Abstract:
Gongzhusi group in Zhongjiang-Penglai gas province is typical deep fractured formation, with unknown vertical depth of leakage point before drilling and specific leakage rate. The existing plugging methods fail to solve the challenges in such formation, and the treatment lasts for long period. In view of this,138 geological engineering factors are extracted from the history data of logging and drilling, the characteristic coefficient of all factors on the leakage rate are calculated and 16 leakage control factors are screened out by the element reduction method. While the formation fracture scale for lab simulation is 0.008~0.130 mm, the contribution coefficient of 6 drilling fluid performance main control factors to the leakage rate is consistent with the on-site value. The test formation fracture scale exceeds the observation range of the core, and the analysis shows the friction and sliding effect of fracture. The nanometer active particles of 0.001~0.010 mm particle size are preferred. While the water base drilling fluid with density 2.24 g/cm
3, adding mass fraction 0.1%~0.8% of particles, is tested, the leakage resistance of drilling fluid in the crack of 0.008 mm wide is increased by 0.01~0.16 MPa. After 0.1%~0.5% particles are added to the bridge plugging system, the pressure on wide 0.008~0.140 mm crack is increased to 7.2~10.9 MPa, showing feasible plugging effect. The leakage point prediction in 2 wells was carried out on site, with actual rate of 80%, and the predicted vertical depth deviation was no more than 31 m, indicating remarkable improvement compared to the traditional methods. Five plugging operations was performed in two wells, the success ratio reached 80%, and one single leakage treatment lasted for 28.25~39.15 h, with averaged value 32.39 h, decreased by 68.60% compared to the existing wells. The filed application shows that prediction of loss characteristics in deep small-scale fractured formations using geological engineering native data is feasible to improve the plugging efficiency.