Early warning system of casing damage for blocks in sandstone oilfield

Concentrated casing damage has appeared in local area of Changyuan Oilfield in Daqing, which severely leads to substantial economic losses. Therefore, a casing damage early-warning model is established after an in-depth analysis of geological and development factors that affect casing damage by means of big data mining. Four geological early-warning indicatorsare selectedby means ofcore observation, petrophysical experiment, casing damage block anatomy, etc. Blocks are categorized into five levels of casing damage risk according to geological early-warning indicators. The block characterized by fault, fossil layer, sand-mud interface development and large formation inclination has high risk of casing damage. Ten development factors of pressure system and injectionproduction parameters are designed and optimized. Limits of early-warning indicators are determined by analyzing the relationship between each indicator and annual rate of casing damage. Early-warning models for respectively each of five geological risk levels are established through SVM method, achieving early warning in both casing damage risk level and index of block. The prediction coincidence rate forcasing damage of block reaches more than 70%.An early warning system platform for the casing damage has been developed, which enables automatic analysis and early warning as well as real-time early warning of casing damage.