Field Development Planning Using Simulated Annealing - Optimal Economic Well Scheduling and Placement

Abstract

A method for optimizing the net present value of a full field development by varying the placement and sequence of production wells is presented. This approach is automated and combines an economics package and Mobil's in-house simulator, PEGASUS, within a simulated annealing optimization engine. A novel framing of the well placement and scheduling problem as a classic "travelling salesman problem" is required before optimization via simulated annealing can be applied practically. Ah example of a full field development using this technique shows that non-uniform well spacings are optimal (from an NPV standpoint) when the effects of well interference and variable reservoir properties are considered. Examples of optimizing field NPV with variable well costs also show that non-uniform wells spacings are optimal. Project NPV increases of 25 to 30 million dollars were shown using the optimal, non-uniform development versus reasonable, uniform developments. The ability of this technology to deduce these non-uniform well spacings opens up many potential applications that should materially impact the economic performance of field developments. Introduction Optimal field-scale production of hydrocarbons must consider many reservoir, operational, and economic inputs and constraints. These inputs and constraints include, among many others, effects on well productivity from well interference, areal variations in reservoir quality, nonuniform well costs and variable oil price projections. These effects on optimal reservoir producibility are spatial and temporal. Drill and complete times place an temporal operational constraint on the availability of hydrocarbon production. Variable cost structures for different well trajectories within a field must be balanced against improved or accelerated production from more expensive wells. Well scheduling constraints may occur if phased development due to capital restrictions is required. Facility throughput limitations may also place operational limits on the number and scheduling of development wells. These multiple competing effects mean that developing an optimum well plan for field development is challenging and that development on uniform spacing may not be optimum when viewed from an economic point of view. In this paper we show how optimization techniques, specifically simulated annealing, can be successfully used to produce economically optimal well placements and schedules. Development scenarios that maximize the entire project NPV show nonuniform well spacings are required when variable well costs and variable reservoir properties are considered. Methodology The well scheduling and placement for optimal field development can be expressed generically as the classical "travelling salesman problem". For this work we define "optimal field development" as the placement and scheduling of wells that maximize a measure of field NPV. Such a realization is a fundamental to the successful application of optimization methods, such as simulated annealing', genetic algorithms or other methods suitable to the travelling salesman problem. P. 209

Design of the Steam and Solvent Injection Strategy in Expanding-Solvent Steam-Assisted Gravity Drainage

“…An overview of SA applied to reservoir description can be found in Ouenes et al (1993). Applications that use simulated annealing for reservoir simulation optimization have been done by Beckner and Song (1995), Ouenes et al (1994), and Sen et al (1995).…”

Application of Assisted Optimization to Aid Oil Exploitation Strategy Selection for Offshore Fields

“…Some research papers present optimization processes based on non-gradient methods such as the stochastic optimization algorithms including Genetic Algorithm (Al-Harthy, 2010), Simulated Annealing (Beckner and Song, 1995), Particle Swarm Optimization, Harmony Search algorithms (HS) (Afshari et al, 2011) and Neural Networks.…”

Well Placement Optimization Using Imperialist Competition Algorithm

“…(1983), is a probabilistic method used for the finding of a global optimum by mimicking the recrystallization process of a heated solid object till it reaches a frozen structure that corresponds to a minimum energy configuration. Beckner and Song (1995) and Norrena and Deutch (2002) applied this method for well placement optimization. Another stochastic optimization method is the partical swarm optimization (PSO).…”