Experimental Design
Using StatTools and @RISK or Crystal
Ball, we collaborated with reservoir simulation experts to develop
proxy equations for cumulative oil and gas production, NPV,
deliverability, and unproved reserves. Fields included giant gas
fields, a heavy oil field, and a new gas discovery. This methodology is
promising and should be applicable to a variety of problems where a
reservoir simulator model is available.
Cost and schedule models
We
built two probabilistic Cost Estimate models, one for a client who
rebuilds drilling rigs to customer specifications and a second from
collaboration with an expert in decommissioning offshore facilities.
Both models account for unplanned events, including weather. This
methodology is a continuation of our several drilling cost models
constructed during the past 10 years.
Probabilistic Production forecast and cashflow models of Barnett and Woodford Shale projects
Detailed
models have been constructed of both the Barnett Shale and the Woodford
Shale. In each case, each of several hundred wells was modeled
separately using hyperbolic decline curves. Revenue interest and
working interest are specified by well. Multi-rig drilling schedules
are constructed. Input distributions include initial gas rate, initial
decline rate, hyperbolic exponent, initial liquid yield and pattern of
yield change over time, Capex and opex, number of wells, probability of
drilling and completion success. Two volumetric models were employed,
one based on length of horizontal section and effectiveness of fracture
stimulation; the other on gas in place per acre-ft.
Probabilistic
outputs include distributions for Gross and Net EUR, Net NPV, annual
production, net revenue , capex and opex as well as summary graphs
(trend charts) representing probabilistic cashflow and production
forecasts.
|
|
