Hydrogen production enhancement using exhaust heat from HCNG engine: ASPEN plus simulation and machine learning prediction

Hydrogen production plays a pivotal role in advancing clean energy solutions for transportation and power generation. However, conventional methods face challenges due to their high energy requirements and inefficiencies. This study investigates a novel strategy to improve hydrogen production by integrating exhaust heat recovery from Hydrogen-enriched Compressed Natural Gas (HCNG) engines with the Steam Methane Reforming (SMR) process. The research analyzes hydrogen production by exhaust heat utilization at a hydrogen ratio of 20 %, an Exhaust Gas Recirculation (EGR) ratio of 24 %, an engine load of 75 %, and an engine speed of 1700 rpm under stoichiometric conditions. Hydrogen production via the SMR process is simulated using ASPEN Plus software, with a detailed evaluation of heat exchanger and reformer component heat duties. Results indicate that at 973 K, increasing the steam-to-methane ratio (S/C) from 1 to 6 leads to a rise in the hydrogen production rate from 4.01 kg/hr to 6.85 kg/hr. Additionally, the maximum heat recovered from the HCNG engine exhaust reaches 89.23 kW out of a total available 133.12 kW under the specified conditions. Another key objective of this study is to predict hydrogen production using machine learning regression models, including Stepwise Linear Regression (SLR), Decision Tree (DT), Linear Support Vector Machine (LSVM), and Boosted Tree (BT). The models are evaluated based on Root Mean Square Error (RMSE), Mean Absolute Error (MAE), coefficient of determination (R²), and computational time across different datasets and input parameters. Among the models, SLR demonstrates superior performance, achieving an RMSE of 0.514 with a single input and 0.074 with four inputs. The SLR showed the minimum MAE is 0.06 and the highest R² value is 0.99, which confirms the strong predictive ability. These findings contribute to the advancement of HCNG engines integrated with SMR and electronic control units, offering a pathway to more efficient and sustainable hydrogen production.