In January 2007, Hyperion acquired Crystal Ball, the Monte Carlo risk analysis plug in to Microsoft Excel. A couple of months later, Hyperion itself was picked up by Oracle Corp. This means that Crystal Ball (CB) for oil and gas is now a component of Oracle’s ‘digital oilfield’ effort. CB for oil and gas users met in Houston last month to hear how CB is used in risk analysis, simulation and optimization and to learn about ‘communicating’ risk and how to get the most from their forecasts and analyses.
Spectra Energy
Ken Jeans described how Spectra Energy is engaged in a large number of capital projects as it expands its natural gas facilities. Crystal Ball is considered key to Spectra’s analysis of project schedules and costs. Spectra is now on its third generation of CB models. These are now aligned with the project execution plan and integrated with capex and schedule models. Optimization now supports risk-based mitigation planning and efficient capital to revenue expenditures. The modeling process includes rigorous validation and discussions with a focus on risk mitigation and optimization. CB’s OptQuest goal seeking function is widely used. Notwithstanding the science, Jeans wonders if there is a ‘conspiracy of optimism,’ and suggests all levels of uncertainty need to be managed, that expectations should be set early and knowledge about risk should be shared by complete disclosure.
Catheart Energy
Robert Merrill (Catheart Energy) explained that shale gas resource evaluation is dependent on a range of shale parameters including density, thickness, total organic carbon, hydrogen index and maturity. CB is used to evaluate recoverable gas, providing outcome probabilities and sensitivity analyses. Investment decisions can then be in the light of statistical estimates of recoverable resource and with an understanding of the associated risks.
Rice
Susan Peterson’s (Rice University) presentation on ‘uncertainty enlightenment’ in field development showed how the decision making process for a marginal offshore field development was facilitated with CB. Issues resolved with probabilistic modeling included arbitrating between work-overs and new wells and drilling sequencing in the light of a FPSO hookup schedule and the nearby discovery of a satellite tie-back. Lease or buy options for the FPSO were analyzed in terms of likely field life and various commercial terms. Probabilistic full field modeling allowed discussions to be centered on quantifiable risks and uncertainties and allowed management development decisions to be based on the uncertainty enlightenment that resulted from those models.
Hoye
Steve Hoye (CB) showed how time series analysis can be used to anticipate future oil prices (it’s not called Crystal Ball for nothing!) by breaking up historical data into periods of relative stability. Price forecasting is then possible using ‘Gaussian mean reversion with jumps.’ CB Predictor, correlation, and distribution fitting tools were used.
© Oil IT Journal - all rights reserved.