Intelligent multiple parameters optimization methods integrating hydraulic model and SIAs with various constraints for extended reach drilling

Abstract

Extended reach drilling (ERD) plays a crucial role in deep water and ultra-deep reservoirs exploitation. Cuttings are prone to deposit to the lower side of wellbore when rate of penetration (ROP) is high in ERD. Drilling problems caused by insufficient wellbore cleaning will increase drilling costs and decrease ROP. Therefore, keeping a high ROP and ensuring wellbore cleaning is very important by optimizing hydraulic parameters. This paper proposes intelligent multiple hydraulic parameters optimization methods integrating accurate hydraulic model and particle swarm optimization algorithm (PSO) as well as sparrow search algorithm (SSA) with various constraints to maximize drill bit hydraulic power, which are abbreviated as MPOM-PSO and MPOM-SSA. Rheological parameters of seven rheological models are calculated regressively and the best rheological model is preferred to improve accuracy of pressure loss. Interrelationship between rock breaking and wellbore cleaning as well as constraints of formation pressure, rated pressure of circulation system, rated flow rate of pump and cuttings bed thickness are considered in MPOM-PSO and MPOM-SSA. It overcomes defects of computation-intensive and inability to perform multi-parameters optimization simultaneously compared to traditional optimization methods. The accuracy of hydraulic model is validated by comparing with results calculated by Landmark. The rheological parameter calculation errors of both Power–Law model and Herschell–Bulkley model are less than 1%. In terms of frictional pressure losses in annulus and in drillstring and standpipe pressure, the average errors are 1.8% and 3.5% for Power-law mode and Herschell–Bulkley mode respectively. The efficacy of MPOM-PSO and MPOM-SSA is proved by Case studies and statistic analysis. The maximum errors of optimal flow rate and density are less than 4% and 1% respectively contrasting to traditional method through 50 simulation experiments. However, the variance of optimal flow rate obtained by MPOM-SSA is larger, demonstrating MPOM-PSO is a litter better than MPOM-SSA. Also the optimization speed of MPOM-PSO is increased by more than 25 times. Through the application of MPOM-PSO and MPOM-SSA, hydraulic parameters can be optimized speedy and drilling efficiency of ERD can be improved.