Mr Elvar Karl Bjarkason
Research | Current
Improved inverse modelling techniques for geothermal reservoir models
One of the most widely used numerical codes for large scale geothermal reservoir simulations is TOUGH2. It is used extensively around the world for large-scale modelling of geothermal systems. When setting up a complex computer model of a geothermal system, the underground structure has to be inferred since direct measurements are only possible in a few boreholes. This process involves assigning certain parameters to each block or element in the model. Then a two-stage calibration process is undertaken to improve the model. In the first stage the parameters are adjusted so that the model results produce a good match to measured pre-exploitation temperatures and in the second stage further adjustments are made so that the model matches the production history.
The current approach to automated calibration of geothermal models is to use gradient based inverse modelling software such as PEST or iTOUGH. The trouble with using PEST or iTOUGH for inverse modelling, is that they require many time-consuming forward runs of the geothermal simulator (TOUGH2) in order to calculate derivatives of the objective function, with respect to each model parameter. This study looks at using a Lagrange multiplier approach to improve the computational efficiency of inversion software used for history matching of geothermal data.
Selected publications and creative works (Research Outputs)
- Bjarkason, E., O'Sullivan JP, Yeh, A., & OSullivan, M. J. (2016). Combined natural state and history matching using the adjoint or direct sensitivity method. Paper presented at 38th New Zealand Geothermal Workshop, Auckland, New Zealand. 23 November - 25 November 2016. Proceedings 38th New Zealand Geothermal Workshop.
Other University of Auckland co-authors: Michael OSullivan, John OSullivan, Angus Yeh
- Bjarkason, E. K., O'Sullivan MJ, O'Sullivan JP, & Yeh, A. (2016). Accelerating calibration of natural state geothermal models. Proceedings, 41st Workshop on Geothermal Reservoir Engineering Stanford, California, USA. Related URL.
Other University of Auckland co-authors: John OSullivan, Angus Yeh