Multilayered Water Injection Zonal Allocation Without Intervention: Maximizing Value of a Smart Well by Harnessing Historical Data in Efforts to Expurgate Imprecise Injection Allocation

Abstract

Field A begun water injection in 2016 via four water injector smart wells, which were equipped with Permanent Downhole Gauges and Inflow Control Valves. The water injection module was housed on a rented MOPU due to space limitation. Amidst the study to revamp the reservoir management plan, the team found multiple discrepancies in the reservoir zonal allocation dating back to start of injection. Inherently, this affects the Voidage Replacement Ratio tracking. Hence, the question remains: How efficient is the water injection in Field A? As Field A injects from a rented facility, the long term RMP strongly influences annual OPEX. This paper explains the journey of reallocating Field A water injection volumes from 2016 until today, and how it affects the outcome of the RMP study. PETRONAS has an offshore monitoring system which visualizes historical pressure and temperature trends at any tagged equipment. Field A water injectors consists of multi-zones completed with ICVs and PDGs. ICVs allow choking and zone changes to happen without intervention, and PDGs show downhole pressure and temperature changes over time. Coupled with the manual database which tracks ICV changes and water injection rates, the team re-modelled the injected volume allocation changes to each zone by anchoring the model on PDG trends, ICV size and choke coefficient, and water injection rates via an advance nodal analysis software. For reservoir characteristics calibration, properties from past FBUS interpreted results were used as a basis. From the modelling journey, at the same injection scheme, results showed that zonal allocation with small PDG pressure changes of less than 5% during stable injection conditions does not significantly affect allocation ratio in the well. Overall, the allocation would change between 0 - 3% in total. As one of the objectives of the exercise was also to gauge expected injected volume allocation to a specific zone when there were obvious pressure changes but no records of changing ICV sizes, this could be achieved via a calibrated model. Once a good anchor was made on reservoir pressure, formation gas-oil ratio, permeability and skin, devoid periods in the past could be modelled for expected ICV sizes by varying the choke size openings till the pressure differential between tubing and annulus pressure was achieved. Hence, modelling the expected zonal allocation during that period. This improved VRR tracking for the injection reservoirs and aided to in the efforts to revamp the reservoir management plan. This paper will explain the lessons learnt of having proper surveillance data as the impact on long term reservoir management plan is significant. In future, fields with smart wells but disorganized data can utilize this alternate method to reallocate production/injection volumes without the need for intervention.