The Best Number of Infill Wells that Maximise the Project Value in Unconventional Mature Oil Fields: Deciding how Many to Drill from Available Candidates

The optimum number of infill wells to be drilled to maximise the field value has been a critical challenge, particularly for mature reservoirs, where the remaining reserves are in isolated pockets, such as heavily fractured, faulted, and tight formations. These reserves are probabilistically independent, and the drilling results at one location do not change our beliefs about the potential profitability of another location. The minimum economic volume MEV is decisive in this decision. MEV is a function of recoverable reserves and price perception. An optimistic perception of future prices can lead to drilling more wells, not all of which may be profitable, while pessimistic estimates risk overlooking promising wells. This study integrates the subsurface uncertainty arising from insufficient sampling of geologic variability with future oil price uncertainty in a consistent decision model for selecting the optimal number of wells. Our decision approach is straightforward. We use a cash flow model with the forward price curve, modelled using a stochastic price process, to calculate MEV. The MEV is used as a yardstick to determine the profitable wells. We assume that the unique geological uncertainty of each proposed infill location is determined using a stochastic-based reservoir simulation model. The uncertainty in each well location is plotted by a three-point discrete probability distribution (low (p10), medium (p50), and high (p90)) values and compared to the MEV. The wells whose average cumulative production value (calculated by the Swanson rule) exceeds the predetermined MEV will be drilled. The proposed approach is implemented in an illustrative example. The results illustrate how a pessimistic outlook on oil prices reduces project NPV by decreasing the proposed number of wells to be drilled. Additionally, relying on p50 is not a good decision-making practice. Instead, using the mean value (the Swanson rule) for a slightly skewed distribution or using either a low (p10) or high estimate (P90) is better than relying on P50. The proposed framework supports a balanced decision that can maximise project NPV by drilling the optimal number of infill wells.