Data-Driven Framework for the Techno-Economic Assessment of Sustainable Aviation Fuel from Pyrolysis

The aviation sector plays a vital role in global transportation, economic growth, and social integration. However, its rapid expansion has led to increased emissions. Sustainable aviation fuel (SAF) provides a promising solution by offering a clean-burning, renewable alternative to conventional jet fuel. SAF can be produced through various processes and feedstocks, significantly reducing the aviation industry’s environmental footprint. Fast pyrolysis (FP) presents a cost-effective and scalable approach for SAF production due to its low-cost feedstocks, rapid reaction times, and simpler technology. However, estimating the economic viability of FP for SAF production is complex and labor-intensive, requiring detailed process models and numerous assumptions. Furthermore, determining the relationship between feedstock properties and the minimum selling price (MSP) of the fuel can be challenging. To address these challenges, this study developed a data-driven framework for the preliminary estimation of SAF's MSP from FP. Synthetic data was generated using Generative Adversarial Networks (GAN) and Variational Autoencoders (VAE), and hyperparameter optimization was performed using Grid Search to enhance model accuracy and predictions. Five surrogate models were evaluated: linear regression, gradient boost regression (GBR), random forest (RF), extreme boost regression (XGBoost), and elastic net. Among these, GBR and RF showed the most promise, based on metrics such as R², RMSE, and MAE for both original and synthetic datasets. Specifically, GBR achieved a Train R² of 0.9999 and a Test R² of 0.9277, while RF recorded Train and Test R² scores of 0.9789 and 0.9255, respectively. The use of data from the VAE further improved model accuracy. Additionally, a publicly accessible graphical user interface was developed, enabling researchers to estimate the MSP of SAF based on biomass properties, plant capacity, and location.