Finding the best decarbonization pathways for refineries

Policymakers and industry leaders in the US are striving for energy security amidst a global increase in energy consumption and global movement away from fossil fuels. The petrochemical industry, one of the most energy-intensive, faces growing pressure to adopt new and more sustainable technologies.

A decision-making tool from researchers at Carnegie Mellon can help individual industries like petrochemicals find their optimal decarbonization strategies.

In Industrial & Engineering Chemistry Research, Ana Inés Torres, Ignacio Grossmann, Sampriti Chattopadhyay, and their collaborators from Shell introduce a framework based on a multiperiod mixed-integer linear programming (MILP) model. It predicts the optimal technology switch strategy for an oil refinery to implement decarbonization technologies. “A company can use our framework to find the best way to transition a refinery toward decarbonization,” says Chattopadhyay, a Ph.D. student in chemical engineering.

Within the US industrial sector, decarbonizing oil refineries is uniquely challenging because of their high efficiency and varied configurations, processes, and energy inputs. Torres, associate professor of chemical engineering, Grossmann, professor of chemical engineering, and Chattopadhyay based their framework on production capacity, plant structure, location, and availability and cost of renewable energy sources. Their approach minimizes cost while considering emissions restrictions.

A company can use our framework to find the best way to transition a refinery toward decarbonization.

Sampriti Chattopadhyay, Ph.D. student, Chemical Engineering

The major sources of emissions in refineries are the combustion of fossil fuels to generate heat, the production of hydrogen, and the processes that break long-chain hydrocarbons into simpler and more useful molecules. Retrofitting refineries to decarbonize hydrogen and steam production targets these emission sources and preserves current equipment, a more economical and environmentally-friendly solution than discarding valuable infrastructure.

Torres, Grossmann, and Chattopadhyay considered electrification for steam generation, green hydrogen production, and carbon capture for blue hydrogen production and other emission sources. Green hydrogen production is powered by renewable electricity. Blue hydrogen relies on carbon capture to offset the emissions from the natural gas used in production.

Industries have many options to reduce their carbon emissions and many methods to assess specific technologies. There are far fewer tools for evaluating transition pathways, which is where Torres, Grossmann, and Chattopadhyay focus. Their framework based on a multiperiod MILP model helps decision-makers navigate uncertainties such as investments in developing technologies, fluctuations in electricity and carbon prices, the availability of renewable energy, and the need to balance existing infrastructure with new technologies.

“There is no one-size-fits-all solution,” says Chattopadhyay. “Every industry has a different way of operating. Each plant has different access to renewable energy, based on its location. This research gives them a tool to find out what is optimal for them. I think that’s a new way of looking at the decarbonization problem.”

Torres, Grossmann, and Chattopadhyay applied their MILP optimization framework to two different refinery configurations. They conducted sensitivity analyses to see how much the results change when the inputs are changed. The two case studies highlight the effect of energy prices and environmental policies.

“Based on the data that we used for projected energy prices and technology costs, we show that natural gas with carbon capture is more economical than the electrification-based alternatives we considered,” says Chattopadhyay.

The framework only favors electrification-based technologies over carbon capture when the price of electricity decreases more than 60% or there are very stringent carbon emissions regulations. The case studies also showed that although carbon taxes or credits motivate earlier adoption of carbon capture technologies, they do not promote electrification.

Torres, Grossmann, and Chattopadhyay’s framework is generalizable beyond oil refineries. During an internship at Shell last summer, Chattopadhyay used a similar mathematical model to build a decarbonization framework for a plastic plant in Monaca, PA.

Although developing technologies hold promise, there are not yet clear alternatives for plastics, aviation fuel, and other petrochemical products on which we rely. Optimizing the retrofit of oil refineries and other high carbon intensity sectors of US industry will accelerate progress toward sustainability and energy security goals.