(130h) Designing Algal Biodiesel Production Networks for Maximum Carbon Sequestration Using Hybrid Mechanistic Machine Learning Approach

The pursuit of sustainable and carbon-neutral energy sources is crucial in addressing climate change and fostering energy security. Algal biodiesel has gained significant attention as a viable renewable biofuel, due to its high lipid content and rapid rate of utilization as Sustainable Aviation Fuel (SAF).¹ However, the optimization of the algal biodiesel production network comprising of several interconnected industrial systems to achieve maximal decarbonization remains a complex challenge. This research introduces an innovative approach that utilizes surrogate modeling and optimization of the algal biodiesel production network via the Sparse Identification of Nonlinear Dynamics (SINDy) algorithm to enhance the decarbonization rate.²

The research methodology encompasses the development of data-driven surrogate models for the algal biodiesel production network, leveraging the ML-based SINDy algorithm. By exploiting sparse regression techniques, the SINDy algorithm effectively identifies generalizable parsimonious models that captures the nonlinear dynamics and interactions within the production network. These surrogate models substantially reduce computational complexity while preserving the essential characteristics of the underlying system, providing a robust foundation for subsequent optimization.^3,4

With the established surrogate models, a sophisticated telecoupled integrative framework is constructed to maximize the decarbonization rate of the algal biodiesel production network. This framework employs the use of numerically integrated values of coupled surrogate models to concurrently optimize economic and environmental objectives of the research, resulting in a sustainable and balanced network. The optimization process pinpoints optimal operating conditions and strategic interventions, including resource allocation and process improvements, to elevate the decarbonization rate.

The outcomes of this study demonstrate the robustness of the SINDy-based surrogate modeling and optimization approach in significantly enhancing the decarbonization rate of the algal biodiesel production network. By incorporating the identified strategies, the algal biodiesel industry can contribute to climate change mitigation, support sustainable energy production, and curtail reliance on fossil fuels.⁵ Thus, this research advances the scientific understanding of sustainable biofuel production and has potential applications in other renewable energy systems.

References:

1. Kandaramath Hari, T., Yaakob, Z. & Binitha, N. N. Aviation biofuel from renewable resources: Routes, opportunities and challenges. Renew. Sustain. Energy Rev. 42, 1234–1244 (2015).
2. Brunton, S. L., Proctor, J. L. & Kutz, J. N. Discovering governing equations from data by sparse identification of nonlinear dynamical systems. Proc. Natl. Acad. Sci. 113, 3932–3937 (2016).
3. Kaptanoglu, A. et al. PySINDy: A comprehensive Python package for robust sparse system identification. J. Open Source Softw. 7, 3994 (2022).
4. Kukreja, S. L., Löfberg, J. & Brenner, M. J. A LEAST ABSOLUTE SHRINKAGE AND SELECTION OPERATOR (LASSO) FOR NONLINEAR SYSTEM IDENTIFICATION. IFAC Proc. Vol. 39, 814–819 (2006).
5. Leite, G. B., Abdelaziz, A. E. M. & Hallenbeck, P. C. Algal biofuels: Challenges and opportunities. Bioresour. Technol. 145, 134–141 (2013).