Fluoride Action Network

Azores: Health hazards and disaster potential of ground gas emissions at Furnas volcano, Sao Miguel

Source: Journal of Volcanology and Geothermal Research 92:95–106. | September 30th, 1999 | Authors: Peter J. Baxter, Jean-Claude Baubron, Rui Coutinho
Industry type: Volcanoes


A health hazard assessment of exposure to soil gases (carbon dioxide and radon) was undertaken in the village of Furnas, located in the caldera of an active volcano. A soil survey to map the area of soil gas flow was undertaken, gas emissions were monitored at fumaroles and in eight houses, and a preliminary radon survey of 23 houses in the main anomaly area was performed. Potential volcanic sources of toxic contamination of air, food, and water were also investigated, and ambient air quality was evaluated. About one-third (41 ha) of the houses were located in areas of elevated carbon dioxide soil degassing. Unventilated, confined spaces in some houses contained levels of carbon dioxide which could cause asphyxiation. Mean indoor radon levels exceeded UK and US action levels in the winter months. A tenfold increase in radon levels in one house over 2 h indicated that large and potentially lethal surges of carbon dioxide could occur without warning. Toxic exposures from the gaseous emissions and from contamination of soil and water were minimal, but sulphur dioxide levels were mildly elevated close to fumaroles. In contrast, evidence of dental fluorosis was manifested in the population of the nearby fishing village of Ribeira Quente where drinking water in the past had contained elevated levels of fluoride. The disaster potential of volcanic carbon dioxide in the area could also be associated with the hydrothermal system storing dissolved carbon dioxide beneath the village. Felt, or unfelt, seismic activity, or magma unrest, especially with a reawakening of explosive volcanic activity (30% probability in the next 100 years) could result in an increase in gas flow or even a gas burst from the hydrothermal system. A survey of all houses in Furnas is advised as structural measures to prevent the ingress of soil gases, including radon, were needed in some of the study houses. Evaluations of the human hazards of volcanic gases should be undertaken in all settlements in volcanic and hydrothermal areas associated with soil gas emissions.

Excerpts from article:

Ambient air pollution from the fumarole emissions appeared unlikely as these contained only very small amounts of sulphur gases, but at times of temperature inversion at night, the gas plume would occasionally be visible lying over the village. We placed sulphur dioxide diffusion tubes (Downing et al., 1994) for a 2-week sampling period in six locations on trees and posts in the vicinity of the fumaroles and in gardens. The presence of persistent, low-level air pollutants in the gases, such as fluoride and heavy metals (mercury, arsenic, cadmium), was investigated by collecting and analysing grasses and lichens (Sp. Parmelia perlata) growing on trees in selected locations in the vicinity of the fumaroles, in fields used by grazing cattle, and in gardens of houses in the village. Lichens have been shown to be useful monitors of fluoride of the plume on Kilauea and Mt. Etna (Davies and Notcutt, 1996) and industrial metal pollution of ambient air (De Wit, 1983). Mercury vapour is known to be emitted from many volcanoes in plumes (Siegel and Siegel, 1984) and soil gas, and it may contaminate the air, water, and vegetation in mineral-enriched geothermal areas (Barghigiani and Ristori, 1994; Edner et al., 1992). Lichens were analysed for the radon daughters 210Po and 210Pb to determine whether long-term emissions of radon into the ambient air could be detected (Kauranen and Miettinen, 1969). As part of the wider study of the volcano, water samples from the numerous springs in the caldera, including those used as sources for drinking water in Furnas and Ribeira Quente, were analysed for a wide range of anions and cations. Spot urine samples were collected from 30 random patients attending each of the family practitioner clinics in Furnas and Ribeira Quente in the same week, and these were analysed for fluoride, mercury, cadmium, arsenic, and thallium

3.6. Lichens and grasses

Fluoride concentrations in lichens were slightly raised near fumaroles and in gardens in the anomaly area (range: 10–75 ppm). Slightly elevated values were also obtained for mercury (0.15 ppm) and lead (10–56 ppm) in lichens from close to the fumaroles, but not elsewhere. Fluoride levels (5–20 ppm) were raised only in grasses on soils where there were high carbon dioxide emissions. The radionuclide levels in eight lichen samples were the same as background regardless of their proximity to the fumaroles: 210Po, 0.31–0.71 Bq/g; 210Pb, 0.38–0.72 Bq/g; dry weight.

3.7. Urine samples

The 30 subjects from Furnas ate fish at least three times a week, with one person eating fish everyday. In Ribeira Quente, 22 out of 29 subjects who took part ate locally caught fish everyday. Constituents of drinking water from spring sources were within WHO guidelines. Fish caught from the sea and from Lake Furnas contained less than 100 ppb mercury. The urine mercury samples were all within the unexposed range (<3 nmol/mmol creatinine). The mean fluoride levels (and standard deviations) were similar in the two communities: 9.17 (5.76) and 10.52 (5.18) ?mol/mmol creatinine), respectively. The higher fish consumption in Ribeira Quente was reflected in the mean urinary arsenic concentrations: 52.2 (48.9) compared to 19.0 (21.6) nmol/mmol creatinine in the Furnas group. The results for cadmium and thallium were all low.

… The ambient air levels of sulphur dioxide were not likely to be a health problem in the village, but close to fumaroles, the averaged concentrations over 2 weeks could conceal peak excursions in excess of 100 ppb and high enough to induce a response in the most susceptible asthma sufferer; the UK air quality standard for sulphur dioxide is 100 ppb over 15 min (Dept. of the Environment, 1995). That fluoride was a potential hazard at this volcano was shown in Ribeira Quente, where most of the inhabitants with secondary dentition had visible evidence of dental fluorosis due to raised fluoride levels in the drinking water (5 ppm); the source of which had been changed to a different spring on the volcano in only the last few years. The low-level contamination of grasses with fluoride from the fumarolic emissions was insufficient for it to be a toxic hazard to cattle and other grazing animals, and fumarolic emissions of lead and mercury were too low to pose a health risk. These findings were confirmed by the negative urine testing.