青海油田柴达木盆地北缘断块带深探井钻井过程中油气层保护存在一系列技术难点:缺少具有代表性的岩心样品和储层岩性、物性资料;存在高压含气、水层;存在多个储层,各储层之间损害机理有一定差异等。采用人工神经网络储层敏感性预测新方法,对该地区储层进行了敏感性预测。在此基础上,按照理想充填方法对暂堵剂进行了优化设计,并通过大量试验确定出了保护储层的低损害聚磺钻井液配方。现场试验表明:该钻井液体系性能优良,流变性易于调整和控制,抗高温能力强,API滤失量和高温高压滤失量很低,泥饼致密、光滑;油气层保护效果显著,储层渗透率恢复率高达82.54%,表皮系数≤1.5;井壁稳定作用明显,未发生井塌、缩径、划眼等井壁失稳的情况,实现了安全、优质、快速钻进。
Many technical challenges exist in formation damage control in the deep exploration wells in the north fault formation of Chaidamu Basin in Qinghai Qilfield, including without typical core data, formation lithology and physical properties, the existence of high pressure formation saturated with gas or water, the existence of several formation with different formation damage theory in different formation. The sensitivity studies for the formation in this area were conducted using Artificial Neural Network. Based on these results, the bridging agents were optimized using the "Ideal Packing Theory", and a kind of polymer-sulfonates drilling fluid formula for formation damage control were determined through a lot of experiments. The field tests show that the drilling fluid has better performance, such as better rheological properties, higher temperature resistance, lower API filtration and lower HTHP filtration rate, thinner and slicker mud cake. The results of formation damage control are obvious. The formation permeability recovered up to 82.54% and the skin factor is less than 1.5. In addition, the wellbore is more stable without any problems such as hole sloughing, tight hole and reaming. The safety, high quality and fast drilling have been achieved.