Moving boundary thermo-hydro-mechanical model for perforation-controlled multi-fracture propagation

Multi-cluster hydraulic fracturing couples fracture interaction and nonuniform fluid allocation with transient cooling of the surrounding formation. This study evaluates fracture competition and its thermal consequences using a sequential moving-boundary framework: a displacement-discontinuity hydro-mechanical model provides fracture geometry, aperture, flow rate, and exposure time, after which rock and fracture-fluid temperatures are calculated on the evolving fracture mesh. The model accounts for evolving fracture geometry and nonlinear leakoff, and its thermal component is validated against the Dysart–Whitsitt analytical solution. Three, four, and five-cluster systems are simulated under uniform and differentiated completion designs. Uniform perforation causes progressive edge-fracture dominance as cluster density increases, leading to strong localization of fracture volume, fluid intake, and reservoir cooling. In the five-cluster case, the final-flow and cooling-area coefficients of variation reach 86% and 55%, respectively. A differentiated design combining nonuniform cluster spacing with higher perforation resistance at edge clusters and lower resistance at internal clusters converts unstable flow capture into bounded dynamic redistribution. For five clusters, this strategy reduces the corresponding coefficients of variation to 21.2% and 11.9%, while restoring the hydraulic and thermal contributions of the internal fractures. Among the investigated cases, four differentiated clusters provide the best compromise between total fracture extent, cooled reservoir contact, and injected-volume efficiency, whereas three clusters provide the highest geometric uniformity and fracture-volume conversion. These results demonstrate that cluster number should be optimized jointly with injection capacity, stress-shadow intensity, cluster spacing, and perforation resistance.