Optimization of In-Situ Hydrogen Production via Cyclic Air-Steam Injection in Heavy Oil Reservoirs

Abstract Amid the global shift towards sustainable energy sources, cyclic air-steam injection has emerged as a groundbreaking method for producing clean hydrogen directly from abundant heavy oil and bitumen reservoirs. The design and optimization of engineering decisions are essential for forecasting and maximizing hydrogen yield economically. In this study, we aim to enhance in-situ hydrogen production by optimizing cyclic air-steam injection operations in a heavy oil reservoir. A two-dimensional (2D) Lloydminster heavy oil reservoir model with two horizontal wells was employed to simulate hydrogen production using CMG STARS. Air and steam were alternated in two-month injection periods over multiple cycles for six years. To identify the optimal injection strategy for maximizing hydrogen production, a differential evolution optimization algorithm was used to fine-tune the air injection rate, steam injection rate, injected air oxygen concentration, and injected steam temperature. Two injection strategies were evaluated: Strategy 1 with fixed parameters for all air and steam injection cycles and Strategy 2 with varying injection parameters for each air and steam cycle. A total of 1000 optimization simulations were conducted for each strategy. Results showed that hydrogen production varied significantly, ranging from 2.44×104 to 1.63×105 m3 from this 2D model. Strategy 2, which involved varying the injection parameters for each air and steam injection cycle, achieved the highest cumulative hydrogen production of 1.63×105 m3 compared to 1.59×105 m3 with Strategy 1, which used fixed parameters for all the air and steam injection cycles. The production trends differed markedly between the two injection approaches. Strategy 1 showed a rapid increase in cumulative hydrogen production compared to Strategy 2, reaching 72% of the total production after just two years. Strategy 1 involved injecting a total of 1×105 m3 of air with a constant 21% oxygen concentration, whereas Strategy 2 injected 2.8×105 m3 of air with oxygen concentrations ranging from 21% to 95%, with a median of 57%. A higher water volume was injected as steam in Strategy 1 (450 m3) compared to Strategy 2 (380 m3). However, injected steam temperatures were higher in Strategy 2 (100–295 °C, with an average of 155 °C) compared to 100 °C for all steam injection cycles in Strategy 1. Despite the slightly higher achieved hydrogen production using Strategy 2, Strategy 1 proved to be more economical, offering higher hydrogen production rates, lower air and oxygen requirements, and reduced power needs for steam generation. The findings provide valuable insights into optimizing cyclic air-steam injection for carbon-zero in-situ hydrogen production, establishing a solid foundation for its future field implementation.