Generating an extra unit of electricity via PV or wind has no cost. One implication of the growth of renewables is that open-market power prices will therefore tend to fall. As the economists say, prices tend to converge on the marginal cost of production. We are seeing this today in electricity markets. This has profound effects.
In this note I look at the impact of the likely continuing fall in open market electricity prices on one important source of GHG emissions. I try to show that hydrogen production, which is currently almost exclusively carried out by a process using methane and steam, will move to being largely based on the electrolysis of water. Much of the commentary on the energy transition is optimistic about the move to electrification of transport and building heating but deeply pessimistic about reducing the fossil fuels used in industrial processes. In the case of hydrogen manufacture this pessimism is mistaken.
More generally, I suggest that hydrogen will become the dominant route to long-term energy storage, not principally as the gas itself but in the form of methane and liquid fuels.
To be clear, I think hydrogen fuel cell cars stand very little chance of competing against battery vehicles. However I do believe that using water electrolysis to make hydrogen, which is then merged with carbon-based molecules (such as CO2) to create synthetic natural gas and substitutes for petrol and aviation fuel is likely to be the central feature of the next phase of world decarbonisation. For the fossil fuel companies trying to find their way out of reliance on oil and gas, synthetic replacements for existing fuels have to be a key focus of their long-term planning. The manufacture of hydrogen, and the creation of renewable fuels that use this hydrogen, is an activity more similar to the core business of oil and gas companies than PV or wind.
I don’t suggest that regulations or international agreements will cause the shift to renewable hydrogen, but rather that simple economics will drive the oil majors, chemical producers and others towards making fuels from electrolysed hydrogen, rather than natural gas or crude oil.
The fall in wholesale electricity prices will continue
The 6th and 7th June 2017 were windy across northern Europe. During the long days, the sun also shone much of the time. In Germany, two thirds of total electricity output at midday on the 7th came from wind and PV. In the UK, gas-fired power stations were throttled back to not much more than 20% of power generation. Coal generators stood completely idle for much of the period.
The impact on power markets was striking. The average spot price for power for near-immediate delivery fell to very low levels. Germany saw negative figures overnight and near-zero figures for much of the day. The average UK price between 3pm Tuesday 6th and 3pm Wednesday 7th was just over £13 a megawatt hour, or 1.3 pence per kilowatt hour. UK short-term prices were below zero for much of the night. Until recently these were very rare events indeed and they still only happen a few times a week.
But as the installed capacity of renewables continues to increase, this pattern will occur increasingly frequently. Both the UK and Germany continue to expand offshore wind, and PV to a lesser extent. The UK has ambitions to have 30 gigawatts of offshore wind by 2030. Full output from offshore will almost cover summer midday demand by itself. The contribution of PV will mean that renewables will cover total electricity need. It is very difficult to see wholesale prices not reflecting this oversupply in a long-run downward shift.
Nevertheless, the UK government continues to forecast sharply rising wholesale retail electricity prices. From an average of £37 per megawatt hour in 2016, the price is expected to increase more than 50% to £56 in 2030. Households are predicted to be facing retail bills equivalent to £180 per megawatt hour by the same date. Let’s put that number against today’s average wholesale price: £13 is just over 7% of £180, an impossibly large gap. The government’s forecasts are frankly delusional: wholesale electricity prices are coming down, and down they will stay. Absent large tax increases, they will never reach £180 for domestic customers.
Importantly, this permanent deflation of electricity prices will inevitably affect the price of fossil fuels. For a generation we have been used to seeing electricity costs as a largely a derivative of fossil fuel prices. Higher gas costs, for example, used to feed automatically into higher wholesale and retail electricity rates. That link is now beginning to work the other way; falling electricity prices are tending to drive natural gas costs down. If less natural gas is used in power production as a result of the growth of renewables, overall demand for the commodity is lower and the price falls. As EVs become more common, the same linkage is being established with oil. Lower power prices make electric vehicles more attractive, reducing the need for petrol and diesel. As time moves on, the price of electricity will therefore become an important determinant of the price of oil.
Electricity’s role as a price-setter for fossil fuels can be seen most clearly by comparing June 6th-7th UK wholesale price with the cost of gas. At £13, the short-term market price was only just above the equivalent price for wholesale gas of around £12.50 a megawatt hour. In other words, for one 24 hour period, electricity, which is usually regarded as the premium source of energy, was just a few percent more expensive than the fuel which is usually used to make it. (By the way, $50 oil is in energy terms equivalent to about £25 a megawatt hour, or twice the price of gas. In the long run, renewables will also restrain the price of oil from upward movements).
Most electricity is bought and sold on contracts several days or months in advance, and these prices will be substantially higher than those experienced in the spot market of the 7th June. But, nevertheless, the short-term indicators are providing a powerful signal to investors thinking of investing in fossil fuel electricity generation. As wind and solar become predominant sources of electricity, the finances of using gas or coal to make power become more and more parlous. For example, new gas-fired generation will require large subsidies across Europe if power stations are to be constructed.
The tight link between fossil fuel prices and renewable costs will become stronger as electricity becomes an ever larger proportion of all energy use. First, I want to illustrate one example of this which I don’t think gets enough attention: the likely switch from the use of methane to water electrolysis as the main route to making hydrogen.
Hydrogen from electrolysis
The world produces about 50 million tonnes a year of hydrogen. (Some sources suggest it is more than this). The gas is used as an additive in oil refineries, as a raw material for making ammonia and for many different industrial processes including, for example, the making of margarine.
Almost all hydrogen is made today from what is known as ‘steam reforming’, usually of methane (the main constituent of natural gas). A stream of gas is mixed with high temperature steam in the presence of a catalyst. The eventual output of the process is a mixture of CO2 and hydrogen. The valuable hydrogen is collected and the CO2 vented to the atmosphere. If my calculations are correct, the hydrogen produced today through the steam reforming process is resulting in approximately 500 million tonnes of emissions a year, or well over 1% of global GHGs. 
Hydrogen can also be made using electrolysis of water. Electricity is used to split the molecule into hydrogen and oxygen. If made using water electrolysis, global hydrogen production would today use about 15% of world electricity generation. When manufacture of H2 is switched from using methane to employing surplus electricity, hydrogen will be an important method of balancing the world’s grids. When power is abundant, the electrolysers will be turned on. Their work will stop when electricity gets scarce.
In the past, electrolysis was very rarely employed because the energy source, electricity, was more expensive than the gas used for steam reforming.
Is this still true? We need to investigate the energy efficiency of steam reforming and its operating and capital costs as well as the relative prices of gas and electricity.
· Very roughly, a new electrolysis plant today delivers energy efficiency of around 80%. That is, the energy value of the hydrogen produced is about 80% of the electricity used to split the water molecule. Steam reforming is around 65% efficient.
· However, the capital costs of a steam reforming system are currently below the price of a new electrolyser of a similar capacity. The project report for the conversion of the Leeds area in Northern England away from natural gas and towards hydrogen for business and residential use suggested a steam reformer cost of about £600,000 per megawatt of capacity. Like much else in the low carbon economy, electrolyser costs are falling fast. Some manufacturers see electrolyser costs of around £700,000 per megawatt within the next year or so. ITM Power, the Sheffield electrolysis manufacturer, says its costs are already below €1m (about £870,000) for each megawatt of capacity. As the size of electrolysers sharply increases - we may see 10 megawatt devices soon – the cost per unit of capacity will fall. Eventually, electrolysers will be significantly cheaper than steam reforming equipment.
· Electrolysers require little maintenance or much administrative labour. Steam reforming has higher operating costs but I have not been able to obtain clear estimates. (If you happen to have a good source, I’d be very grateful to hear about this). So I have ignored this number.
· Whether the hydrogen is made by steam reforming or by electrolysis, both low and high pressure storage will be required. The costs will be equivalent unless, for example, the electrolyser is only run when electricity prices are low. In this case, the electrolysis route will inevitably require more storage.
We can roughly estimate the relative costs of making hydrogen using electrolysis at different electricity prices and comparing this with the average price of hydrogen in Europe today. As far as I can tell, hydrogen from steam reforming currently costs around 5 pence per kilowatt hour’s worth of energy value supplied to an on-site user. This number is without any cost or taxation applied to the CO2 vented to the atmosphere. Even at today’s low carbon prices, this would add to the fully calculated cost of H2.
When will falling electricity prices make it more economic to create hydrogen from electrolysis? Let’s look at the elements that make up the cost of hydrogen from electrolysis
· The capital cost of the electrolyser. I assume a purchase price (including installation) of €700,000 per MW of capacity to take electricity to generate hydrogen. This is lower than the price that would be achieved today but should be possible by 2019/2020. I suggest that the electrolyser will work perhaps 4,000 hours a year, principally when power is cheap because of abundant wind or solar. At a discount rate of 7%, the owner will need to earn €65,000 a year to cover the cost over 20 years. Per MWh of electricity use over 4,000 hours, the cost is €16.25. For simplicity, I will convert this to £14.15 per MWh at today’s £/€ exchange rate
· The running cost. Estimates for this are scarce but the number is not large. I estimate €5 per MWh, or £4.35. I think this is conservative.
· The electricity cost. This is the critical element. Until the recent sharp falls in wholesale electricity prices, the price of electricity made electrolysis seem expensive. I took a reasonably typical day – yesterday, July 4th 2017 – for the analysis. Unlike the days in early June mentioned at the beginning of this article, it wasn’t particularly sunny or windy. I think it is fair to use this day as being representative of the pattern of summer electricity prices. The average price in the short-term balancing market was £35.87 over the 24 hours. However in the lowest-priced 11 hours (22 half hour periods) it was £23.92. Because I assume that the electrolyser runs 11 hours a day (about 4,000 hours a year), I use this average price.
UK 'balancing market' electricity prices for 4th July 2017
· Add these three elements together and we get a total cost for a 1 MW electrolyser running 11 hours a day of £42.42 per MWh of electricity used to make hydrogen.
· This amount of electricity in an 80% efficient electrolyser will generate about 800 kWh of energy value of hydrogen. (This efficiency is slightly better than can be achieved today by ITM’s PEM electrolysers, but not much).
· 800 kWh of hydrogen produced at a cost of £42.42 means a cost of 5.3 pence per kWh of energy. That’s about 5% more than the costs estimated by the H21 project for the conversion of methane to hydrogen to power homes and businesses in the Leeds area of northern England.
· In other words, at today’s power and electrolyser prices, hydrogen from electricity is almost at the same price as hydrogen made via steam reforming (using the assumptions in the H21 project, which employ a slightly higher methane cost than the current UK price).
· As power prices continue to fall, particularly in periods of high wind and sun, and electrolysers get cheaper and more efficient, the relative advantage of using electrolysis will improve. And there is almost no doubt that this will happen. Hydrogen for chemical plants, fertilisers and other uses will be made using cheap electricity, not methane. Air Liquide, one of the three largest hydrogen manufacturers in the world, has already committed to making 50% of its hydrogen for ‘energy uses’ (such as fuel cell cars) from low-carbon sources, including electrolysis, by 2020.
To sum up: hydrogen may or may not be used extensively in cars. Personally, I doubt it. However hydrogen will become a critical vector in the wider low carbon transition. It will be made using water electrolysis when electricity is sufficiently cheap. That will happen more and more frequently particularly in areas of high sun but where natural gas tends to be expensive. (Australia and Chile are examples). That is the first stage. Then the world will move to using hydrogen as a route that allows cheap electricity to be indirectly turned into renewable gases and liquid fuels.
Once we have inexpensive renewable hydrogen, it becomes possible to transform that hydrogen using standard chemical engineering into renewable fuels. It is all a question of price; there is nothing difficult about making aviation fuel, for example, from hydrogen and waste CO2. We just need electricity to be cheap enough. And a quick look at the pricing charts on electricity grids with a high renewables penetration will show just how fast that day is coming.
Electrolysis is like PV fifteen years ago: a promising technology that is still thought to be more expensive than the fossil fuel alternatives. But, as with PV, it is on a steeply declining cost curve. The manufacture of hydrogen from water is a central part of the next phase of the energy transition.
 One molecule of CH4 combined with H2O in the steam reforming reaction creates 4H2 and one molecule of CO2. The molecular weight of one molecule of CO2 is more than five times four molecules of CO2. And the full GHG emissions resulting from steam reforming need to include the heating of steam and other processes.
 http://www.northerngasnetworks.co.uk/wp-content/uploads/2016/07/H21-Report-Interactive-PDF-July-2016.pdf See the figure of £0.0505 at the bottom of page 260.