Nicholas A. Robins

Observational and modeling constraints on global anthropogenic enrichment of mercury.

Centuries of anthropogenic releases have resulted in a global legacy of mercury (Hg) contamination. Here we use a global model to quantify the impact of uncertainty in Hg atmospheric emissions and cycling on anthropogenic enrichment and discuss implications for future Hg levels. The plausibility of sensitivity simulations is evaluated against multiple independent lines of observation, including natural archives and direct measurements of present-day environmental Hg concentrations. It has been previously reported that pre-industrial enrichment recorded in sediment and peat disagree by more than a factor of 10. We find this difference is largely erroneous and caused by comparing peat and sediment against different reference time periods. After correcting this inconsistency, median enrichment in Hg accumulation since pre-industrial 1760 to 1880 is a factor of 4.3 for peat and 3.0 for sediment. Pre-industrial accumulation in peat and sediment is a factor of ∼ 5 greater than the precolonial era (3000 BC to 1550 AD). Model scenarios that omit atmospheric emissions of Hg from early mining are inconsistent with observational constraints on the present-day atmospheric, oceanic, and soil Hg reservoirs, as well as the magnitude of enrichment in archives. Future reductions in anthropogenic emissions will initiate a decline in atmospheric concentrations within 1 year, but stabilization of subsurface and deep ocean Hg levels requires aggressive controls. These findings are robust to the ranges of uncertainty in past emissions and Hg cycling.

Mercury production and use in colonial Andean silver production: emissions and health implications.

Background: Colonial cinnabar mining and refining began in Huancavelica, Peru, in 1564. With a local source of mercury, the amalgamation process was adopted to refine silver in Potosí, Bolivia, in the early 1570s. As a result, large quantities of mercury were released into the environment. Objectives: We used archival, primary, and secondary sources to develop the first estimate of mercury emissions from cinnabar refining in Huancavelica and to revise previous estimates of emissions from silver refining in Potosí during the colonial period (1564–1810). Discussion: Although other estimates of historical mercury emissions have recognized Potosí as a significant source, Huancavelica has been overlooked. In addition, previous estimates of mercury emissions from silver refining under-estimated emissions because of unrecorded (contra-band) production and volatilization of mercury during processing and recovery. Archival descriptions document behavioral and health issues during the colonial period that are consistent with known effects of mercury intoxication. Conclusions: According to our calculations, between 1564 and 1810, an estimated 17,000 metric tons of mercury vapor were emitted from cinnabar smelting in Huancavelica, and an estimated 39,000 metric tons were released as vapor during silver refining operations in Potosí. Huancavelica and Potosí combined contributed > 25% of the 196,000 metric tons of mercury vapor emissions in all of Latin America between 1500 and 1800. The historical record is laden with evidence of mercury intoxication consistent with effects recognized today. Our estimates serve as the foundation of investigations of present-day contamination in Huancavelica and Potosí resulting from historical emissions of mercury.

Estimations of historical atmospheric mercury concentrations from mercury refining and present-day soil concentrations of total mercury in Huancavelica, Peru

Detailed Spanish records of cinnabar mining and mercury production during the colonial period in Huancavelica, Peru were examined to estimate historical health risks to the community from exposure to elemental mercury (Hg) vapor resulting from cinnabar refining operations. Between 1564 and 1810, nearly 17,000 metric tons of Hg were released to the atmosphere in Huancavelica from Hg production. AERMOD was used with estimated emissions and source characteristics to approximate historic atmospheric concentrations of mercury vapor. Modeled 1-hour and long-term concentrations were compared with present-day inhalation reference values for elemental Hg. Estimated 1-hour maximum concentrations for the entire community exceeded present-day occupational inhalation reference values, while some areas closest to the smelters exceeded present-day emergency response guideline levels. Estimated long-term maximum concentrations for the entire community exceeded the EPA Reference Concentration (RfC) by a factor of 30 to 100, with areas closest to the smelters exceeding the RfC by a factor of 300 to 1000. Based on the estimated historical concentrations of Hg vapor in the community, the study also measured the extent of present-day contamination throughout the community through soil sampling and analysis. Total Hg in soils sampled from 20 locations ranged from 1.75 to 698 mg/kg and three adobe brick samples ranging from 47.4 to 284 mg/kg, consistent with other sites of mercury mining and use. The results of the soil sampling indicate that the present-day population of Huancavelica is exposed to levels of mercury from legacy contamination which is currently among the highest worldwide, consequently placing them at potential risk of adverse health outcomes.

Residential mercury contamination in adobe brick homes in Huancavelica, Peru.

This is the first study of adobe brick contamination anywhere in the world. Huancavelica, Peru is the site of historic cinnabar refining and one of the most mercury (Hg) contaminated urban areas in the world. Over 80% of homes in Huancavelica are constructed with adobe bricks made from Hg contaminated soil. In this study we measured total Hg concentrations in adobe brick, dirt floor, surface dust, and air samples from the interior of 60 adobe brick houses located in four neighborhoods. Concentrations of total Hg in adobe bricks, dirt floors, and surface dust ranged from 8.00 to 1070 µg/g, 3.06 to 926 µg/g, and 0.02 to 9.69 µg/wipe, respectively, with statistically significant differences between the four neighborhoods. Concentrations of Hg in adobe brick and dirt floor samples in Huancavelica were orders of magnitude higher than in Ayacucho, a non-mining town in Peru. A strong correlation exists between total Hg concentrations in adobe bricks and dirt floors which confirms that adobe bricks were being made on-site and not purchased from an off-site source. A strong correlation between surface dust and adobe bricks and dirt floors indicates that walls and floors serve as indoor sources of Hg contamination. Elemental Hg vapor concentrations were below detection (<0.5 µg/m3) in most homes; however in homes with detectable levels, concentrations up to 5.1 µg/m3 were observed. No statistically significant differences in Hg vapor measurements were observed between neighborhoods. This study demonstrates that building materials used widely in developing communities, such as adobe bricks, may be a substantial source of residential Hg exposure in silver or gold refining communities where Hg is produced or used for amalgamation in artisanal gold production.

Speciation and bioaccessibility of mercury in adobe bricks and dirt floors in Huancavelica, Peru

Huancavelica, Peru, a historic cinnabar refining site, is one of the most mercury (Hg)-contaminated urban areas in the world. Exposure is amplified because residents build their adobe brick homes from contaminated soil. The objectives of this study were to compare two Hg-leaching procedures, and their application as risk-assessment screening tools in Hg-contaminated adobe brick homes in Huancavelica. The purpose was to evaluate potential health implications, particularly for children, after ingestion of Hg-contaminated particles. Hg was measured in adobe brick and dirt floor samples from 60 households by total Hg extraction, simulated gastric fluid (GF) extraction, and sequential selective extraction (SSE), which provides more detailed data but is resource-intensive. Most of the Hg present in samples was relatively insoluble, although in some households soluble Hg species were present at concentrations that may be of concern after ingestion. A strong correlation was identified between results from simulated GF extraction of adobe bricks and dirt floors and the more soluble fractions of Hg from SSE. Simulated GF extraction data were combined with ingestion and body mass characteristics for small children to compare potential risk of ingestion of Hg-contaminated soil with current health standards. Simulated GF extraction can be used as a risk assessment screening tool for effective allocation of time and resources to households that have measurable concentrations of bioaccessible Hg. Combining simulated GF extraction data with health standards enables intervention strategies targeted at households with the greatest potential health threat from ingestion of Hg-contaminated particles.

Residential metal contamination and potential health risks of exposure in adobe brick houses in Potosí, Bolivia

Potosí, Bolivia, is the site of centuries of historic and present-day mining of the Cerro Rico, a mountain known for its rich polymetallic deposits, and was the site of large-scale Colonial era silver refining operations. In this study, the concentrations of several metal and metalloid elements were quantified in adobe brick, dirt floor, and surface dust samples from 49 houses in Potosí. Median concentrations of total mercury (Hg), lead (Pb), and arsenic (As) were significantly greater than concentrations measured in Sucre, Bolivia, a non-mining town, and exceeded US-based soil screening levels. Adobe brick samples were further analyzed for bioaccessible concentrations of trace elements using a simulated gastric fluid (GF) extraction. Median GF extractable concentrations of Hg, As, and Pb were 0.085, 13.9, and 32.2% of the total element concentration, respectively. Total and GF extractable concentrations of Hg, As, and Pb were used to estimate exposure and potential health risks to children following incidental ingestion of adobe brick particles. Risks were assessed using a range of potential ingestion rates (50-1000mg/day). Overall, the results of the risk assessment show that the majority of households sampled contained concentrations of bioaccessible Pb and As, but not Hg, that represent a potential health risk. Even at the lowest ingestion rate considered, the majority of households exceeded the risk threshold for Pb, indicating that the concentrations of this metal are of particular concern. To our knowledge, this is the first study to quantify key trace elements in building materials in adobe brick houses and the results indicate that these houses are a potential source of exposure to metals and metalloids in South American mining communities. Additional studies are needed to fully characterize personal exposure and to understand potential adverse health outcomes within the community.