Closed-loop feedback control of smart wells for production optimisation using self-potential measurements

Abstract

Closed-loop ‘reactive’ feedback control techniques used for smart well optimisation, triggered by changes in flow (such as unwanted water production) measured at the well can increase the net present value (NPV) and mitigate reservoir uncertainty, as opposed to model-based control strategies, which use models that are rarely predictive at their spatial and temporal scales required to identify optimum control actions. However, the drawback faced with closed-loop ‘reactive’ feedback control is that control actions are only taken after adverse changes of flow occur at the well. We present a modified close-loop ‘proactive’ feedback inflow control approach based on near-well, downhole measurements of self-potential (SP) and quantify the potential benefit of this approach in different well and reservoir settings during waterflooding or aquifer support production. The measurement of SP signals downhole in production wells is an encouraging technique that can be used to image waterfronts and has the potential of detecting water encroachment tens to hundreds of meters away from the well. SP signals arise in order to preserve electrical neutrality when charge separation arises due to gradients in pressure, temperature and chemical concentration of the reservoir brine phase. These gradient effects are commonly encountered during waterflooding processes and can be assessed numerically to predict the SP generated downhole in oil production wells. The numerical modelling of SP can be used as a cheap alternative to carrying out actual field experiments and serve as a proxy for predicting the SP measurements taken during waterflood production. Hence, a closed-loop ‘proactive’ feedback control strategy triggered by downhole SP measurements is developed. We use the NPV of the production wells to measure and compare the performance of the closed-loop feedback control in two different synthetic production cases; the first production case is a simple thin oil-column reservoir with production enabled by a single long horizontal well, and the second more realistic SPE Brugge field model, with production enabled by 20 production wells. The results observed are promising, and suggest that closed-loop control on the basis performance of downhole SP feedback can yield increased gains in NPV, by delaying the production of unwanted fluids compared with water-cut monitoring. These gains are also observed even if the reservoir lies outside the range predicted by reservoir models. Finally, we investigate the potential utility of SP monitoring in analogue real field applications. Overall the results are promising and suggest that SP measurements can be useful in making critical decisions in real field exploration and production applications, and other non-oil related fields such as saline intrusion monitoring in coastal aquifers.