Big data: well performance

August 2015September 2016
Well test data is one example of a dataset that may benefit from more detailed analytics.

Big data: well performance (prefeasibility study)

This research explored the capability of forecasting well test behaviour from the examination of a large number of similar wells. The project provided the foundation for further analytics studies.

CSG resource companies are constantly seeking ways to reduce gas production costs. Current advances in the computer science "Big Data" field means that it is possible to see patterns and analyse streams of unstructured data in text, image and video formats as well as analyse real-time streams of data. For onshore gas industry data this means that there is the ability to combine data from real time well measurements and monitoring, with modelling of reservoir conditions. This will improve the accuracy of forecasting reservoir performance and optimise upstream and downstream engineering operations, which can overall reduce gas production costs. A key element will be to develop analytics for continuous improvement.

Well test data is one example of a dataset that may benefit from more detailed analytics. When used in conjunction with the estimated permeability field and other spatio-temporal information, the additional information streams will improve predictive confidence. Given the high cost of developing coal seam gas due to the large amount of wells required to access higher permeability regions, marginal optimisations in the underlying cost structure can have a significant effect on profitability. The project objective is to explore the capability of forecasting well test behaviour from the examination of a large number of similar wells. This will be done by extracting well test data from a variety of sources, determining appropriate parameterisation of well tests and selecting algorithms for spatial analysis and forecasting as well as the spatial estimation of functional data.

The project provided the foundation for further analytics studies.

PROJECT OUTPUTS

  • Review: A review of available analytics, application and gap analysis for well test analytics.
  • Worked example: of well test analytics.
  • Full program proposal: that described the reasearch required to implement well test analytics.
  • Project status: Complete
  • Project title: Big Data Analytics – Prefeasibility Study (Well Tests)
  • Project leader: Professor Steve Tyson
  • Research group: The University of Queensland Centre for Coal Seam Gas & The University of Queensland School of Earth Sciences
  • Timeframe: August 2015 - September 2016
  • Project funders: APLNG, Arrow Energy, Santos, QGC, University of Queensland