Producing clean syngas from biomass and plastic waste offers a promising path to sustainable energy, but the underlying thermochemical processes are highly complex and difficult to optimize. This study introduces an interpretable machine learning framework that can accurately predict the yield and H2/CO ratio of key syngas components during cogasification. By analyzing the influence of feedstock composition and operating parameters, this model identifies the key factors controlling product distribution and provides mechanistic insights into the reaction system. The findings support improved syngas quality, reduced experimental workload, and more efficient process optimization, providing practical guidance for renewable fuel production and waste utilization.
Synthesis gas, mainly CO, H₂, CH₄, CO₂, and light hydrocarbons (C2-C4) is an essential intermediate for power generation and chemical synthesis. Although biomass gasification is recognized as an alternative to fossil fuel-based energy, the increase in plastic waste has increased interest in co-processing biomass and plastics to increase hydrogen yield and reduce tar production. However, experimental cogasification remains time-consuming and limited by variable feedstock properties and multiple operating parameters. These challenges complicate mechanistic understanding and hinder systematic optimization. Based on these challenges, there is a critical need to develop predictable and interpretable tools to guide cogasification design and operational control.
Researchers from Hainan University and Tianjin University have reported a new interpretable machine learning framework that can predict the composition of syngas during co-gasification of biomass and plastics. This study was published in Waste Disposal & Sustainable Energy in 2025 (DOI: 10.1007/s42768-025-00256-z). The team evaluated four machine learning models and found that CatBoost achieved the highest predictive accuracy. They further applied Shapley additive explanations (SHAP) analysis to reveal how key variables such as temperature, steam-to-fuel ratio, and biomass ratio influence the syngas yield distribution.
The researchers compiled a dataset of 380 experimental data points covering multiple biomass types and three common plastics: polyethylene, PE, polyethylene terephthalate, PET, and polystyrene, PS. The dataset includes 20 input variables such as elemental composition, proximate analysis, temperature, steam/fuel ratio, and equivalence ratio. Four machine learning models were trained and compared: CatBoost, Random Forest, Support Vector Machine, and XGBoost. CatBoost showed the best performance, achieving R² values of 0.80 to 0.94 on the test set across key syngas components.
To increase interpretability, we used SHAP analysis to quantify the contribution of each feature to model predictions. Temperature and steam/fuel ratio were identified as the most influential operational parameters. High temperature promoted the conversion of CH4 and CO2 to CO and H2, and increased steam increased H2 but suppressed CO. The biomass proportion significantly influenced carbon conversion, increasing CO2 while decreasing light hydrocarbons and lowering the H2/CO ratio. Interestingly, plastic ash content emerged as a strong proxy variable reflecting important physicochemical properties that shape syngas composition.
These insights provide practical guidance on feedstock formulation, selection of operating conditions, and catalyst-free optimization strategies.
“The combination of machine learning and interpretable analytical tools provides a new route to decipher the complex reaction behavior in the co-gasification of biomass and plastics,” the authors state. “Rather than completely replacing experiments, this framework supports targeted design and more efficient testing by identifying which variables are most important. It represents a step forward in data-driven clean energy research.”
This framework allows researchers and industry engineers to optimize syngas production without extensive trial-and-error experimentation. This model can guide the selection of biomass and plastic mixtures that maximize hydrogen yield, reduce tar formation, and promote cleaner fuels and waste-to-energy conversion. This approach can be extended to other thermochemical systems where complex feedstock and reaction interactions exist. As countries move towards carbon-neutral energy goals, the integration of machine learning into renewable fuel production pipelines is likely to play an increasingly central role in accelerating the deployment of sustainable technologies.
/Open to the public. This material from the original organization/author may be of a contemporary nature and has been edited for clarity, style, and length. Mirage.News does not take any institutional position or position, and all views, positions, and conclusions expressed herein are those of the authors alone. Read the full text here.
