Introduction

One of the most serious problems faced by the collieries in Upper Silesia, Poland, are the highly saline waters in the carboniferous strata (Różkowski 1995; Różkowski and Wilk 1982). These formation waters often contain elevated concentrations of natural radionuclides, which are mainly radium isotopes. Upper Silesian radium-bearing waters were first described by Sałdan (1965) and later investigated by Tomza and Lebecka (1981). The predominant radionuclides in these waters are radium 226Ra, a member of the uranium series, and radium 228Ra, a member of the thorium series. The 226Ra concentration in the Upper Silesian brines is usually between 0.1 and 8 Bq/L, while 228Ra concentrations usually range from 0.1 to 10 Bq/L.

Natural waters with similar high radium concentrations have been observed elsewhere (Gucało 1964), especially in oil fields (Alley et al. 2011; Eriksen et al. 2009). Hot springs in Iran contained up to 330 Bq/L of 226Ra (Khademi et al. 1980; Sohrabi 1993). In German coal mines (Centeno et al. 2004; Eggeling et al. 2013; Gans et al. 1981), radium-bearing waters were identified with 226Ra concentrations reaching 63 Bq/L. Galhardi and Bonotto (2017) described radium and uranium measurements upstream and downstream of the discharge from the coal mine, and used the activity ratios of 234U/238U and 226Ra/238U to explain the possible leaching of radionuclides from mine tailings. Although several publications have studied elevated 226Ra concentrations in water, there is little information about radium isotopes from the thorium series (228Ra and 224Ra), as they are more difficult to measure (Eriksen et al. 2009; Dickson 1990; Gans et al. 1981).

Daily 226Ra activity in water flowing into underground galleries and discharging from collieries was first estimated in 1987. Since then, the daily release of radium isotopes has been assessed regularly by monitoring radium concentrations in formation waters flowing into underground workings, and in the waters that then discharge to the natural environment. However, only rough estimates were made for individual or groups of mines (Chałupnik et al. 2017). Detailed analyses of radium concentrations in formation waters at all mine levels have not been conducted.

More precise analysis of radium in underground waters became necessary in 1995, when approximately half of the mines that were still active were closed. Some of these collieries were allowed to flood, but waters from others are pumped to the surface directly or through active mines’ dewatering systems (Bukowski 2015). Furthermore, we expect that more mines will be shut down or combined in the forthcoming years. Some coal mines have already or will begin to exploit deeper levels, up to 1100–1200 m below the surface (Dubiński 2013). We expect that radium will appear in inflows of more saturated brines at deeper levels, although we do not expect significant changes in redox potential (Gzyl et al. 2017).

Brines in the Upper Silesian Coal Basin Collieries

Investigation Site

The Upper Silesian Coal Basin (USCB) is located in the southern part of Poland (Fig. 1). There are currently 31 underground coal mines in this region extracting ≈ 72 × 106 metric tons (t) of coal per year. In 1995, 66 coal mines operated in the area, extracting ≈ 150 × 106 t per year. The geological structure of Upper Silesia is very complicated, with numerous faults and other tectonic dislocations (Kotas 1982); see Fig. 2. The geological cross-section of the USCB is shown in Table 1.

Fig. 1
figure 1

Abandoned and active coal mines in the Upper Silesian Coal Basin

Fig. 2
figure 2

Tectonic structural map of Carboniferous (based on Buła and Kotas 1994)

Table 1 Stratigraphical classification of the Carboniferous of the Upper Silesian Coal Basin (based on Buła and Kotas 1994)

There are two hydrological regions in the USCB. The first is located in the southwestern area of the basin, with thick strata of impermeable sediments covering a Carboniferous formation. This overburden (700 m thick) mainly consists of Miocene clays and shales, which render water and gas migration almost impossible.

There are no Miocene clays in the second region. Carboniferous strata are covered by slightly compacted Mesozoic and Quaternary sediments. Meteoric waters, rich in oxygen, can easily reach exploitation zones. Strongly fissured Permian or Triassic limestones are the oldest formations in this area. There are several outcrops of coal seams, enabling easy migration of water and gases. The active and abandoned mines are shown in Fig. 1. Most of the mines in the northern part of the USCB are closed, but due to their interconnectivity, most of them must still be dewatered.

Characteristics of Formation Waters in USCB Collieries

The USCB mine waters contain extremely high concentrations of salts (much higher than oceanic levels), and are sometimes almost fully saturated. The USCB radium-bearing waters are highly mineralised Cl–Na or Cl–Na–Ca brines. The total dissolved concentration is often ≈ 100,000 mg/L, but may reach ≈ 220,000 mg/L (Różkowski 1995). The dominant anion is Cl, which typically does not exceed 100,000 mg/L, while the concentration of HCO3 reaches 10 mg/L (Bondaruk et al. 2015). Although the dominant cation is Na+ (up to 50,000 mg/L), there are significant concentrations of Ca2+ and Mg2+ (up to 10,000 and 7000 mg/L, respectively). Such brines in Silesian collieries contain elevated concentrations of 226Ra and 228Ra, and it has been confirmed that the radium content is correlated with the mine water salinity. We observed 226Ra concentrations in water flowing out from rocks ranging from < 0.002 to 390 Bq/L. While the 228Ra concentrations were sometimes lower, the 228Ra concentrations can exceed that of 226Ra by three times or more (Chałupnik 2007; Wysocka et al. 1998). A similar effect has been observed even in drinking water supplies in Finland, where due to increased salinity near the seashore, radium concentrations were higher there than in inland water supplies (Vesterbacka 2007).

Two types of radium-bearing brine can be distinguished (Tomza et al. 1985)

  • Barium-rich water: contains almost no SO42− ions, but Ba2+ ions are present at relatively high concentrations (up to 6000 mg/L)—not often observed in nature, except in mines;

  • Sulphate-rich water: Ba2+ ions do not occur in these salty waters, but there are SO42− ions at maximum total concentrations of 5000 mg/L.

Although some believe that redox potential is the most important factor affecting radium appearance and behaviour in underground aquifers (Wiegand and Feige 2002) and leaching of radium from rocks, our modelling of radium behaviour (Chałupnik 2008) shows that the only process responsible for radium release into brines is the recoil effect, as was found by others (Dickson 1990; Krishnaswami et al. 1982). Radium activity is further controlled by the salinity of the brines and the presence of barium. Barium can efficiently block the cation exchange centres and prevent radium adsorption. In sulphate-rich brines, sodium (a univalent ion) is not so efficient and there is radium adsorption. This leads to a different activity ratio (226Ra and 228Ra) in barium- and sulphate-rich brines, being 2:1 in the first type of waters, but 1:2 or even 1:3 in the latter one., despite the 1:1 activity ratio of 238U and 232Th in the strata (Krishnaswami and Turekian 1982).

The presence of barium ions greatly influences the behaviour of radium in the environment because it enables the coprecipitation of radium and barium sulphates. Therefore, the radium present in barium-rich water always precipitates, forming radioactive deposits of radio-barite (BaSO4 + RaSO4) (Martin and Akber 1999; Paschoa and Nobrega 1981; Tomza and Lebecka 1981; Wiegand and Feige 2002). This process can occur due to the spontaneous mixing of natural waters from different aquifers, controlled radium and barium removal, pyrite oxidation and release of sulphate ions, or even underground leaching of fly ash and sulphate from backfilled materials (Chałupnik et al. 2001). Precipitation usually occurs in the underground workings, but sometimes takes place in pipelines transporting water to the surface or in streams, small rivers, or main rivers on the surface (Bondaruk et al. 2015). In the early 1990s, we estimated that 20% of the Polish coal mines (11 collieries) contained barium-rich water. Recent assessments showed that barium-rich waters occur in ≈ 10% of the collieries (only 4 mines). The changes in the chemical composition of the formation waters are due to differing geology as mining depths have increased in the active mines and inflows of barium-rich waters in some of the mines that have closed (Chałupnik et al. 2016).

Radium-rich waters without barium ions are common in the mines in the northern and central areas of the USCB. The hydraulic conductivity of the overburden is relatively high due to the high intergranular and fissure porosity. The permeability of overlying dolomite and limestone is enhanced due to the effects of historical mining and karstic development. Water conditions in the coal measures are determined by the geological structure of the coal basin and anthropogenic factors. The most important factor is the drainage of water through the mine workings (Różkowski 1995; Wagner 1996). Sulphate rich brines are also present in mines with barium and radium-bearing waters. Both types of aquifers are interlaced and mixing of both types of waters leads to spontaneous coprecipitation of sediments with high radium content (Tomza and Lebecka 1981).

The characteristics of Upper Silesian radium-bearing waters are given in Table 2 which is based on the database of the Silesian Centre of Environmental Radioactivity in the Central Mining Institute. The highest radium concentrations are observed in barium and radium-bearing waters, where 226Ra concentrations typically exceed 6.5 Bq/L and sometimes exceed 100 Bq/L. In barium-rich waters, 226Ra concentrations are always higher than 228Ra. In sulphate-rich waters, the predominant radium isotope is the thorium-series 228Ra. However, the concentrations of both radium isotopes in the latter type of waters are significantly lower and have not exceeded 12 Bq/L in recent years, although they previously reached 20 Bq/L (Chałupnik 2007).

Table 2 Characteristics of typical radium-bearing waters from the Upper Silesian Coal Basin; K = 1000

The assessment of the daily activity of radium isotopes in waters flowing into underground mine galleries is not very precise due to the reasons described earlier. Water inflows with different salinities, flow rates, chemical compositions, and radium concentrations often enter the same mine. It is especially difficult to assess daily activity in mines where barium-rich and sulphate-rich water flow into the same mine, as spontaneous coprecipitation of radium and barium sulphates occurs immediately after the different waters mix underground. Nonetheless, even a rough approximation provides a better understanding of the scale of the problem.

Methods

Sampling

Water samples were collected by the mining staff during routine underground and surface monitoring, as required by Polish regulations. Chemical and radiological analyses were conducted to measure the concentrations of radium isotopes and other particular ions. Water directly flows into the underground workings from rocks or from boreholes, and the water pumped out from specific mining levels were also sampled in all mines. Radiochemical analyses were conducted by the Central Mining Institute.

Routinely, inflows into the underground workings are sampled once per year, and in mines with radium concentrations above 1 Bq/L, sampling is conducted more frequently. The flow rates are estimated in situ and therefore are not very precise. The level of uncertainty for the flow rate assessment of the inflows from the strata was usually between 15 and 30%, accordingly to data from mines. Assessments of the flow rates from particular horizons in the coal mines were more precise, as the volume of released waters is determined using the flow rate meters of pumps. Therefore, the uncertainty of radium activity in outflowing waters is mostly due to the uncertainty of radium activity concentration measurements (8% for 226Ra and 15% for 228Ra). The surface settling ponds and downgradient brooks and rivers were routinely sampled once a year. Mines with effluent radium concentrations exceeding 1 Bq/L were sampled four times a year. However, assessments of radium activity in surface waters is hampered because flow rates in rivers are usually unstable. The latter values were obtained from the subtraction of radium activity in outflows from the total radium activity in inflows.

Determination of Radium Isotopes

226Ra and 228Ra were determined by liquid scintillation counting after chemical separation of RaSO4 as a barium carrier, following the method developed by Chałupnik and Lebecka (1993). This method allows the simultaneous determination of 226Ra and 228Ra. 228Ra is determined by direct measurements of the low-energy beta particles it emits, while 226Ra is determined by measuring the alpha particles that this radionuclide and its daughter products emit. As the beta spectrum is continuous, beta particles emitted by 226Ra daughter products are measured in the same energy range as 228Ra. Therefore, this effect must be corrected for. The lower limit of detection (LLD) for 226Ra (0.002 Bq/L) is achieved when a low background liquid scintillation spectrometer QUANTULUS model 1220, Packard Company, is used to analyse the initial 1 L sample, with a counting time of 1 h. The corresponding LLD value for 228Ra is approximately 0.004 Bq/L.

The analyses of natural radionuclides in deposits have been done using low background gamma spectrometry. Sediment samples were dried, pulverized, and homogenized; after that, 0.6 L of the final sample was placed in the Marinelli beaker and measured by high resolution gamma spectrometry. The detection limit for radium isotopes was less than 1 Bq/kg (Michalik 2008).

Results

The main goal of our investigations was to estimate the changes of radium activity in inflows into the USCB collieries as well as in discharges from these mines. To prepare such a balance, two data sets were necessary: the flow rates and radium concentrations. The flow rate(s) of each significant inflow into mine workings and the volumes of water pumped daily to the surface from mines were obtained from every mine, together with water samples. The radium activity balance in the inflows is a sum of products, resulting from multiplication of daily water volumes and radium isotopes concentrations in every investigated inflow. The same procedure was applied for the discharge waters.

As mentioned earlier, the total activity of radium in deposits is the result of subtracting the activity values for the effluents from that of the inflows. Only general water volume data are presented here for the USCB collieries, which changed over time due to the restructuring of the Polish mining industry. For example, in the early 1990s, the total daily inflow of water was ≈ 850,000 m3. The volume of inflowing water gradually decreased to ≈ 700,000 m3 per day in 1999. In 2010, the total daily inflow of natural water into mines was 657,500 m3, of which 35% (almost 230,000 m3 per day) was pumped from abandoned mines (Bukowski 2015). According to our 2016 assessment, the total daily inflow of water into mines reached ≈ 630,000 m3. Closing the mines not only caused a decrease in the influent water, but also a change in the water’s chemical composition. For example, according to Bukowski and Augustyniak (2013), the total release of chloride and sulphate ions decreased from 8000 t per day in 1991 to 5000 t per day in 2010. We also observed changes in the total activity of radium entering Silesian mines daily in the same period.

The results of the assessment conducted in 1995 showed that the approximate amount of 226Ra in water flowing into the USCB coal mines reached 625 MBq/day (230 GBq per year), while that of the 228Ra was ≈ 700 MBq/day (255 GBq per year). Although radium concentrations in sulphate-rich waters were usually less than in barium-rich waters, the total activity of radium in inflows to mines with sulphate-rich radium-bearing waters was much greater (Skowronek et al. 1998; Wysocka et al. 2017). It should be pointed out that significant concentrations of radium isotopes in waters can only be found in active mines; in closed mines, the mixing of inflows from different aquifers causes the concentration of radium isotopes to not exceed 0,5 Bq/L. Moreover, the daily volume of dewatering is typically much less in abandoned mines than in active mines.

The calculated daily radium activity of mine inflows in 2016 differed from that calculated for 1995; the approximate calculated amount of 226Ra in water flowing into coal mines in the USCB reached 483 MBq/day (176 GBq per year), and 715 MBq/day (260 GBq per year) for 228Ra. Comparing the assessments, the daily activity of 226Ra in inflows was less in 2016 than in 1995, while the activity of 228Ra was comparable. The reasons for the changes in the formation waters are complex. The most important are:

  • Several coal mines in which barium-rich waters with elevated concentrations of 226Ra were closed;

  • An increase in the inflows of water with higher concentrations of 228Ra than of 226Ra (by a factor of approximately 2) was observed in the three coal mines that are the most important contributors to the daily release of 228Ra;

  • Deepening of the exploitation level allowed access to new coal seams where formation waters with elevated radium concentrations occur. The best example is that the radium level in one of the collieries prior to 2010 did not exceed 0.2 Bq/L, but as a result of exploitation at the deeper horizon, the radium activity in inflowing waters increased to 5–7 Bq/L, while the radium content in the discharge waters was 1.5–2.5 Bq/L (Skowronek et al. 1998; Wysocka et al. 2017).

  • A reduction in radium removal from underground installation water due to technical and economic problems.

The highest inflow of 226Ra and 228Ra isotopes to collieries with sulphate-rich waters was approximately 245 MBq/day (80 MBq/day of 226Ra and 145 MBq/day of 228Ra). For collieries with barium-rich waters, the corresponding activity was 225 MBq/day (150 MBq/day of 226Ra and 75 MBq/day of 228Ra). The total daily inflow of water into collieries with sulphate-rich waters was ≈ 3.5 times higher than that to collieries with inflows of barium-rich waters.

The daily activity of radium isotopes in inflows and discharge from mines for 1995 and 2016 are compared in Fig. 3. Additionally, we compared the activities of the radium in water flowing into mines to the activities of radium in water discharged from coal mines to the environment.

Fig. 3
figure 3

Comparison of the assessments of the daily activity of radium isotope inflows to coal mines, in discharge waters and sediments in 1995 and 2016 [MBq/day]

Radium accumulates in sediments due to the coprecipitation of barium and radium sulphates and adsorption of radium to bottom sediments. We assumed that the difference between the daily activity of radium in inflows and discharges were a rough estimate of the radium activity in the sediments remaining underground. It is not possible to prove this though, as deposition typically takes place in exploited zones in an uncontrolled way.

The distribution of the 226Ra and 228Ra activities in water flowing into the mines is shown in Fig. 4. In most of the coal mines, the daily radium activity in the inflowing water ranges from 1 to 10 MBq. In the second-largest group of mines, daily radium activity ≤ 1 MBq. However, in four mines, the daily activity of radium in the inflows > 100 MBq. There are thus large differences in the total radium activity within the groups. The lowest total activity for one of the colliery groups equals 6.4 MBq/day, while the highest one exceeds 1000 MBq/day. The contribution of radium activity from these four mines is thus a crucial part of the radium activity in mine waters. Surprisingly, the radium concentrations in the water of those mines is not that high (reaching ≈ 17 Bq/L of 226Ra and 228Ra), but the volume of inflowing formation water is significant, reaching 30,000 m3/day. This is much higher than that observed in the other coal mines, and the impact of these waters is responsible for approximately 92% of the total daily radium activity release in the entire USCB. However, the highest concentration of radium isotopes in underground inflows (49.70 ± 1.69 Bq/L of 226Ra and 38.05 ± 3.69 Bq/L of 228Ra) was measured as seepage from rocks in 2016.

Fig. 4
figure 4

Distribution of the daily activity of radium 226Ra + 228Ra in water flowing into coal mines and total radium activity within each group of coal mine

Environmental pollution caused by the release of radium with the brine from the coal mines is an important issue. The radium (450 MBq/day in 2016) being discharged with mine effluents into surface settling ponds, which are only designed to remove suspended material, is significant, and adsorption of radium to the sediment in these ponds only removes a small portion of the radium activity (5–10%). The remaining activity is released downstream into streams and rivers with the brine, where the radium concentrations gradually decrease due to dilution and slow radium adsorption to the bottom sediments. Elevated radium concentrations, up to 0.05 Bq/L can be seen as far as 100 km downstream from discharge points, as the typical radium concentration in surface and river waters do not exceed 0.008 Bq/L (Koster et al. 1992; Wardaszko et al. 2001). This process leads to the creation of deposits with radium concentrations that often exceed 1 Bq/g.

The typical concentrations of 226Ra and 228Ra in the USCB soils are similar to the world average (25 Bq/kg for both nuclides), while their concentrations in the strata adjacent to the coal seams do not exceed 120 Bq/kg and 100 Bq/kg, respectively (Michalik 2005; Wysocka and Skowronek 1991). This value has been established as the “exemption level” in the Euratom/59/2013 EU Directive (European Council 2014). Therefore, the dewatering of coal mines will require a decision by the National Atomic Energy Agency for a method to solve the problem. Radiation hazards to inhabitants of the Upper Silesia region are another issue. In our experience, the most likely pathway of this hazard is external radiation from the sediments because this water is unsuitable for drinking due to its salinity. However, these sediments are typically covered by water (brooks and rivers) and may sometimes appear on river banks due to floods or the dredging of watercourses. We estimate that additional doses for inhabitants should not exceed 0.1 mSv/year, which is well below the 1 mSv/year threshold (Skubacz et al. 1992; Michalik 2008, 2011). Nonetheless, due to the long half-lives of radium isotopes, particularly 226Ra, the deposition of radium on the river beds and contamination of adjacent areas by released brines is, in our opinion, an important environmental problem in the USCB (Chałupnik et al. 2017).

Conclusions

The results of long-term monitoring of radium activity in Polish collieries have shown that overall radium activity being discharged from coal mines have decreased during the last two decades. Nonetheless, the release of underground water from the mines with high radium isotope concentrations may lead to considerably increased radioactivity in the environment.

In 1995, the maximum 226Ra concentration was as high as 25 Bq/L, while for 228Ra it was about 13 Bq/L. In 2016, the maximum concentration of 226Ra was up to 7 Bq/L, while for 228Ra it was about 10 Bq/L (maximum total activity 17 Bq/L). The approximate calculated activity of 226Ra in water flowing into the USCB mines in 2016 reached 483 MBq/day (176 GBq/year), while 228Ra was approximately 715 MBq/day (260 GBq/year). A comparison of assessments conducted in 1995 and 2016 shows that the activity of 226Ra in daily inflows was about 30% less in 2016 than in 1995, while the activity of 228Ra was comparable in both years. The reasons for the changes in the 226Ra/228Ra ratio of formation waters are complex, and include changes in the geological conditions at the mine’s deeper horizons and technical measures due to the restructuring of the mining industry.

The radium concentrations in the mine water discharges were, and are still, enhanced. Due to dilution, there is a significant decrease in radium concentrations, especially in big rivers like Vistula. On the other hand, during seasons with low precipitation, the enhanced radium activity can be seen few dozens or even 100 km downstream from the mines’ discharges.

Radium is gradually removed from river water by adsorption onto bottom sediments. Although the impact of radium-bearing water from the Upper Silesian coal mines on river water and bottom sediments is clear, the radiation doses received by inhabitants were low. Evaluations conducted by scientists from Central Mining Institute showed that annual doses received from wastewater with enhanced radioactivity and the presence of radionuclides in waste rocks released by the mines are well below the 1 mSv threshold.

A comparison of radium activity in inflows and discharges from collieries in 1995 and 2016 showed that even though roughly 50% of the coal mines in the USCB closed over the last two decades, and radium removal techniques have been applied in underground galleries, only a 20% decrease of total radium activity has been observed in colliery inflows and mine effluents. This is because of deeper exploitation in active mines and the fact that water from abandoned mines that are adjacent to active ones, must still be pumped out. Thus, it remains an important environmental problem in the USCB in Poland.