I have been writing a lot about energy recently, partly because there is a lot of energy news, and also I find it helpful to focus on a topic for a while to help me see how all the various pieces fit together. It is often easy to see how individual components of energy and other technology might operate in isolation, and more challenging to see how they will all work together as part of a complete system. But the system approach is critical. We can calculate the carbon and energy efficiency of installing solar panels on a home and easily understand how they would function. A completely different analysis is required to imagine solar power supplying 50% or more of our energy – now we have to consider things like grid function, grid storage, energy balancing, regulations for consumers and industry, sourcing raw material, disposal and recycling.

This is why just letting the market sort everything out will likely not result in optimal outcomes. Market decisions tend to be individual and short term. When EVs and solar panels are cost effective, everyone will want one. Demand is likely to outstrip supply. Supply chains could bottleneck. The grid won’t support rooftop solar beyond the early adopters. And where is everyone going to charge their EVs? At scale, widespread change in technology often requires new infrastructure and sometimes systems planning.

This can create an infrastructure dilemma – what future technology to you build for? You can take the “build it and they will come” approach, which assumes that infrastructure investment will affect and even determine the future direction of the market. California discovered the limits of this approach when they tried to bootstrap a hydrogen vehicle revolution by building a hydrogen infrastructure. Or you can backfill infrastructure as technology requires it, but this doesn’t quite work either. People won’t buy cars until there are roads, and won’t want to invest in roads until lots of people have cars. At some point you have to bet on future technology – just be flexible and willing to change course as technology evolves.

In the meantime, it is helpful for experts to analyze what a top-down infrastructure might look like in order to optimize cost and carbon efficiency given certain technologies. A recent such effort caught my eye, by MIT researchers, looking into coupling the electricity and hydrogen sectors. They argue that doing so could yield some benefits in reducing the cost of getting to a decarbonization of the energy sector and other industries.

Despite the fact that I think hydrogen vehicles have essentially lost the battle with EVs (because EVs are just more efficient), hydrogen remains a very important energy-carrying molecule. I just think the best use of hydrogen is not zipping around in cars carrying highly pressurized tanks of the stuff. While there still may be a niche for hydrogen when it comes to trains or buses (large vehicles with defined routes), hydrogen will likely have its real impact for grid storage and industrial use.

For grid storage, the goal is to make hydrogen from water using green energy (like solar) to power electrolysis. The hydrogen is then stored and burned back with oxygen to generate energy as needed. (Right now most hydrogen is sourced from fossil fuels, which almost entirely eliminates the benefit.) The main limiting factor of hydrogen as grid storage is that the round trip efficiency is modest, 18-46%. Pumped hydro is the best at 70-85%, while Li ion batteries have an efficiency of about 95%. This may make it seem like Li ion batteries for grid storage are a no-brainer, and from this perspective they are. But if we try to add the massive amounts of grid storage we will need for significant intermittent energy source use, the up front costs can be huge.

If, however, we could get the efficiency of hydrogen up to 50-70% by making the process of creating it more efficient (currently the limiting factor), and make it using currently untapped solar power, it could be an efficient way to get to grid scale storage. But even at current efficiencies, the researchers find that the overall system can be highly cost effective, if we connect hydrogen production for grid storage to the hydrogen pipeline for industrial use. This would have the advantage of displacing fossil fuel-based hydrogen with green hydrogen from renewable energy. Further, hydrogen as an energy-carrying molecule is highly useful in many carbon-intensive industries.

Once you have hydrogen, you can make all sorts of high-energy substances without further using a lot of energy or relying of fossil fuels. You could, for example, make fertilizer or biofuels. Perhaps the key insight from the recent analysis is that it is better to use hydrogen for industry than for grid storage. They write:

We apply this sector-coupling framework to the U.S. Northeast under a range of technology cost and carbon price scenarios and find greater value of power-to-H2 (P2G) vs. P2G2P (power to gas to power) routes. Specifically, P2G provides grid flexibility to support VRE (variable renewable energy) integration without the round-trip efficiency penalty and additional cost incurred by P2G2P routes. This form of sector coupling leads to: (a) VRE generation increase by 13–56%, and (b) total system cost (and levelized costs of energy) reduction by 7–16% under deep decarbonization scenarios. Both effects increase as H2 demand for other end-uses increases, more than doubling for a 97% decarbonization scenario as H2 demand quadruples.

If, therefore, we couple hydrogen production to green energy sources (wind and solar) this increases the efficiency of those sources because they can use their excess energy production to make hydrogen. This becomes more important as the share of renewable energy increases. Further, the greater the demand for industrial hydrogen the greater the efficiency gain from this system. It’s likely that increased availability of cost effective and truly green hydrogen may also lead to greater industrial demand (again, hydrogen is a useful high-energy molecule).  These industries are also difficult to decarbonize without the use of hydrogen. Hydrogen is also useful for carbon capture, so that can be integrated into the system as well.

Their analysis also shows that investment in this hydrogen production-renewable energy infrastructure gets more cost-effective when you increase the price of carbon. This is where a specific regulation might have a positive impact.

The benefits of this plan may be as simple as building hydrogen production plants next to wind and solar farms. We will also likely need to beef up our hydrogen infrastructure – we need efficient ways of transporting all that hydrogen and/or industries that use hydrogen can be built alongside the energy and hydrogen production. It is likely that hydrogen will be playing a significant role in our decarbonized future, but mainly for industrial use.