Intelligent well technology use Inflow Control Devices (ICDs), Inflow Control Valves (ICVs), and measurement devices that provide opportunity for monitoring and control of production from different reservoirs from the same wellbore. The common value drivers for the deployment of intelligent wells include reduction in well count during field development; accelerated ultimate recovery (UR); and accelerated production. Efficient and accurate modelling is therefore critical for the realisation of the full benefits of intelligent wells in addition to other technologies like the use of geochemical finger printing.
In this paper, a simple workflow for modelling the performance of intelligent wells is presented. This workflow identifies the limitation of current standalone PROSPER model and provides a window to easily match the model to actual well test results, providing multiple calibration points and ensuring full utilisation of the data made available by the intelligent well accessories like the permanent downhole guage and downhole flowmeter. The workflow has been applied to carryout nodal analysis of both oil and gas completions in the Niger Delta, Nigeria.
The modelling workflow is divided into two (2) aspects – the inflow modeling, for each inflow zone, and the outflow modelling that captures the commingled production. The commngled outflow model is used to generate the lift table for the GAP model.
The well deliverability and PQ curves are generated and plotted using the Openserver utility in IPM, which can be used also to view the performance of the model against validated well test results.
Results of applying the workflow to two case studies in the Niger Delta were analysed. The model was used to predict the expected performance of the wells at different surface choke sizes and ICV settings. The model results (flow rate, flowing tubing head pressures, flowing bottomhole pressures) matched closely (with maximum of 10% deviation) with the actual measured results, confirming the accuracy of the recommended workflow.