Establishing effective sentinels - Setting the baseline for shale gas
Abstract
The UK has a nascent shale gas industry and, unlike the US we have the opportunity to establish structures both physical and regulatory to reassure the public that any impact of a developing shale gas will be .properly licensed, regulated, monitored and, if necessary, mitigated. To assess and indeed demonstrate an impact of any activity, let alone those of shale gas exploitation, it is necessary to show, within a reasonable level of certainty, that the industry has changed a environmental state over and above that which was true without the activity present. The need for demonstrating impact not only means that a baseline needs to be established but that the baseline needs to be robustly established within a statistical and probabilistic framework so that certainty of impact can be demonstrated. A number of technologies have been proposed for monitoring the water quality impacts of shale gas developments, however, to be an effective and robust sentinel of change the parameter should have several properties: it should be a lead indicator and not a lag indicator of change; it should have a high contrast with the normal or background activity; it should show a high specificity for the activity of concern and not be associated with other activities; and it should readily deployed in time and space. By far the greatest difference between the waters arising from a shale gas well pad and surface waters is nothing more than salinity or its associated determinds. The salinity of flowback water and deep formation water can be many times greater than seawater let alone greater than the salinity of most UK surface waters. Therefore, we have built a probabilistic model of the salinity of English surface waters. We have developed a generalised linear model of the existing salinity data available for English surface waters. Generalised linear modelling means that we can use all the existing data, the approach is entirely data driven; it does not require parameterisation; and can include existing factorial and covariate information. The model was developed in a Bayesian hierarchical framework. The model creates a dynamic baseline against which it is possible to assess whether an observation is within that expected for that river under those temporal and hydroclimatic conditions. The model is tested for the Vale of Pickering gasfield.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2017
- Bibcode:
- 2017AGUFM.H53A1429W
- Keywords:
-
- 1803 Anthropogenic effects;
- HYDROLOGY;
- 1829 Groundwater hydrology;
- HYDROLOGY;
- 1831 Groundwater quality;
- HYDROLOGY;
- 1878 Water/energy interactions;
- HYDROLOGY