The application of risk-based methods to multilateral well design

Abstract

Horizontal and multilateral oil and gas wells are used to maximise hydrocarbon recovery while reducing the required well count and associated costs. Presently, lateral lengths are designed using semi-quantitative methods. Guided by a desire to minimise the risk of poor well deliverability, the tendency is to design producing lengths longer than required, with the rationale that the well connects with sufficient hydrocarbon bearing reservoir to provide good deliverability. Drilling long producing lengths, however, is expensive and generates a higher risk of drilling and lifecycle (intervention and workover) problems. Furthermore, attempting to increase deliverability by extending the producing length encounters the law of diminishing returns as the flow becomes constrained by tubing friction loss. This paper seeks to quantify the optimal length for a horizontal well for a given range of reservoir conditions through multiphase fluid modelling and stochastic analysis. A discretised horizontal well model was created, which shows how changing the well length transforms the probability density function of the production rate for the well. A parametric case study was conducted, which demonstrates the evolution of the optimal well length and production rate with parameters including well diameter, fluid viscosity and well flowing bottomhole pressure. A simplified economic analysis illustrates the incremental change in discounted cash flow and quantified risk from drilling a longer well. The model also considered the influence of inflow control devices (ICDs) to adjust the inflow to match permeability and even-out inflows along the producing length, thus reducing the risk of gas and water coning, and improving hydrocarbon recovery.