Water Air Soil Pollut (2011) 218:371–386 DOI 10.1007/s11270-010-0652-1

Assessment the Health Hazard from 222Rn in Old Metalliferous Mines in San Luis, Argentina A. A. R. da Silva & D. L. Valladares & R. M. Anjos & H. Velasco & M. Rizzotto & E. M. Yoshimura

Received: 14 May 2010 / Accepted: 29 September 2010 / Published online: 30 October 2010 # Springer Science+Business Media B.V. 2010

Abstract Radon levels in two old mines in San Luis, Argentina, are reported and analyzed. The radiation dose and environmental health risk of 222Rn concentrations to both guides and visitors were estimated. CR-39 nuclear track detectors were used for this purpose. The values for the 222Rn concentration at each monitoring site ranged from 0.43±0.04 to 1.48± 0.12 kBq m−3 in the Los Cóndores wolfram mine and from 1.8±0.1 to 6.0±0.5 kBq·m−3 in the La Carolina gold mine, indicating that, in this mine, the radon levels exceed up to four times the action level of A. A. R. da Silva : E. M. Yoshimura Instituto de Física, Universidade de São Paulo, P.O.Box 66318, 05314-970 São Paulo, SP, Brazil D. L. Valladares : H. Velasco : M. Rizzotto GEA, Instituto de Matemática Aplicada San Luis (IMASL), Consejo Nacional de Investigaciones Científicas y Técnicas, Universidad Nacional de San Luis, Ej. de los Andes 950, D5700HHW San Luis, Argentina R. M. Anjos (*) Instituto de Física, Universidade Federal Fluminense, Av. Gal Milton Tavares de Souza, s/n°, Gragoatá, 24210-340 Niterói, RJ, Brazil e-mail: [email protected] A. A. R. da Silva Serviço Especializado em Engenharia de Segurança e Medicina do Trabalho, Departamento de Saúde, Universidade de São Paulo, Rua da Reitoria, 109, 05508-900 São Paulo, SP, Brazil

1.5 kBq m−3 recommended by the International Commission on Radiological Protection. The patterns of the radon transport process revealed that the La Carolina gold mine can be interpreted as a gas confined into a single tube with constant crosssection and air velocity. Patterns of radon activity, taking into account the chimney-effect winds, were used to detect tributary currents of air from shafts or larger fissures along the main adit of the Los Cóndores mine, showing that radon can be used as an important tracer of tributary air currents stream out from fissures and smaller voids in the rock of the mine. Keywords Radon . Natural radionuclides . Geology . Gold and tungsten mines . Health . Risk . Hazard

1 Introduction Underground environments are traps for radon. This colorless, odorless, tasteless, chemically inert, radioactive, and mobile noble gas results from the decay of uranium and thorium series, and is produced continuously in bedrocks and sediments. The natural radon isotopes, 222Rn, 220Rn, and 219Rn, are part of the 238 U, 232Th, and 235U decay series, respectively. In the environmental context, the isotopes 220Rn and 219 Rn are of negligible importance, because of their lower abundance and shorter half-lives, compared with 222Rn with a half-life of 3.82 days. Therefore, in

372

this work the term radon refers always to 222Rn. The radon decay products (progeny) are also radioactive, and two of them, 214Po and 218Po, are also α emitters making a substantial contribution to the total radiation dose received by the population from the environmental radioactivity (Gillmore et al. 2001). When 226Ra disintegrates, a 222Rn atom and an αparticle are formed. The α-particle is ejected and the radon atom recoils and can be displaced from the mineral lattice or molecule where 226Ra disintegrated. The distance which the 222Rn atom can move ranges between 0.02 and 0.07 μm in a mineral grain, according to its density. Through this movement, the atom might leave the mineral grain and enter the porous between the grains. This sort of escaping of 222 Rn from the mineral grain is possible only if the atom is very close to the surface and the recoil occurs in an outward direction (Cigna 2005). The high mobility compared to other members of the radioactive decay series is a characteristic feature of 222Rn. It is able, in a short time, to escape from the mineral in which it is born into the pore spaces. Besides the direct ejection by recoil from α-emission, the diffusion through damaged channels after chemical dissolution in the pore water is also an important mechanism for radon mobility. From the pore space radon atoms migrate along microcracks, fractures, and tunnel volumes either by diffusion or forced flow (Hakl et al. 1992, 1997). Variations in the flux of radon are caused only by physical factors, since it is not a reactive species. Hence, the characterization of the radon flux in an environment could give valuable information on dynamical transport processes (Viñas et al. 2007). In addition, 238U can be oxidized and mobilized by groundwater flow. Once reducing conditions are encountered, the uranium is readily precipitated from solution. This leaching fixation process leads to the enrichment of uranium (and therefore radon) in adjacent deposits. This secondary transport and enrichment process is important in underground environments, as fractures in rocks can increase the surface area interacting with water. Radon tends to disperse in the external environment, but its levels can build up in areas with low air movement, particularly in enclosed spaces, such as mines, caves, and some built environments (Hakl et al. 1992, 1997). Understanding the properties of 222Rn and its decay products in the underground environment is also of great importance due to the effects of radon on

Water Air Soil Pollut (2011) 218:371–386

human health, especially regarding people who work in this ambient as miners in active mines and tourist guides in old mines explored as tourist attractions. The human exposure to high concentrations of radon and progeny can result in an increased risk of developing lung cancer. The majority of the radiation dose to the respiratory tract comes from the progeny rather than from radon itself (Gillmore et al. 2001). Since the days of the Spanish and Portuguese conquerors, South America has been closely associated with the metalliferous ore mining. In the earliest days of mining, gold and silver were the prospected metals. Over the last century, significant new metalliferous mineral resources have been explored such as tin, lead, tungsten, nickel, copper, and palladium ores. In addition, there has also been the development and promotion of other economical activities relating to mining, such as underground mine tourism. Published works are scarce on radon levels in South American mines. In this study, we propose to investigate the radon transport process and its health hazard in two exhausted and abandoned mines in San Luis Province, Argentina. These mines were chosen because they have different physical configurations in their cavities, features which can affect the airflow patterns and radon concentrations. La Carolina gold mine is currently a blind end system, corresponding to a horizontal excavation into the side of a mountain, with only a main adit. Los Cóndores wolfram mine is also a horizontal excavation into the side of a mountain, but has a vertical output (a shaft) at the end of the main gallery.

2 Material and Methods 2.1 Site Description and Its Geological Features The tungsten and gold mining regions of Argentina are confined to the Pampa Range, which runs almost north and south between latitudes 25 and 35 south, and is about 250 miles east of the Andes chain (Miller and Singewald 1919). Geologically, this range is composed of granitic gneiss and schists striking north and south with a westerly dip. These old rocks were invaded at some later time by granites accompanied by tourmaline-bearing pegmatites and veins and lenses of quartz, carrying ores of tungsten and other metals, and a later series of barren quartz veins (Rastall and Wilcockson 1920). Figure 1 shows a

Water Air Soil Pollut (2011) 218:371–386

geological illustration of the study area. The country rock of the gold veins consists of the metamorphic rocks, either pre-Cambrian or early Paleozoic, and many of the veins lie parallel to the schistosity. Young eruptive rocks are present in some districts, in others they are lacking and the mineralization appears to be related to deep-seated rocks of probable Paleozoic age, as seems to be the case in the Pampa Ranges of San Luis and Córdoba. Auriferous quartz is the principal filling of the veins with which is associated more or less auriferous pyrite. Sulfides of copper, lead, and zinc are present in smaller amounts. The wall rock is often impregnated with auriferous pyrite. Most of the mining has been confined to the enriched oxidation zone in which the gold occurs chiefly in the

373

native state. On reaching the leaner primary ores, operations have usually been suspended. On the whole, the veins seem to yield rather low-grade ores (Miller and Singewald 1919). An important lode district is the La Carolina of San Luis Province, on the west slope of the Cerro Tomolasta. The deposits consisted of a principal vein with a north–south strike and dip of 50° to 80° east and several parallel veins in the footwall, enclosed in black and gray slates extensively impregnated with pyrite. The width of the mineralized zone was 125 to 150 m. The ores were auriferous pyrite irregularly distributed in zones in quartz (Miller and Singewald 1919). The La Carolina gold mine (32°48′0″ S, 66°60′0″ W) belonged to West Argentine Gold Company, a British

Fig. 1 Geological map of the study area (modified from López de Luchi et al. 2003)

374

company that, in 1882, drove a 380 m adit from a point on the western side of the Carolina Mountain to the eastern direction, intersecting the four principal veins. In total, about 500 m of tunnels were excavated and several shafts were sunk. They continued in operation until 1894, when a catastrophic collapse of the main adit killed 30 miners, resulting in the closure of the mine (Hoskold 1904). After its abandonment, several secondary tunnels have been buried. Currently, this mine shows a horizontal excavation into the side of a mountain, with a main adit (Fig. 2). Tungsten deposits are situated partly in the province of Catamarca in the north, and in the provinces of Cordoba and San Luis in the south. In this region, the most important mines are located, distributed in the ranges of Sierra de Cordoba and the Sierra de San Luis. The ore was chiefly wolfram rich, with lower contents of scheelite, hubnerite, and tungstic ochre accompanied by pyrites, copper ores, galena, bismuthinite, a little molybdenite, occasionally cassiterite, and, in some places, notable amounts of niobium and tantalum minerals. The gangue is mainly quartz, sometimes with white mica and occasionally topaz and fluorspar (Rastall and Wilcockson 1920; Werner et al. 1998). The Los Fig. 2 Topographical illustration of the La Carolina mine and position of detectors. The numbers indicate the sites where the detectors were placed

Water Air Soil Pollut (2011) 218:371–386

Cóndores was the major tungsten quarry in the country and the second one in South America, being exploited by German and American miners from 1898 to 1965. The lode was brecciated and composed of anastomosing veins. This mine is situated in the Sierra de San Luis, near a small town called Concarán in San Luis Province (32°33′25″ S, 65°15′20″ W). The lode cropped out on the side of a hill and was worked by adits. At the level of the main adit, it was displaced by a series of step faults, which necessitated frequent cross-cutting. The ore was, then, exhausted down to this level, and the mine has been further developed by a shaft sunk to a depth of 90 m. After its abandonment, several secondary adits have been buried. Currently, what remains of the mine appears as a 450 m main adit with a shaft at the end of this adit. Figure 3 shows the topographical illustration of the Los Cóndores mine. 2.2 Experimental Procedure Three different experimental methodologies were used in this work. The distribution of natural radionuclide activities (40K, 232Th, and 238U) was determined from rock samples collected along the

Water Air Soil Pollut (2011) 218:371–386

375

Fig. 3 Topographical illustration of the Los Cóndores mine and position of detectors. The numbers indicate the sites where the detectors were placed

main adits of the La Carolina and Los Cóndores mines using laboratory-based gamma ray spectrometry. The external gamma dose rate due to natural radionuclides present in the mines was evaluated using a portable survey meter. Radon concentration measurements were performed by CR-39 nuclear track detectors during the summer season. The radon monitors were exposed for 105 days in the La Carolina mine and 42 days in the Los Cóndores mine The exposure time in Los Cóndores, which was monitored after La Carolina, was reduced to about 40 days in order to facilitate the track detector analysis, since the CR-39 detectors exposed in La Carolina mine showed very high track densities. The CR-39 plastic track detectors (1.7 cm2 area and 0.9 mm thick) were enclosed in a radon diffusion chamber (NRPB/SSI type monitors) (Orlando et al. 2002). They were deployed at 14 locations within the adits of the La Carolina gold mine and 20 locations of

the Los Cóndores wolfram mine, as illustrated in Figs. 2 and 3, respectively. The detectors were placed at a distance of 20 cm from the wall and 1 m above the ground. The determination of radon concentrations was performed at the Dosimetry Laboratory of Nuclear Physics Department of São Paulo University (Da Silva and Yoshimura 2005). For this determination, two main assumptions are used: radon in the environmental air rapidly reaches equilibrium with that inside the monitor; both radon progeny and shorter half-life radon isotopes are prevented from entering the diffusion chamber (Shweikani and Durrani 1995). The mean radon concentration during the whole period of exposure is directly proportional to the track density in the plastic detector and the time of exposure. The conversion factor of track density to radon concentration was experimentally determined as 2.8±0.2 (track cm−2)/(kBq m−3 h) and its value is known to be very constant for this combination of

376

monitor and plastic detector (Orlando et al. 2002; Howarth and Miles 2003). After a chemical etching (in a 30% KOH solution at 80°C for 330 min) CR-39 detector track densities were evaluated with a semiautomated optical microscopy analysis system described in a previous work (Da Silva and Yoshimura 2005). The uncertainties of the concentration activities are derived from the counting statistical error. A portable survey meter, a Geiger–Mueller (GM) radiation monitor was used for in situ measurements. The GM detector was used as an area monitor and it was calibrated to within ±10% full scale accuracy up to 100 mSv h−1 ambient dose equivalent rates. Approximately 200 measurements, uniformly distributed over the adits of the La Carolina mine, were performed. During the measurements, the equipment was kept at a distance of 1 m above the ground. The same procedure was adopted in the Los Cóndores mine. Twenty rock samples from the adit walls of the two mines were collected. The sampling sites selection was made on the basis of accessibility and in order to cover equally well the adits. The sample analysis allowed evaluating the mean behavior of the natural radionuclides distribution for each mine. In the laboratory, all rocks samples were ground to powder, dried, and packed in cylindrical plastic containers, weighed and sealed. About 150 g of sample material was used for the measurements of each sample. Before the measurements, the containers were kept sealed during 4 weeks, in order to reach the equilibrium of the 238U and 232Th series and their respective progeny. It was assumed that 220Rn and 222 Rn could not escape from the sealed containers after closure. Amounts of thorium, uranium, and potassium concentrations were obtained using the conventional technique of gamma ray spectrometry (Anjos et al. 2005, 2006, 2010). Samples were prepared and analyzed at the Laboratory of Radioecology of the Physics Institute of the Federal Fluminense University. A 55% efficiency high-purity Germanium detector was used, in order to obtain the concentrations of 238 U, 232Th, and 40 K. Energy spectra from each sample were accumulated during 24 h. After each measurement, one water-filled plastic cylindrical container was placed in the detection system during the same counting period, in order to collect the background count rates. The system calibration was performed using International Atomic Energy Agency

Water Air Soil Pollut (2011) 218:371–386

reference materials for K, U, and Th activity determination: RGK-1, RGU-1, and RGTh-1, respectively (IAEA 1987). The energy resolution at the 60Co gamma ray line (1,332.46 keV) was 2.0 keV. The determination of 40K, 238U, and 232Th was based on the measurements of total energy peaks from 40K (1,460.8 keV), 214Bi (1,764.5 keV), and 208Tl (2,614.4 keV), respectively. The standard IAEA samples were measured at the same conditions and geometry. Each radionuclide absolute activity was determined by comparing the peak intensity of the sample to the respective standard material peak intensity. This methodology has the advantage of including effects of coincident summing in the calibration procedure. The peak intensity is defined as the difference between the peak area for a sample spectrum and the background contribution. As 228Ac, a decay product of 232Th, produces 1,459.2 keV γ-rays, which interfere with the 40K emission (1,460.8 keV), the derivation of the 40K activities had a special attention during the spectral analysis. Whenever necessary, the thorium contribution to the potassium line was calculated, using the branching ratios of the γ-decay, the counting time and the detector efficiency. Nevertheless, the sediment samples were composed by low concentrations of heavy minerals such that the 40K activities could be properly corrected for the 232Th admixture (Anjos et al. 2006). With the determined activities and the sample dry weights, the 232Th, 238U, and 40K concentrations could be expressed in activity per unit mass (Bq kg−1). Since the correlations between Th, U, and K are usually given in equivalent soil concentrations (Anjos et al. 2007, 2010), the concentrations of 232Th and 238U were expressed in parts per million and 40K in percent, so that the specific parent activities of samples containing 1 ppm of 232Th or 1 ppm of natural U are 4.08 and 12.3 Bq kg−1, respectively. For natural potassium, a concentration of 1% by weight of sample corresponds to a 40K specific activity of 317 Bq kg−1. Systematic uncertainties come from several sources: the assumption that the samples are in secular equilibrium, the mass and volume of the samples, efficiency calibrations, peak area determination, activity to mass conversion, and branching ratios. Random uncertainties are associated with background and sample counts. All these errors were estimated to be of the order of 5% for the eTh, eU, and K contents. The lower limits of detection were estimated to be

Water Air Soil Pollut (2011) 218:371–386

1.0 ppm (4.1 Bq kg−1) for Th, 0.2 ppm (2.5 Bq kg−1) for U, and 0.02% (6.3 Bq kg−1) for K. 2.3 Natural Radionuclides as Tracer of Geological Process Natural gamma ray spectrometry has been used for a wide range of geological applications including the lithological mapping of rocks and unconsolidated sediments, mineral exploration, sediment transport studies and investigation on discharged materials, and also in radiation protection, as in the risk analysis to workers exposed to radionuclides present in respirable sediments (Anjos et al. 2006, 2007; Estellita et al. 2010). Radioactive minerals often occur in relatively small concentrations in granitic and sedimentary rocks. Even shales typically contain only 30% to 70% radioactive clay minerals (Schlumberger 1986). The basis of the radiometric techniques is the measurement of gamma ray emissions from the primordial radionuclides. For practical proposes, the natural gamma emitter isotopes comprise 40K and most notably 214Bi and 208Ti, decay products of the 238U and 232Th series, respectively. Thorium, uranium, and potassium concentrations in granitic and sedimentary rocks are intimately related to their mineral compositions and general petrologic features (Whitfield et al. 1959; Rogers and Ragland 1961; Anjos et al. 2005, 2006). Uranium and thorium in igneous and metamorphic rocks are usually found in a few accessory minerals such as apatite, sphene and zircon. Other highly radioactive minerals, like monazite, allanite, uraninite, thorite, and pyrochlore, are widespread in nature, but they are very minor constituents of rocks. Uranium tends to be highly mobile near the surface of the rock whereas thorium is relatively stable. Uranium is easily oxidized to a watersoluble form and can be readily leached from pegmatites and granites and re-deposited in sediments at large distances from the source rock. Thorium is much less soluble than uranium and potassium and does not move except by mechanical means such as wind and erosion processes. Both thorium and uranium contents tend to be high in felsitic rocks and to increase with alkalinity or acidity, the highest concentrations being found in pegmatites. The potassium content of rocks also increases with acidity. Potassium is usually found in potash feldspars, such as microcline and orthoclase, or in micas, like muscovite and biotite. Rocks that are free of these minerals have very low K activity.

377

The petrologic features of granitic and sedimentary rocks associated with effects of weathering and metamorphism produce expressive alterations in the relationship between the natural radionuclides (Th, U, K, Th/U, and Th/K). Consequently, the measurements of thorium, uranium, and potassium concentrations of different rock samples result in individual differentiation sequences. Unless there is a complex mixture of radioactive minerals in the formation of the granitic and sedimentary rocks, the natural radionuclides correlations can be used to identify the most common ones. Nowadays, the potassium–thorium cross-plot is widely used for the recognition of heavy and clay mineral associations and the discrimination of micas and feldspars. As both thorium (by adsorption) and potassium (chemical composition) are associated with clay minerals, the ratio Th/K expresses only relative potassium enrichment as an indicator of clay mineral species, and as such might be the diagnostic of other radioactive minerals (Anjos et al. 2005, 2006, 2010). 2.4 Effective Dose Estimates The health risk associated with radon arises from the inhalation of the short-lived decay products (218Po, 214 Pb, 214Bi, and 214Po) and the consequent dose to critical cells of the respiratory tract. The activity of radon decay products is described by a quantity called the potential alpha energy concentration (PAEC). The PAEC of a radionuclide is the sum of alpha energies emitted during the decay of all products of this atom up to a long-lived daughter. The PAEC of any mixture of short-lived radon daughters is the sum of the potential alpha energy of these atoms present per unit of volume of air. Owing to the fact that the direct measurement of concentrations of all radon daughters is difficult, they are estimated from considerations of equilibrium between Rn and its decay products. An equilibrium equivalent radon concentration (symbol EEC, units Bq m−3) is defined as the 222Rn concentration in secular equilibrium having the same PAEC as the mixture of radon daughters. Both concentrations, EEC and the actual indoor concentration of Rn (C0), are related by the equilibrium factor (F): F¼

EEC ¼ 0:105f218Po þ 0:515f214Pb C0 þ 0:380f214Bi þ 6  108 f214Po

ð1Þ

378

Water Air Soil Pollut (2011) 218:371–386

where fi are the fractional concentration of decay progeny defined as the ratio of the individual species concentration to that of radon, and the constants are the fractional contributions of each decay product to the total potential alpha energy due to the decay of a unit activity of the gas. By means of factor F, a measured radon concentration can be converted to an equilibrium equivalent concentration. F is a function of three independent variables: the aerosol attachment ratio of the progeny nuclides to airborne aerosol particles, the deposition rate of radionuclides to indoor surfaces, and the air exchange rate or ventilation rate. These variables are highly dependent on room dimensions, aerosol concentration, and other ambient conditions (Eappen et al. 2006). The value of F decreases with increasing ventilation as the radon progeny is removed by the air flow. Many measurements have been made of radon and its decay progeny in order to estimate the factor F. These results indicate that the typical outdoor equilibrium factor ranges between 0.5 and 0.7 (UNSCEAR 2000). However, mine environments are very different and results in the literature indicate that F may vary spatially and temporally, and has values in the range from 0.04 to 0.95, depending on the environment in the site. According to a series of published results, the average value of F obtained with 880 measurements was calculated to be 0.57 (Hakl et al. 1997; Gillmore et al. 2001; Cigna 2005; Lario et al. 2005). In this work, we have assumed F to be 0.5 as a representative value of all mine sites examined. A better estimation of effective doses will be possible when a local estimation of F is made, in a future study. For occupational exposure to inhaled 222Rn decay products, the ICRP (1993) recommends the use of a single factor conversion to relate the radon concentration to effective dose to an individual. Thus, the effective dose (E) for inhalation of radon decay products can be calculated through: E ¼ F  C0  DCF  t

ð2Þ

where C0 is the actual air radon concentration (Bq m−3), t is the exposure time (h), and DCF (nSv Bq−1 h−1 m3) is the dose conversion factor that takes into account the radon progeny (dose per exposure). Estimations of DCF have been obtained by epidemiological studies and by means of dosimetric

models, as in the International Commission on Radiological Protection (ICRP) Human Respiratory Tract model (ICRP 1994). Dosimetric evaluations of DCF gave values in the range 6–15 nSv Bq−1 h−1 m3, and the epidemiological approach provides a value of 6 nSv Bq−1 h−1 m3 (UNSCEAR 2000). UNSCEAR recommends that a DCF value of 9 nSv Bq−1 h−1 m3 continues to be used in dose evaluations (UNSCEAR 2000, 2006), being closer to the International Commission on Radiological Protection recommended value of 7.95 nSv Bq−1 h−1 m3. Both values might be used for workers and public in general (ICRP 1993). The ICRP expects to publish new dose coefficients that will replace the ICRP Publication 65 dose conversion convention and recommends that the current dose conversion of 7.95 nSv Bq−1 h−1 m3 may continue to be used until new dose coefficients are available. ICRP advises that the change is likely to result in an increase in DCF of around a factor of two (ICRP 2009). For this assessment of effective dose to public members and workers, we used the dose conversion factor from the ICRP Publication 65, namely, 7.95 nSv Bq−1 h−1 m3 (ICRP 1993).

3 Results and Discussion Experimental histogram plots of relative frequency of the outdoor dose rates for natural gamma emitters from the main adits of the La Carolina and Los Cóndores mines are shown in Fig. 4. Each histogram has been fitted to a Gaussian distribution, for which the mean values of external gamma dose rate in air (Xc ± ΔXc) were obtained from the distribution maxima and from the full width of the peak at half of maximum frequency—FWHM (Anjos et al. 2008). According to this figure, the natural radiation dose rate measured inside both mines shows values similar to average terrestrial outdoor gamma dose rate levels, since the main adit of the La Carolina mine has Xc =0.11±0.06 μSv h−1 and Los Cóndores mine has Xc =0.13±0.03 μSv h−1. The dose rates due to external exposure of gamma rays depend on the concentrations of the radionuclides present in the walls of the mines. The greatest part of the gamma radiation comes from terrestrial radionuclides. As the natural radionuclide concentrations from rocks samples from the La Carolina and Los Cóndores adits were also deter-

Water Air Soil Pollut (2011) 218:371–386

Fig. 4 Experimental histogram plots of the relative intensity of the gamma ambient dose equivalent rates in the main adits of the a La Carolina gold mine and b Los Cóndores wolfram mine. The solid curves represent Gaussian fits

mined, the dose rate results can be compared to the activity concentration of the 40 K, 232Th, and 238U. Table 1 shows the activity concentrations, suggesting that the radioactive minerals present in the rock samples show small activity concentrations of 40 K, 232 Th, and 238U, consistent with the values of the outdoor dose rates for natural gamma emitters observed. This fact can also be corroborated by analysis of a binary diagram of thorium and potassium contents, illustrated in Fig. 5, which can be used to provide qualitative mineralogical information about the rocks that make up the tunnels of the two mines. Th (scaled in parts per million) and K (in percent) have a direct relation, since the data can be classified in zones or clusters with distinct values of the Th/K

379

ratio, each of which representing different mineral associations. Using the mineral identification chart proposed by Schlumberger (1986), also plotted in Fig. 5, our results suggest that the rocks that compose the two mines have similar mineral formations, having small activity concentrations of natural radionuclides, consistent with the dose rate distributions observed in Fig. 4. Radon can emanate from underground at a rate which is influenced by a wide number of factors, such as the lithology of the bedrock, structural features such as faulting and jointing, soil physical and chemical properties, mode of transport (diffusion or fluid flow), and atmospheric influences such as precipitations, wind, and temperature (Sharma 1997). Radon can be transported by molecular diffusion and by advective air flow in the underground environment. In general, molecular diffusion is the dominant mechanism in the intergranular channels, capillaries, and small pores. In the larger pores and fractures, advective air flow may become important or even dominant. The heterogeneity of the geological material is a source of large variation of the radon diffusion in a porous medium, with respect to the theoretical assumptions. Mica and vermiculite, which are flaky minerals, have a shape factor that causes the diffusion coefficient to be one-half to one-third of the theoretical value. Clay and shales contain significant proportions of flaky minerals, usually oriented so as to impede vertical movement. They retard diffusion to a greater extent than a porous medium having the same porosity but consisting of spherical particles (Cigna 2005). The advective transport process, resulting from temperature differences between the underground and the surface, is of greater importance than molecular diffusion, it gives 222Rn the chance to travel over large distances in the subsurface before its decay (Sharma 1997). In horizontal caves, the daily mean radon concentration generally has a seasonal variation with a maximum in summer and a minimum in winter reflecting the main direction of underground air flows (Hakl et al. 1992, 1997). Taking into account the fact that the measurements from the La Carolina and Los Cóndores mines were performed during the summer season, the temperature differences between the mines and outside were similar, the walls that compose the two mines have similar rock formations, then, it could be expected that the radon concentration distributions were also

380

Water Air Soil Pollut (2011) 218:371–386

Table 1 Mean values of the 40K, 232Th, and 238U activity concentrations in rock samples from the main adits of the La Carolina and Los Cóndores mines La Carolina gold mine Sample position

Los Cóndores wolfram mine

40

K (Bq kg−1)

232

Th (Bq kg−1)

238

U (Bq kg−1)

Sample position

40

K (Bq kg−1)

232

Th (Bq kg−1)

238

U (Bq kg−1)

S1

567±32

40±4

31±4

C1

1,109±66

43±3

38±2

S2

558±28

28±3

19±3

C3

550±33

46±4

31±2

S3

720±40

46±4

34±4

C5

650±39

25±2

20±1

S4

1,020±50

60±6

57±6

C7

930±56

45±3

64±4

S9

320±30

60±6

45±6

C9

553±33

31±2

32±2

S10

215±20

19±5

12±2

C12

1,295±78

49±3

28±2

S12

1,550±100

85±9

73±7

C14

819±49

42±3

30±2

S14

250±20

40±6

35±5

C18

800±48

37±2

26±2

Uncertainty values represent the standard deviations of the mean

similar in the two mines. However, significant differences in the concentration of radon between the two mines were found, as illustrated in Table 2. Hence, a more detailed modeling of the air exchange processes in the mine adits must be taken into account to understand the spatial variations of the radon concen-

Fig. 5 Correlations between thorium and potassium contents from rock samples of La Carolina and Los Cóndores mines. The gray lines were obtained from the mineral identification chart proposed by Schlumberger (1986)

tration. The radon concentration varied significantly throughout the La Carolina gold mine. The lowest values can be observed near the mine entrance. Moving towards its interior, the concentration increases rapidly first, then more slowly, reaching a plateau at the end of the main adit (around the

Water Air Soil Pollut (2011) 218:371–386 Table 2 Radon concentration in La Carolina and Los Cóndores mines

381

La Carolina gold mine Detector position

Los Cóndores wolfram mine Rn (kBq m−3)

222

Detector position

Rn (kBq m−3)

222

S1

1.8±0.1

C1

0.44±0.04

S2

3.8±0.3

C2

0.47±0.04

S3

4.6±0.4

C3

0.44±0.04

S4

4.5±0.4

C4

0.46±0.04

S5

4.5±0.4

C5

0.50±0.04

S6

5.0±0.4

C6

0.48±0.04

S7

5.1±0.4

C7

0.69±0.06

S8

5.2±0.4

C8

0.65±0.05

S9

5.6±0.4

C9

0.53±0.04

S10

5.3±0.4

C10

0.65±0.05

S11

5.0±0.4

C11

0.58±0.05

S12

5.6±0.4

C12

0.62±0.05

S13

5.8±0.5

C13

0.48±0.04

S14

6.0±0.5

C14

0.43±0.04

C15

0.47±0.04

C16

0.48±0.04

C17

0.77±0.06

C18

0.68±0.06

C19

1.43±0.12

C20

1.48±0.12

detector S9 in Fig. 2). This value remains practically unchanged in the secondary adits (S10 to S14). This behavior allows us to interpret the main adit of La Carolina mine as a blind end system. On the other hand, at the Los Cóndores wolfram mine the radon concentration varies evenly along the main adit (C1 to C12 in Fig. 3) and the shaft area (C13 to 16). The highest values are observed in the secondary adits C17–C18 and C19–C20. The radon transport processes depend essentially on the configuration and the connection of the underground cavities, passages, shafts, and other communications to the exterior, such as fissures and fractures. In the case of blind end systems, the difference of external and internal air density is the main control parameter, which are superimposed by convective air exchange due to temperature differences between the mine and the outside. For two or more entrance systems, which have openings to the exterior situated at two different levels at least, chimney-effect winds may dominantly

govern the radon transport (Wigley 1967; Hakl et al. 1996; Hakl 1997). In this case, the pressure exerted at the entrance by the column of air inside the mine will differ from the pressure of external air because the density of the air depends upon temperature. The pressure difference, in first approximation, is proportional to the temperature difference between the mine and outside (Atkinson et al. 1983; Hakl et al. 1996; Hakl 1997). According to this relationship, when the mine air is colder than the outside air, the pressure exerted by the mine air will be greater than the pressure outside and the air will blow out through the lower entrance. When the outside conditions are colder than inside, the air flows into the mine through the lower entrance. The air temperatures in mines and caves are usually nearly constant, while external temperatures vary. Then, the chimney-effect winds reverse direction seasonally and also might reverse daily at some times of the year (Atkinson et al. 1983; Hakl et al. 1996; Hakl 1997).

382

Water Air Soil Pollut (2011) 218:371–386

the release of radon into the mine environment. Under steady conditions and assuming the contour condition [C(0,t)=0], the radon concentration activity at any distance, x, from the entrance is given by:   CðxÞ ¼ C1 1  elx=v ð4Þ

Fig. 6 Radon concentration versus distance from entrance of the La Carolina gold mine. The dotted curve represents a fit using Eq. 4. The best-fit was obtained by the least squares minimization (R2 =0.97). The labels indicate the detector positions, according to Figs. 2 and 3

The pressure difference, originated by the temperature difference between the outside air and the air inside the mine, has an important effect on the radon concentration profile due to the advection transport mechanism. However, variations in atmospheric pressure have no an important effect on the advection process. Figure 6 shows the radon levels from the La Carolina gold mine, which increase with the distance from the mine entrance. In the foregoing description, a model taking into account the advective air flow, molecular diffusion, and the radon production and decay is used, assuming that La Carolina mine can be treated as if it was a simple tube with constant crosssection and air velocity. Underground radon transport can subsequently be described by the following diffusion–convection equation (Wigley 1967; Atkinson et al. 1983; Nazaroff 1992): @C ðx; t Þ @ 2 C ðx; tÞ  Deff þ vC ðx; t Þ ¼ l C ðx; t Þ þ f @t @x2 ð3Þ

where C [Bq m−3] is the radon concentration, Deff [m2 h−1] is the effective diffusion coefficient of radon, v [m h−1] is the air velocity, l=7.54×10−3 h−1 is the decay constant of radon, and f [Bq m−3 h−1] is the source term. The second term on the left hand side corresponds to normal diffusion while the third describes convection or advection. On the right hand side, the terms describe the radon decay process and

where C1 ¼ f=l. The Eq. 4 was used to fit the data of radon concentration (C) as function of the distance (x) from the entrance, and is plotted as solid line in Fig. 6. There is a close agreement between the model curve and the data (R2 =0.97). The curve shows that the radon concentration increases to a plateau, at which the rate of addition of radon is just balanced by its decay. The fit parameter values are C∞ =5.7±0.1 kBq m−3, v=0.75± 0.09 m h−1, and f=43±2 Bq m−3 h−1. The arrangement of the main adit of Los Cóndores mine is a clear example of underground environments which have openings to the exterior situated at two different levels at least, where the airflow would be controlled by the chimney-effect winds. As the mine air is colder than outside air, the pressure exerted by the mine air is greater than the pressure outside and the wind blows out of the lower entrance (level 0 in Fig. 3). This way, the radon levels at Los Cóndores mine were studied taking into account the airflow direction. The variation of radon concentrations with the distance from the Los Cóndores mine entrance (shaft) is shown in Fig. 7, which reveals that the radon concentration varies irregularly, unlike the pattern observed in Fig. 6. The presence of a shaft in the Los Cóndores mine might explain the overall lower radon concentrations as compared to La Carolina mine. However, only in the first 50 m from shaft (C10 to C12) does the expected behavior occur, and even there it is punctuated by low values around the detectors C9 and C6 (see Fig. 3). According to Atkinson et al. (1983), the irregular decreases of radon activity could be interpreted as being due to effects of several tributaries of air which have lower radon content than the main adit air, in the manner indicated by the dotted line in Fig. 7a. Therefore, the application of the model described by Eq. 4 to the Los Cóndores mine would be an oversimplification; its pattern of radon concentration is better explained if secondary airflow tributaries are taken into account. This effect is shown in Fig. 7b.

Water Air Soil Pollut (2011) 218:371–386

Fig. 7 a Theoretical profiles of radon activity in a long tubular mine, modified from Atkinson et al. (1983), where plot 1 represents a constant flow along the tube, plot 2 the effect of a tributary with lower radon concentration at X1, and plot 3 the effects of several tributaries; b radon concentration versus distance from shaft of the Los Cóndores tungsten mine. The dotted curves represent the radon profile with a sharp decrease due possibly to the presence of secondary tributaries with low content of radon. The labels indicate the detector positions, according to Figs. 2 and 3

By the time the mine was operating, several secondary tunnels were opened from the main adit, and they were buried after the abandonment. Although the data suggest the entry points of low-radon tributaries at C9 and C6, the plot in Fig. 7b indicates that the low value measured by detector C9 could be interpreted as due to the monitor being situated in the path of much smaller stream airflow before it has been fully mixed with the main flow. The sharp decreases of radon concentration at the central part of the main adit (C6) could be interpreted as a radon dilution due to the entrance of air with low radon content from secondary tributaries or larger fissures,

383

not taken explicitly into account in our description of the mine. The dashed line curve shown in Fig. 7b would represent, with this assumption, the actual trend of radon concentration in the main adit. In both cases, the radon concentration values provide new information on the structure of the mine, acting then as a tracer. With regard to the magnitude of possible environmental health risk by reporting the high levels of radon that have been found in these two abandoned mines, the effective dose to workers was also estimated. Values of activity concentration of radon from La Carolina mine ranged from 1.8±0.1 to 6.0± 0.5 kBq m−3. The mean radon concentration found (4.8 ± 0.8 kBq m −3) suggests the existence of significant health implications for both casual and occupational mine explorations (visitors and tourist guides, respectively) since this value exceeds by about three times the upper limit of the action level of 1.5 kBq m −3 recommended by the International Commission on Radiological Protection for workplaces (ICRP 2008). It is worth noting that the same publication suggests a lower action level (0.60 kBq m−3) for homes, but the exposure in mine ambient is more suitably described as a work than domestic exposure, in our opinion. By the other hand, radon data from Los Cóndores mine ranged from 0.43±0.04 to 1.48±0.12 kBq m−3, with a mean radon concentration of 0.6±0.2 kBq m−3. These results reinforce the importance of the introduction of forced ventilation (exhausting systems) in underground working environments, if the natural ventilation is not enough. The mines in Argentina are visited by a significant number of mining enthusiasts. Although the mean value of radon activity concentration obtained from La Carolina mine suggests that the risk posed by elevated radon levels to casual mine explorers may be significant, it is necessary to estimate the amount of time spent in the mines by various groups before the risk can be quantified. According to Gillmore et al. (2001), casual explorers may be expected to spend 2 h underground at a time in many mines in England, perhaps on ten occasions per year, totaling annual exposure duration of 20 h. For dedicated/occupational mine explorers, however, the annual duration is much longer. Similar behavior occurs in the Argentine tourist mines.

384

Water Air Soil Pollut (2011) 218:371–386

Table 3 Radon and gamma contributions to the effective dose of mine explorers (two classifications) and of workers with two possibilities of occupational regimes from La Carolina and Los Cóndores mines La Carolina mine Classification (number of trips)

Accumulated annual time underground (h)

Los Cóndores mine

Gamma dose (mSv)

Radon dose (mSv)

Gamma dose (mSv)

Radon dose (mSv)

Casual (10)

20

<0.01

0.38±0.07

<0.01

0.05±0.02

Sporting (30)

60

<0.01

1.1±0.2

<0.01

0.15±0.05

600

0.07±0.03

11±2

0.08±0.02

1.45±0.5

1,560a

0.16±0.07

30±6

0.20±0.03

4±1

Occupational (300) Occupational (fulltime) a

Assuming that a person works 6.5 h per day inside the mine during 240 days per year

From Eq. 2, it is possible to make an estimation of the effective dose due to inhalation of radon by visitors and tourist guides. Similarly, it is possible to estimate the contribution of the external gamma radiation to the effective dose. Table 3 shows the results for three classifications of mine explorers, showing that the external gamma doses are always much smaller than the doses due to inhalation of radon. Moreover, as, according to UNSCEAR (2000), the contribution of the external gamma dose to the average worldwide annual effective dose ranges typically from 0.3 to 0.6 mSv, the gamma radiation is not an issue even for a fulltime worker inside the mines. This is not the same concerning the radon contribution for the effective dose: the UNSCEAR publication reports the annual average contribution of 222 Rn and 220 Rn as 1.15 and 0.10 mSv, respectively, and Table 3 gives predictions of 11 mSv year−1 or 30 mSv year−1 for the tourist guides, depending on the daily shift in La Carolina mine. This is a significant result in relation to the occupational dose limit of 20 mSv year−1 suggested by ICRP (2008). As the exposure of the guides is not yet considered as occupational, they do not have a radiation monitoring program. Additional investigation, taking into account the use of personal monitors during work time would clarify better this issue.

4 Conclusions The results obtained from this study suggest that workers of La Carolina gold mine can be exposed to high concentrations of radon. The values obtained at

each monitoring site ranged from 1.8±0.1 to 6.0± 0.5 kBq m−3, with a mean value of 4.8±0.8 kBq m−3, indicating that these measurements exceed by about three times the upper limit of the action level of 1.5 kBq m−3 recommended by the ICRP for workplaces. Remedial actions should be taken in this workplace in order to reduce the health risks to tourist guides. The results confirm that the same problem can be observed for miners that work in underground environments that are poorly ventilated. Typical mitigating actions such as the introduction of forced ventilation or reduction of daily employee’s shift could be adopted. However, before adopting remedial actions, we suggest a study of the daily variation of radon concentration and the typical equilibrium factor in the mine tunnel in order to evaluate precisely the impact of radon. A test program of individual monitoring of the respirable radon and progeny, and the training of the employees in radiation protection could also be performed. Taking into account that the radon transport process into La Carolina gold mine can be interpreted as a gas confined into a tunnel with only one entrance and that there are no sharp variations in the radon concentration as a function of the distance to the entrance, we infer that this mine has no large cracks or fissures that could cause radon dilution. On the other hand, the irregular decreases of radon activity from Los Cóndores wolfram mine suggest the presence of secondary tributaries or larger fissures along its main adit. These results have demonstrated the usefulness of the radon as a suitable tool for geological studies, such as in the understanding of the mine regimes or mapping of active faults.

Water Air Soil Pollut (2011) 218:371–386 Acknowledgments The authors would like to thank the Brazilian (CNPq, PROSUL/CNPq, and FAPERJ) and Argentine (CONICET and UNSL) funding agencies for their financial support. We are grateful to workers of the La Carolina and Los Cóndores mines who assisted us with the radon measurements.

References Anjos, R. M., Veiga, R., Soares, T., Santos, A. M. A., Aguiar, J. G., Frascá, M. H. B. O., et al. (2005). Natural radionuclide distribution in Brazilian commercial granites. Radiation Measurements, 39, 245–253. Anjos, R. M., Veiga, R., Macario, K., Carvalho, C., Sanches, N., Bastos, J., et al. (2006). Radiometric analysis of quaternary deposits from the southeastern Brazilian coast. Marine Geology, 229, 29–43. Anjos, R. M., Veiga, R., Carvalho, C., Macario, K., & Gomes, P. R. S. (2007). Geological provenance of quaternary deposits from the southeastern Brazilian coast. Nuclear Physics A, 787, 642c–647c. Anjos, R. M., Veiga, R., Carvalho, C., Sanches, N., Estellita, L., Zanuto, P., et al. (2008). Natural sources of radiation exposure and the teaching of radioecology. Physics Education, 43, 423–428. Anjos, R. M., Macario, K. D., Lima, T. A., Veiga, R., Carvalho, C., Fernandes, P. J. F., et al. (2010). Correlations between radiometric analysis of quaternary deposits and the chronology of prehistoric settlements from the southeastern Brazilian coast. Journal of Environmental Radioactivity, 101, 75–81. Atkinson, T. C., Smart, P. L., & Wigley, T. M. L. (1983). Climate and natural radon levels in Castleguard Cave, Columbia Icefields, Alberta, Canada. Arctic and Alpine, 15, 487–502. Cigna, A. A. (2005). Radon in caves. International Journal of Speleology, 34, 1–18. Da Silva, A. A. R., & Yoshimura, E. M. (2005). Track analysis system for application in alpha particle detection with plastic detectors. Radiation Measurements, 39, 621–625. Eappen, K. P., Mayya, Y. S., Patnaik, R. L., & Kushwaha, H. S. (2006). Estimation of radon progeny equilibrium factors and their uncertainty bounds using solid state nuclear track detectors. Radiation Measurements, 41, 342–348. Estellita, L., Santos, A. M. A., Anjos, R. M., Yoshimura, E. M., Velasco, H., da Silva, A. A. R., et al. (2010). Analysis and risk estimates to workers of Brazilian granitic industries and sandblasters exposed to respirable crystalline silica and natural radionuclides. Radiation Measurements, 45, 196–203. Gillmore, G. K., Phillips, P., Denman, A., Sperrin, M., & Pearce, G. (2001). Radon levels in abandoned metalliferous mines, Devon, Southwest England. Ecotoxicology and Environmental Safety, 49, 281–292. Hakl, J. (1997). Application of radon-222, as a natural tracer in environmental studies. Ph.D. thesis. Lajos Kossuth University. http://ganymedes.lib.unideb.hu:8080/dea/bit stream/2437/80292/2/ertekezes.pdf Hakl, J., Hunyadi, I., Csige, I., Géczy, G., Lénárt, L., & Törőcsik, I. (1992). Outline of natural radon occurences on

385 karstic terrains of Hungary. Radiation Protection Dosimetry, 45, 183–186. Hakl, J., Csige, I., Hunyadi, I., Várhegyi, A., & Géczy, G. (1996). Radon transport in fractured porous media— Experimental study in caves. Environment International, 22, S433–S437. Hakl, J., Hunyadi, I., Csige, I., Géczy, G., Lénárt, L., & Várhegyi, A. (1997). Radon transport phenomena studied in karst caves—International experiences on radon levels and exposures. Radiation Measurements, 28, 675–684. Hoskold, H. D. (1904). Official Report upon the Mines, Mining, Metallurgy and Mining Laws of the Argentine Republic. Buenos Aires: Ministry of Agriculture, Commerce and Industries. Howarth, C. B., & Miles, J. C. H. (2003). NRPB Report W-44: Results of the 2002 NRPB Intercomparison of Passive Radon Detectors. Chilton: NRPB, National Radiological Protection Board. IAEA, (1987). Preparation and certification of IAEA gamma spectrometry reference materials, RGU-1, RGTh-1 and RGK-1. International Atomic Energy Agency. ReportIAEA/RL/148. ICRP. (1993). International Commission on Radiological Protection, Protection Against Radon-222 at Home and at Work. ICRP Publication 65. Oxford: Pergamon. ICRP. (1994). International Commission on Radiological Protection. Human respiratory tract model for radiological protection. ICRP Publication 66. Oxford: Pergamon. ICRP, (2008). The 2007 recommendations of the international commission on radiological protection. In: ICRP Publication 103, vol. 37. Annals of the ICRP. 2–4. ICRP, (2009). International Commission on Radiological Protection. International commission on radiological protection statement on radon. ICRP Ref 00/902/09. http:// www.icrp.org/icrp_radon.asp. Lario, J., Sánchez-Moral, S., Canaveras, J. C., Cuezva, S., & Soler, V. (2005). Radon continuous monitoring in Altamira Cave (northern Spain) to assess user’s annual effective dose. Journal of Environmental Radioactivity, 80, 161–174. López de Luchi, M. G., Cerredo, M. E., Siegesmund, S., Steenken, A., & Wemmer, K. (2003). Provenance and tectonic setting of the protoliths of the Metamorphic Complexes of Sierra de San Luis. Revista de la Asociación Geológica Argentina, 58, 525–540. Miller, B. L., & Singewald, J. T. (1919). The Mineral Deposits of South America (p. 598). New York: McGraw-Hill Book Co. Nazaroff, W. W. (1992). Radon transport from soil to air. Reviews of Geophysics, 30, 137–160. Orlando, C., Orlando, P., Patrizii, L., Tommasino, L., Tonnarini, S., Trevisi, R., et al. (2002). A passive radon dosemeter suitable for workplaces. Radiation Protection Dosimetry, 102, 163–168. Rastall, R. H., & Wilcockson, M. A. (1920). Tungsten ores: London, Imperial Institute Monograph on Mineral Resources with Special Reference to the British Empire, John Murray, 81 p. Rogers, J. J. W., & Ragland, P. C. (1961). Variation of thorium and uranium in selected granitic rocks. Geochimica et Cosmochimica Acta, 25, 99–109.

386 Schlumberger. (1986). Log Interpretation Charts (p. 122). Houston: Schlumberger Well Services. Sharma, P. V. (1997). Environmental and Engineering Geophysics (p. 372). Cambridge: Cambridge University Press. Shweikani, R., & Durrani, S. A. (1995). Thoron contributions in radon measurements in the environments. Radiation Measurements, 25, 615–616. UNSCEAR. (2000). Sources and Effects of Ionizing Radiation. United Nations: United Nations Scientific Committee on the Effects of Atomic Radiation. UNSCEAR. (2006). Effects of Ionizing Radiation—Annex E: Sources-to-Effects Assessment for Radon in Home and Workplaces. United Nations: United Nations Scientific Committee on the Effects of Atomic Radiation.

Water Air Soil Pollut (2011) 218:371–386 Viñas, R., Eff-Darwich, A., Soler, V., Martín-Luis, M. C., Quesada, M. L., & Nuez, J. (2007). Processing of radon time series in underground environments: Implications for volcanic surveillance in the island of Tenerife, Canary Islands, Spain. Radiation Measurements, 42, 101–115. Werner, A. B. T., Sinclair, W. D., & Amey, E. B. (1998). International strategic mineral issues summary report— Tungsten. US Geological Survey Circular 930-O. Whitfield, J. M., Rogers, J. J. W., & Adams, J. A. S. (1959). The relationship between the petrology and the thorium and uranium contents of some granitic rocks. Geochimica et Cosmochimica Acta, 17, 248–271. Wigley, T. M. L. (1967). Non-steady flow through porous medium and cave breathing. Journal of Geophysical Research, 72, 3199–3205.