1 Introduction

The 222Rn in groundwater is mainly due to the available concentration of the source 226Ra in groundwater. The hard rock aquifers have groundwater accumulated in cracks and interconnected joints; the movement of groundwater is very slow which indirectly enhances the dissolution of the aquifer matrix. The increase of the residence time in hard rock aquifers is generally reflected in the higher Electrical conductivity (EC) (Chidambaram 2000; Srinivasamoorthy et al. 2008). Having a short half-life, 222Rn in groundwater attains a secular equilibrium with 226Ra in the aquifer matrix in the regions with low velocity or with increased residence time. Similar to that of the major ions (e.g., Cl−, SO4 2−, HCO3 −, Na+, Ca2+, Mg2+, etc.), 222Rn can also migrate in groundwater from its source along the direction of movement of water (Przylibski et al. 2004). It has been inferred by earlier researchers that the grade of metamorphism plays a significant role in controlling the activity concentration of 222Rn in groundwater (Butsaert et al. 1981).

In the hard rock aquifers during non-monsoon periods or due to the failure of monsoon, surface water sources area dries up. As a result, groundwater serves as the main source of water for drinking and domestic purposes. In this case, if the available groundwater is rich in 222Rn and its radioactive daughter products, it will create health hazards when entering the human chain. Though 222Rn has a very short half-life, (3.82 days) the time interval between the pumping of groundwater and its consumption is generally well within this half-life period. There have been several studies to understand the impacts of 222Rn on human health. These include lung cancer (Cothern et al. 1986), an increased incidence of childhood leukaemia and possibly gastric cancers (Akerblom 1994) for users of 222Rn-rich groundwater.

The direction and velocity of groundwater movement has been traced by studying 222Rn in regional and local scales (Miller et al. 1990; Cable et al. 1996; Crotwell and Moore 2003; Schwartz 2003; Desideri et al. 2006). Further, 222Rn has also proved to be an indicator of seismic and volcanic activities (King 1986; Virk et al. 1997; Wu et al. 2003). The persistence of openness of fracture can be marked by means of 222Rn results (Wu 2007); moreover, it also helps to determine the degree of fracture openness. According to Mulligan and Charette (2006) the radon-based estimate of SGD (Submarine groundwater discharge) can help us to measure the discharge of fresh groundwater from the coastal aquifers to the sea.

Since no study exists to understand the 222Rn activity concentration in the groundwaters of Madurai region, an attempt has been made in this paper to study the seasonal variation of 222Rn and its distribution in groundwater, and to correlate them with various field parameters.

2 Methods and Materials

2.1 Study Area

The study area is Madurai district located in the southern part of Tamilnadu, which is bounded between Latitude 9°30′.00 and 10°30′.00 N and Longtitude 77°00 and 78°30′E and covers an area of about 3,741 km2. It is bordered to the east by Sivaganga, to the north by Dindigul and Thiruchirapalli, Theni to the west, and Virudhunagar district to the south (Fig. 1). The study area is predominantly covered by crystalline formations comprising significantly fissile Hornblende Biotite Gneiss followed by Charnockite, Granitic intrusions, Quartzite and flood plain alluvium (Fig. 1). Graphite bearing hornblende gneiss and calc-granulites of the study area are reported to have higher amounts of uranium (12–28 ppm) than schist (Pandey and Krishnamurthy 1995). The flood plain alluvium occurs along the Vaigai river course. Groundwater occurs in unconfined conditions in regions with higher weathered thickness, and is also reported to occur in two different fracture zones namely, shallow and deeper zone. In the deeper fracture zone of the study area, groundwater is reported to exist under semi-confined to confined conditions (CGWB 2007). It was observed that more than 30 % of the wells yield between 1 and 3 L/s and 29 % of the wells yield below 1 L/s (CGWB 2007). In general, the groundwater flow is inferred to be from the NW to SE (Thivya 2013). The yield is mainly governed by the distribution of fractures in the rock, or the degree of interconnectivity. Vagai and Mullaiperiyar are the major rivers of the study area. The study of a long-term data on rainfall from 1901 to 2004 shows that north east monsoon represents 47 % and the south west monsoon 32 % of the total rainfall received. The temperature ranges from 15 to 41 °C and the climate is subtropical. The relative humidity during north east monsoon is high and varies from 45 to 85 %.

Fig. 1
figure 1

Lithology and sampling points of the study area (modified after Thivya et al. 2013a, b)

Full size image

2.2 Collection of Samples

A total of about 42 samples were collected from hand pumps and borewater representing the entire district for two seasons, i.e., South West Monsoon (SWM) in the month of August and North East Monsoon (NEM) in the month of November. A portable water analysis kit was used to for insitu measurements of pH, temperature and TDS (Thivya et al. 2013a, b). A Thermo Orion 3 star bench top meter was used for measuring pH and temperature. The pH electrode was calibrated by three buffers (pH 4.01, 7.00 and 10.00) to get accurate results. The accuracy of pH and temperature was ±0.002 and ±0.1 °C, respectively. TDS was measured by the Eutech handheld instruments. The accuracy was ±1 % with range of 100, 1,000 ppm and 10.00 ppt. 250 mL water samples were collected separately in glass bottles for 222Rn analysis. The sample was allowed to flow for 10 min before the collection in order to avoid the interference of old water which is already present in the well casing. The atmospheric contact of the water in casing would have resulted in loss of 222Rn by decay or due to outgassing. The 222Rn sampling was carried out carefully by the fact that the gas easily escapes from water. Care was also taken to collect the sample without bubbling. The collected samples were analyzed in the field within 1–3 h of collection; according to the time of analysis, a decay correction was also carried out.

2.3 Radon Activity Concentration Measurements

RAD7 of Durridge Company, USA, was used in determining 222Rn concentration in water. The RAD7’s solid state alpha particle detector is almost completely insensitive to beta or gamma radiation, so there was no interference from beta-emitting gases or from gamma radiation fields. The most likely effect of high levels of beta or gamma radiation will probably be an increase in detector leakage current and increased alpha peak width. Typical environmental levels of beta and gamma emitters have absolutely no effect on the RAD7. The overall calibration accuracy of RAD7 is about ±5 %.

The RAD7 (Fig. 2) setup consists of three components: 1. Vial with sample; 2. the desiccant tube; and 3. the Alpha detector. The RAD7 instrument has a convenient stand for the desiccant tube placed between two clasps present in the lid, with provision for the vial in the foam cavity. In principle, it is a closed loop system, where the flow rate is independent of the volume of air and water; but the volume of air and water are constant. The 222Rn extracted from the sprayed water is recirculated as air until a state of equilibrium is developed. The system attains the state of equilibrium when no more 222Rn can be extracted from the water, and this process persists for 5 min. More than 95 % of available 222Rn in water is removed before the state of equilibrium (Singh et al. 2010).

Fig. 2
figure 2

Radon meter (RAD7) and Schematic diagram of the measurement system, the dotted line is for a bypass loop which is used after air- water equilibration is reached (after Singh et al. 2010)

Full size image

3 Results and Discussion

3.1 Quality and Status

There is a seasonal variation in higher concentration levels of 222Rn. In SWM, the highest value is observed as 60 Bq/L, whereas in NEM the value is 211.6 Bq/L at the same sampling location (Karungalakudi) in the study area (Table 1). The USEPA permissible limit of 222Rn in drinking water is 11 Bq/L (USEPA 2003). In the study area, 16 % of the samples exceeded the permissible value in SWM and 11 % in NEM. In SWM, 7 samples (Chinnasooragundu, Othakadai, Karngalakudi, Kottampatti, Thanichiyam, near Thiruvadhavoor and Poonjuthi) were above the permissible limit, whereas in NEM 6 samples (Chinnasooragundu, Othakadai, Vadipatti, Alanganallur, Karungalakudi and Poonjuthi) were above the permissible limit.

Table 1 The value of 222Rn, pH, TDS and temperature (Temp) for the sample of both seasons with measurement accuracy of 222Rn for each sample
Full size table

Several attempts have been made in India to study 222Rn concentration in groundwater (Table 2). The various studies show that highest 222Rn concentration is observed in Budhakedar, Gharwal Himalaya (Prasad et al. 2009). The study area Madurai presents the second highest 222Rn concentration up to 211.6 Bq/L (Table 2). The concentrations of 222Rn in the regions like Himachal Pradesh, Uttarkasi, Doon valley and Siwaliks pertain to the tectonically active sedimentary terrain. Furthermore, those present in Punjab are related to higher uranium concentration in groundwater. In the present study, the 222Rn concentration pertains to hard rock aquifer with higher uranium and with good fracture intensity (Thivya 2013). The study also identified that a maximum uranium value of 157 μg/L was observed in this region. 222Rn, in any form, either in water or rock, can accumulate in open fractures. CGWB (2007) observations showed that there are two prominent fracture zones: 1. Less than 50 m; and 2. More than 100 m. In each of these zones, 2–3 fracture forms were identified (CGWB 2007).

Table 2 A comparison of 222Rn in groundwater in different parts of India with the present study area
Full size table

Figure 3 presents the spatial distribution of 222Rn in groundwater during SWM, which shows that 2,442 km2 of the area has concentrations below 0.69 Bq/L. The higher 222Rn activity concentrations above 21 Bq/L, extends in an area of about 28 km2 in the north eastern part of the study area. Concentrations between 0.70 Bq/L and 20.06 Bq/L extend for an area of about 1,245 km2. It is believed that the geology is the most important factor controlling the source and distribution of 222Rn, because it is controlled by the chemical composition of the rocks and soils from which 222Rn originates. Relatively high levels of 222Rn are associated with granites, phosphatic rocks, ironstones and aluminum shales (Appleton 2008). In some granites, most uranium is found as uraninite, which is relatively easily weathered to radium from which 222Rn can escape into groundwater (Appleton 2008).

Fig. 3
figure 3

Spatial distribution of 222Rn in (a) SWM and (b) NEM

Full size image

Concentration values lower than 0.2 Bq/L were found in about 2,712 km2 during NEM in the study area. The spatial representation of highest 222Rn concentrations above 42.8 Bq/L covers an area of about 89 km2 in the north eastern part of the study area. The concentration between 0.3 and 42.7 Bq/L covers an area of about 916 km2 (Fig. 3). There is a gradual increase of 222Rn activity concentration along the north eastern part of the study area which is represented by pink granitic intrusions.

Hence it is inferred that regions with elevated levels of 222Rn are represented in the granitic terrains in both seasons. Water travelling through granitic aquifers was found to have higher levels of 222Rn, uranium and thorium series of radionuclides (Otwoma and Mustapha 1998). The samples were collected through hand pumps and it is believed that the 222Rn contents in hand pumps have a sensible relationship to geological and geohydrological patterns (Prasad et al. 2009). The transport of groundwater is enhanced during NEM due to the recharge after the rainfall of this or previous monsoon. The dissolution and chemical reaction favoring the release of major ions and 222Rn is enhanced immediately after the monsoon (Prasad et al. 2008). The study was carried out during the monsoon period and it has to be noted that in India SWM follows summer and NEM follows SWM. Hence, the precipitation of salt along the available pore space is more prominent during summer and the dissolution starts during SWM and subsequently the gas gets enhanced during NEM.

The lithology wise characterization of the 222Rn distribution in groundwater shows that higher concentration was observed in the granitic terrain followed by quartzite in both seasons (Table 3). The order of dominance of 222Rn in groundwater samples representing both seasons are as follows: granite > quartzite > charnockite > hornblende biotite gneiss > flood plain alluvium.

Table 3 Lithology wise 222Rn concentration in Madurai
Full size table

Groundwater from quartzites shows higher concentrations of 222Rn next to that of granitic aquifer. It is known that quartzites have many small fractures and cracks (Kumar et al. 2014). The nearby granites serve as the source of 222Rn to quarzites, as they occur along the movement of water. 222Rn gets enhanced as it passes through nano-pores (Rama and Moore 1984). 222Rn from the rocks through the process of alpha recoil accumulates by dissolution in the interstitial water. Furthermore, these gases travel rapidly through the microfractures to larger fractures and ultimately reach the sampling holes. The activity of 222Rn can be higher in groundwater systems when the fracture width or aperture becomes smaller. Since the formation does not have huge interconnected pore spaces, it does not serve as a good aquifer (Kumar et al. 2014) and hence has a lesser volume of water. It is also inferred that the volume of water present in the aquifer is inversely proportional to the concentration of 222Rn in the formation (Butsaert et al. 1981). The diffusion rate of 222Rn varied according to the nature of fracture. In microfractures, the activity of 222Rn is higher than in larger fractures. But the openness of the interconnected fractures can be identified only if the 222Rn level attains its detection limit when sampled, in hard rock terrains.

The review of 222Rn concentrations in groundwater from granitic terrains show higher values (Table 4). The observably enhanced concentrations of 222Rn in waters of granitic aquifer are due to the moving out of 222Rn gas from the parent rock and dissolving in the surrounding water under geologic pressure; it then tends to get released into the atmosphere at normal atmospheric conditions (Somashekar and Ravikumar 2010). In general, it is documented that larger concentrations of 222Rn are typically associated with granitic rocks that contain elevated concentrations of 238U (typically ten or more ppm). Hall et al. (1987) measured 222Rn concentrations in waters from the granitic regions and inferred to have excess values (3.7 to 104 Bq/L) than that of other formations. Larger than average 222Rn have also been measured in other 238U naturally enriched sites, such as phosphate bearing rocks. Uranium in the hard rock is also reported in the study area which is mainly hosted by magnetite and allanite occurring as independent grains with flaky graphite and also with inclusions of quartz (Pandey and Krishnamurthy 1995).

Table 4 Few studies on 222Rn distribution in different granitic terrain of the world
Full size table

3.2 Correlation

To obtain an overall interrelationship between the various sample parameters, correlation analysis was attempted between different field parameters, i.e., 222Rn, pH, TDS and temperature. In SWM, the 222Rn was negatively correlated with temperature, pH and TDS. In NEM, the 222Rn showed a poorly positive correlation with temperature and a negative correlation with the pH and TDS (Table 5). It is also inferred that 222Rn emanation and the rock waters interaction processes are favored by high temperatures. The 222Rn concentration was found to be negatively correlated (−0.70) with the water temperature in few places in India (Prasad et al. 2009), but in the present study the relationship is site-specific depending on depth of fracture and the type of the aquifer. In general, the temperature study reveals the fact that it is not a controlling factor for the release of 222Rn.

Table 5 correlation matrix of SWM and NEM
Full size table

In SWM, the maximum and minimum temperature was observed in Charnockite rock type (26.5 °C and 36.5 °C, respectively). The lowest 222Rn concentration was related to the highest temperature in SWM. In NEM, the maximum value was observed in fissile hornblende biotite gneissic rocks, whereas the minimum concentration was observed in quartzite. The lower concentration of 222Rn in NEM occurred at the temperature of 31.7 °C. The highest concentration of 222Rn in SWM occurred at the temperature of 31.1 °C, whereas in NEM the respective value was at 30.7 °C (Fig. 4). When the data are studied in total in the correlation matrix, 222Rn content in groundwater has no significant relation to temperature, but when represented graphically with respect to lithology, it can be clearly seen that the samples of charnockite aquifer have a negative correlation to the 222Rn concentration. The residence time of the water play a significant role in determining the temperature (Karmegam et al. 2010) in the Charnockite formation. The figure shows that there is no definite correlation of temperature to the concentration of 222Rn in other formation, which indicates that in certain aquifers and site-specific location this relationship persists.

Fig. 4
figure 4

Bivariate plot of temperature and 222Rn in SWM and NEM

Full size image

In SWM, the highest and lowest pH values were observed in floodplain alluvium. In NEM, the highest pH value was observed in fissile hornblende biotite gneiss and the lowest value in flood plain alluvium. In SWM, the highest 222Rn concentration occurred at pH value of 6.9 and the lowest at pH value of 7.5 (Fig. 5). In NEM, the highest 222Rn concentration was related to a pH value of 7.0 whereas the lowest to a pH value of 7.3.

Fig. 5
figure 5

Bivariate plot of pH and 222Rn in SWM and NEM

Full size image

The highest concentration of TDS was observed in floodplain alluvium in both seasons. The lowest concentration of TDS was observed in quartzite, in both SWM and NEM. The lowest 222Rn concentration was related to a TDS value of 1,089 mg/L in SWM, whereas, the highest observed TDS value was 1,059 mg/L (Fig. 6). In NEM, the lowest 222Rn concentration had a TDS value of 790 mg/L and the highest 222Rn concentration had a TDS value of 1,060 mg/L.

Fig. 6
figure 6

Bivariate plot of TDS and 222Rn in SWM and NEM

Full size image

The correlation was carried out using all samples, which did not provide a clear picture, and hence, an attempt was made to correlate the 222Rn values of the samples to lithology, temperature, TDS and pH. The temperature showed lower R2 values, indicating that the 222Rn has no or lesser relationship to temperature in certain lithologies of the study area. The R2 value of 222Rn and pH showed that good correlation exist between these parameters for the samples collected in quartzite and sedimentary formation (floodplain alluvium). Similarly, good correlation of TDS was observed with 222Rn in the quartzite formation in both seasons (Table 6). Good correlation between 222Rn in granite formation and TDS was observed in NEM season. Though the occurrence of 222Rn was reported to be higher in the granitic formation of the study area and the relation between the occurrences of this isotope to observed field parameter shows that 222Rn collected from the samples of quartzitic terrain had good correlation with pH and temperature.

Table 6 R2 value of 222Rn (Bq/L) with temperature (°C), pH and TDS (μg/L)
Full size table

4 Conclusions

The study on the 222Rn activity concentration in groundwater of this region shows that there is a seasonal variation in higher concentration levels in both seasons. 14 % of samples during NEM have higher value and they are noted in the granitic formations of the study area. These higher values are represented in NE part of the study area. The order of dominance of 222Rn in groundwater is as follows granite > quartzite > charnockite > hornblende biotite gneiss > floodplain alluvium. The higher value in granite is mainly due to the presence of uranium bearing minerals. The source of 222Rn in quartzite is mainly due to the 222Rn enriched groundwater flowing from the granitic terrain, as quartzite formation is along the groundwater flow direction. But since it moves through microfractures and the volume of water is lesser it shows an enhanced 222Rn by representing the second highest value in this aquifer of the study area. It shows less correlation with other observed parameters like TDS, pH and temperature, and negative correlation is exhibited in few samples of charnockites. The seasonal variation may be due to the source, fracture systems, and leaching processes occurring in the granite bed rock where leaching reaches maximum during NEM. The consequences of these processes lead to the high 222Rn concentrations in water extracted from the bedrock. The study on the cations and anions, along with the isotope concentrations can help in identifying the complete processes behind the variations of higher 222Rn concentrations.