By David B Watson Chartered electrical engineer; formerly manager of projects at Foster Wheeler Energy,
The arguments in favour of using hydrogen as an energy source - the most abundant gas in the universe, releasing only water upon combustion - appear persuasive. They weaken, however, under closer engineering examination.
Hydrogen doesn’t exist to any useful extent in a free form on Earth; every molecule has to be manufactured, and the cheapest and most widespread method is by steaming natural gas. The resulting hydrogen is known as ‘blue’ if the process incorporates carbon capture, ‘grey’ if it doesn’t.
Depending on the technique used, this steam reforming requires between 33 and 50 per cent more energy than is produced by the resultant hydrogen and produces large volumes of carbon dioxide as a by-product. At utility scale, the UK government Committee on Climate Change in its 2018 report ‘Hydrogen in a low-carbon economy’ described the need “to demonstrate that hydrogen production from [carbon capture and storage] will be sufficiently low-carbon to play a significant role” as it is presently “not close to zero-carbon” due to residual emissions of 20-30 per cent. This will not be plug and play, and perpetuates the use of fossil fuels for transport.
The alternative process of electrolysing water is expensive and needs 20-30 per cent more energy than the hydrogen energy value obtained. This method is described as ‘green’ and promoted by wind farm owners as good use of ‘free’ power that isn’t required by the Grid, but will always cost more than the net cost of the electrical energy used to produce and transport it. And the wind farms have to be built in the first place.
Some new wind farm developments incorporate production of hydrogen by electrolysis when their power output exceeds grid demand, an approach seen as a means to ultimately provide energy storage via the UK gas grid. It is also claimed that that the technique can support electricity grid stability by allowing wind infrastructure to run continuously rather than being constrained off during periods of low electrical demand. This arguably enables them quickly to support the grid when power demand suddenly rises, addressing the threat to frequency control whose rate of change might necessitate load shedding to secure recovery to within legal limits and retain system stability. Some argue that such a set-up would also reduce the level of stabilising rotational inertia required on the grid.
There are some issues with this however. Wind turbines are not directly connected to the grid so they have no inertia. They are connected through power electronics hence have no synchronously rotating mass. Developing deployable ‘synthetic inertia’ may elude us. Also, there has to be wind blowing at the time of any surge in demand, and the speed of response to support grid recovery has to be the order of a very few seconds. Large-scale inertia is always available from thermal generation such as nuclear, coal and CCGT, which can almost instantaneously support grid stability as a shock absorber; here we are introducing uncertainty into the system parameters, an unacceptable probability risk and inherently less dependability.
In global meteorological terms the combined areas of the UK and North Sea are small and will frequently experience almost the same weather with, on many days of the year, no output of consequence from either onshore or offshore wind.
The world’s largest offshore wind farm, the 1.4GW Hornsea 2 off the east coast of England that is due on line in 2022, will also produce ‘green’ hydrogen. The 2MW Dolphyn prototype ‘wind-to-hydrogen’ floating wind farm 15km off Aberdeen, the first of several similar projects, will have no electrical connection to shore and deliver hydrogen only via a pipeline from 2024, with an additional 10 MW turbine planned for 2027.
‘Net Zero: Leading the Way’, a Future Energy Scenario published in 2020 by UK National Grid ESO, anticipates that by 2050 that there will need to be 73GW of electrolysis capacity ramping up and down according to renewable generation availability. That’s three times the UK’s current installed wind capacity, and a considerable investment and huge build rate bearing in mind that many existing wind farms will have reached the end of their working lives by the middle of the century.
Hydrogen promoters point to it having an energy density of around 33kWh per kg, the highest of any known substance. Compare this with petrol at around 13kWh/kg and diesel at 12.6kWh/kg and the numbers seem promising. However, due to its low ambient temperature density, hydrogen when compressed has an energy density of only 1.4kWh/litre at 700 bar assuming perfect combustion, compared to petrol’s 10kWh/litre and diesel’s 10.6 kWh/litre.
Combined with this low energy density, internal combustion drive trains also have much lower efficiency than electric drives, which severely challenges hydrogen ICE vehicle viability. In addition, hydrogen when burnt in air in ICEs produces up to six times the NOx emissions of natural gas, and exhaust gas recirculation is an added expense. Research development continues.
Hydrogen is a low-density gas and extremely light. To store 1kg of it in a fuel cell electric car (FCEV) at atmospheric pressure would need a tank roughly 3 metres by 2 metres by 2 metres - about the size of a small car. So we have to compress it, which requires a lot of energy, expense and infrastructure, all of which affects the overall energy performance. A typical FCEV car available today can achieve a range of about 400 miles with 5kg of hydrogen stored at 700 times atmospheric pressure with a 125-litre tankage capacity (pressure on the wreck of Titanic on the sea bed is only 416 times atmospheric pressure) and travel around 2.85km/kWh. A diesel car travelling 16km/litre achieves 1.5km/kWh, petrol around 1.25km/kWh and fully electric cars typically 4.0-6.5km/kWh. That’s more than twice the motive energy efficiency of FCEVs.
There will more than likely be a future place for hydrogen fuel cells in heavy goods vehicles, taxis and perhaps, less so, some public transport as refill times similar to petrol and diesel keep timetables practicable by sidestepping the problem of slow-charging batteries. But vehicle and transport fixed costs will rise significantly and we need to build the infrastructure. Aberdeen already has a small fleet of FCEV buses that represent an investment of around £500,000 per vehicle. And price tags around £65,000 put current FCEV family cars well out of the range of the majority of the motoring public.
At the huge pressures involved, leakage is an issue. Hydrogen is both an asphyxiant and at 1/14th the weight of air disperses quickly, rises and creates a mixture with a wide flammability range from around 4 per cent to 75 per cent, has very low ignition energy, high combustion energy and a near invisible flame that can reach 2000 degrees C.
Detailed leakage testing carried out last year on two of the principal FCEV cars on the market showed that emissions of hydrogen varied considerably between models. There was concern that whilst in-car equipment leakage plus the purging of hydrogen through the exhaust on start-up and shutdown, which is a common design feature, gave ppm levels that were within the limits recommended by the Society of Automotive Engineers under the short-duration and limited-cycle test regime, the study concluded: “hydrogen emissions performance of vehicles within confined spaces should be paid much more attention”. The control strategy for the purge processes was their main concern.
If we have a home with an integral or even an adjoined garage, buildings insurance companies are prospectively likely to make a ‘cold eyes’ assessment of whether to allow you to keep a hydrogen car in it.
Analysis by academics like Professor Tom Baxter of the University of Aberdeen and manufacturers like Volkswagen of the wind turbine to wheel efficiency of using ‘green’ hydrogen to power vehicles have reached similar indisputably discouraging conclusions. The emerging design approach is for the electricity from the wind turbine to run the electrolysis process offshore. Ignoring the significant energy required to first purify the sea water then, broadly, addressing each process step, electrolysis is about 75 per cent energy efficient. The hydrogen then has to be compressed and transported onshore. This will be about 90 per cent efficient if piped but less if tankered. Around 7 per cent of the energy will be lost in refuelling a 700-bar car however, as stations will compress to 825-1000 bar and pre-cool to around -40 degrees C to prevent the car tanks exceeding 85 degrees C. FCEVs use hydrogen to produce DC electricity which is typically up to 60 per cent efficient then to run the vehicle’s electric motor and to charge its high voltage battery is around 95 per cent efficient. Overall, therefore, the operation from wind turbine to wheel is around 35 per cent energy-efficient or lower.
With pure electric cars (PEVs) supplied from the same wind turbine, around 5 per cent of energy will be lost in transmission due to grid network electrical impedance. Another 15 per cent is lost in charging/discharging of the car’s large-capacity battery and the electric motor drivetrain is about 95 per cent efficient, meaning that PEVs show around 75 per cent efficiency from generation to turning the wheels. That’s twice what is currently achievable with FCEVs, and PEVs travel between 150 and 200 per cent per cent further per kWh of energy.
Our net-zero strategy has to be about less energy consumption, not more. To quote the late Prof David Mackay, “The climate change problem is an energy problem”.
Although the Volkswagen Group has come down firmly on the side of batteries, other manufacturers like Volvo have established that due to the initial higher energy building needs of pure electric cars and their consequently “relatively large carbon footprint” the carbon dioxide break-even point where electric cars become cleaner than internal combustion cars, if supplied 100 per cent with wind-generated electricity, is 31,000 miles. If the charging power is from the current EU mix of sources that rises to 48,500. The challenges are obvious, not least when battery supply chain sustainability and resourcing ethics are also considered.
So it’s not all good news regarding decarbonising our surface transport energy demand toward net zero. Hydrogen ICE and FCEVs aren’t realistic competitors in the private car energy chain and remain both significantly energy-inefficient and very expensive for haulage and public transport.
Cars account for 61 per cent of surface transport emissions, HGVs only 17 per cent, buses 3 per cent, and rail 2 per cent (CCC, December 2020) so for cost/benefit it cannot be worthwhile switching to hydrogen fuel cell buses and trains. Through any impartial lens of engineering science, hydrogen fuel cell cars do not appear to be a transport winner and the Government should revisit decisions it has made about related funding. But then there is political virtue signalling.
The opinions expressed are those of the author and do not necessarily reflect the views of IES.