A thermally stable polyampholyte thinner for high-density water-based drilling fluids and a machine-learning-based model for rheological prediction

Maintaining rheological stability of high-density water-based drilling fluids (WDFs) under high temperature conditions remains a formidable challenge for deep-well drilling operations. In this study, a zwitterionic polymer thinner (ZPT-3) was synthesized and systematically evaluated in both weighted mud (WM) and weighted polymer mud (WPM) systems through comprehensive rheological testing. Total organic carbon (TOC) analysis, focused beam reflectance measurement (FBRM), and optical microscopy were employed to investigate the mechanism of ZPT-3 in regulating the rheological behavior of high-density WDFs. ZPT-3 demonstrated exceptional thermal stability and thinning efficiency, yielding a viscosity reduction rate of 66.3% in complex WPM at 2 wt% after aging at 200 °C, significantly surpassing conventional commercial thinners (XY27 and SF260). To facilitate practical engineering applications, a multi-output machine learning (ML) model was developed to accurately predict rheological parameters of WM systems under varying additive concentrations and aging temperatures. These findings highlight ZPT-3 as a promising and robust thinner for high-density WDFs, with the ML predictive model providing a reliable digital assistant for the design and optimization of drilling fluid formulations.