Various standard and some novel performance metrics have been developed to understand the efficacy of an undergoing waterflood. These metrics provide insights about spatial relationships of production performance. Further, these metrics pictorially depict if disproportionate water injection influence happens in one part of the reservoir compared to the other. Specialized group functions have been devised which are cross plotted to differentiate and rank wells and sectors. The temporal relationships of well performance during an injection period are also captured. Overall, this technique is routinely applied to initiate a detailed waterflood study. This along with CRM provides immediate feedback about the health of the waterflood and quick-fixes e.g. where and how much injected water needs to re-distributed to optimize production.
Capacitance Resistance Model (CRM)
This technique, developed by the University of Texas at Austin, is utilized in conjunction with the above mentioned Production Diagnostics metrics. This intuitive and numerically fast technique operates on well production and pressure data alone. It exhibits the strengths of injector-producer connections without geological bias. Not only it is helpful in highlighting sectors of underperformance but also can operate in parallel to geomodel building process. CRM computes two parameters, Gains Matrixand Time Constant. Gains Matrix is a measure of the injector-producerconnection strength i.e. higher the Gains Matrix better is the injection influence. The Time Constant suggests the degree of depletion. Higher the Time Constant longer does it take for the injection pulse to arrive at the producer and therefore lesser is the injection influence on the producer. This methodology is used to capture water injection lost and suggest quick-fixes to re-distribute water to optimize production.
The recent progress of Streamline technology to encapsulate the influence of gravity, well changes, compressibility effects and rapidly morphing streamline geometries can be harnessed in conjunction with a predictive modeli.e. finite difference simulation to optimize volumetric and sweep efficiencies of a displacement process. In particular, streamlines can compute injector efficiencies (IE) and well allocation factors (WAF) for a given pattern flood and injector-producer connection. These reflect all the complexities impacting the dynamic behavior of the reservoir model, including the spatial permeability and porosity distributions, fault locations, underlying computational grid, relative permeability data, PVT and historical rates. With the aid of this crucial and valuable information, a predictive model can then be tuned to optimize flood variables e.g. reallocating injection water from low-efficiency to high-efficiency injectors constrained by surface-handling capability.
Feasibility of waterflood was investigated in onefractured carbonate reservoir. Available data of cores, logs, production, some pressures and previous geologic and engineering study reports formed the basis of the current study.
The methodology encompasses geo-scientific work, engineering analysis, production diagnostics and reservoir simulation. Classicclustering analysis is used to augment results of production diagnostics to capture the effects of fracture dominated flow. Further, material balance utilizes early production and pressure history , during both understaurated and saturated conditions, to provide bounds of in-place fluids. Petrophysical work analyzes available well logs to obtain matrix attributes and classify reservoir and non-reservoir types.
The static model devises empirical transforms to distinguish various facies types. The facies types are stochastically populated in the interwell region.A possible realization of the reservoir based on total-property approach is constructed. Thereservoir model is then calibrated based on 25 years of production and some pressure history.Finally, various development scenarios are investigated.