Validation of Full Transient Hole Cleaning Model, with At-Scale Datasets, Implementation into an Automated Advisory System

Abstract

AbstractTransport of drilled cuttings, from bit to outlet at surface, is one of the most fundamental processes underpinning well construction, and yet the industry is long overdue proliferation of fit-for-purpose models in this area. Currently, the majority of operations globally use steady-state models based on empirical correlations or mechanistic models to determine acceptable operational parameters. Annually, NPT and ILT associated with inadequate hole cleaning costs the industry millions, which furthermore reinforces the urgent need for comprehensive transient hole cleaning models to become more widely utilized. Lack of deployment of transient models in real-time is not due to lack of understanding of the physical process, the barriers have been accuracy and robustness of developed algorithms, as well as computational processing power constraints which have made real-time deployments challenging.This paper details the features of a transient hole cleaning model capable of accurately solving the temporal and spatial course of both pressure and cuttings concentration along the wellbore. Specifically, for the latter, a drift-flux-model is employed, populated with coefficients from high resolution 2D and 3D CFD simulations.For primary validation of the model, output of cuttings mass exiting the wellbore is compared against measured mass from drilled cuttings waste (skips). Secondary validation is through comparison of simulated stand-pipe pressure, including the effect of modelled spatial distribution of both static beds and concentrations, with the measured pressure on surface. Results show a good fit between model outputs and measured data successfully validating the model. A sensitivity analysis is detailed demonstrating the relative importance of various input variables including cuttings size, drilled lithology density and signal source for ROP.Finally, integration of the transient hole model into an advisory software-service is described, highlighting the importance of UX to concisely and intuitively convey simulation data to end-users, ultimately facilitating de-risking of well construction operations.