Click to view page image in pdf format.
7th Middle East Geosciences Conference and Exhibition
Manama, Bahrain
March 27-29, 2006
1 EP Solutions, Shell, 200 North Dairy Ashford, Houston, TX 77079, phone: ++1-7134321963,
[email protected]
2 EP Solution, Shell, 200 North Dairy Ashford, Houston, TX 77079
The paper presents an integrated-iterative workflow to manage uncertainties and sequentially improving predictability. It is
illustrated at a microporous, fractured limestone reservoir, covered by poor quality seismic
. The workflow includes: i)
Uncertainty Identification (Chart) ii) Uncertainty Quantification and Ranking (Experimental Design, Monte Carlo
Modeling
,
Distribution Curve, Tornado Chart) iii) Uncertainty Mitigation (Decision Tree, Data Acquisition) iv) Revised Uncertainty
Ranking and Mitigation (repeat step ii and iii) Field performance and previous models were used to estimate the key
uncertainties on STOIIP and recovery. Several simple static models were constructed and simulated, changing ONE
uncertainty at the time. Historic production was used to select meaningful models. Simulation also revealed the key
uncertainties: fracture properties, saturation behaviour and top structure. Experimental design provided a matrix to guide
construction of deterministic models by combining uncertainties in a statistically meaningful way. Alongside a workflow was
set up to vary theses uncertainties probabilistically (Monte Carlo). A STOIIP distribution curve was generated and
P15/50/85 values determined. Deterministic models with STOIIP's close to the P15/50/85 values and additional models
were simulation to provide an production forecasts envelope. A decision tree was established to guide data acquisition in
new wells. Well results revealed good predication of saturation and fractures except along faults. Top structure uncertainties
were under-predicted, particularly along faults. Well results were used to update models. New wells will probe structure and
fractures at fault but disregard saturation. The workflow is an ‘evergreen' method that provides statistically meaningful
production forecast and iteratively reduces uncertainties to optimize field development.