OR/18/020 Water monitoring
| R S Ward1, G Allen2, B J Baptie1, L Bateson1, R A Bell1, A S Butcher1, Z Daraktchieva3, R Dunmore4, R E Fisher5, A Horleston6, C H Howarth3, D G Jones1, C J Jordan1, M Kendall6, A Lewis4, D Lowry5, C A Miller3, C J Milne1, A Novellino1, J Pitt2, R M Purvis4, P L Smedley1 and J M Wasikiewicz3. 2018. Preliminary assessment of the environmental baseline in the Fylde, Lancashire. British Geological Survey Internal Report, OR/18/020. |
Water monitoring network
A monitoring network of groundwater and surface water sites has been established across the Fylde, including areas proximal to PNR and Roseacre Wood, in order to evaluate baseline water quality and allow for evaluating any future changes in conditions should shale-gas operations take place. The constructed hydrogeological model (Figure 9) was used to inform the selection of sites suitable for monitoring and further investigation. Sites for surface water and groundwater monitoring were chosen within approximately 10 km of the proposed shale-gas exploration sites, with effective barriers imposed by the Rivers Wyre and Ribble and the Fylde west coast. An inventory of groundwater sites was collated from information in available well databases augmented by field visits. Groundwater sources are mainly boreholes owned by local small businesses and households, although some public water supplies exist in the area of interest. First-order stream sites were sought for surface-water monitoring to limit the size of the influencing catchment area.
A shortlist of suitable sites was drawn up on the combined basis of site access, source condition (status and representativeness) and site safety. Investigations established that availability of suitable groundwater sites was limited in the area of interest and so all groundwater sites deemed to be suitable, accessible and representative for sampling have been included in the resulting water monitoring network. The established monitoring network comprises 19 groundwater sources and 11 surface waters, the latter all streams (Figure 10 and Table 1). Groundwater sources include 9 sites from the Superficial (Quaternary) aquifer and 10 sites from the Sherwood Sandstone to the east of the Woodsfold Fault. All groundwater sites in the initial water monitoring network are owned by third parties.
BGS boreholes
As the water monitoring network relied on pre-existing groundwater sources which were in some cases not located ideally, a decision was made to purpose-drill a number of water boreholes closer to the proposed Cuadrilla sites for the purposes of detailed investigation and monitoring, with a view to final reinstatement for private water supply pending further investigations. These were located within ca.1 km of the PNR or Roseacre Wood proposed sites (Figure 10 and Table 1). Shallow boreholes completed within the Quaternary (Superficial) aquifer were installed in pairs, one ‘deep’ and one ‘shallow’ (though all <50 m below ground level. Completions were according to site conditions, favouring screens within sandy horizons to provide best-available groundwater yields.
A single borehole was also completed within the Sherwood Sandstone at 500 m depth, at some 2 km distance west of the Roseacre Wood site. This borehole was cased off from surface to 350 m depth with an open-hole section, wholly within the Sherwood Sandstone, thereafter.
| Site Number | Site Type | Monitoring Type |
| Site 1–24 | Groundwater | Initial network |
| Site 25–35 | Surface water | Initial network |
| Site 36–48 | Groundwater | BGS borehole |
| Site 49–53 | Groundwater | Cuadrilla borehole |
Sampling and analysis
Sampling of groundwater and surface water in the monitoring network began in February 2015. Sites have been sampled quarterly in accordance with BS ISO 5667-11 (2009). Groundwater sources were purged before sampling and where possible, flow cells were set up inline to monitor redox characteristics. These, along with pH, temperature and specific electrical conductance (SEC), were monitored until stable readings were obtained, at which point, total alkalinity was determined onsite by titration and samples were collected for subsequent laboratory analysis. Samples for analysis of major cations, anions and trace metals were filtered (0.22 µm) and collected in factory-new LDPE bottles, pre-rinsed with filtered sample. Aliquots for determination of major-cation and trace-metal concentrations were acidified (1% v/v) with pure HNO3 (and subsequently in the laboratory with 0.5% v/v pure HCl). Samples for determination of non-purgeable organic carbon (NPOC) were filtered (0.45 µm) using Ag-impregnated filters and collected in acid-washed glass bottles. Samples for dissolved gases were collected inline at pump pressure in steel bombs. Samples were also collected periodically for organic compounds (total petroleum hydrocarbons, TPH; polycyclic aromatic hydrocarbons, PAH; volatile (and semi-volatile) organic compounds, VOC/SVOC).
Major cations and trace elements were analysed by ICP-MS and anions by ion chromatography. The analysing laboratory holds ISO 17025:2005 certification for analysis of environmental materials (including water) using these methods. The laboratory operates an AQC regime including use of International Standard Reference Materials and operates an Aquacheck proficiency testing scheme.
Samples for TPH were solvent-extracted and analysed using a modified USEPA 8015B method. Solvent-extractable petroleum hydrocarbons with carbon banding in the range C8-C40 were determined by GC-FID. Gasoline Range Organics (GRO) in the carbon chain range of C4-12 were determined by headspace GC-FID. Samples for SVOC were solvent-extracted and analysed using a modified USEPA 8270 method, by GC-MS. The testing laboratory holds ISO 17025 certification for TPH CWG and SVOC in surface water. VOC determinations were by direct aqueous injection using purge-and-trap GC-MS and PAH determination by solvent extraction, GC-MS. The testing laboratory for VOC, PAH determinations holds ISO 17025:2005 certification for these methods in water.
In addition, a non-accredited (semi-quantitative) analysis of water samples by target-based screening has also been determined using GC-MS and LC-MS.
Real-time monitoring
Real-time monitoring for a restricted number of water-quality parameters has been carried out within six of the shallow BGS boreholes located around the PNR and Roseacre Wood sites. Monitoring includes water temperature, water level, pH and electrical conductivity. These are monitored hourly to obtain high-resolution data for characterisation of the groundwater baseline conditions and to act as early-warning indicators of any future environmental change. Data are telemetered to BGS and displayed on the BGS website (www.bgs.ac.uk/lancashire).
Water quality results
Superficial aquifer
Water-quality data for groundwater from the Quaternary (Superficial) aquifer indicate an overwhelmingly anoxic condition (lacking dissolved oxygen), with pH values near neutral. Dominant ions are Ca, Na and HCO3, though SO4 concentrations are relatively high in some samples (Figure 11). Most have low conductivity although some groundwater samples have values more than 1000 µS/cm (Figure 11).
As a result of the anoxic condition of the groundwater, concentrations of nitrate (NO3) are low and concentrations of dissolved iron (Fe) and manganese (Mn) are high (Figure 12). Concentrations of arsenic (As) and ammonium (NH4) are relatively high in some. Methane (CH4) is also often detected, though rarely at high concentrations. The site in the Superficial aquifer with the highest observed concentrations of CH4 (of the order of 3 mg/L) has a depleted δ13CCH4 isotopic signature (-73.1 ‰ VPDB) consistent with a biogenic methane origin, possibly linked to degradation of peat within the superficial deposits present at the site.
Monitoring of groundwater in the Superficial aquifer (Figure 13) has shown that the chemical characteristics have been broadly consistent over time, although variations are greater at some sites (e.g. Site 3) than others. Monitoring at Site 4 ceased in Summer 2016 due to the installation of water-treatment equipment at the site, negating access to raw groundwater for sampling.
Monitoring of groundwater from the BGS boreholes in the Superficial aquifer (Figure 14) shows a greater heterogeneity of water types. Two sites show relatively high concentrations of Ca and SO4 (greater than 500 mg/L and 1500 g/L respectively); one of these shows high concentrations of Na and Cl (ca.1400 mg/L) relative to samples from the other BGS sites. Variability is likely due to spatial variability in mineral-dissolution and ion-exchange reactions.
Spatial variation in dissolved methane concentration in the Superficial groundwater is sporadic (Figure 15). One of the samples with the highest observed concentration occurs in the south-western part of the Fylde. This roughly coincides with occurrences of peat deposits in the shallow Quaternary sediments (Figure 3). One of the Cuadrilla boreholes on the PNR well pad also has groundwater with comparatively high CH4 concentrations.
Activities of dissolved radon gas are low in the groundwaters, typically <10 Bq/L (Figure 16). This is consistent with clay-dominated superficial deposits in the UK, and with the Public Health England/BGS map of radon potential in the Fylde: www.ukradon.org/information/ukmaps
Sherwood Sandstone aquifer
Groundwater from the samples in the Sherwood Sandstone aquifer in the eastern part of the study area has a much less variable chemical composition. Here, Ca and HCO3 are the dominant ions, pH is neutral to slightly alkaline, and all samples have low SEC values (<1000 µS/cm; Figure 11). The groundwater in the Sherwood Sandstone is also anoxic because of the confining condition imposed on the aquifer by the presence of overlying Quaternary Superficial deposits. Groundwater in the Sherwood Sandstone has correspondingly low concentrations of NO3 and high concentrations of Fe and Mn (Figure 12); NH4 concentrations are relatively high in some samples. Methane is almost invariably detectable, though concentrations are low and in the µg/L range (Figure 12). Temporal variability has been more limited over the period of monitoring than for the Superficial aquifer (Figure 14). Spatial variations are sporadic, although few data exist to make a detailed assessment (Figure 15).
Activities of dissolved radon in the Sherwood Sandstone groundwater typically have a larger range than those in groundwater from the Superficial aquifer, although values are still not high. Activities up to 15 Bq/L were recorded in one representative sampling round (Figure 16). These values are consistently below the national parametric value for Rn in drinking water of 100 Bq/L.
Stream water
Stream compositions are also dominated by Ca and HCO3 and low SEC values (<1000 µS/cm) are indicated (Figure 11), although relatively high concentrations of Cl and SO4 are observed infrequently (Figure 18). The greater variability in observed compositions over time could be due to varied rainfall and contributions from runoff. Methane has not been measured in the stream water because rapid degassing is likely and concentrations are expected to be low. Figure 18 indicates some high (and variable) concentrations of NO3, especially in the earliest phases of monitoring. This is likely due to inputs from surface pollutants (agricultural and/or domestic discharges).
Organic compounds
In samples from the monitoring network monitored over the period February 2015 to May 2017, concentrations of organic compounds measured by quantitative methods are low, almost invariably being non-detectable.
Semi-quantitative determinations of compounds from a representative sampling round are shown in Figure 19. The presence of some pesticides is indicated, along with components of plastics (including Bisphenol A, BPA) and a number of perfluorinated compounds. Concentrations of these are also low however, almost all <1 µg/L. The only exceptions are benzenesulphonamide, DEHP and BPA. The low concentrations of PAH (<0.1 µg/L, Figure 19) support the non-detects determined for PAH by quantitative methods (e.g. fluorene and pyrene).
Deep Sherwood Sandstone groundwater
Sampling of groundwater from the deep (500 m) borehole in the Sherwood Sandstone close to Roseacre Wood has shown that the groundwater in the borehole is saline and salinity is stratified with depth (Table 2). The dominant ions in seawater are Na and Cl (ca.10 500 and 19 000 mg/L respectively). The concentrations in the borehole water show that salinity values are more concentrated than solutes in seawater, in the case of the first analysis from a sample at 500 m depth, around fivefold more. Concentrations in the second round of sampling at 360 m depth increased slightly relative to the first analysis.
The salinity values observed indicate that the groundwater quality is notably different from that in the Sherwood Sandstone to the east of the Woodsfold Fault and suggest a lack of any significant hydraulic connection between the two. The high salinity also suggests a lack of significant groundwater flow below the Mercia Mudstone Group. The high salinity, albeit from only one borehole source, suggests that the deep Sherwood Sandstone in the Fylde (west of the Woodsfold Fault) is unlikely to be suitable for drinking water.
| Sample depth (m) | Sample date | Ca | Na | Cl | SO4 |
| 360 | 12-05-16 | 448 | 15 200 | 24 000 | 1,530 |
| 500 | 12-05-16 | 1,120 | 56 100 | 92 000 | 3,100 |
| 360 | 08-05-17 | 795 | 41 100 | 62 700 | 2,150 |
| 430 | 08-05-17 | 835 | 43 000 | 67 100 | 2,240 |
Real-time monitoring
Sondes have been inserted into a sub-set of the new BGS boreholes in the Fylde, monitoring pH, SEC, water temperature and water level hourly, with data transfer to BGS by telemetry. Five sondes (EBM 1–5) have been monitoring since early 2016, a further one (EBM 12) since May 2017. Results of the monitoring are shown in Figure 20. The vertical spikes in data represent the short intervals when sondes were removed from the boreholes for maintenance or calibration. The plots demonstrate the instrument drift in pH values over time and the need for frequent recalibration. SEC shows broad consistency over time in each borehole. Water temperature varies most in EBM 2, consistent with it being the shallowest deployed sonde and hence reflecting response to surface air temperature.
Summary
Collated water-quality data for water from the monitoring network, BGS boreholes and sondes, indicate some distinct compositional differences between groundwater from the Superficial aquifer, from the Sherwood Sandstone and from streams. Groundwater from the Superficial aquifer is typically of Ca-Na-HCO3 type, pH-neutral, and almost universally anoxic. Compositions are consistent with interaction with clay minerals in the aquifer. Concentrations of dissolved methane are typically low (µg/L) but up to some 2–3 mg/L at a small number of locations. Occurrence of methane is consistent with derivation by natural reactions with organic material, e.g. peat, in the Superficial deposits. Activities of dissolved Rn are low. Monitoring shows much consistency over time.
Groundwater from the Sherwood Sandstone aquifer to the east of the Woodsfold Fault is more typically a Ca-HCO3 water, also pH-neutral and also anoxic because of the continuous layer of Superficial deposits overlying the Sandstone aquifer. Concentrations of dissolved methane are universally low (µg/L range). Activities of dissolved Rn are higher than in the Superficial groundwater but still low (<100 Bq/L). Monitoring also shows some consistency in compositions over time.
Data for one deep borehole (500 m) into the Sherwood Sandstone aquifer below the Fylde (Roseacre Wood area), indicate that the groundwater below the thick Mercia Mudstone (ca.350 m) is highly saline, increasing with depth, and is unsuitable for water supply.
Streamwaters show greater variation in compositions over time than the groundwaters, as expected as a response to varying rainfall and surface runoff.
Concentrations of analysed organic compounds (VOCs, SVOCs, TPH, PAHs) are usually low in the groundwater and stream samples investigated, almost all being repeatedly non-detectable. Semi-quantitative analysis by GC-MS and LC-MS has detected some occurrences of organic contaminants, including pesticides, plasticisers, and perfluorinated compounds. These are at low concentrations, usually much less than 1 µg/L, but nonetheless, indicate effects of human activity on shallow groundwater and streamwater.
The water monitoring data collated over the 2015–2017 sampling period are facilitating the detailed characterisation of the baseline condition of the waters in the Fylde area. This work will be completed once a full set of baseline data has been compiled.