Radioactivity and hydrochemical properties of certain thermal Turkish spa waters

Turkey is very rich in thermal waters that originate from rocks in different chemical composition and age, and these waters commonly have been used in spas (thermal springs) since ancient times. The physical and major chemical properties of some natural thermal waters have already been studied while their radioactivity is less known.

Radionuclides are the main reason of radiation exposure of human beings and constitute background radiation levels (Bozkurt et al. 2007). The presence of radioactivity in nature is related to the radionuclides sourced by naturally decay chains, cosmic rays and artificial radionuclides.

The determination of radionuclide dispersion in the environment and calculation of the harmful effects of radiation exposure from the background is required. Natural radioactivity levels of a certain environment radionuclide depend on concentrations in air, water and rock that vary relative to geological and geochemical features of the source rocks. Cosmic rays from space also contribute to the background relating to altitude of the environment. Determination of the radionuclide concentrations of the thermal water samples in spas is very important for human health because of the diversity of the background radiation. After the Chernobyl accident, studies about determination of environmental radioactivity levels have been performed, especially in the northern parts of Turkey (TAEK 1998).

The determination of radioactivity is important in waters which have a significant role on dispersion of radionuclides in nature. Natural waters are known as alpha, beta and gamma emitters and in a wide range of concentrations (Akyıl et al. 2009; Zorer et al. 2009; Janković et al. 2012; Görür & Camgöz 2014; Kuluöztürk & Doğru 2015). Radiation emitters in water are accountable for a small rate of the total dose exposed from natural and artificial radioactivity (UNSCEAR 2000). Transition of the chemical elements from rocks into the water depends on the geochemical characteristics of aquifers and period of the interaction between water and rocks (Janković et al. 2012).

Several thermal spring waters having different characteristics formed throughout the western and eastern parts of Turkey, located mostly around active volcanism and fault zones. The aim of the present study is to determine the radioactivity of thermal spa waters with gross alpha (GA)/gross beta (GB) counting, and gamma emitter radionuclide activity concentrations; and also to determine some physical and hydrochemical characteristics of the waters and find a relationship with radioactivity properties. In accordance with this purpose, a total of 31 thermal water samples were collected at spas from different parts of Turkey. The activity concentrations of ²²⁶Ra, ²³²Th, ⁴⁰K and gross α/β in the samples were determined and annual effective dose (AED) was calculated. Additionally, the concentrations of main cations, anions and physical parameters of water samples were determined.

The radioactivity analyses were made at the Science and Technology Application and Research Center of Bitlis Eren University (Bitlis/Turkey). The GA and GB activity concentrations in the water samples were measured by ultra-low level α/β counter MPC 9604-1 (Protean Instrument Corporation). The detectors were calibrated for α and β energies using ²⁴¹Am (185 Bq) and ⁹⁰Sr (172 Bq) standard sources, respectively. Background counting was performed with empty steel planchets at 720 min intervals for each detector. Sample counting was made at 600 min intervals for all samples (Kuluöztürk & Doğru 2015).

The studied thermal waters were collected around the tectonically active zones, geothermal areas and young volcanic centres. Most of the thermal springs are located roughly parallel to active fault systems, e.g., normal, oblique, horst-graben, and around Neogene aged volcanic areas (Figure 1). In the western part of Anatolia, the main tectonic features are extensional E–W-trending horst-graben systems (Bozkurt 2003). All of the thermal sources in western Anatolia, low temperature springs (>65 °C), are related to the low-angle detachment fault while high temperature geothermal fields developed along E–W-trending high angle normal faults (Karakuş & Şimşek 2013). Precambrian aged basement rocks, which are the reservoir of the springs, are composed of gneiss, metagranite, schist, paragneiss and metagabbro, whereas marble and schist cover Paleozoic to Early Tertiary aged rocks (Karakuş & Şimşek 2013 and references therein). The basement rocks are composed of marble and schist of the Paleozoic age in the western part of central Anatolia. Neogene carbonates unconformably cover the basement rocks and they are the reservoir rocks of the samples W-1 and W-20 together with Paleozoic aged rocks (Mutlu & Güleç 1998). Tertiary granodioritic rocks are covered by Neogene rocks which are formed from volcanics, different sized detrital sediments, basaltic rocks and Quaternary alluvium which are the reservoir rocks of the samples W-9 to 12. Paleozoic to Early Mesozoic aged metamorphics, e.g., gneiss, schist, marble and ophiolites and Upper Cretaceous mélange form the reservoir rocks while Upper Miocene-Pliocene detrital sediments form the cover rocks of the reservoir for the samples W-2 to 6 and W-13 to 15 (Mutlu 2007). The occurrence of thermal springs (W-7 and W-8) in the study area is linked to the young normal faults (Gökgöz & Tarcan 2006 and references therein). The authors indicate that karstic limestones and Lycian nappes are the reservoir rocks and there is no cap rock for geothermal systems of the samples W-7 and W-8. The reservoir rocks of sample W-16 formed from Paleozoic to Mesozoic aged metamorphic rocks and Miocene to Pliocene aged sedimentary and carbonate rocks. The Pliocene-Quaternary volcanic and volcanoclastic rocks are the cap rocks of the sample (Pasvanoğlu & Güler 2010). The Upper Miocene units formed from basaltic and andesitic lavas and volcanoclastic rocks are the oldest units and are overlain unconformably by the Pliocene units (Kalkan et al. 2012 references therein). The spring water sample W-17 lies on the Late Miocene fault zone, which is one of the most active fault belts of the Eastern Anatolian Region. Sample W-18 represents outflow of deep groundwater that has been recharged and circulated in a possible fracture zone of flysch and sandstone of the Eocene age (Saner 1978). The Paleozoic aged detrital sedimentary rocks, limestones and granitic rocks which form the reservoir rocks of the sample W-19, show poor or very poor aquifer characteristics, covered unconformably by Pleistocene to Quaternary sediments (Yalçın et al. 2007 and reference therein).

The total dissolved solids (TDS) of the studied waters varies in a wide range between 590 and 29,212 mg/L and may be related to long residence time and circulation. The pH of the waters range from 6.40 to 9.21, the EC values range from 0.88 to 43.6 mS/cm, and temperatures of the thermal water samples vary from 26.5 to 87.0 °C (Table 1). The wide range variation of the above-mentioned properties may be related to distance from the main fault zone, penetrating depth, circulation time and/or source rocks' temperature. The thermal waters are slightly acidic/neutral to alkaline in character. From the hydrogeochemical point of view, the following four water types were defined as ⁠, Na⁺-Cl⁻, ⁠, Mg²⁺ and/or (Table 1). EC values and Cl content are high in especially samples (W-7, 8, 9 and 19) taken from near the coast which may reflect mixing with sea water or deep water circulation and partially long residence time (Table 1). Also, Gökgöz & Tarcan (2006) suggested that Cl-rich waters were sourced from the sea water contribution. On the other hand, Mutlu & Güleç (1998) indicated that the Cl-rich character related to the presence of connate fossil waters at depth. Additionally, Gökgöz & Tarcan (2006), who studied in the same area from which came the samples W-7 and W-8, suggested that these waters are of meteoric origin and sea waters percolated to the reservoir through the karstic voids and fractures. Additionally, the spring waters are mixed water which are heated at depth and ascend to the surface via major faults. High HCO₃ content with Na were determined in especially W-17, 18 and 20, and may be related to the reaction of cold meteoric water with carbonate rocks and ion exchange in the aquifers (Özen et al. 2012). Therefore, variations of HCO₃ concentrations among the studied waters may be related to dissolution and precipitation of carbonate minerals in reservoir and cover rocks. Additionally, Tarcan et al. (2009) stated that enrichment of thermal waters with Na reflects rock dissolution and ion exchange reactions in deep aquifers at high temperatures. Although a strong positive correlation (r = 0.97) was determined between Ca and SO₄ in samples W-1, 2, 7-9, 16 and 19, sulphate is not determined as the main anion in any sample. In addition, high levels of sulphate concentration may be related to oxidation of metallic sulphides and/or escape of H₂S from a deep hot-water system and dissolution of sulfate minerals. The processes mentioned regarding rising sulfate concentration were not thoroughly developed in the studied reservoir rocks. The negative trend of Ca with SO₄ can be attributed to calcite and/or aragonite dissolution in parallel with gypsum and/or anhydrite precipitation.

The characteristics of the waters were affected by many parameters, e.g., extensive volcanism, different active fault systems, mixing sea water, penetrating time and depth, type of the reservoir/cap rock, etc. Some water samples having nearly the same reservoir and cap rocks show different physical, chemical and radioactivity properties (2, 5 and 6, and 11 and 12, Figure 1). Additionally, different concentrations of natural radioelements in groundwater can be related to temperature, dissolved inorganic salts, geological composition of the rocks and other factors such as conductivity and pH, etc.

The highest ⁴⁰K was measured for peloid samples P-11, 20 and 20/1 and is 1,698, 1,516 and 1,041 Bq/kg, respectively (Karakaya et al. 2015). Content of K₂O % wt and ⁴⁰K of the peloids shows moderately positive correlation (r = 0.65), and K (ppm) and also ⁴⁰K activity of the waters presents positive correlation (r = 0.79) which indicates that the K activity originates from source rocks (Table 2). However, there is no correlation between K₂O % and ⁴⁰K activity of the peloids with some parameters of the water samples. As well, no statistically significant correlation was found between GA and GB activities when compared to concentrations of ²²⁶Ra, ²³²Th and ⁴⁰K measured in peloids and waters. The lack of correlation may be related to geochemical composition of the source rocks, redox conditions and circulation time due to easy dissolution of some nuclides (Vesterbacka 2007). The highest ⁴⁰K activities were determined in the chloride-rich waters samples W-7, 8, 9 and 19, and in these samples strong positive correlation (r = 0.98) is found between ⁴⁰K and Cl⁻. Thorium activity concentration of the thermal waters varies from 0.089 to 3,477 Bql⁻¹ (mean 1.027 Bql⁻¹) and is usually lower than other radionuclides. The low concentration of thorium is due to being a relatively insoluble element in natural waters and found generally within soils or rocks (Labidi et al. 2002).

It was determined that GA and GB activity concentrations are generally lower than recommended values for drinking water by WHO (2004), and the measured GA and GB values are generally in similar ranges of other spring waters in the world (Table 2, Figure 3(b)).

Recommended activity limit values for gamma emitters (32, 45 and 420 Bql⁻¹ for ²²⁶Ra, ²³²Th and ⁴⁰K, respectively) by UNSCEAR were not exceeded (Figure 3(a)). Recommended activity limit values for GA and GB by WHO (2004) were exceeded by 82% and 21% of samples, respectively. Total calculated AED values were given as the sum of the ²²⁶Ra, ²³²Th, ¹³⁷Cs, ⁴⁰K, GA and GB dose values (Table 3). The largest and lowest contribution to the total dose was sourced from ²²⁶Ra and ⁴⁰K, respectively. The total AED is between 0.040 and 0.173 mSvy⁻¹ and mean value is 0.086 mSvy⁻¹. The recommended AED value is 0.1 mSvy⁻¹ (WHO 2011) and 33% of samples exceed this value (Tables 2 and 3).

In this study, physical properties, major element compositions and radioactivity profiles (²²⁶Ra, ²³²Th, ¹³⁷Cs, ⁴⁰K, GA and GB) of some therapeutic spa waters in Turkey were investigated. According to their major anion and cation content, water types were classified as ⁠, Na⁺-Cl⁻, ⁠, Ca²⁺ and/or ⁠. The investigated properties of the waters vary over a wide range, depending on the nature of the aquifer, e.g., mixing sea water, penetrating time and depth, conductivity, temperature, etc. The water types and the wide range of TDS (190 to 29,200 mg/L) and EC values (0.88 to 43.6 mS/cm) of the waters indicate that the thermal waters originated from different geochemical processes in Turkey. The water samples with high Cl concentration showed the highest mean value of ²²⁶Ra concentration resulting from seawater contribution and leading to proper conditions of ²²⁶Ra mobilization. At the same time, these waters also are enriched with ⁴⁰K and ²²⁶Ra isotopes, and GA and GB activities are mostly sourced from the ²²⁶Ra. Close affinity of Cl-TDS indicated that the nuclides formed from dissolved minerals in these waters.

Most of the GA and some of the GB radioactivity concentrations are higher than the recommended guideline activity concentrations by WHO (1993) for drinking water. Values of GA in 82% and GB in 21% of the samples were higher than the recommended limit values. 33% of the samples exceed the recommended activity and AED values. ²²⁶Ra is the main contributor to the AED. The thermal waters are not used for drinking purposes, however, based on the radiological properties of the investigated waters, some problems can be caused when using therapeutic treatments with peloids.

The project was funded by The Scientific and Technological Research Council of Turkey (TÜBİTAK 110Y033) and the Selçuk University Scientific Research Projects support program (BAP 11401045).