The battle of the greens

Is it better to design a zero emission power system based on 37 years of weather data, or on one year?

A few weeks ago, a skirmish broke out in the fantasy world of UK climate change and net zero.

Sir Christopher Llewellyn-Smith, criticized the Climate Change Committee because they based their climate change plan on only one year of historical climate data.

Llewellyn-Smith, a Fellow of The Royal Society and professor at Oxford University was the leader of a group of scientists who were tasked with evaluating the energy storage needs for the UK in a future net zero scenario in which all energy use was electrified and all electricity provided by renewables. Their 100-page report along with a shorter policy briefing was published in September of 2023.

The report calculated the storage requirements based on 37 years of historical solar and wind data and came up with a total storage requirement of 100 TWh using hydrogen as the primary energy storage medium. 100 TWh is the energy content of the hydrogen, not the amount of electricity that can be generated from the hydrogen. It is equivalent to 55 TWh of electricity using their assumed efficiency of 55%.

55 TWh is a massive amount of storage, and the Royal Society quite rightly concluded that batteries and medium-term storage systems such as pumped hydro and compressed air were not going to do the job. Their recommendation was to use hydrogen, generated from surplus renewables and stored in underground salt caverns.

The sheer size and number of salt caverns required to store all that energy in the form of hydrogen prompted the Royal Society to recommend in their briefing report that construction of the salt caverns begin as soon as possible.

The Climate Change Committee produced their report titled “Delivering a reliable decarbonized power system” in March of 2023. Their remit was not the same as that of The Royal Society, they were tasked with evaluating and proposing an emission-free power system by 2035. Their final proposal includes other forms of generation, nuclear, fossil fuel with carbon capture, biomass and a limited amount of hydroelectric power as well as imports and exports.

However, the CCC only used one year of weather data for their study. They chose the year 2012, to represent a normal year.

Instinctively, most of us would assume that a study based on 37 years of data would give us a better answer than one based on just one year of data. In fact, the Royal Society approach is the method I used when I looked at the same problem more than a year ago in my first Substack article. I have since put a little more thought into it and I have realized that the approach is wrong.

The chart below comes from the Royal Society report, it shows the annual surpluses and deficits in electricity generation for the 37 years of the study period. The blue bars are the annual figures and the yellow line is the cumulative total, it represents the amount of energy that is input or extracted from storage at any time.

During the 37-year period of the study data, there is a sixteen-year period from 1985 to 2001 when winds are above average, and the storage is being filled. After 2001 the storage is gradually emptied, most of that drawdown takes place during a period of exceptionally low winds starting in November of 2009 and ending in the spring of 2011.

Without the 16 years of above-average wind, there is no way to fill the storage. Since there is no way to control the weather, there is no way to ensure that the exceptional period of low wind does not happen when the storage is empty, or near empty.

The Royal Society approach could leave the country desperately short of electricity at any time.

A system based on a study of a long period of weather data with storage to match that data only works for that particular period, it cannot be reliable because the future weather will never be the same as the study period.

The system must be designed for the worst-case situation that can reasonably be expected to occur. The design should be based on an exceptional low wind period that occurs when the storage is empty.  The massive storage proposed by The Royal Society is pointless, the only way to make the system reliable is to massively overbuild so that the wind can generate enough energy during the long periods of low wind, or to provide generators that are not weather dependent.

The CCC proposal is primarily focused on providing a zero-emissions electricity system by 2035, but one of the reports does include a proposed power system for 2050. The 2050 system includes higher demand from the conversion of all building heat and industrial heat to electricity. The 2050 demand data used by the CCC and The Royal Society are the same.

The report, by AFRY Management Consulting, evaluates the flexibility of the proposed power system.

The system has been designed using an average year (2012) and verified for five selected weather years. There is also a check for the low wind year (2010) combined with a 30-day wind drought. However, that check showed a deficit which would need an extra 20 GW of dispatchable backup that does not appear to have been included in the proposed installed capacity mix shown in the chart below. It seems the CCC has checked the worst-case situation, but has not included sufficient generating capacity to cover it.

There is a lot to criticize in the CCC proposal, not least of which is the massive overbuild of capacity, 369 GW of installed capacity for a system with an average demand of 65 GW. There are no credible cost estimates included in the reports, but that massive overbuild can only result in massively increased costs.

The proposed system is not carbon-free. It relies on carbon capture on the exhaust from biomass generation, backup gas generators and some hydrogen production (hydrogen is a mix of electrolysis and autothermal reduction using natural gas). CCS does not capture 100% of the emissions. It also includes 12 GW of unabated natural gas generators, and it relies on demand-side response to trim 35 GW of peak demand.

It is also questionable whether the 28 GW of interconnectors can be reliable if countries at the other end of those connectors are also building weather-dependent systems that will probably be short of capacity when it is needed.

The non-weather dependent generation included in the proposed system is 68 GW, plus another 28 GW of interconnector capacity. It could meet the demand without any of the wind or solar, which leads me to ask why we need wind and solar at all.

I will do a more detailed evaluation of the CCC proposal in a later article.

To answer the question of whether the Royal Society or the CCC approach is right. I would have to say that the CCC approach is better, but it must provide a system that can work in a worst-case period, not an average period.