Flow-Control Optimization Maximizes Accuracy of Multiphase-Flow Rate Allocation

Abstract

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 173873, “Flow-Control Optimization To Maximize the Accuracy of Multiphase-Flow Rate Allocation,” by Reza Malakooti, Khafiz Muradov, and David Davies, Heriot-Watt University, and Alexander Kuznetsov, Weatherford, prepared for the 2015 SPE Bergen One Day Seminar, Bergen, Norway, 22 April. The paper has not been peer reviewed. The value added by intelligent wells (I-wells) derives from real-time reservoir and production-performance monitoring together with zonal downhole flow control. Unfortunately, downhole sensors that directly measure the flow rates and phase cuts required for optimal control of the well’s producing zones are not typically installed. This paper describes an “active” monitoring technique that uses a direct search method to optimize the sequence of interval-control-valve (ICV) positions during a routine multirate test in an I-well. Introduction I-well completions include both downhole flow-control tools and monitoring sensors. Reservoir inflow is controlled by means of passive devices [inflow- control devices (ICDs)], active valves (ICVs), and self-adjusting devices ( autonomous ICDs and autonomous ICVs). Reservoir and well-performance properties should be monitored regularly to support flow- control decisions efficiently. Various physical quantities, including temperature, pressure, flow rate, acoustics, strain, and seismic, are currently measured by electronic, radioactive, and fiber-optic sensors to assist the multiphase-flow rate allocation in I-wells. Producing from only a single well or zone during a test is a common practice. This delivers a single flow-rate value measured at a particular time. However, this approach is inefficient, resulting in lost production and possible difficulties when trying to restart the closed-in wells or zones. A costly production-logging well intervention is a possible alternative; it does provide an inflow rate and a phase-cut profile, but only at a specific time without the possibility of delivering the current phase cut and production rate continuously. The alternative is to use (continuous) passive indirect measurements. Such soft-sensing techniques are used widely in (surface) facilities and process plants where extensive, complex networks of sensors can be implemented relatively easily. They are now being introduced for reservoir and well applications. A multiphase soft sensor consists of (1) an estimation technique, (2) a multiphase-flow model, and (3) the measurements. The multiphase-flow model relates the measurements to the parameters required for the estimation of downhole flow rates. The estimation technique is used to compute these parameters by minimization of the mismatch between measured information and predicted values. Published soft sensors have used both stochastically and deterministically based optimization methods as the estimation technique. The complete paper provides a review of recent research and developments in this field. Soft-sensing techniques are subjected to multiphase-flow-model errors and measurement uncertainties. The latter refers to variance of the values attributed to a measured quantity. For this reason, a new approach is required to overcome the limitation of the passive soft- sensing techniques used to allocate downhole flow rates. To accomplish this goal, an active multiphase-flow-rate soft-sensing method has been developed to design an optimum series of flow tests. The soft sensor obtains the most accurate estimate of the zonal properties by use of the deformed-configuration (DC) method to optimize the flow-test conditions (the ICV flow area in multizone I-wells or the wellhead-choke position for multiple conventional wells). To the authors’ knowledge, this work is the first publication to demonstrate the applicability of the DC optimization method in well soft-sensing problems.