Power-to-gas: the remaining critical ingredient in the energy transition

A windy week in Germany produced the expected result. Wholesale electricity prices from 19th to 26th February 2017 dipped below zero four times and much of the weekend saw figures below €25 a megawatt hour. This pattern is increasingly frequent across many electricity markets. As the Economist pointed out last week, the arrival of large scale renewables with zero operating cost is eating away at the businesses of those companies reliant on selling on the open market. €25 does not pay for the cost of the gas to generate a megawatt hour in a power station.

German electricity production

(Prices are the wavy lines at the bottom of the chart. Electricity production from wind is the light green area)

Source: Energy-charts.de. (Best site in the world for full public information about a power market!)

Source: Energy-charts.de. (Best site in the world for full public information about a power market!)

In the US, NRG, which is the largest independent producer of power, summed up the problem by saying its business model was now ‘obsolete’. Lower and lower prices are making it impossible to produce electricity from gas or coal in markets increasingly captured by solar and wind. Equally, no-one can raise the finance to build new power stations, even in those countries with ageing fleets, such as the UK, because of low prices and fewer and fewer hours of operation. This problem will get worse.

Whether you are an enthusiast for a fast transition to a renewables-based energy system or are sceptical about the pace of change, the destruction of the traditional utility by the eating away of wholesale prices is not good news. It increases the possibility that the increasingly rapid switch to renewables around the world will be brought to a shuddering halt by governments worried about the security of energy supply because of the intermittency of wind and solar. Although we can make huge progress in adjusting electricity use to varying supply, ‘demand response’ will never be enough to deal with weeks of low wind speed and little sun in northern countries.

I want to put forward the view that there is only one way to deal with this problem. When power is in surplus, it needs to be turned into natural gas. This will reduce the amount of excess electricity and provide renewable gas for burning in power stations when renewables are in short supply. ‘Power-to-gas’ is the critical remaining ingredient of the energy transition. Can I put this as strongly as I can? Without a rapid and whole-hearted commitment to this technology, the renewables revolution may ultimately fail.

Power to gas

Electricity can be used to split water into hydrogen and oxygen in the reaction known as electrolysis. The hydrogen is then combined with carbon dioxide, either using biological techniques or through the conventional Sabatier process. This generates methane, the main part of natural gas. If the CO2 used in the reaction is derived from organic sources, such from anaerobic digestion, it is ‘renewable’.

What is the net impact of this transformation of electricity to natural gas? First, the surplus of electricity is reduced. Second, the energy in the electricity is largely transferred to the energy in methane. This methane can be indefinitely kept in natural gas networks, which generally have a capacity for storage vastly greater than the batteries are ever likely to possess. Although Britain has relatively little gas storage, other countries often have months of capacity. They can make gas when electricity is abundant and then use that gas to generate power when the wind and sun are not available.

The energy economics of power to hydrogen

Large amounts of hydrogen are generated today around the world. The gas is almost entirely created through a process known as ‘steam reforming’ which takes methane and water creating hydrogen and carbon dioxide. The CO2 is vented to the atmosphere, thus adding to global emissions. Very approximately, hydrogen made from methane costs about twice the cost of natural gas per unit of energy carried. So if natural gas (mostly methane) costs 1.6 pence (2.0 US cents) per kilowatt hour, which is approximately the current wholesale rate in the UK, then producing a kilowatt hour of hydrogen will cost about 3.2 pence (4.0 cents).

The alternative way to produce hydrogen is through water electrolysis. This uses electricity and until recently the conversion process has been less than 70% efficient. And, generally speaking, electricity has been several times expensive than natural gas per kilowatt hour. A commercial customer might have bought electricity at 8 pence a kilowatt hour or more, meaning that at 70% efficiency hydrogen costs about 12 pence per kilowatt hour (14.6 cents) or almost four times as much as gas produced from methane. Clearly, no-one produces hydrogen using electrolysis unless they are remote from steam reforming plants.

Electrolysers are getting much cheaper and more efficient. We will see electrolysis costs fall to around $400/kilowatt and efficiencies rise above 80%. However making hydrogen from power will still be usually more expensive than from steam reforming of natural gas.

But look again at the chart of German prices above. Anybody owning an electrolyser that could work when electricity prices are low would have been able to make hydrogen for much less than from methane for much of last week. Very roughly, at any time the German power price was below €25, an electrolyser could make hydrogen more cheaply from electricity than from gas. That is, if the electrolyser owner could get access to inexpensive wholesale power, it could absorb cheap electricity. I reckon – but do not have the numbers to prove this – that German prices were below €25 per megawatt hour for at least 30% of last week.

This is a complicated area so please let me labour this point. The evolution of power markets is pushing the typical short-term wholesale price of electricity down to historically unprecedented levels. At the same time, the commercial and household price of power is rising as subsidy and electricity network costs rise as the renewables revolution takes hold. The low wholesale price of power at times of wind or of strong sun means that making hydrogen from electrolysis is often cheaper than using natural gas. And as wind and solar capacity rises, this reversal of usual pricing differences is going to happen far more frequently.

Of course most business do not buy power through a wholesale market, and almost everybody has to pay grid distribution charges. So the logical place to put these electrolysers is next to wind farms or solar parks which can use power at no direct cost. When these entities are expecting to get very low power prices they will swing over to making hydrogen instead.

Hydrogen to methane

Hydrogen is useful and will grow in importance. But moving it around is complicated and expensive. So I think it will be used predominantly at the point of production, either for chemical products, fuelling fuel cell cars or making methane. In my view, it is making methane that offers by far the most important opportunity because it can be stored and transported so much more efficiently than hydrogen.

Methane (CH4) can be made from hydrogen and CO2 in one of two main ways. The traditional Sabatier process offers a simple route, albeit with substantial energy loss. That is, one kilowatt hour of hydrogen (you’d get this by burning about 25 grams of the gas) turns into about 0.75 kilowatt hours of methane. The rest is lost as heat. The second is biological. Some microbes in the class called Archaea can absorb hydrogen and CO2 and exude methane as a waste product. Their efficiency is about the same, or slightly better, turning up to 80% of the energy in hydrogen into methane. They can make the transformation quickly and in relatively low cost production systems. As I say in The Switch, the leading contender is a German company called Electrochaea which operates its first 1 megawatt plant near Copenhagen getting its CO2 from a stream of biogas out of a wastewater treatment plant. The CO2 is free. In fact it should have a negative cost since it allows the whole stream of biogas to be feed into the natural gas grid rather than inefficiently burnt in gas turbines on site.

Think of methane as identical to natural gas, although the gas in pipelines also contains varying amounts of longer molecules. If we use surplus electricity to make hydrogen and then combine it with CO2 to make methane, then we are losing energy at two different stages: electrolysis and methanation. Very roughly, the best we can hope for is to obtain 65% of the energy in electricity out of the process in the form of methane.

Natural gas trades at about 1.6 pence (2.0 US cents) per kilowatt hour at the central trading point in the UK. How cheap does electricity have to be to make it financially attractive to use it to make ‘renewable’ methane? Very roughly, and before the operating costs of the machines, it has to be 1.6 pence times 65% or just over 1 pence per kilowatt hour (1.25 US cents).

The German market operated at less than this price for about 35 hours last week, or one fifth of the time. In all those periods, an electrolyser could have been profitably making hydrogen to be converted back into methane. The methane – which has very low greenhouse gas emissions because it has been made from renewable electricity and the CO2 from organic waste – can be pumped into the gas grid. It can then be used to make power in a gas turbine when electricity is in short supply.


To most people in the utility industry, the idea that it can possibly make sense to use valuable electricity to make cheap natural gas still seems absurd. They aren’t looking at the charts, I say. As wind and solar electricity grows in importance, the cost of power will inevitably drift towards zero. (First year economics tells us that prices always edge towards the marginal cost of production). Electricity will become cheaper than gas. On a windy weekend night in the North Sea offshore turbines will produce more electricity than northern Europe needs at some date in the not-to-distant future. Negative wholesale electricity prices will become increasingly prevalent.

We really need this to happen. First, it means we can happily heat buildings with low carbon electricity, even without the advantages of heat pumps. More important, it means that instead of using fossil natural gas for power and heat generation, we can use renewable natural gas instead, particularly when power is costly because of lack of wind and sun.

The central argument of this article is thus that the right way to ‘fix the broken utility model’ that the Economist talks about is to link the gas and electricity markets through large-scale application of power-to-gas technologies. Big utilities talk about understanding the need for decentralisation but the reality is that they will be terrible at moving away from centralised production plants. What they would be good at is running large scale electrolysis and methanation operations that allow them to continue to run CCGT power plants when electricity is scarce. We will not need capacity payments or other complex subsidies and incentive schemes. By creating a continuing role for CCGT we will have found a way to keep our energy supply secure without threatening decarbonisation objectives. 




1.     With many thanks indeed to Vyas Adhikari for his help understanding some of the questions of chemistry and energy transformations involved. Errors are all mine.

2.     The material in the piece above is highly compressed. I’m happy to provide more analysis and back-up if anyone is interested.