MIT researchers outline a path for scaling clean hydrogen production

January 10, 2024

MIT NEWS: Hydrogen is an integral component for the manufacture of steel, fertilizer, and a number of chemicals. Producing hydrogen using renewable electricity offers a way to clean up these and many other hard-to-decarbonize industries.

But supporting the nascent clean hydrogen industry while ensuring it grows into a true force for decarbonization is complicated, in large part because of the challenges of sourcing clean electricity. To assist regulators and to clarify disagreements in the field, MIT researchers published a paper January 8th in Nature Energy that outlines a path to scale the clean hydrogen industry while limiting emissions.

Right now, U.S. electric grids are mainly powered by fossil fuels, so if scaling hydrogen production translates to greater electricity use, it could result in a major emissions increase. There is also the risk that “low-carbon” hydrogen projects could end up siphoning renewable energy that would have been built anyway for the grid. It is therefore critical to ensure that low-carbon hydrogen procures electricity from “additional” renewables, especially when hydrogen production is supported by public subsidies. The challenge is allowing hydrogen producers to procure renewable electricity in a cost-effective way that helps the industry grow, while minimizing the risk of high emissions.

U.S. regulators have been tasked with sorting out this complexity. The Inflation Reduction Act (IRA) is offering generous production tax credits for low-carbon hydrogen. But the law didn’t specify exactly how hydrogen’s carbon footprint should be judged.

To this end, the paper proposes a phased approach to qualify for the tax credits. In the first phase, hydrogen created from grid electricity can receive the credits under looser standards as the industry gets its footing. Once electricity demand for hydrogen production grows, the industry should be required to adhere to stricter standards for ensuring the electricity is coming from renewable sources. Finally, many years from now when the grid is mainly powered by renewable energy, the standards can loosen again.

The researchers say the nuanced approach ensures the law supports the growth of clean hydrogen without coming at the expense of emissions.

“If we can scale low-carbon hydrogen production, we can cut some significant sources of existing emissions and enable decarbonization of other critical industries,” says paper co-author Michael Giovanniello, a graduate student in MIT’s Technology and Policy Program. “At the same time, there’s a real risk of implementing the wrong requirements and wasting lots of money to subsidize carbon-intensive hydrogen production. So, you have to balance scaling the industry with reducing the risk of emissions. I hope there’s clarity and foresight in how this policy is implemented, and I hope our paper makes the argument clear for policymakers.”

Giovanniello’s co-authors on the paper are MIT Energy Initiative (MITEI) Principal Research Scientist Dharik Mallapragada, MITEI Research Assistant Anna Cybulsky, and MIT Sloan School of Management Senior Lecturer Tim Schittekatte.

On definitions and disagreements

When renewable electricity from a wind farm or solar array flows through the grid, it’s mixed with electricity from fossil fuels. The situation raises a question worth billions of dollars in federal tax credits: What are the carbon dioxide emissions of grid users who are also signing agreements to procure electricity from renewables?

One way to answer this question is via energy system models that can simulate various scenarios related to technology configurations and qualifying requirements for receiving the credit.

To date, many studies using such models have come up with very different emissions estimates for electrolytic hydrogen production. One source of disagreement is over “time matching,” which refers to how strictly to align the timing of electric hydrogen production with the generation of clean electricity. One proposed approach, known as hourly time matching, would require that electricity consumption to produce hydrogen is accounted for by procured clean electricity at every hour.

A less stringent approach, called annual time matching, would offer more flexibility in hourly electricity consumption for hydrogen production, so long as the annual consumption matches the annual generation from the procured clean electricity generation. The added flexibility could reduce the cost of hydrogen production, which is critical for scaling its use, but could lead to greater emissions per unit of hydrogen produced.

Another point of disagreement stems from how hydrogen producers purchase renewable electricity. If an electricity user procures energy from an existing solar farm, it’s simply increasing overall electricity demand and taking clean energy away from other users. But if the tax credits only go to electric hydrogen producers that sign power purchase agreements with new renewable suppliers, they’re supporting clean electricity that wouldn’t have otherwise been contributing to the grid. This concept is known as “additionality.”

The researchers analyzed previous studies that reached conflicting conclusions, and identified different interpretations of additionality underlying their methodologies. One interpretation of additionality is that new electrolytic hydrogen projects do not compete with nonhydrogen demand for renewable energy resources. The other assumes that they do compete for all newly deployed renewables — and, because of low-carbon hydrogen subsidies, the electrolyzers take priority.

Using DOLPHYN, an open-source energy systems model, the researchers tested how these two interpretations of additionality (the “compete” and “noncompete” scenarios) impact the cost and emissions of the alternative time-matching requirements (hourly and annual) associated with grid-interconnected hydrogen production. They modeled two regional U.S. grids — in Texas and Florida — which represent the high and low end of renewables deployment. They further tested the interaction of four critical policy factors with the hydrogen tax credits, including renewable portfolio standards, constraints of renewables and energy storage deployment, limits on hydrogen electrolyzer capacity factors, and competition with natural gas-based hydrogen with carbon capture.

They show that the different modeling interpretations of additionality are the primary factor explaining the vastly different estimates of emissions from electrolyzer hydrogen under annual time-matching.

Getting policy right

The paper concludes that the right way to implement the production tax credit qualifying requirements depends on whether you believe we live in a “compete” or “noncompete” world. But reality is not so binary.

“What framework is more appropriate is going to change with time as we deploy more hydrogen and the grid decarbonizes, so therefore the policy has to be adaptive to those changes,” Mallapragada says. “It’s an evolving story that’s tied to what’s happening in the rest of the energy system, and in particular the electric grid, both from the technological as policy perspective.”

Today, renewables deployment is driven, in part, by binding factors, such as state renewable portfolio standards and corporate clean-energy commitments, as well as by purely market forces. Since the electrolyzer is so nascent, and today resembles a “noncompete” world, the researchers argue for starting with the less strict annual requirement. But as hydrogen demand for renewable electricity grows, and market competition drives an increasing quantity of renewables deployment, transitioning to hourly matching will be necessary to avoid high emissions.

This phased approach necessitates deliberate, long-term planning from regulators. “If regulators make a decision and don’t outline when they’ll reassess that decision, they might never reassess that decision, so we might get locked into a bad policy,” Giovanniello explains. In particular, the paper highlights the risk of locking in an annual time-matching requirement that leads to significant emissions in future.

The researchers hope their findings will contribute to upcoming policy decisions around the Inflation Reduction Act’s tax credits. They started looking into this question around a year ago, making it a quick turnaround by academic standards.

“There was definitely a sense to be timely in our analysis so as to be responsive to the needs of policy,” Mallapragada says.

The researchers say the paper can also help policymakers understand the emissions impacts of companies procuring renewable energy credits to meet net-zero targets and electricity suppliers attempting to sell “green” electricity.

“This question is relevant in a lot of different domains,” Schittekatte says. “Other popular examples are the emission impacts of data centers that procure green power, or even the emission impacts of your own electric car sourcing power from your rooftop solar and the grid. There are obviously differences based on the technology in question, but the underlying research question we’ve answered is the same. This is an extremely important topic for the energy transition.”