We Have To Stop Kidding Ourselves About The True Cost Of Renewables

There is a feedback loop among wind energy proponents in which estimates that are more wishful thinking than reality are used in studies, spreading misinformation until it becomes a false fact.

The Royal Society, founded in 1660, awarded a royal charter by Charles II in 1662, and boasting a long list of illustrious fellows including Isaac Newton and Stephen Hawking, is the oldest continuously operating scientific body of learning in the world. Prestigious scientists from all disciplines are represented amongst its 1700 fellows.

When the Royal Society publishes an evaluation of energy storage methods along with a Policy Briefing document, with contributions from no less than fifty PhDs and Professors, I would expect that document to carry significant weight among those who are making energy policy decisions in the UK. Certainly, it will carry more weight than my own evaluation published last year on Substack and read by about 500 people.

So, when I saw in the paper from The Royal Society that they have used basically the same methods as I did, and the same data source for wind and solar capacity factors, but they came up with a range of £52-£92/MWh for the cost of electricity generation, compared to my own estimate of $183 (£150)/MWh, my first reaction is to check my own work to see where I might have gone wrong.

The Royal Society publication considers various technologies for long-term energy storage and settles on hydrogen, stored in underground salt caverns as the preferred option, with a very small amount of battery storage used only for frequency modulation. I also used hydrogen storage but without the batteries. Their estimates for the cost of electricity in 2040 are shown in the chart below:

For the purpose of this comparison, I use £68 ($83)/ MWh, the middle case, and the mid-point between the green bar (5% discount rate) and the orange bar (10% rate).

Renewables provide all the generation, with a split of 80% wind (split 70:30 offshore to onshore) and 20% solar compared to my own calculation which had 55% wind, 45% solar with the wind split evenly between onshore and offshore. On running my analysis with the 80/20 wind-to-solar split I find that projected electricity costs only change by 1%. The analysis is not sensitive to the wind/solar split.

Total hydrogen storage requirements are projected at 100 TWh, but that figure is the energy content of the hydrogen in storage, the electricity that can be generated from that hydrogen is only 55 TWh, which amounts to 35 days of average demand. My own estimate was 30 TWh of produced electricity, but I used existing demand rather than projected 2050 demand, storage requirements work out to 38 days of average demand, very similar to the Royal Society estimates.

The real difference between our analyses boils down to the assumptions we made about costs and efficiencies. In the table below I have compared those assumptions and added a column showing the impact that each assumption would have on my projection of electricity costs.

Using the cost assumptions from the Royal Society paper, I arrive at an electricity cost of $93.77 (£76.89)/MWh, lower than my original $183/MW, still higher than the Royal Society estimate but not wildly different.

The next step is to check the sources of those cost assumptions and see if they are reasonable.

Hydrogen storage costs:

I found a better and more detailed cost estimate for the storage of hydrogen in salt caverns in a paper published by the Argonne National Laboratory, summarised in the chart below:

The estimated cost is $35 to $38/Kg of stored hydrogen, which amounts to just over $1,000/MWh of stored energy, higher than the Royal Society estimate but lower than the $2,000 I had assumed in my previous analysis.

Hydrolyser costs and efficiencies

The sources of the Royal Society’s hydrolyser cost data are documents produced by IRENA, and by the International Energy Agency, organizations that promote the adoption and use of renewable energy. Those estimates are shown in the table below.

We both chose PEM electrolyzers because of their ability to work with widely varying power inputs, but where I chose to use today’s real costs and efficiencies, the Royal Society report uses estimates of future costs and efficiencies that may or may not be achievable.

Costs for hydrolyzer projects today are running around $1,800/KW, four times the mid-range figure used in the Royal Society study. Although costs will no doubt come down in the future as the technology is manufactured at scale, it seems wildly optimistic to assume that costs will come down by as much as 75%. Total project costs must include water treatment, drying, compressors, electrical systems, and all the buildings, foundations, and infrastructure, those costs are unlikely to fall drastically even if the costs of the hydrolyzers come down. It seems to me that the IRENA and IEA forecasts are more wishful thinking rather than realistic projections.

Wind and solar costs

Technology improvements and the relocation of a large part of the supply chain to China have drastically reduced the cost of solar panels over the last decade. My own estimate of UK solar costs is probably high, even after considering that UK solar achieves less than 11% capacity factors. However, solar costs are not a major component of the projected overall electricity costs, so I won’t investigate them further.

The biggest difference between my estimates of future renewables costs and the Royal Society study is in the cost of wind power. The mid-point of the Royal Society estimate uses a figure of £35 ($42.70)/MWh for the cost of renewables in 2040, including solar and both onshore and offshore wind, based on 2021 costs.

The data comes from a cost projection published by the UK Department for Business, Energy and Industrial Strategy (BEIS), table below:

The table shows projected costs as far as the year 2040.

Earlier this year, the UK government held an auction for contracts for renewable power generation. Strike prices for solar and onshore wind were £47 and £52.29/MWh respectively in 2012 prices. That translates to £54.57 for solar and £60.71 for wind in 2021 prices, well above the price quoted in the above chart and used in the Royal Society study. There were no bids for offshore wind, no one was willing to supply offshore wind at the set maximum price.

Some auctions for offshore wind had lower prices, but the contracts were not mandatory and no one has been providing power at those prices. Companies were awarded the permits to build but did not take up the contracts because they could get higher prices in the marketplace. Some of the recently bid projects have been cancelled because the bid prices cannot support the projects, and wind power companies are trying to renegotiate prices.

You can find the real revenues for offshore wind at the TGS website. The prices range from a low of around £95 to over £200. Average prices are three times higher than the prices used in the Royal Society study.

Wind power costs have come down over the past decade, mostly because of volume production and the use of larger turbines with higher hub heights that get higher wind speeds. However, there is a diminishing return on higher hub heights that require heavier foundations and stronger supports. We are probably close to the limit of what can be achieved by making ever-larger turbines.

Turbine failures have been more frequent in the newer larger turbines and those failures have increased insurance and maintenance costs. Turbine suppliers have been hit with major losses and the cost of turbines has increased by 38% in the last two years.

Wind turbines are mostly concrete, steel, and balsa wood wrapped in fiberglass, with copper windings and cables, and rare-earth magnets. There is no compelling reason to assume that any of those costs will come down over the next 25 years.

There is, however, a compelling reason to assume costs will rise. The best sites are already taken, wind farms will be forced into increasingly deeper water and further from shore. Costs must rise, there is no reason to assume that costs in 2040 will fall anywhere close to the levels assumed in the BEIS publication and used in the Royal Society study.

Cost and efficiency of hydrogen generators

Finally, the cost of generators that burn hydrogen is probably underestimated. The Royal Society study assumes the use of reciprocating internal combustion engines, which is probably the right choice for their low capital cost and ability to ramp up and down quickly.

The assumption of 55% efficiency is optimistic. The Wartsila 32DF generator holds the world record for the most efficient diesel generator at 51%, it is overly optimistic to assume an efficiency any higher.

Generator costs are around $350/MWh but that is only the generator. The power plant needs foundations, buildings, cranes, electrical systems, and infrastructure. The best reference I can find for complete installations comes from the US IEA at $1,300/KWh., more than 3 times higher than the cost assumed in the Royal Society study.

Conclusion

The Royal Society is projecting a cost for electricity generation from renewables and hydrogen storage that has been estimated using data that are not realistic. Each of the cost assumptions alone has a small to medium impact on the final price, but the combination of underestimated costs together is a major impact causing the final price to be underestimated by a factor of 2 or more.

This is a chart from the Royal Society study:

It purports to show that a system with 100% wind and solar (green line on the chart) will provide lower costs in 2040 than a system that includes nuclear power (orange and blue lines).

However, if realistic costs were to be input into the models, the green line would be off the charts, much higher than the orange and blue lines.

What is happening here is that governments and bodies such as IRENA are producing wildly optimistic forecasts of future renewables costs. Those costs get used in academic studies that are then issued to policymakers and we have this feedback loop of misinformation that results in wrong-headed policies.

If we don’t stop kidding ourselves about the true costs, we are heading for disaster.

This is the first in a three-part series of articles on this subject. In the next article, I will look closely at the wind power cost forecasts put out by BEIS and show you why they are wrong. In the third article, I will address the timeline for the energy transition and show how nuclear power can be built faster than this proposed conglomeration of wind, solar, and hydrogen.

Thank you for taking the time to read this, it is too bad that it will be read by only a few hundred people and it is unlikely to be noticed by any of those in a position to have any influence on final decisions.

Comments are welcome and appreciated.