Abstract
Ageing fields are often in a race against depreciating Net Present Value (NPV) and decreasing production, advent of rising water contacts threatening daily and cumulative production. Static geological models form the primary input model into dynamic models which define and determine the development strategy of fields. Typically, dynamic models built from incipient seismic and geological models comprising of well data from early appraisal wells. In time, incorporation of production data refines the dynamic behaviors of field and in turn, these dynamic models are then used for subsequent re-development of the ageing fields. However, the incipient seismic models rarely receive similar updates as seismic acquisition are too costly to justify in an ageing asset with ever-depreciating NPV.
This paper illustrates the process undertaken by subsurface team in identifying potential sites constraints by seismic and geological studies, in comparison with initial oilfield production incorporating pressure data to identify sweep efficiency. Starting with understanding a reservoir that experienced poor injection support, has led to unearthing the presence of faults that were not identified in previous seismic interpretations, and thus compartmentalizing the reservoir. The presence of the ‘unseen’ barriers in between an injector and an oil producer has hampered the injection effort towards the producer, and thus has been detrimental to the health of the reservoir.
In lieu of the renewed understanding of the reservoir with regards to field compartmentalization which is recognized as key threat to water injection performance in Reservoir-L, usage of artificial intelligence machine-learning generated fault volumes helped define compartments better. The results indicate the NNW-SSE conjugate (shear) faults formed during the Mid-Late Miocene times which were not clearly demarcated on earlier seismic, play key contributing role with regards to connectivity and intra-field communication. This in turn led to revised understanding on areas of reduced production that led to model updates in line with production behaviors.
Based on the new model, well location optimizations were undertaken for two water injectors in the infill drilling campaign for shallower Reservoir-K of Field S to avoid potential compartmentalization that would impact the sweep efficiency. Relocation of the water injectors into the same reservoir compartment as the target oil producer was expected to improve the EUR by ensuring direct sweep and good injectivity to the producer.
In summary, by revisiting field models in advent of updating and recalibrating the dynamic model in view of historical production data, the team was able to update the model contributing towards field revitalization and field management in ensuring the field model delivers to the field development plan. The hard work from the team has paid out through the recent production performance of the reservoir.
