A promising feature of renewable energy projects in Brazil is the potential to include the sale of decarbonization credits (CBIOs) within the supply chain models. These credits, which represent a verified reduction in carbon dioxide emissions, can be traded to generate additional revenue, making decarbonization financially attractive. In biofuel production, the commercialization of carbon credits provides an economic incentive that supports the adoption of sustainable practices. Previous studies (Carpio et al., 2021; Machado and Abreu, 2024) have shown that incorporating CBIOs into supply chain models not only increases the economic viability of biofuel production but also encourage stakeholder preferences for sustainable practices, besides supporting the alignment of environmental and economic objectives.
This study presents an optimization model for the RLG supply chain, considering four decision layers: feedstock, production, storage, and market distribution. The model is formulated as a mixed integer linear program, using a multi-period approach, integrating spatial and temporal decisions to maximize the Net Present Value (NPV) over a 20-year horizon. The supply chain network includes the selection of feedstock sources, the installation of production and storage facilities, the choice of conversion technologies, and the allocation of product flows to different market regions. The optimization considers investment and operational costs, transportation expenses, feedstock availability, and market demand constraints. Additionally, a life cycle assessment based on the Brazilian carbon intensity method (Matsuura et al., 2018) is incorporated to estimate the carbon intensity of the produced RLG, allowing the monetization of decarbonization credits.
The feedstock decision layer represents the material flows for purchasing and transporting raw material, while considering its annual availability. Once acquired, the feedstock is transported to a production center, incurring both transportation and purchase costs. Transportation costs depend on the unitary cost and the distance between the source and destination locations. In the production decision layer, the feedstock is converted into products and co-products using specific technology, which requires the installation of a production facility at selected location. The processed feedstock quantity depends on the facility's operational status and is constrained by the technology's processing capacity. The production process also involves the consumption of utilities and inputs, as well as the generation of residues, contributing to the operational costs. The storage decision layer defines storage centers, represented by a binary variable indicating whether a storage facility is installed. A mass balance constraint controls the inventory levels of RLG at each period, accounting for inflows from production, transfers between storage centers, sales to markets, and stock from previous periods. Finally, the market decision layer ensures that the demand for RLG in each market is supplied. The model incorporates tax policies related to origin and destination. The profit for each period is calculated as revenue minus costs, with an additional deduction for profit taxes. In the objective function, profits are discounted using a real interest rate to determine the NPV are discounted in the initial period.
The model was applied to a case study on RLG production from glycerol to supply major consumers, represented by state capitals. Glycerol is a residue of biodiesel production, and given Brazil's extensive biodiesel industry, a great glycerol volume is generated, creating an opportunity for its conversion into renewable fuels. The supply chain is only economically viable if the RLG price is 70% higher than LPG. Feedstock acquisition represents the main NPV cost and since the conversion efficiency is 28%, the production of RLG requires a large quantity of glycerol. Investment costs have a great impact on NPV and are minimized by installing a single large-capacity center instead of multiple smaller ones. Low transportation costs, accounting for 6% of negative components of NPV justify the centralized network. All the available glycerol in Brazil was used to replace 33% of the demand of the main consumer, São Paulo capital, and fully supply the demand in other cities. The sales of CBIO credits accounted for 1% of total revenues and the carbon intensity of RLG produced by the optimal network, using hydrogen produced by steam methane reforming, was 42% lower than LPG. Emissions resulting from hydrogen production account for 75% of the total life cycle emissions of RLG, while the emissions from distribution to the end consumer represent only 4% of the total. When adopting hydrogen produced from steam methane reforming with carbon capture, the carbon intensity of RLG is reduced by 71% compared to LPG. Despite the additional revenue from CBIO sales, the NPV decreases by 58% due to the higher costs of hydrogen production with carbon capture compared to conventional hydrogen production without carbon capture. When hydrogen produced via alkaline electrolysis using renewable energy is employed, the selling price of RLG must be 90% higher than LPG to ensure an economically viable supply chain, explained by the higher cost of hydrogen production. Additionally, with hydrogen from alkaline electrolysis, RLG distribution emissions become the major contributor to its life cycle emissions. The optimized network remains unchanged for all scenarios evaluated with different hydrogen sources, indicating the resilience of the proposed supply chain.
This study highlights the potential of RLG as a viable alternative to LPG in the Brazilian market, where an abundance of glycerol can support its production. However, the results indicate that the economic viability of RLG is sensitive to production costs, feedstock acquisition, and logistical challenges. The high cost of feedstock, driven by the low conversion efficiency of the glycerol route, proved to be one of the main challenges in exploring RLG. Advancing the technological development of the glycerol route could lower costs, enhance profitability and facilitate the RLG market integration. Additionally, the role of hydrogen in RLG production significantly impacts both costs and emissions. While cleaner hydrogen production methods can reduce carbon intensity, they also introduce financial trade-offs that affect the overall feasibility of the supply chain. The inclusion of carbon credits provides an economic incentive to offset some of the costs, but their contribution remains limited compared to the primary cost drivers. Future efforts should focus on improving process efficiency, optimizing hydrogen sourcing strategies, and exploring policy mechanisms that enhance the competitiveness of renewable fuels. By addressing these challenges, RLG can become a more viable and sustainable option for replacing fossil-based LPG.
References
Carpio, R. R., de Carvalho Miyoshi, S., Elias, A. M., Furlan, F. F., de Campos Giordano, R., & Secchi, A. R. (2021). Multi-objective optimization of a 1G-2G biorefinery: A tool towards economic and environmental viability. Journal of Cleaner Production, 284, 125431.
Machado, R. L., & Abreu, M. R. (2024). Multi-objective optimization of the first and second-generation ethanol supply chain in Brazil using the water-energy-food-land nexus approach. Renewable and Sustainable Energy Reviews, 193, 114299.
Matsuura, M. I. S. F., Scachetti, M. T., Chagas, M. F., Seabra, J., Moreira, M. M. R., Bonomi, A., ... & Novaes, R. M. L. (2018). RenovaCalcMD: Método e Ferramenta Para a Contabilidade da Intensidade de Carbono de Biocombustíveis no Programa RenovaBio. 2018. Available in: http://www. anp. gov. br/images/Consultas_publicas, (10).
Theozzo, B., & dos Santos, M. T. (2021). A MILP framework for optimal biorefinery design that accounts for forest biomass dynamics. Computers & Chemical Engineering, 146, 107201.