Shale wellbore instability poses a considerable challenge in terms of drilling efficiency. However, previous studies typically assumed single-phase flow in the seepage field, neglecting the influences of water saturation and temperature on rock mechanical properties. Furthermore, there has been a lack of quantitative analysis on the degree of correlation between various factors and collapse pressure. To address these gaps, this research introduces an oil–water two-phase seepage model to describe the water saturation distribution surrounding the wellbore,and also constructed a wellbore temperature field model that accounts for mechanical and hydraulic heat sources and validated its accuracy. Furthermore, this paper examines the effects of temperature and seepage on shale mechanical parameters. By incorporating a rock failure criterion that accounts for water saturation, a shale collapse model in thermal-seepage coupling is established, which has also been validated. Finally, this paper utilises the grey relational analysis to quantitatively assess the correlation between different variables and collapse pressure. The results indicate that the temperature field and collapse models are consistent with field data. As drilling time increases, water saturation and invasion distance grow, while the seepage radius hardly increases anymore. When porosity exceeds 0.05, its influence on water saturation weakens. Under unbalanced conditions, capillary force drives drilling fluid into the formation. The maximum wellbore temperature is found near the drill bit and exceeds the formation temperature. The bottom-hole temperature difference increases with the increase in well depth and drilling time, sometimes exceeding 20 °C. Among drilling engineering parameters, pump rate has greater influence on bottomhole temperature than WOB, RPM and ROP. Regarding drilling fluid property parameters, density, flow index and consistency coefficient considerably influence bottom-hole temperature. The influence of seepage on collapse pressure is greater than that of the temperature field. Grey relational analysis reveals that pump rate and water saturation have the highest correlation with collapse pressure. As drilling time increases, the correlation degree of pressure difference, horizontal section length and weight on bit gradually increase, while that of the mud behaviour index, consistency coefficient, rock thermal conductivity and mud yield stress slightly decreases.