steel production. This category is a significant source of Hg air emissions largely because Hg-containing switches are in the scrap metal (such as cars) used to make steel. In the United States, a National Vehicle Mercury Switch Recovery Program (NVMSRP) was established in 2006. The NVMSRP, along with a few state Hg switch programs, is expected to reduce Hg emissions by approximately 34 t over the duration of the program, which represents the Hg content in approximately 61 million switches. The program is designed to remove Hgcontaining switches from scrap vehicles before they are recycled in secondary steel mills. At this time, the precise cost-effectiveness of this program is unknown, although components of the costs include outreach, education, and design efforts that typically do not require significant ongoing monetary investment. However, the voluntary effort to remove switches provides an incentive of about $1.00 per switch. Although this may not reflect the actual cost of removing the switch (some states have proposed incentives of up to $7.00 per switch), it still costs significantly less than installing end-of-pipe controls to capture Hg at the furnace. A major review of information on the costs of abatement for combustion of coal and other economic sectors was carried out within the EU ESPREME (http://espreme.ier.uni-stuttgart.de) and DROPS (http://drops.nilu.no) projects. Table 4 presents the annualized investment and operational costs for installations that are used to remove Hg, including ESPs, FFs, FGD, and “add on” measures just for Hg removal. The accuracy of cost estimates in Table 4 is within ⫾50% as calculated within the EU ESPREME project. These costs are given in relation to the production of 1 t of specific production and are indicated in Table 4 as a specific activity indicator (SAI). Information on the efficiency of Hg removal using these installations is also included in Table 4. Volume 60 March 2010

Industrial sources of Hg emissions often include small industrial boilers. Because of the economy of scale, different control strategies may be needed for these small industrial boilers than for large industrial sources or for coal-burning utility power plants. A major assessment of costs and environmental effectiveness of options for reducing Hg emissions to air from small-scale combustion installations (SCIs) (⬍50 MWh) was prepared for the European Commission by Pye et al.29 It was concluded that among the most cost-effective options were preventive options (e.g., options before combustion to minimize emissions), such as coal washing and fuel switching. Such options require the use of a better quality, cleaner fuel within the same fuel type, or switching to a different type of fuel with lower Hg content and resultant emissions. Another preventive option is reduction in energy consumption through energy efficiency, as discussed before for the coal combustion sector. Only limited technical abatement options (e.g., removal of Hg from flue gases after combustion) were identified for SCIs. An assessment of abatement costs for reduction of HMs, including Hg within various industries, was carried out for the HM emission reduction protocol of the U.N. ECE LRTAP convention.16 The results of this assessment are similar to the data presented in Table 4. Benefits of Hg Emissions Abatement Information on monetary valuation of environmental and human health benefits related to the reduction of Hg emissions from individual industrial sources is largely missing in the literature. Societal benefits related to the decrease of the 2005 Hg emissions from industrial sources worldwide until the year 2020 are estimated by Pacyna et al.10 as a part of an assessment of socioeconomic costs of continuing the status quo of Hg pollution from all major anthropogenic sources. For the metal industry and cement manufacturing, it is suggested that annual damage costs to society due to ingestion of Hg-contaminated food in the year 2020 can be as high as $0.7 billion along the assumptions defined for the SQ scenario. The damage costs to society due to inhalation of Hg-polluted air were estimated as insignificant compared with the damage costs due to ingestion. Thus, the societal benefits due to reduction of Hg emissions from the metal industry and cement manufacturing in the year 2020 were estimated to be approximately $0.4 billion annually.10 REDUCTION OF MERCURY EMISSIONS TO AIR FROM MERCURY-CONTAINING WASTE Emissions of Hg to air coming from Hg-containing waste can be abated by nontechnological measures through (1) reduction of generation of wastes that contain Hg; (2) promotion of separate collection and treatment of Hgcontaining wastes; or (3) technological measures through reduction of Hg emissions from medical, municipal, and hazardous waste incinerators as well as reduction of migration and emission of Hg from landfills. Hg Abatement Efficiency and Costs Within the overall cost assessment, abatement of Hg emissions from Hg- containing wastes is regarded as small to large, depending on the abatement technique. The cost Journal of the Air & Waste Management Association 309

310 Journal of the Air & Waste Management Association

Primary lead

Primary lead

FF–retrofitted

Dry ESP–optimized

Cast iron

Steel

FF–medium efficiency Dry ESP–medium efficiency FF

Wet FGD–optimized Dry ESP–optimized

Coke Coke

Cast iron Cast iron Cast iron

FF Dry ESP–medium efficiency Wet FGD

Coke Coke Coke

Notes: aThe accuracy of cost estimates in the table is within ⫾50%.

Electric arc furnace steel

Iron and steel foundry

Coke production

Primary copper

Primary zinc

FF–state-of-the-art

ACI ⫹ FF ⫹ FGD

Primary lead

Primary copper

90

FF

Sinter

Dry ESP–medium efficiency FF FF–medium efficiency

10

Calcium hydroxide-impregnated adsorbents Dry ESP–medium efficiency

Sinter

Primary zinc Primary zinc Primary copper

80

ACI ⫹ FF

Sinter

Primary lead

70

Dry ESP–optimized

Sinter

Sintering

98

70

5 5 98

40 70

5 5 30

10

5 10 5

5

100

5

Dry ESP–medium efficiency

SAI (per 1 t)

Hg Reduction (%)

Sector

Emission Control Technology

0.3

69.8

10.8 38.1 12.5

3.0 1.4

0.5 0.8 2.8

3.9

0.1 4.5 1.8

2.5

0.1

0.1

1.1

2.1

0.2

0.1

Annual Investment Costs

1.4

79.0

82.8 55.5 71.1

2.8 1.6

3.1 1.1 1.9

25.7

0.06 1.1 13.8

1.3

1.1

0.04

1.2

1.1

0.2

0.1

Annual Operating Costs

1.7

148.8

93.6 93.6 83.6

5.8 3.0

3.5 1.9 4.7

29.5

0.2 5.6 15.6

3.8

1.2

0.1

2.3

3.2

0.4

0.2

Annual Total Costs

Annual costs (US$ 2008/SAI)

Basic oxygen furnace steel

Pig iron production

Coke production

Cement production

Secondary copper

Secondary zinc

Secondary lead

Sector

Steel

Steel

Cast iron Cast iron Steel

Cast iron Cast iron

Steel Steel Steel

Coke

Cement Cement Coke

Secondary lead Secondary lead Secondary zinc Secondary zinc Secondary copper Secondary copper Cement

SAI (per 1 t)

FF–medium efficiency FF–optimized Wet FGD Use of raw materials with low HM content FF–medium efficiency Dry ESP–optimized FF Dry ESP–medium efficiency FF Dry ESP–medium efficiency Dry ESP –retrofitted Dry ESP–optimized Dry ESP–medium efficiency Wet scrubber–Venturi Dry ESP–optimized

FF

Dry ESP

FFs

Dry ESP

Dry ESP–medium efficiency FF

Emission Control Technology

Table 4. Abatement cost for installations used to reduce Hg emissions from various industrial processes as calculated within the EU ESPREME project.a

70

8

72 70 5

5 5

70 5 5

5

98 90 5

5

10

5

10

5

10

5

Hg Reduction (%)

4.3

5.7

1.3 2.8 1.2

0.2 1.5

1.4 0.2 0.8

0.2

0.4 1.4 0.00

0.2

6.6

10.9

0.1

0.1

6.8

0.1

Annual Investment Costs

4.5

0.5

0.9 3.2 3.0

0.8 2.2

1.6 1.7 1.1

1.7

0.4 0.5 0.02

0.2

44.0

15.9

1.4

0.06

1.1

0.06

Annual Operating Costs

8.8

6.2

2.2 6.0 4.2

1.0 3.8

3.0 1.9 1.9

1.9

0.8 1.8 0.02

0.4

50.6

26.8

1.5

0.2

7.9

0.2

Annual Total Costs

Annual costs (US$ 2008/SAI)

Pacyna et al.

Volume 60 March 2010

Pacyna et al. for reduction of generation of wastes that contain Hg is variable from medium to large, depending on the management technique (e.g., recycling, landfilling, or incineration). Promotion of separate collection and treatment of Hg-containing wastes is regarded as small to medium cost (at least in developing countries), whereas the costs are large to medium in the case of incineration and landfilling, respectively.8 Reduction of Generation of Wastes that Contain Hg. There are several sources of waste containing Hg. These sources may differ from one region to another, and the quantity of Hg waste from different sources may be correlated with lifestyle and the level of economic development in the different countries and regions. Because the sources are different and the emissions from these sources are local and region-specific, the costs to reduce generation of waste that contains Hg and the measures (as well as their implementation) to reduce the emissions from these sources differ depending on whether the source is in a developed country or a developing one. To reduce the generation of waste that contains Hg in developed countries, different policy instruments (e.g., regulations, market-based instruments, and information) are used. The policy instruments have led to the implementation of different measures and waste management of wastes containing Hg, such as recycling, landfilling, and incineration. However, although the use of regulations and market-based instruments have led to different technical measures that are moderate in the case of landfills and high in the case of incineration (which has led in some cases to the export of hazardous waste to less developed countries), the most cost-effective measures are nontechnical, being the results of good information highlighting the consequences of Hg emissions on the environment and human health. On the other hand, in developing countries, the environment issue became an interesting subject quite recently, and many countries lack well-formulated guidelines and policy structures regarding waste in general and Hg-containing waste in particular. However, although technical and nontechnical measures are used in developed countries, the measures in developing countries are most often artisanal and in many cases chaotic. Uncontrolled dumping of wastes on the outskirts of towns and cities has created overflowing landfills, which are not only impossible to reclaim because of the haphazard manner of dumping, but also because they have serious environmental implications.30 Nevertheless, when it comes to abatement costs in situations when policy instruments are in place, these may be high if they are technical and if transaction costs including monitoring are included in the estimations. However, costs to reduce Hg-containing waste may be low and cost-effective if the policy instruments are based on guidelines and information. Promotion of Separate Collection and Treatment of Hg-Containing Wastes. Promotion of hg management is warranted at levels including households, industries, and the public sector. However, except technical measures that Volume 60 March 2010

may be used to reduce emissions of Hg, other measures such as substitution would be more cost-effective. Where Hg-containing products are used, promotion of separate collection and treatment of Hg-containing waste is likely to be most effective in limiting releases of Hg. Although promotion may give results in the developed world, this strategy may be more challenging in the developing world where there is often no differentiation between municipal, hazardous, and medical waste in terms of applied techniques or achievable emission limits. The costs of Hg management may be small to medium in the case of developed countries (e.g., costs of collection, transportation, and recycling of switches in the United States range from $0.004 to $1.00 per switch). Developing countries import considerable quantities of electronic waste. Reduction of Hg Emissions to Air from Medical, Municipal, and Hazardous Waste Incinerators and Reduction of Migration and Emission of Hg from Landfills. For properly operated incinerators, the destruction and removal efficiency exceed the 99.99% requirement for hazardous waste and can be operated to meet the 99.99% requirement for polychlorinated biphenyl (PCB) and dioxins. Off gases and combustion residuals generally require treatment. The abatement cost for installations used to reduce Hg emissions from waste incineration and cremation processes is presented in Table 5. Investment and maintenance costs of appropriate landfills are relatively lower. Landfill controls can be implemented to limit Hg release and will also benefit management of many other hazardous wastes. As an example, the costs for the thermal treatment application at Lipari Landfill (Lipari site) in New Jersey included $430,000 in capital costs and $5,019,292 in O&M costs. The unit cost for this application was $67/t on the basis of treating 80,000 t of soil. The landfill in Lipari was used for disposal of various household, chemical, and other industrial wastes.31 Because the costs related to appropriate incineration and landfillings are high and medium, respectively, in developed countries these settings are economically hard to manage. On the basis of these high costs, opportunities to substitute Hg-free alternatives may be the most preferable option. Benefits of Hg Emissions Abatement The benefits when reducing generation of wastes that contain Hg are very high relative to the abatement costs if management is in place. Separate collection and treatment of Hg-containing waste leads to relatively large benefits compared with the costs of abatement, including technical and substitution measures. The benefits of incinerating or landfilling can be very high compared with the total costs of these management technologies. Using the abatement cost of $67/t for landfill and $1,047/t for incineration (assuming 1 m3 ⫽ 1 t), these costs are lower than the damage avoided or the benefits reached if these management technologies are in use. Hence, the benefits of incinerating or landfilling are large compared with the total costs of these management technologies. Journal of the Air & Waste Management Association 311

Pacyna et al. Table 5. Abatement cost for installations used to reduce Hg emissions from waste incineration and cremation processes as calculated within the EU ESPREME project.a Annual Costs (US$ 2008/t of waste)

Emission Control Technology

Hg Reduction (%)

Investment Costs

Operating Costs

Total Costs

20

0.1

0.1

0.2

60 70 99

0.6 1.8 2.3

0.6 7.0 2.5

1.2 8.8 4.8

90 80 95 99

2.3 2.2 5.3 5.8

1.8 4.0 6.2 7.1

4.1 6.2 11.4 12.7

Wet scrubber with alkaline additives –medium efficiency if emission control Waste separation–medium Dry ESP ESP ⫹ wet scrubber ⫹ activated carbon with lime ⫹ FF Two-stage scrubber ⫹ wet ESP ACI ⫹ FF ACI ⫹ Venturi scrubber ⫹ ESP – ACI ⫹ Venturi scrubber with lime milk ⫹ caustic soda ⫹ FF Notes: aThe accuracy of cost estimates in the table is within ⫾50%.

Because the origins of Hg-containing waste differ and the emissions from these sources are local and/or regionspecific, the costs to reduce the generation of wastes differ depending on whether the source is in a developed country or in a less developed one. The preliminary qualitative cost assessment reveals that these costs are variable depending on the management technique, such as incineration and landfilling. Although the introduction of various emission control measures may give results in the developed world, the outcome of this strategy may not be very positive in the developing world where there is often no differentiation between municipal, hazardous, and medical waste in terms of applied techniques or achievable emission limits. Therefore, emphasis in the developing countries should be placed on developing adequate policy instruments to mitigate Hg releases. REDUCTION OF MERCURY CONSUMPTION IN CHLOR-ALKALI PRODUCTION There are three different processes for chlor-alkali production. The first two processes, the Hg method and diaphragm technique, date from the end of the 19th century. The third process, the membrane technique, was developed on an industrial scale in the 1970s. Membrane cells release less hazardous substances and are more energyefficient than the older techniques.32 For chlor-alkali, the global trend for conversion to non-Hg cell technology or reductions in Hg use and emissions has been established. This industry is declining substantially in the world, with chlorine and caustic soda now being produced using more efficient, environmentally friendly, non-Hg processes. As of 2004, there were approximately 150 chlor-alkali plants worldwide that still use Hg cell technology. There were approximately 14 U.S. facilities in the mid-1990s using the Hg-cell process (so-called mercury cell chlor-alkali plants [MCCAPs]); this year only 5 such facilities will still be in operation. The existing U.S. facilities are subject to a technology-based emissions standard (the maximum achievable control technology [MACT]) regulation) that requires controls and emissions limits for 312 Journal of the Air & Waste Management Association

process vents and relatively stringent work practice standards or a cell room monitoring program to minimize fugitive emissions from the cell rooms. Hg use by the U.S. chlor-alkali sector was reduced by 94% from 1995 to 2005, from approximately 160 t in 1995 to 10 t in 2005. Emissions were reduced approximately 50% from 1990 to 2002 (from ⬃10 to 5 t) and were expected to decrease to 2.5 t by 2008. These numbers suggest that the benefit of reducing Hg consumption in chlor-alkali production can be quite high, with little opportunity cost given industry trends. According to Euro Chlor information, there remained at the beginning of 2005 over 50 MCCAPs in Europe that continue to use the Hg process to produce chlorine.33 Hg consumption and releases have been greatly reduced from the 500 –1000 t/yr estimated in the 1970s. However, the average age of the EU plants is nearly 35 yr, and further efforts to reduce Hg releases below present levels may challenge the technical limits of what is possible without converting to a Hg-free process. Unacceptably high Hg emissions before and into the 1980s pressed the member countries of OSPAR (the Oslo and Paris Convention for the Protection of the North Sea and North-East Atlantic) to recommend in 1990 that the Hg cell chlor-alkali process should be phased out by 2010. The implementation of the 1990 OSPAR decision and the IPCC directive are ultimately the responsibility of each of the countries concerned. However, the countries’ uneven response to the OSPAR decision and flexible interpretation of the 2007 IPCC deadline reflect the diverse and shifting political and economic priorities of different countries within the EU.34 The chemical industry self-imposed a target for 2007 of 1 g of Hg/t of chlorine capacity. A discussion is now being carried out that this limit can be lowered to 0.75 g Hg/t of chlorine capacity by 2012. It should be noted that the best performing EU MCCAPs report emissions in the range from 0.2 to 0.5 g Hg/t of chlorine capacity, and this lower range of emission is reflected in the best available technique reference document on chlor-alkali production. The phase-out of Hg in the chlor-alkali industry is Volume 60 March 2010

Pacyna et al. Table 6. Annual investment and operating costs for chlor-alkali industry as calculated within the EU ESPREME project.a Annual Costs (US$ 2008/t chlorine)

Sector Chlorine production (Hg cell plants)

Emission Control Measures Good practices during maintenance and repair Improvements of the Hg cells Wet scrubber with chlorinated brine or hypochloride addings ACI ⫹ FF–optimized Technology switching to diaphragm or membrane cells

Hg Reduction (%)

Annual Investment Costs

Annual Operating Costs

Annual Total Costs

20 15 60

0.02 0.1 1.7

0.02 0.02 1.4

0.04 0.1 3.0

98 100

– 37.0

4.3 0.00

– 37.0

Notes: aThe accuracy of cost estimates in the table is within ⫾50%.

expected to be a fairly straight-line phase-out of remaining Hg cell capacity by 2020.33 Hg emissions from MCCAPs in regions other than North America and Europe seem to be higher. Srivastava31 reported that the Hg consumption in Indian companies is at least 50 times higher than in the world’s best companies. This high consumption of Hg results in emission of approximately 47 g of Hg/t of caustic soda produced— one of the highest emission factors ever noted for this industry. Srivastava35 called for a serious effort by the Indian chlor-alkali industry to move toward membrane cell technology. Hg Abatement Efficiency and Costs The existing MCCAPs use various control techniques to reduce Hg emissions, including (1) gas-stream cooling, (2) mist eliminators, (3) scrubbers, and (4) adsorption on activated carbon or molecular sieves.36 Gas-stream cooling is often used as the primary Hg control technique or as a preliminary removal step to be followed by a more efficient control device. Mist eliminators can be used to remove Hg droplets, water droplets, or PM from the cooled gas streams. Scrubbers are used to chemically absorb the Hg from the hydrogen stream and the end box ventilation streams. Sulfur- and iodine-impregnated carbon adsorption systems are commonly used to reduce the Hg levels in the hydrogen gas stream if high removal efficiencies are desired. This method requires pretreatment of the gas stream by primary or secondary cooling followed by mist eliminators to remove approximately 90% of the Hg content from the gas stream. A major review of information on the costs of abatement for existing MCCAPs was carried out within the EU ESPREME and DROPS projects. The results for chlorine production are summarized in Table 6. EPA promulgated an emissions standard on the basis of MACT in December of 2003 to limit Hg emissions from this industry. The MACT rule requires controls and emissions limits for process vents and relatively stringent work practice standards or a cell room monitoring program to minimize fugitive emissions from the cell rooms. The total estimated capital cost of the final rule for the nine Hg cell chlor-alkali plants was approximately $1.6 million, and the total estimated annual cost is approximately $1.4 million/yr. Plant-specific annual costs in the authors’ Volume 60 March 2010

estimate range from approximately $130,000 for the leastimpacted plant to about $260,000 for the worst-impacted plant. The final rule will reduce Hg air emissions from existing emission points within Hg cell chlor-alkali plants by 675 kg/yr, a 74% reduction from current levels. The final rule also requires rigorous work practice standards such as periodically washing down work floors and covering waste containers. These requirements will reduce Hg emissions from so called “fugitive sources” throughout the plants. Although EPA is not able to accurately quantify the reductions associated with these work practice standards, these requirements will reduce Hg air emissions throughout the industry. By any accounting, the costs of implementing the MACT rule are significantly less than facility conversion to non-Hg cell technologies. Benefits of Hg Emissions Abatement Recent studies on the costs and benefits of reducing Hg emissions from U.S. coal combustion facilities were used to derive a conservatively estimated annual EU health benefit of some $39 –$47/g of MCCAP atmospheric Hg emissions eliminated.30 Concorde33 also analyzed the costs and benefits (especially energy savings, reduced costs of Hg monitoring, and waste disposal, etc.) to industry of converting a typical MCCAP to the membrane process. There are various cases of actual conversions that have generated an attractive 2- to 3-yr return on investment. However, it was pointed out that a EU industry investment on average in conversion of the MCCAP process to the membrane process may not show an attractive bottom-line return for almost 10 yr. The Concorde study33 concludes that when combining the considerable bottom-line benefits of MCCAP conversion with a conservative estimate of the public health benefits (even when accumulated over only 5 yr), it can be expected that the overall benefits are nearly twice the costs associated with the technology transition. Therefore, the conversion of MCCAPs should be regarded with a high priority when discussing the whole range of public health and other benefits associated with industrial development of the chemical industry. DISCUSSION A qualitative cost-benefit assessment is useful for obtaining a general overview on whether various measures are Journal of the Air & Waste Management Association 313

Pacyna et al. economically viable to society, and when combined with quantitative information it can thus be a useful tool to establish criteria for government intervention. From a welfare economics perspective, if a high standard of wellbeing is to be maintained, the concerns for human health and a healthy environment must be balanced against requirements of economic growth. Therefore, a project is regarded as beneficial if no member of the society becomes worse off and at least one becomes better off. In economic literature this is often referred to as “Pareto optimality.” It is presently uncommon for countries to invest in technologies to reduce only Hg from the emissions stream. Instead, countries usually use a multipollutant approach, which is much more cost-effective. For example, approaches and technologies for controlling conventional air pollutants (including PM, SO2, and NOx) typically result in some reduction of Hg emissions as a cobenefit. In most countries, Hg controls are contingent on controls for conventional pollutants, although the degree of Hg capture by various technologies varies widely. In this context, the incremental cost of adding a Hg reduction effort to a national strategy is much smaller. Efficient, nontechnological measures and pretreatment methods are also available for the reduction of Hg releases from various uses of products containing Hg. These measures include bans on use and substitution of products containing Hg and cleaning of raw materials before their use (e.g., coal cleaning). These measures also include energy conservation options, such as energy taxes, consumer information, energy management, and improvement of efficiency of energy production through a co-generation of electricity and heat in coal-fired power plants. Other potential measures affecting Hg emissions also comprise prevention options aimed at reducing Hg in wastes, material separation, labeling of Hg-containing products, and input taxes on the use of Hg in products. The message from the review of abatement installations for reduction of Hg emissions from various anthropogenic sources from UNEP’s second Ad Hoc Open-Ended Working Group on Mercury is that there are several technological and nontechnological solutions available than could be presently used to reduce Hg emissions by the year 2020. Of course, it is expected that even more technological measures will be available in the near future, particularly in the field of application of renewable sources of energy production and the improvement of Hg removal using the “add on” measures in addition to ESPs/ FFs combined with FGD. There is also a great potential for improvement of nontechnological measures such as a decrease in the use of Hg in the future and development of incentives for applications of measures aimed at reduction of Hg emissions to the environment. CONCLUSIONS This qualitative assessment shows that costs vary from medium to large for reduction of Hg from coal usage and industrial processes. For reduction of Hg from waste and the chlor-alkali industry, the costs vary from small to large. Measures with the application of technology (e.g., implementation of installations to remove Hg from the flue gases in electric power plants, waste incinerators, and 314 Journal of the Air & Waste Management Association

smelters) are typically rather expensive (medium to large costs) compared with nontechnological measures (e.g., prevention activity, capacity building, and promotion of Hg-containing waste separation) (small to medium costs). Both groups of measures would result in medium or large benefits. This indicates that the technological and nontechnological solutions for Hg emission and exposure reductions can be carried out in parallel. More emphasis on technological measures can be placed on the developed countries, whereas the process of emission and exposure reduction in the developing countries may start with nontechnological solutions. Technological solutions may be gradually introduced in these countries as a follow-up process after nontechnological solutions are in place. ACKNOWLEDGMENTS The information presented in this paper has been prepared within the UNEP Chemicals project on costs and benefits for each of the strategic objectives required for reducing the environmental and health risks from Hg set out in Annex 1 of the report from the first meeting of the Governing Council of UNEP’s Ad Hoc Open-Ended Working Group on Mercury. The authors are grateful for financial support for this work by UNEP Chemicals. The authors also acknowledge the contribution of Dr. Damian Panasiuk and Anna Glodek of NILU Polska, Katowice, Poland to this paper. REFERENCES 1. Arctic Pollution, 2006, Acidification and Arctic Haze; Arctic Monitoring and Assessment Programme: Oslo, Norway, 2006. 2. Jakus P.; McGuinness M.; Krupnick, A. The Benefits and Costs of Fish Consumption Advisories for Mercury in the Chesapeake Bay. Resources for the Future; Discussion Paper 02-55; Resources for the Future: Washington, DC, 2002; available at http://www.rff.org/Documents/RFF-DP-0255.pdf. 3. Kjellstrom, T.; Kennedy, P.; Wallis, S.; Mantell, C. Physical and Mental Development of Children with Prenatal Exposure to Mercury from Fish. Stage I: Preliminary Tests at Age 4; Report 3080; National Swedish Environmental Protection Board: Solna, Sweden, 1986. 4. Myers, G.J.; Davidson, P.W.; Cox, C.; Shamlaye, C.F.; Palumbo, D.; Cernichiari, E.; Sloane-Reeves, J.; Wilding, G.E.; Kost, J.; Huang, L.S.; Clarkson, T.W. Prenatal Methylmercury Exposure from the Ocean Fish Consumption in the Seychelles Child Development Study; Lancet 2003, 361, 1686-1692. 5. Grandjean, P.; Weihe, P.; White, R.F.; Debes, F.; Araki, S.; Yokoyama, K.; Murata, K.; Sørensen, N.; Dahl, R.; Jørgensen, P.J. Cognitive Deficit in 7-Year-Old Children with Prenatal Exposure to Methylmercury; Neurotoxicol. Teratol. 1997, 19, 417-428. 6. Trasande, L.; Landrigan, P.J.; Schechter, C. Public Health and Economic Consequences of Methyl Mercury Toxicity to the Developing Brain; Environ. Health Perspect. 2005, 113, 590-596. 7. Spadaro, V.J.; Rabl, A. Global Health Impacts and Costs Due to Mercury Emissions; Risk Anal. 2008, 28, 603-613. 8. UNEP Report on a General Qualitative Assessment of the Potential Costs and Benefits Associated with Each of the Strategic Objectives Set out in Annex 1 of the Report of the First Meeting of the Open Ended Working Group; U.N. Environmental Programme: Geneva, Switzerland, 2008; available at http://www.chem.unep.ch/mercury/OEWG2/documents/e5)/ English/OEWG_2_5_REV1.pdf (accessed 2009). 9. Pacyna, J.M.; Munthe, J.; Wilson, S. Global Atmospheric Mercury Assessment: Sources, Emissions and Transport; U.N. Environmental Programme: Geneva, Switzerland, 2008 10. Pacyna, J.M.; Sundseth, K.; Pacyna, E.P.; Munthe, J.; Belhaj, M.; Astrom, S.; Panasiuk, D.; Glodek, A. Socio-Economic Costs of Continuing the Status-Quo of Mercury Pollution; Tema Nord 2008:580; Report to the Nordic Council of Ministers: Oslo, Norway, 2008; available at http:// www.norden.org/pub/sk/showpub.asp?pubnr⫽2008:580 (accessed 2009). 11. Kraus, K.; Wenzel, S.; Howland, G.; Kutschera, U.; Hlawiczka, S.; Peeters Weem, A.; French, C. Assessments of Technological Developments: Best Available Techniques (BAT) and Limit Values; Report for the Task Force on Heavy Metals, the U.N. Economic Commission for Volume 60 March 2010

Pacyna et al. Europe Convention on Long-Range Transboundary Transport of Air Pollution: Geneva, Switzerland, 2006. 12. Senior, C. Mercury Reduction from Coal via Washing. Paper presented at the Symposium on Mercury Emissions, Research Triangle Park, NC, 2009. 13. Effective Mercury Reduction Strategy for Western Coal/K-Fuel Technology; Report prepared for the U.S. Environmental Protection Agency by Black & Veatch: Chesterfield, MO, 2003. 13a. Enhancing Mercury Control on Coal-Fired Boilers with SCR, Oxidation Catalyst, and FGD; Institute of Clean Air Companies (ICAC): Washington, DC, 2009; available at http://www.icac.com (accessed 2010). 14. Serre, S.; Lee, C.W.; Chu, P.; Hastings, T. Evaluation of the Impact of Chlorine on Mercury Oxidation in a Pilot-Scale Coal Combustor: the Effect of Coal Blending; Paper presented at the MEGA Symposium, Baltimore, MD, 2008. 15. Rini, M.J.; Vosteen, B.W. Full-Scale Test Results from a 600 MW PRBFired Unit Using Alstom’s KNX Technology for Mercury Control. Paper presented at the MEC6 Conference, Ljubljana, Slovenia, 2009. 16. Visschedijk, A.J.H.; Denier van der Gon, H.A.C.; van het Bolscher, M.; Zandveld, P.Y.J. Study to the Effectiveness of the UNECE Heavy Metals (HM) Protocol and Cost of Additional Measures; TNO Report 2006-AR0087/B; TNO: Apeldorn, The Netherlands, 2006 17. Control of Mercury Emissions from Coal-Fired Electric Utility Boilers: an Update; U.S. Environmental Protection Agency; Air Pollution Prevention and Control Division; National Risk Management Research Laboratory: Research Triangle Park, NC, 2005. 18. Feeley, T.J., III; Brickett, L.A.; O’Palko, A.; Jones A.P. DOE/NETL’s Mercury Control Technology R&D Program: Taking Technology from Concept to Commercial Reality. Paper presented at the MEGA Symposium, Baltimore, MD, 2008. 19. Rice, G.; Hammitt, J.K. Economic Valuation of Human Health benefits of Controlling Mercury Emissions from U.S. Coal-fired Power Plants; Northeast States for Coordinated Air Use Management: Boston, MA, 2005. 20. Pacyna, J.M. Publishable Final Activity Report of the EU DROPS Project; Norwegian Institute for Air Research: Kjeller, Norway, 2008. 21. Pacyna, J.M. In Toxic Metals in the Atmosphere. Advances in Environmental Science and Technology; Nriagu, J.O., Davidson, C.I., Eds.; John Wiley & Sons: Chichester, United Kingdom, 1986. 22. Joint EMEP/CORINAIR Atmospheric Emission Inventory Guidebook; U.N. Economic Commission for Europe: Geneva, Switzerland, 2000. 23. Pacyna, J.M. Habilitation Thesis, Wroclaw University of Technology, Wroclaw, Poland, 1980. 24. Characterization of Mercury-Enriched Coal Combustion Residues from Electric Utilities Using Enhanced Sorbents for Mercury Control; EPA-600/ R-06/008; U.S. Environmental Protection Agency; National Risk Management Laboratory: Research Triangle Park, NC, 2006. 25. Characterization of Coal Combustion Residues from Electric Utilities Using Wet Scrubbers for Multi-Pollutant Control; EPA-600/R-08/077; U.S. Environmental Protection Agency; National Risk Management Laboratory: Research Triangle Park, NC, 2008. 26. Pacyna J.M.; Zwozdziak A.; Zwozdziak J.; Matyniak Z.; Kuklinski A.; Kmiec, G. Zagadnienia Ochrony Powietrza Atmosferycznego dla Stanu Istniejacego i Perspektywicznego na Obszarze Pilotowym Wojewodztwa Legnickiego; Report SPR 7/81; Instytut Inzynierii Ochrony Srodowiska Politechniki Wroclawskiej: Wroclaw, Poland, 1981. 27. Pacyna J.M.; Pacyna, E.G. Assessment of Global and Regional Emissions of Trace Metals to the Atmosphere from Anthropogenic Sources Worldwide; Can. J. Environ. Rev. 2001, 9, 269-298.

Volume 60 March 2010

28. Locating and Estimating Air Emissions from Sources of Mercury and Mercury Compounds; Report EPA-454/R-93-23; U.S. Environmental Protection Agency: Research Triangle Park, NC, 1993. 29. Pye, S.; Jones, G.; Tewart, R.; Woodfield, M.; Kubica, K.; Kubica, R.; Pacyna, J.M. Costs and Environmental Effectiveness of Options for Reducing Mercury Emissions to Air from Small-Scale Combustion Installations; Report AEAT/ED48706; AEA Technology/NILU Polska: Harwell, U.K., 2005. 30. Gupta, S.; Mohan, K.; Prasad, R.; Gupta, S.; Kansal A. Solid Waste Management in India: Options and Opportunities; Res. Conserv. Recyc. 1998, 24, 137-154. 31. Thermal Desorption at the Lipari Landfill, Operable Unit 3, Pitman, New Jersey; Federal Remediation Technologies Roundtable; available at http://costperformance.org/profile.cfm?ID⫽137&CaseID⫽137 (accessed January 12, 2008). 32. Mercury: Investigation of a General Ban; KEMI Report 4/04; Swedish Chemical Inspectorate (KEMI): Stockholm, Sweden, 2004. 33. Status Report: Mercury Cell Chlor-Alkali Plants in Europe; Report prepared by Concorde East/West Sprl for the European Environmental Bureau: Brussels, Belgium, 2006. 34. Proposal for a Regulation of the European Parliament and the Council on the Banning of Exports and the Safe Storage of Metallic Mercury; Document COM(2006)636; Final SEC(2006) 1370; European Commission: Brussels, Belgium, 2006. 35. Srivastava R.C. Guidance and Awareness Raising Materials under New UNEP Mercury Programs; Center for Environmental Pollution Monitoring and Mitigation: Nirala Nagar, India, 2008. 36. Mercury Study Report to Congress; EPA-600/P-94/002a; U.S. Environmental Protection Agency: Washington, DC, 1995. 37. Feeley, T.J., III; Brickett, L.A.; O’Palko, A.; Murphy J.T. Field Testing of Mercury Control Technologies for Coal-Fired Power Plants. Paper presented at the Mercury R&D Review Meeting, December 2005. 38. Kilgroe, J.; Sedman, C.B.; Srivastava, R.; Ryan, J.; Lee, C.; Thorneloe, S. Control of Hg Emissions from Coal-Fired Electric Utility Boilers: Interim Report; EPA-600/SR-01-109; U.S. Environmental Protection Agency: Washington, DC, 2002.

About the Authors Jozef M. Pacyna is a director and professor, and Kyrre Sundseth and Elisabeth G. Pacyna are research scientists at the Norwegian Institute of Air Research in Kjeller, Norway. Jozef M. Pacyna is also a professor II at the Gdansk University of Technology. John Munthe is a director and research scientist, and Mohammed Belhaj and Stefan Astrom are research scientists at the IVL Swedish Environmental Research Institute in Gothenburg, Sweden. Please address correspondence to: Jozef M. Pacyna, Norwegian Institute for Air Research, Instituttveien 18, P.O.Box 100, NO-2027 Kjeller, Norway; phone: ⫹47-63-89-81-55; fax: ⫹47-63-89-80-50; e-mail: [email protected].

Journal of the Air & Waste Management Association 315