Field-scale evaluation of collection methods for dissolved methane samples in groundwaters

Bowes, M.J.; Rivett, M.O.; Bell, R.A.. 2022 Field-scale evaluation of collection methods for dissolved methane samples in groundwaters. Nottingham, UK, British Geological Survey, 58pp. (OR/22/035) (Unpublished)

Before downloading, please read NORA policies.

Download (6MB) | Preview


This report presents the findings of a jointly funded project by the British Geological Survey (BGS) and Environment Agency (EA project SC210014) that addresses some of the research needs identified in the EA project SC190007 “Methods for sampling and analysing methane in groundwater: a review of current research and best practice”. Primary field sampling allowed comparison of sample collection techniques for dissolved methane in groundwaters, to provide a field evidence base to help establish good practice guidelines (or protocols). This included evaluation of purging protocols, groundwater sampling devices, surface collection protocols, and the influence of methane concentration. The field study used two boreholes previously shown to have groundwater of contrasting low methane concentration (~1mg/l) ‘Site A’, and high methane concentration (~25 mg/l) ‘Site B’ in close proximity in the Vale of Pickering. The boreholes were previously used for hydrochemical baseline monitoring and were similar in construction and hydrogeological setting, each having shallow (~ 1 m depth) water table and a conventional 3-m long well screen sampling a fractured Kimmeridge Clay unit with methane naturally present from elevated organic matter contents. A stage 1 zero-purge passive sampling device was used to obtain initial samples, followed by a low-flow methodology with parallel use of submersible, bladder and peristaltic pumped samples, with stage 2 sampled after purging a single screen volume, and stage 3 sampled after purging to hydrochemical parameter stabilisation over 5.7 – 7.5 pumped screen volumes. Finally, a post-purge stage 4 sample was obtained with the same passive sampling device. Sample collection protocols tested were open (direct fill vial), semi-closed inverted (submerged-inverted vial), semi-closed upright (submerged-upright vial) and closed (double valve cylinder closed to atmosphere). All samples were obtained in triplicate from each pump during stages 2 and 3, but in stages 1 and 4 only open samples were possible from the passive sampling device. Data interpretation was supported by logged hydrochemical borehole groundwater depth profiles before and after the sampling programme, and by the historical methane baseline record. Methane concentrations measured at Site A are challenging to interpret: they are highly sensitive to purging volume, with a decrease in mean concentration of 72% over the purging stages. This, and the time required to obtain multiple samples, obscured specific sensitivity of methane concentration to pump and sample collection protocol variables at Site A. Although the differences in concentrations seen between pumps and between collection protocol are not statistically significant, the high variability in Site A data overall, 52-117% relative standard deviation (RSD), mean these data are generally not useful for developing good practice proposals. Site B, with high methane concentration, provided more consistent data that allowed meaningful comparisons of methane sensitivity between purging volume, pump type and collection methods with findings that are generally consistent with existing literature. Methane concentrations had a lower sensitivity to purging than at Site A (21% mean concentration declines with ~30 % RSD). Most of the conclusions made from Site B data can reasonably be expected to also apply to sites with lower concentrations. In isolation, pump selection - bladder, submersible or peristaltic pump - makes little difference to sampled methane concentrations. The HydasleeveTM passive sampler consistently produced the lowest concentrations, but is very likely to have sampled different water in the borehole to that sampled mid-screen by the active pumps. However, bladder and peristaltic pump closed samples yield higher concentrations, which is attributed to their capacity to provide increased pressure at low flow, thereby reducing degassing potential. The bladder pump is preferred for this use, due to its lower closed sample concentration variability, which is attributed to the controllability of the bladder pump pressure. The peristaltic pump was tested under favourable shallow water table conditions, and a negative concentration bias that is generally expected from suction pressure was not evident, but this is acknowledged as a concern, especially for deeper water tables, where its use may need more caution. Closed sampling at Site B consistently yielded the highest methane concentrations across all pumps with lowest variability, and is the recommended sample collection protocol. Commercial availability of closed sample analysis is, however, limited. The semi-closed (inverted and upright) methods yielded intermediate concentrations between closed and passive samples, with inverted methods generally giving higher concentrations than semi-closed. When using the inverted protocol, exsolving gas headspace accumulation leads to uncertainties in concentration measurements, meaning that the semi-closed upright protocol is often preferred, especially where degassing is evident or expected although results in this study do not directly support this. Open samples consistently produced the lowest concentrations and should only be used where other protocols are impractical, e.g. sampling from a non-pumped collection device. Switching protocols from open sampling to upright sampling requires minimal investment, and is likely to produce more robust concentration data and/or reduced variability, although results from this study do not provide undisputable evidence of this. The observed sensitivities of measured methane concentrations to different parts of the sampling methodology underline the central importance of using an identical protocol with specific pumps, similar deployments, identical purging volumes or stabilisation criteria, and specific sample collection protocol, in order to generate robust temporal records. Reliable cross comparison of data produced by different organisations requires sampling protocols to be as identical as possible. Any protocol change should be done using an overlap period in which both old and new protocols are used simultaneously, to prevent a sampling-related step change in data. This study also indicates that extended purging of any borehole leads to lower concentration samples over time, which critically has the potential to underestimate methane risk. Combining the findings of this study and wider literature reviewed under EA project SC190007, a concise ‘lookup’ sheet is presented as a non-prescriptive aid to assist practitioners in ‘Groundwater methane sampling protocol development’. It covers: site selection, pump/sampler selection/deployment, supporting reconnaissance measurements, sample collection and protocol, data management and wider use. Finally, outstanding field research needs are indicated. The foremost of these is comparative field testing of down-hole devices for obtaining closed system samples at in-situ groundwater pressure.

Item Type: Publication - Report
Funders/Sponsors: British Geological Survey, Environment Agency
Additional Information. Not used in RCUK Gateway to Research.: This item has been internally reviewed, but not externally peer-reviewed.
Additional Keywords: Groundwater, GroundwaterBGS
Date made live: 07 Mar 2023 13:32 +0 (UTC)

Actions (login required)

View Item View Item

Document Downloads

Downloads for past 30 days

Downloads per month over past year

More statistics for this item...