Wajahat Waheed Kazmi

Hydrogen-assisted metal–support balance over TiO ₂ –ZrO ₂ /Ni enables selective chain scission and rapid hydrogenation of LDPE, suppressing side reactions and yielding linear long-chain hydrocarbons (lube-oil range).

Biomass pyrolysis produces three major product fractions—bio-oil, syngas, and biochar—whose yields are governed by nonlinear interactions among feedstock composition, catalyst properties, and operating conditions. Accurate prediction of these product yields remains challenging because literature-derived pyrolysis data are heterogeneous and unevenly distributed across the experimental space. In this study, an interpretable machine-learning framework was developed to predict biomass pyrolysis product yields using readily accessible descriptors from proximate/ultimate analysis, catalyst characterization, and process operation. A curated dataset containing 297 experimental records and 14 input descriptors was used to benchmark multiple regression algorithms, including linear and regularized linear models, support vector regression, k-nearest-neighbour regression, Random Forest, Extra Trees, gradient boosting regression trees, XGBoost, and CatBoost. Target-specific models were developed for bio-oil, syngas, and biochar to maximize the use of available yield data without imputing missing target values. The results showed that nonlinear tree-based ensembles outperform linear and distance-based approaches, while the optimal model depends on the product fraction. SHAP and feature-importance analyses further reveal distinct descriptor–yield relationships. Bio-oil prediction is mainly influenced by catalyst acidity, BET surface area, volatile matter, ash content, and temperature; syngas prediction is dominated by temperature and feedstock composition; and biochar prediction is strongly associated with heating rate, residence time, ash, fixed carbon, and oxygen content. This work provides a transparent data-driven tool for product-yield prediction and descriptor prioritization in biomass pyrolysis. The results showed that nonlinear tree-based ensembles consistently outperformed linear and distance-based approaches. On the independent test set, XGBoost achieved R² = 0.834, RMSE = 5.503 wt%, and MAE = 4.371 wt% for bio-oil, and R² = 0.790, RMSE = 5.427 wt%, and MAE = 3.791 wt% for syngas. GBRT achieved R² = 0.739, RMSE = 3.966 wt%, and MAE = 2.706 wt% for biochar. Repeated five-fold cross-validation further produced mean R² values of 0.783, 0.698, and 0.745 for bio-oil, syngas, and biochar, respectively.

Plasma gasification is a sustainable and advanced technology for the safe and efficient treatment of municipal solid waste (MSW). In this process, a plasma torch serves as the primary heating source to convert MSW into syngas and inert vitrified slag. The produced syngas can be used for various downstream applications, including power generation. In this study, an updraft plasma gasifier is modeled using the Aspen Plus process simulator, with municipal solid waste from Lahore, Pakistan, used as the feedstock. Air is selected as a plasma-forming gas due to its low cost and widespread availability. The primary aim of this research is to analyze the effect of specific torch power and the air-to-feed mass flow ratio on syngas molar composition, syngas higher heating value (HHV), and cold gas efficiency (CGE), and to maximize gasifier performance. CGE of the gasifier is optimized using a surrogate-based model integrated with a genetic algorithm (GA). An artificial neural network (ANN) is employed as the surrogate model for the optimization of CGE. The novelty of this work lies in two key aspects: firstly, this is among the first studies to specifically model and simulate plasma gasification of Lahore’s MSW, capturing its unique waste composition characteristics; and secondly, the integration of process simulation with a data-driven optimization framework using an ANN surrogate model. A total of 1521 data points were generated from the Aspen Plus simulation to train the ANN model and perform optimization in MATLAB. The optimized CGE was found to be 90.6%. Validation of the ANN-GA optimization was carried out by implementing the optimized input parameters in the Aspen Plus gasifier model. The resulting CGE shows a percent relative error of only 0.11% compared to the MATLAB-predicted value, confirming the accuracy of the surrogate model. Furthermore, comparison with the base case simulation reveals that the optimized operating conditions lead to an 8.6% increase in cold gas efficiency, demonstrating the effectiveness of the proposed optimization approach.