On the Air

Three Decades of Mercury Levels in the Tennessee River System

The Tennessee Valley Authority (TVA) is a major electricity producer with much of this power produced from coal-fired power plants as can be seen in Figure 1. All coal contains trace levels of mercury that is released when the coal is burned. One would have expected mercury concentrations in both sediment and fish to increase if, indeed, local deposition were the primary contributing source. Based on TVA monitoring data from the 1970's to present (as described below), there has been no increase in mercury concentrations in the sediment or fish, while coal combustion has increased.

Although there is much current interest in atmospheric deposition, particularly from local and regional sources, it should be noted that the source of most atmospheric mercury is not local. Instead it comes from a variety of natural and human sources worldwide, each contributing to the total mercury content of the atmosphere where it continuously circles the globe. Mercury deposited from the atmosphere, even decades ago, is re-emitted continuously to the atmosphere by evaporation from soil, water, and other surfaces. The total mercury content of this environmental cycle of evaporation and deposition is termed the global pool. New mercury emissions from industrial processes, including the burning of coal that contains mercury, add to the total mercury content of the atmosphere. While the burning of coal by U.S. electric utilities is one source of mercury emissions, recent inventories of worldwide sources by the Electric Power Research Institute (EPRI) and others indicate that U.S. utilities account for only about 1 percent of new global mercury emissions.

Figure 1. Coal-fired power production by TVA.

The Tennessee Valley Authority has a long history of working with state agencies to determine the overall condition of streams and lakes as well as the safety of fish for consumption. TVA is a multi-purpose agency with environmental stewardship responsibilities that include the oversight and management of 45 reservoirs with 11,000 miles of shoreline. As part of its stewardship responsibilities, TVA collects sediment and fish tissue samples for analysis for mercury and other pollutants in Tennessee Valley streams and reservoirs, and shares this information with state agencies. States combine this information with their own monitoring results to determine if fish consumption advisories are appropriate. Mutual efforts by TVA, state agencies, and U.S. Environmental Protection Agency (EPA) that are specifically related to mercury began in 1970 in response to the discovery of mercury-contaminated fish in the Great Lakes and the adoption of guidelines for mercury in fish by the U.S. Food and Drug Administration. Results from the initial studies in the 1970’s provide the backdrop for comparing results in recent years to examine trends in mercury concentrations in the Tennessee Valley.

During the testing of fish species in the early 1970's, two locations, Pickwick Reservoir and the North Fork of the Holston River, were identified at that time as having "elevated" levels of mercury. Both locations were found to have received regular discharges from chlor-alkali plants using a production process known to require large amounts of mercury.

Much of the work from the 1970's and 1980's detailing mercury sampling in the Tennessee River system has been reviewed in order to pinpoint possible factors controlling the accumulation of mercury in aquatic systems. Principal sources included industrial discharges, agricultural and sediment runoff, geologic sources, and atmospheric deposition.

Once in the environment, mercury must be converted into a form that can enter and accumulate in the food chain. Five parameters were identified as having the greatest probability of influencing mercury accumulation rates in an aquatic ecosystem: the acidity of surface water, photosynthetic productivity, the rate at which sediment forms, the percentage of surface area covered by water or wetlands, and the concentration of organic matter in runoff. Most of the Tennessee River system ranks low in these factors.

Data Sources

Fish and sediment samples have been collected within the Tennessee Valley region from 1969 to the present. Through the 1970's and continuing into the early 1980's, analysis of mercury in fish or sediment was evaluated on an as-needed basis to address specific concerns. These efforts have resulted in a more complete data record for the large reservoirs on the Tennessee River than for tributary reservoirs and for the 1970's than the 1980's.

Standardization of sampling frequency, technique, and laboratory analysis did not begin until the establishment of the Vital Signs Monitoring Program in 1990. The Vital Signs Monitoring Program was developed to evaluate the ecological health of all rivers and reservoirs. The current Fish Tissue Program, originating in 1987, was included in this monitoring program and was used as a source for this paper.

Tennessee River and Reservoirs

Valley-wide mean mercury levels in sediments declined substantially during the period from 1970 to 2000; whereas mean concentrations in fish have remained relatively stable (Figure 2). The large reductions in sediment mercury occurred in the 1970's as many large industrial users of mercury changed production methods to less mercury-intensive operations. As an example, all chlor-alkali plants within the Tennessee Valley region converted from the mercury-cell process to more modern and less mercury-intensive technologies. In locations without historical industrial sources of mercury, no increases have been observed in sediment mercury concentrations either. Levels of mercury in sediment decrease as new sediment containing less mercury covers the older sediment, effectively capping the previous layers. Disruptions in sediments due to construction activities, however, could cause the mercury in these layers to again become available for biological uptake.

Figure 2. Mean mercury levels in fish and sediments of the Tennessee River reservoirs.

The rate of decline in mercury levels in Tennessee River reservoirs slowed substantially during the 1980's and the 1990's for sediment and especially for fish (Figure 2). Since 1970, the mercury concentrations in largemouth bass and channel catfish have remained relatively steady with no apparent increase. In addition to these observations, data from 23 tributary reservoirs and 7 tributary rivers within the Tennessee Valley show a decrease in sediment mercury levels to an overall average of 0.2 mg/kg, with most of the decrease occurring between 1970 and 1980. However, there has been no corresponding trend in fish tissue mercury levels. Mercury levels in fish tissue fluctuate throughout the period averaging between 0.1 to 0.3 mg/kg based on species and across tributaries. These data exclude the North Fork of the Holston River because of the very high levels there from historical industrial activity.

Historical Contamination

The North Fork Holston River has historically had the highest levels of sediment mercury of any sampled location (Figure 3). Almost from the first date of operation in the late 1890's, the Saltville, Virginia chlor-alkali plant released mercury into the North Fork of the Holston River. The Saltville plant eventually closed in 1972 but is still an EPA Superfund site monitored by EPA and state agencies. Figure 3 provides sediment data for both the North Fork of the Holston River and the Holston River. The impact of the mercury releases from the chlor-alkali plant on the North Fork is readily apparent in the sediment data for the 1970's. These very high sediment levels have declined significantly since the closure of the chlor-alkali plant.

Figure 3. Mercury levels in sediments of the North Fork of the Holston River and the Holston River.

Another industrial case involves a chlor-alkali plant that was built on the Pickwick Reservoir in Muscle Shoals, Alabama in 1954 with discharge near Tennessee River mile 258. There were high mercury levels noted in the sediment samples (up to 20 mg/kg) during the 1970's. By 1980, the plant completed conversion to a membrane process, which uses significantly less mercury. Since that conversion, the data indicate decreases in both fish and sediment mercury levels downstream from this location.

Conclusions

TVA's Vital Signs Monitoring Program has systematically monitored mercury levels since 1990. Analysis of these data sources indicate that mercury levels in the sediment of the Tennessee Valley region have declined over the 1970's, 1980's and 1990's to average levels below 0.5 mg/kg in the main river and below 0.2 mg/kg in the tributary reservoirs, excluding areas of known historical industrial mercury releases. Those sites with so-called mercury "hot spots" have been the result of industrial activity, particularly the chlor-alkali industry. Those chlor-alkali plants have either shut down or converted to a less mercury-intensive process. Mercury levels in fish tissue have declined in main stem reservoirs along the Tennessee River where known industrial sources have existed, but the same trend has not been noted for the tributary reservoirs where three decades of data exist. This decline in mercury levels in the Tennessee River system has occurred over a period of time when coal combustion has increased substantially, providing at least circumstantial indication that fossil power plant contributions to local mercury are not significant. Future work for this ongoing effort will include re-designing the database and search tools, as well as continuing the effort to locate additional sources of mercury data within the Tennessee Valley region.

Contacts

Mary F. Eubanks, 423-751-4316, mfeubanks@tva.gov
Thomas A. Burnett, 423-751-3938, taburnett@tva.gov
P. Alan Mays, 865-632-1634, pamays@tva.gov

If you would like additional information on important air quality topics, please contact Jeanie Ashe by telephone (256-386-2033), E-mail (jbashe@tva.gov), facsimile (256-386-2499), or TVA mail at CTR 1K-M, Muscle Shoals, Alabama 35662.