RD282 - Virginia Mercury Study, Revised October 21, 2008


Executive Summary:
This report was replaced in its entirety by the Department of Environmental Quality on October 27, 2008.

This report has been prepared pursuant to the requirements of Chapter 867 of the 2006 Acts of Assembly (House Bill 1055). The Act directs the Department of Environmental Quality (DEQ) to conduct a detailed assessment of mercury deposition in Virginia in order to determine whether particular circumstances exist that justify, from a health and cost and benefit perspective, requiring additional steps to be taken to control mercury emissions within Virginia. The assessment included (i) an evaluation of the state of mercury control technology for coal fired boilers, including the technical and economic feasibility of such technology and (ii) an assessment of the mercury reductions and benefits expected to be achieved by the implementation of the Clean Air Interstate Rule (CAIR) and Clean Air Mercury Rule (CAMR) regulations. An interim report was provided by DEQ in October 2007 that provided a status report on the assessment. The interim report is available at http://www.deq.virginia.gov/export/sites/default/regulations/pdf/2007statusofhgstudy.pdf.

DEQ used a contractor experienced with performing mercury deposition modeling to assist with identifying the mercury reductions and benefits to be achieved in Virginia as a result of implementation of the CAIR and CAMR. The analysis DEQ performed differed from the analysis the U.S. Environmental Protection Agency (EPA) performed for the CAMR. As part of Virginia’s study, the emission inventory for sources in Virginia was reviewed and modified to reflect the most up-to-date information concerning mercury emissions from stationary sources located within Virginia. Additionally, ICF worked with electric generating units (EGUs) to obtain information on the specific pollution control equipment industry plans to install in the future and the predicted emission reductions related to the installation and operation of those pollution control tools. In contrast, EPA’s analysis made general assumptions concerning future controls and associated mercury reductions without obtaining information on facilities’ future plans from industry. Virginia’s report also focuses more closely on impacts to Virginia fish, the number of fish consumption advisories issued for Virginia fish and the potential for reduced fish advisories in the future as a result of less mercury deposition occurring in Virginia waters.

This study began in 2006 once the regulatory details of CAIR and CAMR were known. In February 2008, the U. S. Circuit Court of Appeals for the District of Columbia issued an opinion vacating CAMR. In July 2008 the U. S. Circuit Court of Appeals for the District of Columbia issued an opinion vacating the CAIR. Although the D.C. Circuit recently issued opinions vacating CAIR and CAMR, the agency has continued to move forward with completion of the report pursuant to the requirements and direction of House Bill 1055. As directed, this report examines modeling results anticipated to be achieved through the implementation of CAIR and CAMR requirements. Any reductions of mercury deposition and average mercury fish tissue concentrations identified in this report are based on modeling results and may not ultimately be achieved.

Mercury Deposition Modeling

The mercury deposition modeling conducted by ICF used data from the years 2001 and 2002 to develop a baseline year estimate for mercury deposition occurring in Virginia and surrounding states. This baseline year estimates the mercury deposition occurring before implementation of CAIR and CAMR. Modeling was performed to estimate the deposition of mercury occurring in 2018, after CAIR and CAMR had been implemented. The modeling conducted for this study indicates overall mercury deposition for Virginia would be lower by 20.4 percent for 2018, when compared to the base year. The greatest reduction in deposition comes from EGU sources located outside of Virginia (in the 12-km modeling domain that encompasses several nearby states), and 61 percent of the reduction in mercury deposition for Virginia is attributable to reductions in emissions from EGU sources in these nearby states. In addition, 7.2 percent of the overall simulated mercury reduction for Virginia is attributable to reductions in the emissions from EGU sources located within the state, 5.7 percent is attributable to reductions in the emissions from non-EGU sources in the state, 4.6 percent is attributable to reductions in non-EGU sources in nearby states, and 2.8 percent is attributable to emissions reductions in the remainder of the United States.

Fish Tissue Impacts

After examining the reductions of mercury deposition predicted to occur in Virginia as a result of implementation of measures to comply with CAIR and CAMR, there may be reductions in the number of mercury fish consumption advisories in place within Virginia. Of the 13 mercury-sensitive waterbodies in Virginia with current fish consumption advisories due to mercury contamination, the fish mercury levels may be lowered enough in the future (to below the 0.5 parts per million (ppm) mercury level currently used by the Virginia Department of Health (VDH)) such that three or four of these advisories may no longer be warranted. In all but two of the advisory areas, at least one species of fish may have reduced mercury levels in the future that could allow for its removal from the fish consumption advisory and in one case, (Dismal Swamp Canal), the advisory area may be reduced. Under the projected reduced air deposition rates for the future, nine to 10 of the current fish consumption advisories will likely remain in place for at least one species of fish.

It will take time for any reductions in mercury deposition to be reflected in fish tissue samples because the ecosystem must readjust to the lower mercury levels in the environment. Each individual water body will react slightly differently due to natural variances in the chemical and physical conditions and differences in food web structure. Lakes are expected to respond quickest (within a few years to decades) to reduced mercury deposition, with wetlands requiring more time to equilibrate to the lowered mercury inputs.

The DEQ has proposed the adoption of a fish tissue criterion for mercury of 0.30 ppm, which is lower than the fish tissue mercury level used by the VDH to determine when fish consumption advisories are issued. If the State Water Control Board adopts this criterion, waterbodies with average fish concentrations greater than 0.30 ppm will be classified as impaired. Even though reductions in mercury deposition may occur and some fish consumption advisories may be removed, the waterbodies examined in this study could remain classified as impaired by DEQ if average mercury concentrations for at least one species of fish remain higher than 0.30 ppm.

Cost Benefit Analysis

Virginia coal-fired power plants vary in the amount and type of mercury control equipment installed. Currently, all Virginia coal-fired power plants burn a low-sulfur, low-mercury, and high-chlorine bituminous coal, and most of the plants also burn coal that has been initially washed and processed after mining. Furthermore, some of the plants have technologies already in place to control nitrogen oxide (NOx), sulfur dioxide (SO2) and particulate matter (PM). As a result, a certain level of mercury (Hg) removal is achieved as a co-benefit of these controls; this report attempts to capture the costs of mercury control (costs of control technologies and also possible costs of control levels).

The costs of mercury control at coal-fired power plants are affected by a number of parameters, including what technologies are chosen, what regulations are in place, and the market-based determination of demand versus supply of energy. A number of options for reducing mercury emissions from coal-fired power plants are commercially available, and others are being developed. A number of control technologies for the reduction of mercury are available to coal-fired power plants, allowing the facility to choose the best fit in terms of cost-effectiveness. The DEQ cost assessment was based on a thorough review of existing and future projected mercury controls by Virginia-based electric generating units. Specifically, best available information on control technologies (performance, constraints, market prices of inputs and by-product disposal estimates) was used in this analysis. The results support the view, which is widely held by EPA, the U.S. Department of Energy (DOE), industry research and other state agencies, that mercury control is more cost-effective if coal-fired power plants adopt a multi-pollutant, post-combustion control technology sequence.

Fish Consumption Trends in Virginia’s Waterways

As part of this study, DEQ contracted with Virginia Commonwealth University’s Center for Environmental Studies (VCU-CES) to obtain Virginia-specific fish consumption data collected in areas where mercury fish consumption advisories are in effect. Additionally, VCU-CES was tasked with estimating the associated health risks from resulting methylmercury exposures. VCU-CES developed a fish consumption survey, and worked with DEQ staff to identify the launching and fishing locations where anglers could be surveyed. The survey was designed to obtain information on fishing behaviors, fish consumption, and demographic data on the anglers and families. During the summer of 2007, a team from VCU-CES administered the survey to 158 anglers at boat launching and fishing sites. Surveys were completed for anglers who were fishing at 17 locations on 5 rivers: the James River below Richmond, the Chickahominy, Pamunkey, Mattaponi, and upper Piankatank rivers. These rivers are affected by methylmercury contamination, have been surveyed in previous, similar investigations and are used by anglers for recreational fishing.

The surveys were administered to anglers predominantly on Friday, Saturday or Sunday. Approximately 44 percent of all respondents and their families consume the fish that they catch from these waters. Half (50 percent) of the anglers only, not family members, consume some fish that they catch, and more men (54 percent) than women (43 percent) were reported to consume the fish with elevated methylmercury levels. The most commonly consumed fish were catfish, spot or croaker, sunfish and largemouth bass; catfish and largemouth bass are two of the species on the fish consumption advisory. Catfish also represented the largest number of meals and total amount of self-caught fish consumed per year.

The data on fish consumption were analyzed with DEQ data on methylmercury concentrations in fish that had been collected in previous years to estimate the amount of methylmercury consumed in fish yearly. In order to estimate total methylmercury from all fish consumption, canned tuna and purchased fish consumption were added to mercury exposures from self-caught fish. Mercury levels in tuna and purchased fish were taken from national data.

The methylmercury exposures determined from survey data and DEQ fish tissue levels were compared to the dose of mercury exposure that EPA has set (and the VDH uses) as the dose without appreciable health risks.

The analysis of the fish consumption and fish tissue concentrations was performed using a probabilistic computer program that is used for risk assessments. This program randomly selects certain values, as defined, to use in the equations for determining total mercury from all fish consumed. The analysis indicates that a significant number of anglers who regularly catch and consume significant amounts of catfish and large mouth bass from the affected waters are exposed to methylmercury at levels above the EPA reference dose.

Using the information obtained from various statistical methods, VCU-CES modeled the loss of IQ points from prenatal exposure to methylmercury through the maternal diet, specifically mercury from consumption of mercury-contaminated fish. To model the loss of IQ points from prenatal exposure to methylmercury through the maternal diet, the target population of interest is women of childbearing age. With the survey results and fish mercury concentrations from DEQ’s fish tissue database, a probability distribution of ingested doses was created. Based upon the estimated maternal exposure to current fish mercury concentrations, the VCU-CES study estimated future levels of IQ changes due to 2010 and 2018 levels of mercury controls to result in average (mean) avoided IQ deficits of 0.03 IQ points.

Monetization of Human Health Risk Effects (IQ level)

This report attempts to quantify and monetize, to the extent feasible, the economic benefits associated with modeled avoided IQ deficits due to reduced exposure from the consumption of recreationally caught freshwater fish. The monetization of the human health risk effects (IQ being the human health effects of measurement) builds upon the findings of the VCU-CES study (Appendix B) and adopts the approach used by EPA to conduct the economic benefit analysis at the federal level (U.S. EPA 2005). This regional assessment focused on estimating the changes in exposures to women of childbearing age because adverse health effects in children have been linked to prenatal mercury exposures (Sorenson et al. 1999). This report builds on the VCU-CES study that focused on select counties of eastern Virginia where fish advisories for mercury existed and using consumption surveys, where IQ losses were estimated. IQ losses were then monetized to evaluate the economic benefit of mercury emission controls (or impacts of no reduction in emissions).

EPA’s CAMR analysis indicated a monetized impact of $15 million solely due to power plant emissions over the entire United States (3 percent discount rate and Year 2000 dollars); however, such an analysis is not representative of Virginia, Virginia-specific individual consumption patterns and DEQ’s fish tissue data. The DEQ assessment used 10 years of birth data for only the select counties where fish consumption patterns were surveyed to quantify economic impacts associated with average avoided IQ deficits of 0.03 IQ points found in the VCU-CES study and associated with methylmercury consumption through 2010 and 2018. Economic losses to the exposed populations of interest involved an assessment of two scenarios – worst-case and most likely. Under the worst case scenario, the estimated net per capita income earning loss to children is $337.00, or $4.8 million across all 14,364 children born in the select counties. Under the “most likely” scenario, it was estimated that 6,104 pre-natal children (i.e., less than half of the 14,364 children born in the select counties) would be exposed to methylmercury and would thus have net income losses totaling $2.05 million. The two monetized scenarios are estimates of impacts for areas where risk assessment of methylmercury exposure due to fish consumption was undertaken.

Conclusions

As a result of conducting this study, specific information concerning mercury deposition in Virginia was obtained. Excluding background and natural sources of mercury, the largest percentage of mercury deposition within Virginia originates from EGUs in surrounding states (54 percent). The next largest geographic source contributing to mercury deposition in Virginia is EGUs located within Virginia (14 percent). Non-EGUs in surrounding states contribute to 13 percent of the deposition occurring within Virginia, and in-state non-EGUs contribute to 12 percent of the deposition occurring within Virginia.

As part of the mercury modeling conducted by ICF, emissions and deposition information from the 15 largest mercury emitters in the state was modeled using the AERMOD model to examine the direct impact these facilities have on the area within a three km area surrounding each source. This analysis yielded three key findings: (1) dry deposition is greater than wet deposition for all facilities, (2) maximum wet deposition tends to occur at locations closest to the facility, and (3) maximum dry deposition tends to occur farther away from the facility location. The AERMOD model also corroborated the findings of the regional-scale modeling. Specifically, individual facilities located in Virginia contribute to mercury deposition within the state, and the greatest impacts from the in-state sources are simulated near the source locations. This includes EGU sources and non-EGU sources.

As mercury deposition into waterbodies is reduced, each individual waterbody is expected to react slightly differently due to natural variances in the chemical and physical conditions and differences in food web structure. Lakes are expected to respond the most quickly (within a few years to decades) to reduced mercury deposition, with wetlands requiring more time to equilibrate to the lowered mercury inputs.

The VDH issues fish consumption advisories when average concentrations of mercury in fish exceed 0.50 ppm. Under the projected reduced mercury air deposition rates for the future, nine to 10 of the current fish consumption advisories will likely remain in place for at least one species of fish. The DEQ has recently proposed the adoption of a fish tissue criterion for mercury of 0.30 ppm, which is lower than the threshold concentration used by the VDH to issue fish consumption advisories. If the State Water Control Board adopts this fish tissue criterion for mercury, in the future DEQ may classify some waterbodies as impaired due to elevated mercury contamination in fish before the VDH would find it necessary to issue a fish consumption advisory.