Automated Optimization of Interdependent Artificial Lifting Fields Using Integrated Response Surface Model and Hybrid Heuristic-Gradient Algorithm: A Field Case Study

Abstract

The generic characteristic of producing fields in Gulf of Thailand (GOT) is that oil wells are also developed on the same facilities as gas wells when possible. Different efforts are taken to optimize both oil and gas wells together. Gas-lift is one of the examples of artificial lift that is installed in oil wells. For gas wells, booster compressor (BC) is common facility on platforms to reduce back pressure to each producer routed to the unit. The situation can get even more complex when gas from the BC is used as a source for lifting gas in an oil well. This paper extends the novel automated optimization workflow of gas field optimization (Ekkawong P. et.al. SPE-173450) and oil field with gas-lift optimization (Kritsadativud P. et.al.SPE-174039) to maximize production from more complicated gas/oil fields with various combined artificial lifting systems. Firstly, to improve calculation efficiency, the production performance estimated by NODAL analysis is transformed into a multi-dimensional response surface model (RSM) as a relationship between flowing pressure, gas injection rate and oil production. This transformation significantly eases the computation burden. The RSM is then used as the well performance prediction tool to be used in a production optimization workflow using hybrid optimization algorithm, the combination of heuristic and gradient based optimization. Genetic Algorithm is roughly searching for the global optimum solution, and then the gradient based method is used to further refine the solution. The process is run repeatedly with different initial guesses to ensure the optimality of the obtained solution. The benefit of this method is demonstrated in a synthetic field that illustrates the complex offshore field in GOT. This field features oil and gas wells with interdependent artificial lift equipment: gas-lift and BC. The objective is to maximize field oil/condensate production under other field operating constraints. The controllable parameters are: amount of gas to be injected into each oil well, well to be routed to BC, and wellhead flowing pressure from each well. The result of this proposed workflow shows increase in oil production honoring all practical constraints. Additionally, the computation time is also significantly reduced from the conventional approach using manual fine-tuning and the integrated NODAL analysis modeling. This automated workflow simplifies the more complex process to be a much more computationally efficient and easy to understand. Hence, it can be used as the tool to optimize day-to-day production operations in the complex fields that the conventional tool is obstructed by the computational burden.