Application of principal component analysis on water flooding effect evaluation in natural edge-bottom water reservoir

摘要

Water flooding effect evaluation is considered as the basic work to formulate comprehensive adjustment measures and improve the effectiveness of oilfield development. However, natural edge-bottom water energy is seldom considered in the conventional evaluation method. So, it cannot reflect the comprehensive effect of both natural edge-bottom water and injected water. Principal component analysis is a kind of multivariate statistical analysis method, which has been widely used in social science and other fields. Based on this method, the water flooding effect of 5 edge-bottom water reservoirs is comprehensively evaluated. First, 11 indicators are selected from four aspects, including natural edge-bottom water energy, production change, water injection development and utilization, energy maintenance and deficit compensation. Then, the selection of principal components is optimized. Based on the consideration of keeping as much information as possible to get more convincing results, three principal components are obtained. Finally, take five oilfields as examples to realize comprehensive evaluation. Results indicate that the natural energy of B oilfield is quite sufficient and water injection is timely in the later stage of development. So the water flooding effect is the best among five oilfields and the comprehensive principal component value is 1.434. That of A and C oilfields are 0.527 and 1.021, respectively, ranking 3 and 2. Although D oilfield has quite sufficient natural energy, water injection is not timely. So the water flooding effect is poor and the comprehensive principal component value is 0.259. That of E oilfield is ??3.241, indicating that it has the worst water flooding effect. The ranking results of five oilfields are consistent based on principal component analysis and Tong's chart, which are both B, C, A, D and E oilfield, verifying this method’s feasibility and practicability. Additionally, compared with the single index, it can reflect the comprehensive water flooding effect of both natural edge-bottom water and injected water. Specific oilfield cases are evaluated by the proposed method, which help for better understanding its application potential for evaluating the water flooding effect of natural edge-bottom water reservoirs.