Investigation of sulphate sulphur isotope variations in the Skerne Magnesian Limestone water body

This report presents the results of a sulphur isotope investigation undertaken in the Skerne catchment, located in County Durham, north of Darlington, to investigate the source of groundwater sulphate in the Magnesian Limestone Aquifer. Groundwater and surface waters in the catchment are at risk from a number of current and historic anthropogenic activities. Sulphate is the biggest risk to the public water supplies; as there is currently no cost-effective treatment available and it could render supplies unusable. The elevated sulphate could be both naturally occurring, due to the presence of gypsum or anhydrite bands in the Magnesian Limestone, or it could be due to abandoned coal mine water, or even saline intrusion pollution. Because of the large difference in the sulphate sulphur isotope composition expected between “marine sulphate”, including sulphate derived from marine evaporites, and “non-marine sulphate” derived from the oxidation of sulphide in the coal seams and mine workings, sulphur isotopes were considered promising tracers to discern mine water sources from natural Permian evaporite sources of sulphate. A survey was carried out at 28 sites where groundwater was sampled in July 2018 from boreholes in the Magnesian Limestone Aquifer and in the Coal Measures, following a pilot study comprising 7 boreholes in July 2017. A small number of surface waters, hyporheic zone waters, springs, and soil leachates, sampled during 2017-2018, were also analysed for sulphur isotopes to complement the borehole data. This has allowed the characterisation of the sulphur isotope composition of potential sources of dissolved sulphate. Most of the Magnesian Limestone aquifer groundwaters cluster close to the Global Meteoric Water Line (GMWL) on the dual water δ 18O and δ2H graph with no evidence of mixing with Narich coal mine water, the latter being more depleted in 18O and 2H; there is a small number of boreholes immediately in proximity of the coal seam boreholes, clearly showing signs of water mixing. With higher δ18O and δ2H than the main Magnesian Limestone group, and slightly offset from the GMWL, is also a small group of Magnesian Limestone boreholes. Repeated sampling would better discern the different recharge paths suggested by this single sampling event in July 2018. Groundwaters associated with the worked and unworked coal seam boreholes in this study are of two water types: sodium sulphate (Na–SO4) and sodium bicarbonate (Na–HCO3) waters, variably enriched in dissolved sulphate. Two δ 34S measurements of the dissolved sulphate in the Na–SO4 coal seam boreholes are +13.1‰ and +23.4‰. The lack of the more typical 34S-depleted sulphate derived from the oxidation of pyrite is hence apparent. A similar range of high sulphate δ 34S values has been described in recent studies, and attributed to deep coal mine systems. From a review of published δ34S values for marine evaporites, groundwaters containing sulphate solely derived from the dissolution of Permian marine evaporites are characterised by 34Senriched sulphate (δ34S values range from +8.2 to +11.1‰). There is, therefore, less of a contrasting isotope signature between potential “evaporite” and “coal mine water” end-members. For example, one sample of coal mine water with δ 34S values of +13.1‰ is not too dissimilar to the average Permian evaporite sulphate with δ 34S value of around +10‰. This makes discrimination of the dissolved sulphate sources based on sulphur isotope less certain, especially at low sulphate concentrations. To help the data interpretation, we have modelled the sulphate and sulphur isotope compositions of mixtures of hypothetical end-members and used the evidence from these simulations to constrain possible groundwater contributions and mixing. In particular we simulate how the HARDWICK HALL borehole, representing the Magnesian Limestone aquifer background, with a sulphate concentration of 89 mg/l, and a δ34S value of +1.0‰, evolves during mixing with the following end-members: i) the coal mine waters in this study, ii) a Permian evaporite source, iii) seawater and iv) acid mine drainage. A summary of the data interpretation based on the above modelling is as follows. Over the mine plume area, inputs of coal mine water-derived sulphate are significant in at least one Magnesian Limestone borehole, and detectable in others, supported by the water isotope δ 18O and δ2H data, indicating for these samples water mixing between the coal mine water and the Magnesian Limestone aquifer. Among the Magnesian Limestone boreholes, where gypsum or anhydrite were noted in the borehole logs, only DALTON PIERCY NO 3 and NO 6 boreholes have high sulphate concentrations and display constant δ 34S values of +10.2‰. Given how close this value is to the Permian evaporites’ δ34S values, it could be plausibly explained by a gypsum dissolution source, although a “coal mine water” contribution with a δ34S signature of +13‰ cannot be totally excluded, as shown by the mixing curves. Many of the Magnesian Limestone boreholes with a sulphate concentration around 100 mg/l (range 85–130 mg/l) are characterised instead by a low δ 34S range (-0.7 to +7.2‰). For most of these low sulphate Magnesian Limestone boreholes, uncertainties in discriminating the source of sulphate are higher. The contribution of sulphate from seawater is difficult to discern in the present data for the saline waters of HART RESERVOIR and HARTLEPOOL IND ESTATE REPLACEMENT boreholes, with similar δ34S values of +21.1‰ and +27‰, as they fall far away from the Seawater–Magnesian Limestone mixing line. Many samples fall far outside of these mixing envelopes, suggesting non-conservative behaviour of the sulphate. The very high δ34S and low sulphate concentrations can be interpreted as a possible sign of reduction of sulphates and enrichment in the heavier 34S isotope of the residual (low concentration) sulphate. Additional samples obtained during this study include: i) A spring in the Ford Formation from AYCLIFFE QUARRY to the south east of Aycliffe Village which provides an additional background sample characterised for sulphur isotopes. The water has a SO4 of 69 mg/l and a δ34S value of +2.3‰ and well resembles the composition of HARDWICK HALL borehole. ii) A Mg–SO4 spring, sampled in Woodham Burn and described in previous studies for its impact on the surface water quality because of its high sulphate concentrations of ~800 mg/l. It has a stable δ34S value of ~ +5.5‰. iii) a surface water impacted by mine water inflow with a Mg–SO4 composition, and a δ34S value of +6.9‰. The δ34S value of +5.5‰ of the Mg-SO4 spring at Woodham Burn points to a contribution of low δ 34S-sulphate, as expected from the oxidation of pyrite. These data support the mechanism, hypothesised in Palumbo-Roe et al. (2020) to account for the spring composition, of dissolution of dolomite in the presence of acidic water, where the source of acidity comes from coal mine water due to the oxidation of pyrite. There is a much narrower and lower range of δ 34S in surface water compared to the groundwater samples. With most δ 34S values less than +7‰, none of the high values measured in the boreholes were noted in the surface water, hyporheic zone or soil leachate samples, except for two samples in the hyporheic zone of Woodham Burn with δ 34S +36.3‰ and +13.4‰, values taken as further evidence of the sulphate reduction during the 2018 summer indicated by the hydrochemistry. Recommendations for future work, building upon these findings, are suggested.