Shale gas characterization and production forecasting
: a case study on UK Bowland Shale and the US Barnett Shale

  • Michael Kenomore

Student thesis: Doctoral Thesis

Abstract

Based on the recent reports from Oil and Gas Authority (OGA), a declining trend of hydrocarbon production capability is expected for UK, in upcoming years. In order to keep a relative self-sufficiency in terms of oil and gas supplies, access for new hydrocarbon resources seems necessary.
The giant unconventional reservoirs, such as Bowland shale gas, have been recently discovered in the Central Britain and there is a big potential for them to be the future gas sources in UK. However, the development of shale gas reservoirs is still challenging, there are several shale gas reservoirs around the world developed successfully, which provide us with valuable knowledge and experiences. Most of the efforts in discovery and development of these reservoirs, were performed in USA. A combination of advanced techniques, horizontal drilling and hydraulic fracturing, has enabled profitable extraction of shale gas trapped in lowpermeability formations.
Development of shale gas reservoirs, besides of a comprehensive study on reservoir’s properties, demands for a precise forecast of reservoir’s production, in order to make the project economical.
In this study, the Bowland shale gas, as a high potential candidate for future developments in UK, have been analysed. We conducted a comprehensive study on every aspects of the reservoir, including geological, petrophysical, geochemical, and geomechanical properties. Meanwhile, Barnett shale gas, as a huge shale gas reservoir, which already developed in USA, has reviewed. The results used as a guidance and subsidiary information for appraising
future development of Bowland shale gas.
Various methods of production forecasting, comprising numerical, analytical, and advanced machine learning techniques have been examined. Numerical and analytical methods provided favourable results in production forecasting of Barnet reservoir, compared with real production test data. Also, the use of time series methods, as a branch of machine learning techniques, provided us with even more precise forecasts. The numerical and time series methods, both considered to be reliable methods in production forecasting of Bowland shale gas.
The results of this study open up an executive workflow for any future development of Bowland reservoir. The workflow by accounting any available data from the reservoir and utilizing experiences from already developed similar reservoirs, will provide a clear projection of future behaviour of the reservoir. The high variety of available data, the use of advanced forecasting methods, and successful experience of hydraulic fracturing operation in Barnett, regarding the similarities between the two reservoirs, give us the required feedback for any future development stage of Bowland.
Date of AwardJun 2020
Original languageEnglish
SupervisorMohamed Galal Hassan Sayed (Supervisor), Amjad Shah (Supervisor) & Hom Dhakal (Supervisor)

Cite this

'