There is so much talk about energy efficiency in the electricity system, but very little about time efficiency. In this post Richard Kemp-Harper discusses the case for hydrogen in the time domain.
In Physics, Energy and Time are flip sides of the same coin. There is even an "uncertainty" relationship between the two: ΔEΔt ≥ ℏ/2. I am a bit sensitive to this as I spent my PhD and subsequent science career converting between time and energy domains to such extent that I became an expert at Fourier Transforms by eye.
Electricity and energy are often talked about interchangeably, particularly with all the talk of "electrification of everything", but this is a bit lazy. Electricity is not energy. It is power that becomes energy when it is turned into work by conversion devices. Power is instantaneous and electricity is very tweaky in the time domain. Generation must be met by demand, and an electricity network manager's main role is to manage this balancing.
Electrification is generally efficient in energy, but you pay for that in managing electricity in time.
This is made most visible in the management of grid frequency within the regulatory boundaries of 1% of 50Hz either way, and in reality much tighter to avoid breaching the limits. Using the analogy from my science specialism I can tell you that maintaining a national network to such a narrow band of frequency means power generation needs to be very broad and smooth in the time domain. And that's in response to rapidly fluctuating demand
The result is a whole heap of cost and complexity. There is a national organisation set up to manage this aspect of the network, the Electricity System Operator. The ESO manages a multiplicity of technologies and approaches with different time-domain responses, from generator inertia to reactive demand response.
The problem becomes more acute as we move to more renewable generation where we are dependent on what wind and sun provides. And hence a wide range of innovation in energy storage, demand response and grid management services all designed to smooth the time domain. Competing energy storage technologies that need assessment, integration and management. Or cooperative demand response approaches such as vehicle to grid that have behavioural aspects. All specific technologies designed to solve electricity's time problem and all adding cost and complexity.
The National Grid ESO Control Room
That chart again
Electrified commercial vehicles make the problem worse. With cars, smart charging can fix some of the challenges of grid management as utilisation is low. But commercial vehicles that have a day job need to be charged to be ready whether the wind is blowing or not. So you can try to paint a picture of renewables for vehicles as being efficient, but in reality you need two or three wind farms in different places to provide the power when it's needed.
This is the problem with the infamous hydrogen vehicle efficiency chart. The paper it comes from calculates the renewable energy requirement to power transport: "How many wind farms do we need?" It assumes 100% renewable energy generation and compares the energy efficiency of the transfer and use in vehicles. But it fails to take into account the time domain problem and misses the fact that to solve for time you need to build out more renewables: create the smooth time-domain response from multiple sources . Using the ( frankly unrealistic!) logic from the paper, this means that for commercial vehicles the wind farm to vehicle efficiency is only 50% at best, and so the chart should show the overall efficiency of hydrogen and electric vehicles to be roughly the same.
Time for hydrogen
In contrast to electricity, while hydrogen production from renewables is less energy efficient, it is much better in the time domain. Hydrogen systems provide both demand response and energy storage. Electrolysis can provide fast frequency response and longer term power demand cycling and hydrogen can be stored for, well...as long as you like. One piece of kit, great flexibility in the time domain. Some energy losses, but great gains in managing energy across time, and there is a lot of value in that.
In contrast to other technologies, hydrogen systems provide this at the same time as doing their main job; such as providing fuel for transport or chemical feedstock for industry. It's not as efficient as other technologies designed specifically for energy storage and almost certainly not as cheap. But you get the function of energy storage for free as a by-product. One technology, but a multiplicity of uses.
This means reduced complexity in the energy system as one technology does many things and crucially reduced cost. Or rather the cost of hydrogen systems is spread across all of its applications. I see endless studies calculating the cost of renewable hydrogen production for transport, industry, or heat and few, if any, factor in the energy system cost savings of hydrogen production.
Putting this another way, imagine spending the energy system budgets for energy storage and frequency response on electrolysis and hydrogen storage at scale. Oh, and as a sideline, then selling that hydrogen to end users in transport or industry. The business case for other energy storage technologies is based on the arbitrage between energy prices arising from supply and demand mismatches. Hydrogen systems cross-subsidise energy storage with other revenue streams and vice versa.
The end of competition
Electricity is energy efficient, but constrained in time. Hydrogen is less efficient in energy, but efficient and effective in the time domain. So hydrogen as a parallel energy vector is a friend of electrification, not an enemy. You can imagine an energy system with just renewables and hydrogen - mad but possible. Hydrogen can support the electricity network to maintain its working parameters. It can solve the time problem for electricity. It can enable greater renewable penetration. It can enable storage and movement of energy to places to reduce the need for grid upgrades.
So can we please stop treating them as competing?
Leave a Comment