Application of isotopic tools to understand the hydrological processes and groundwater dynamics of Mianwali area-Punjab, Pakistan

Increasing population and industrialization have put worldwide pressure on water resources, under natural and anthropogenic conditions. The structure, status, and processes of the groundwater system, which can only be acquired through scientific research efforts, are critical aspects of water resource management. In this regard, stable and radioactive isotope data provide essential tools in support of water resources management (Chen et al. 2019). Many of the present-day methodologies in water science have developed with limited data and data processing procedures. Due to the growth and improvement of modern technologies, a new generation of data has been made available to water resource management, which have a much higher spatial and temporal resolution (Peters et al. 2018). An application of this includes tracing the evolution of a water mass from its origin as precipitation, surface water through its recharge processes, and ending at its appearance in an aquifer. Besides this, isotopes used to determine the origin of specific solutes in groundwater. The inventory of isotopes, which has significant implications for water resource management, has grown in recent years (Shi et al. 2017). Some researchers are using models by applying isotope tools such as stable (¹⁶O, ¹⁷O, ¹⁸O, ¹H, ²H) and radioactive (³H) isotopes in water to understand the hydrological processes (Polona & Zoltan 2020). Some researchers applied environmental isotopes along with hydro-chemical data as powerful tools for investigating groundwater flow conditions (Doveri & Mussi 2014; Zhong et al. 2019), including hydrogeological investigations of alluvial formations (Liu & Yamanaka 2012; Yoshioka et al. 2016). A multifaceted analysis using a suite of tracers (multi-tracer approach) can effectively be used to interpret complex groundwater flow conditions, which cannot be observed directly. Multivariate analysis methods, such as principal component analysis and factor analysis, which allow for the dimensionality of multiple hydro-chemical indicators to be reduced and features of groundwater flow conditions to be extracted, have been widely applied in hydrological system analysis (Valder et al. 2012). Understanding of groundwater recharge sources, and flow pathways and rates (dynamics) is key to sustainable management of aquifers including preservation of drinking water security and integrity of the aquifers and surface waters that receive inflows from groundwater. Three main modelling approaches to provide a statistical description of groundwater age (i.e. residence, transit or travel time) are described: (1) lumped parameter models (LPMs); (2) mixing cell models; and (3) direct age models. The mean of which can be combined with a flow path length to estimate a lateral groundwater flow velocity and flux across an interface (such as groundwater recharge or discharge) (Cook & Böhlke 2000). Tracer LPMs (used in present studies) have most commonly been used to estimate mean groundwater ages (i.e. mean residence times), which can be used to estimate fluxes when combined with a flow path length and a representative porosity value. The LPM approach uses a statistical description of the variability of groundwater age, which is represented by the weighting function used in convolution. Stable isotopes can be used in such models to determine the altitude of recharge, flow distances and the hydraulic gradient (Chris & Brian 2014).

Tracers such as isotopes and chlorofluorocarbons (CFCs) for tracing and dating groundwater are important tools in groundwater research and in sustainable management of groundwater resources. Important applications in shallow groundwater include estimation of groundwater recharge and evaluation of the fate of contaminants, because meaningful groundwater ages give the possibility to determine the residence time of groundwater and the dissolved contaminants (Cook et al. 1995; Johnston et al. 1998). Recently the use of CFC gases (CFC-11, CFC-12, and CFC-113) had a revival as age dating tools due to improved sampling and analysis techniques (Busenberg & Plummer 1992). This has led to a strong increase in the research on CFC-gases as tracers and age-dating tools in groundwater often in combination with isotopes (Hinsby et al. 1992; Dunkle et al. 1993; Ekwurzel et al. 1994; Oster et al. 1996).

Pakistan, like many other developing countries, is suffering from huge shortfall of drinking water. Conservation of groundwater resources managed by gathering the critical information about hydrological system of the area.

The study area is an important area as it has a Pakistani huge nuclear power plant complex comprising of five operational nuclear power plants (NPPs). Limited studies about understanding of groundwater dynamics and hydrological process in this area have carried out. Previous hydrological studies in this area have focused on spatio-temporal variations in the groundwater quality, groundwater nitrate dynamics (Akhtar et al. 2019; Lytton et al. 2021), and the groundwater quality mapping by using Electric Resistivity Survey (ERS) (Samreen 2011). Geochemical data of water to identify hydrological processes in the Mianwali area were gathered (Ghaffar & Naveed 2020), but how recharge sources influence the flow of groundwater in the aquifer is still unclear. There is not a single study carried out in the area to investigate the groundwater dynamics and hydrological processes by applying the advance isotopic tools (stable and radiotracer) and groundwater dating tools such as H-3 and CFCs. Therefore, further research is needed to provide information for hydrological processes in the area to enable sustainable use and conservation of groundwater sources.

This study is the first comprehensive research work in the Mianwali area to understand groundwater dynamics and hydrological processes. This hydrological investigation comprises identification of water recharge source, groundwater residence time and groundwater dynamics carried out in the area by applying stable isotopes, radiotracer studies and groundwater dating tools to estimate groundwater recharge sources, groundwater real velocity, filtration velocity and in-situ effective porosity. The analytical and experimental data and findings of this study will help to understand the hydrological system near NPP site and will provide basic information used to predict the fate of radioactive contaminants released into groundwater in case of an accident and/or during normal operations.

The Los Gatos Research (LGR) Liquid Water Isotope Analyzer was used for stable isotope analysis. The LGR Liquid Water Isotope Analyzer provides measurements of δ¹⁸O, and δ²H of water in liquid with matchless performance.

Tritium analysis of water samples performed by Liquid Scintillation Counter (LSC), Quantulus-1220 followed by electrolytic enrichment. The uncertainty in measurement was about 0.8 TU. All the samples distilled on routine basis to remove impurities. About 01 g sodium peroxide added to 250 ml distillate and solution thus prepared transferred to electrolysis cells for electrolytic enrichment. All the 20 cells placed in a specially fabricated refrigerator and connected in series. A total charge of 700 AH was passed to get about 14–15 ml of enriched sample. The refrigerator is set to maintain a temperature of 0 ± 1 °C around the cells. Three spikes were included in each run of electrolysis to determine the degree of enrichment. The average enrichment factor is about 16.

On completion of electrolysis, the electrolyte was neutralized by adding 04 g lead chloride. The samples again distilled to recover the sample water. 8 ml of enriched sample was taken in 20 ml polyethylene vial and 12 ml scintillation cocktail (Ultima Gold LLT) was added to it. The contents of the vial were mixed up thoroughly by vigorous manual shaking. A batch of samples comprising 3 background samples, 3 standard samples, 3 enriched spikes, 3 un-enriched spikes and 17 enriched samples having unknown tritium concentration were placed in the LSC. Unquenched standards of tritium (Packard Instrument) were used to measure and optimize the counting efficiency of LSC. All the samples were counted for a preset time of 50 minutes and the whole batch cycled 10 times. Finally counting data was statistically evaluated and processed by excel template to calculate the tritium activity (TU). This template comprises the calculation of the mean background count rate, net count rate of enriched spike, net mean count rate of standard and net count rate of enriched unknown samples.

Groundwater samples were collected for CFCs during the sampling campaigns. The sampling bottle was purged according to the USGS National Field Manual. The bottle was placed into a 3 liter SS beaker and then inserted the end of the copper tubing from the pump all the way into the bottom of the sampling bottle. The sample was allowed to pass into the bottles and continued to overflow the bottle, while sitting in the outer beaker, until one liter of water was flushed through the bottle. The bottle was tapped to dislodge air bubbles and tightly capped while submerged in water.

The radon concentration of water samples collected around the NPPs and measured using RAD-7 to assess groundwater discharge phenomenon in the area. Twenty water samples were collected in 250-ml vials from different sources i.e. boreholes, tube wells, canals, hand pumps as well as the outfalls of power plants. CFC concentrations in groundwater samples were carried out using GC-EC equipped with a dedicated CFC sample preparation system. The sampling bottle purged according to the USGS national field manual to avoid any air bubble entrapped in sample.

About 280 mCi iodine-131 (¹³¹I) in the form of sodium iodide solution was injected in the injection well on 3rd November, 2019 at 16:45 hours. Background radiation in air and water measured well before the injection of radiotracer (¹³¹I). Monitoring of radioactivity in all the detection wells was initiated just after the injection of the tracer and continued until the count rate reached in the limit of three times the standard deviation (σ) in the detection wells. An additional borehole for point dilution technique (PDT) also constructed at 500 cm distance from central borehole in general groundwater flow direction in the rear (Northeast) direction of the setup. A rheometer probe equipped with multiple accessories such as control unit, data logger, computer, tracer injection system, homogenization system and a radiation detection system employed to determine the filtration velocity at specified location. The filtration velocity depends on two variables; the temporal response of the detector and the extent of distortion of the groundwater flow field caused by the presence of the well (Tazioli 1973; Udong 2007).

Soil samples from each borehole collected for grain size analysis. For estimation of hydraulic conductivity approach of Salarashayeri was adopted which incorporate entire particle distribution. (Salarashayeri & Siosemarde 2012).

The control points to interpret the isotopic data of any area in terms of recharge mechanism are meteoric water line, rain index and river system index in the area under investigation. Extensive samplings over a long period are required to establish these control points or indices. The weighted average ‘δ¹⁸O: −3.62 ‰, δ²H: −19.14‰’ of Chashma rain data was taken as rain index while the average ‘δ¹⁸O: −11.38 ‰, δ²H: −75.31‰’ monthly isotopic data of same period of river system in the study area was taken as isotopic river index. The local meteoric water line (LMWL) shown in Figure 4 was established during 2005–2007 using actual slope and intercept of all rain samples (non-weighted) 7.58 and 6.97 respectively was taken as reference. The average isotopic signatures ‘δ¹⁸O: −11.9 ‰, δ²H: −83.4‰’ of the river Indus at Basham, 300 km upstream at altitude of 625 m were also included for reference to compare the isotopic values with some higher altitude location.

Seasonal effect reveals that during pre-monsoon, the isotopes are enriched in comparison to the post-monsoon. This attributed to the enrichment of heavy isotopes in pre-monsoon season (Figure 4(a)) by two key factors: temperature effect and the amount effect. Temperature effect cause isotope fractionation, which may lead to the enrichment of surface water with respect to heavier isotopes (δ¹⁸O & δ²H) during pre-monsoon season. Secondly, the amount of effect does not contribute considerably for modification of isotopic values, owing to few rain events in pre-monsoon season. Conversely, depletion of isotopes (δ¹⁸O and δ²H) during post-monsoon season (Figure 4(b)) is primarily due to heavy rains and high tides condition in river Indus.

Additionally, depletion of isotopic values was ascribed to isotopically depleted snow/glacier melt coming from the high mountains of the Himalayas contributing to river Indus, a major recharge source for the study area. Similar trends for the transition period (Figure 4(c)) were observed, as the temperature is substantially low in the winter season, which results in higher fractionation of isotopes in both surface water and groundwater. Therefore, it concluded that river and canal water is the main recharge source for the study area.

During the dry period, the river Indus near the study area behaves like a gaining stream. Only for a short period when the river flows at full swing, it contributes to nearby groundwater. Correlation of δ¹⁸O and δ²H and spatial distribution of δ¹⁸O (Figure 5) clearly indicate this phenomenon during different seasons.

The mean residence time (MRT) of groundwater is a critical hydrological parameter required for the roundwater dynamics studies. Tritium, CFCs and noble gas isotope techniques were applied for determination of MRT. CFCs analysis of collected groundwater samples reveals that CFC-ll range from 0.02 to 12.96 pmol/kg, CFC-12 varies from 0.13 to 2.63 pmol/kg, and CFC-113 concentrations spans from 0.02 to 6.35 pmol/kg.

Tritium concentration in shallow groundwater during various seasons exhibits diverse range, extends from 2–20 TU. During pre-monsoon, post-monsoon and transition period, 72% (n = 29), 69% (n = 25) and 75% (n = 30) of the shallow groundwater samples range from 8–14 TU, respectively. Likewise, 17% (n = 05), 25% (n = 09) and 20% (n = 08) of pre-monsoon, post-monsoon, and transition period, respectively exhibits tritium concentration ranges from 2–6 TU. Whereas, only a few shallow groundwater lie in the range of 16–20 TU, with 10% (n = 03) of pre-monsoon, 8% (n = 03) of post-monsoon and 5% (n = 02) of transition period.

Comparing histograms with the spatial distribution diagrams (Figure 6) it is observable that most the samples having similarity of tritium concentration with the surface waters lie near the water channels indicate quick recharge while a smaller population having lower tritium concentration are relatively farther from surface water resources have higher residence time. The stable isotope values and the groundwater direction depicted by water table contours also support this finding. Few samples having extraordinary higher tritium concentration may be attributed to the river Indus.

A frequency histogram shows that radon concentration in surface water samples ranges from 50 to 100 pCi/L, while a broad range exist in groundwater samples extends from 100 to 350 pCi/L. The spatial plot of radon activity reveals that surface water in the form of canals and lake has influential impact on groundwater recharge of the study area, which is manifest by low radon concentrations in groundwater samples located near the surface water sources and approximately Chashma NPPs. However, samples located in the south-west quadrant of the spatial map indicate groundwater discharge to the river Indus, which characterized by relatively high radon values near the river.

Average concentration of radon in samples is given in Table 2.

The tracer was first appeared in detection well (Figure 3) located in the southwest of injection well on November 12, 2019 at 16:30 hours. The breakthrough time calculated as 223:8 hours (9.32 days). The monitoring of radiotracer continued in and all other detection wells in the inner semicircle. The peak value was observed in on November 19, 2019 at 22:30 hours indicating transit time for maximum activity of 395.8 hours (16.5 days). The groundwater velocity for waterfront was calculated as 21.65 cm/day. After the breakthrough in all the wells in second semicircle also continuously monitored for tracer presence. Finally, the tracer appeared in detection well in the outer circle after 527.8 hours (21.99 days) of injection on November 25, 2019 at 10:30 hours, indicating breakthrough velocity of 22.74 cm/day in the same direction as determined by appearance of tracer in ⁠. The appearance of tracer in any other boreholes in second semicircle was not detected as the count rate remained within and three times the background count rate (Table 3).

This value is in good agreement with the porosity of medium to fine sand.

I = hydraulic gradient = (h₁ − h₂)/d = 1.01/1,506 = 0.000761

h₁ = water table in BH-1 = 190.12 m

h₂ = water table in BH-2 = 189.11 m

d = distance (BH-1 and BH-2) = 1,506 m

Filtration velocity = 10.94 cm/day = 0.1094 m/day

Hydraulic conductivity = K = 0.1094/0.000761 = 163 m/day

Thickness of alluvium = >300 m (NESPAK 1992)

Transmissivity (T) = K × Thickness of aquifer = 163 × 300 = 48,900 m²/day

The hydraulic conductivity estimated by both approaches is in good agreement.

The study was conducted to investigate hydrological processes and groundwater dynamics including groundwater recharge and discharge patterns, flow paths, surface water-groundwater interaction and velocities using conventional, common isotopic techniques and emerging techniques.

Groundwater age investigated by the Tracer LPM and CFCs suggests the groundwater age in the range of 14–59 years. Tritium concentration reveals that most of the groundwater samples have fresh recharge and classified as young water. Although, a few samples located near the river Indus exhibits relatively high tritium (15–19 TU), suggest major contribution of glacier melt with high tritium from Himalayas.

The water table contours and multiwell tracer experiment indicated that the groundwater generally flows towards river in south-west direction. The flow pattern in the aquifer system therefore suggested that the southwestern region of the study area is susceptible to groundwater contamination that may release from NPPs in case of any leakage.

A combination of point dilution and multiwall radioactive tracer technique suggest that the real and filtration velocity of groundwater in the area are 22.2 cm/day and 10.94 cm/day, respectively, whereas the estimated in-situ effective porosity of the aquifer is 49.3%, which is in good agreement with the porosity of fine to medium sand. The hydraulic conductivity of the aquifer varies from 48 to 278 m/day (avg. 139 m/day) with increasing trend from northeast to south-west direction in the study area. The NPPs operation did not impose any destructive impact on groundwater regime as there is no significant change in groundwater flow patterns in study site except a slight lowering of the water table, which may be due to decline in river flow, rainfall patterns and/or increased water abstraction for regional agriculture activities

Author is thankful to competent authorities of National Engineering Services Pakistan (NESPAK) for sharing the geographical data of site and CHASNUPP (Chashma Nuclear Power Plant: Pakistan Atomic Energy Commission) for providing their technical and logistic support in the area.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.