Q3 2020: How to handle 40 GW of wind?
Download PDF
by Dr Iain Staffell – Imperial College London
Headline annual figures mask the huge variation in wind output that will be experienced in future.
Day-to-day management of Britain’s power system will be very different with double our current wind capacity. The charts below show the daily generation mix over the course of a year, both as it happened with the 2019 generation mix, and how it would look with 2030’s generation mix.
National Grid ESO is comfortable managing the current levels of fluctuation in electricity supply and demand. The darker blue area in the top chart shows the amount of demand left over to be met after renewables and nuclear typically ranges from 10 to 30 GW throughout the year. Fast forward just 10 years to a country with 40 GW of offshore wind and the situation is very different. This range expands all the way from 30 GW down to minus 30 GW, meaning nuclear and renewables producing substantially more than total national demand. This situation of ‘over-production’ has been faced by Denmark for many years, and more recently by Scotland, but these are small countries with interconnections to much larger neighbours.
The bumpy ride shown in these charts is the daily average of demand and output – the hourly variation is even stronger, but this could more easily be accommodated by battery storage, pumped hydro and electric vehicles. The variations seen over multiple days of high or low winds will test the system further though.
The charts below show the situation in more detail, zooming in on three specific weeks which would challenge the limits of system operation. On the left is a particularly windy and sunny week, where intermittent renewables and nuclear produce 55% more than demand. In the middle is a calm week, where a third of electricity demand has to come from other sources. On the right, nuclear, wind and solar provide almost the entire demand averaged over the week, but their share ranges from just 30% of demand to over 200% in specific hours. Synchronous generators (in darker blue) must flex up and down by over 30 GW in the space of just 8 hours.
Taken across the whole of 2030, 34 TWh of electricity needs to be generated from non-renewable sources to satisfy the residual electricity demand. At other times, there is 37 TWh of excess electricity production that can be used for other applications. So, what could be done to balance out this surplus and shortfall?
Britain could export its excess wind to neighbouring countries. Interconnection capacity is expected to expand rapidly over the coming decade, perhaps tripling to over 15 GW if all potential projects are built. That said, weather patterns are much larger than the British Isles, so if it is windy here, it is typically windy across most of Northern Europe. In the hours when Britain’s wind farms operate at over 80% capacity factor, those in France, Netherlands, Ireland and Belgium are typically in the range of 50–85%. When British farms deliver under 20% capacity factors, the neighbouring countries’ farms average just 6–20%.
Storage is another option, but it would need to have very long duration to cope with entire weeks of excess or shortfall (shown above). Britain’s pumped hydro stations offer about 8-24 hours of storage, while battery storage systems can handle only 1-4 hours. With several days of surplus or shortfall, we find a total of 28 TWh of storage would be needed to capture all of that excess. This is around 1000 times the current electricity storage capacity available in Britain, but it is comparable to the total natural gas storage the UK has in the form of underground salt caverns.
Perhaps some of these wind farms should produce something other than electricity. Electrolysers can be used to turn electricity and water into hydrogen. The excess electricity production in 2030 could be used to make 670 million kg of hydrogen. That would be enough to fill 133 million fuel tanks in fuel cell vehicles such as the Toyota Mirai, or to heat nearly 2 million homes.
Some could make hydrogen directly out at sea by using their electricity to split water. Especially as farms move further from shore, grid connections can contribute up to one-third of the total cost. Hydrogen could potentially be hauled to shore at lower cost, piggy backing off the existing oil and gas pipelines, which will see limited use as the North Sea fields start to wind down. Hydrogen production from just 4 GW of offshore farms could potentially feed one million homes with zero carbon heat – or supply the high-temperature heat needed for to decarbonise British industry.
The large-scale adoption of hydrogen will require changes by energy users, not to mention a nationwide transmission and distribution system. The cost of electrolysers will affect its economics, particularly given that they won’t be able to run full-time. Offshore plants can only run when it is windy, while grid-connected electrolysers will need to turn down as the price of electricity rises on low-wind days. But these are challenges that might be overcome. Simply plugging massive numbers of wind farms into the existing grid and hoping is much less likely to be a successful approach.
[1] We model the 2030 situation by combining annual generation by technology from National Grid’s ‘Leading the Way’ scenario with hourly capacity factors for the future wind and solar fleets from Renewables.ninja. By 2030, National Grid expects that onshore wind and solar PV capacity will also have grown by 75%, and together will produce 60 TWh per year (compared to 170 TWh from offshore wind). While this will make balancing harder, they expect nuclear output will have fallen by one third due to the ageing fleet, which will make balancing easier. The actual situation in 2030 will be complicated further by changing demand patterns due to uptake of electric heating and vehicles.