Maximizing oil recovery in a naturally fractured carbonate reservoir using computational intelligence based on particle swarm optimization

Mohammad Rasheed Khan, Ahmed Sadeed, Abdul Asad, Zeeshan Tariq, Muhammad Tauqeer

Research output: Chapter in Book/Report/Conference proceedingConference contribution

16 Scopus citations


A major challenge in carbonate reservoirs is the highly-fractured nature of the rock. The flow rate may be high or low depending on the targeted fracture clusters. In addition, it is possible that flow rates vary from one region of the reservoir to another. Smart wells furnished with smart completion strategy presents great prospects to produce such reservoirs intelligently, thereby, helping to deal with heterogeneities rather smartly. It is established that early water break-through occurs when multi-lateral wells are completed with constant choke settings, and therefore one way to mitigate this problem is using smart completions that manage the unexpected production through fractures, thereby increasing ultimate recovery. The early water breakthrough is obvious because if a lateral section intersects a clusters of fracture zone, there is a possibility that these fractures may connect with the water zone that may trigger the breakthrough. This can be managed by preferentially regulating production from manifold laterals. The evident communication among the various laterals of the mother bore raises difficulty in optimizing the production from the variable productivity intervals. In theory, the optimization scheme of smart completion involves different constraints, nevertheless, the settings of the smart inflow control valve (ICV) is the single most important parameter that may prove to be the differentiating factor between a high producing well to a poorly producing one. This study engrosses its effort on the reservoir engineering characteristics of finding the optimum choke setting that would lead to maximum recovery. Computational Intelligence through Particle Swarm Optimization (PSO) is utilized as the integral algorithm to determine the optimal ICV configuration for a fishbone well in a naturally fractured carbonate reservoir. A commercial black oil simulator was used to determine the objective function; whose role here is to evaluate the fitness of a configuration of the choke; this was carried out under a workflow programmed in the MATLAB programming language that coupled the optimization algorithm with the numerical simulator. A single fishbone well, having 15 laterals was studied in order tot see the effect of the fracture network on the water breakthrough and consequent impact on recovery. Three different scenarios are developed to see the impact of optimization; a base case employing only multilateral well technology without the smart well completion, a smart well completion scheme with no optimization and finally the optimized smart well completion. The results very sequentially clarify the need for not only optimization but also highlights the role of intelligent completions for wells in the reservoir being studied. It is evident that without using smart wells, the water breakthrough is relatively earlier and produces less hydrocarbons, but as the use of smart wells is incorporated, the results start improving and for complete optimization scheme of the ICVs, it is observed that the recovery has increased by almost 80% from 21% to 38%. Moreover, the time to water breakthrough and eventually the cumulative water cut has also been managed quiet significantly.
Original languageEnglish (US)
Title of host publicationSociety of Petroleum Engineers - PAPG/SPE Pakistan Section Annual Technical Conference and Exhibition 2018, PATC 2018
PublisherSociety of Petroleum Engineers
ISBN (Print)9781613996829
StatePublished - Jan 1 2018
Externally publishedYes

Bibliographical note

Generated from Scopus record by KAUST IRTS on 2023-09-20


Dive into the research topics of 'Maximizing oil recovery in a naturally fractured carbonate reservoir using computational intelligence based on particle swarm optimization'. Together they form a unique fingerprint.

Cite this