Affiliation:
1. State University of Campinas
Abstract
Abstract
The exploitation of an oil field in deep water presents many challenges related to high water production, high cost of frequent well interventions and many uncertainties. One of the technologies available, which can overcome these problems, is the use of intelligent wells (IW), which are capable of reducing water production rates, to avoid intervention in the well and to add operational flexibility to mitigate risk. However, the real benefits of this technology are not always clear due to the lack of a consolidated methodology in the literature. Moreover, there are also two main ways of controlling valves, i.e., reactive and proactive controls, making it necessary to better understand them to extract advantages and disadvantages from each one. Therefore, the objective of this work is the comparison between conventional wells (CW) and IW, using reactive and proactive controls. The first control is simpler to be used and quicker to be optimized but the second type can be more profitable, although more difficult to optimize. The optimization method used to solve the problem is an evolutionary algorithm, which is coupled to a commercial simulator to search for the maximum net present value (NPV), based on the ‘shut in’ water cut to determine the optimum time in which to close each valve and the well, in all types of controls. This work employs a model using an inverted five-spot configuration of wells to represent a part of a reservoir under a waterflooding recovery method. Some case studies are used considering different reservoir heterogeneities, type of oil and under economic uncertainty. The conclusion shows that IWs are able to increase production time, oil recovery and the NPV; as a consequence total water production is also increased. The results also show higher benefits in cases with more heterogeneity and light oil. Moreover, IWs using proactive control is better than IWs with reactive control and using either of them is better than CWs.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献