(432c) Optimal Flowsheets for Electrification of Oil Refineries through Superstructure Optimization

The chemical industry is investing in renewable resources and exploring various initiatives to achieve energy transition and decarbonization goals, such as using green hydrogen, carbon capture and storage technologies, electrification of heat, and electrochemical conversion. However, uncertainties around new technologies, electricity and carbon prices, and capital investment make navigating the transition complex. The industry must change to meet decarbonization standards, but valuable infrastructure should not be prematurely discarded for emerging technologies. There is a dilemma of when to transition and which initiative to adopt, as switching too early leads to inefficiencies and switching too late increases the carbon footprint.

In this work we address the problem of determining the optimal retrofit electrification of an existing oil refinery. We propose a generic process refinery flowsheet that represents the current operations of an oil refinery. We then consider several electrification alternatives that have been identified from the literature. These electrification alternatives together with the refinery flowsheet are included in the superstructure that is to be optimized. Examples of the alternative electrification technologies include e-boilers, alternative pathways for hydrogen generation, connection to decarbonized electricity grids and battery storage. The different electrification alternatives are combined with the initial flowsheet to create a network superstructure of alternative flowsheets in which existing units of the refinery can be removed, replaced by new units, or reassigned to perform a different task, which in turn may require new interconnections of. the units. Data on the technological readiness level, capital and operating costs, and uncertainties associated with each alternative have been obtained from various sources.

The optimization of the superstructure to find the minimum cost electrified refinery is formulated as a Mixed-integer Linear programming model that includes simplified process equations, as well as logical and feasibility constraints for removal and addition of units and interconnections. The proposed objective function includes capital and operating expenses of the potential technologies to be selected, as well as taxes for CO₂ emissions to promote sustainability goals. This MILP model is also extended to a multi-period optimization problem in order to consider long-term planning horizons, e.g. a ten-year long plan for a given forecast of product demands, and that can determine the optimal timeline for switching to the decarbonized options. Numerical results of the proposed MILP models are presented for nominal values of the parameters, as well as for several scenarios that include changes in the costs and performance of electrification technologies, the price of crude oil, and the carbon tax. The results show the potential economic advantages associated the optimal electrification for the retrofit of oil refineries. Finally, we briefly discuss how the proposed MILP models could be extended with stochastic programming techniques to explicitly account for uncertainties in the parameters of the MILP models.