The Emirates Shows Us How Not To Build Solar

Implications of the U.A.E.'s solar megaproject, and how solar could still help

Illustration. Photo by Antonio Garcia on Unsplash

The Emirates has recently commenced construction on a $6 billion solar project, combining a 5.2 GW solar plant with a 19 GWh battery storage. The system aims to provide one gigawatt of continuous, “renewable” energy on a 24/7 basis. But why does one need to store so much electricity in one of the sunniest regions of the world to do that? And what does this tells us about the rest of the planet trying to switch to wind and solar? There is a lot to unpack here and plenty of lessons to be learned, including one on how solar could still help us on our long way down the Hubbert-curve… Because, like it or not, when the Gulf Arabs start spending billions on solar panels, you can be sure that the slippery downslope of world oil production is clearly in sight.

Thank you for reading The Honest Sorcerer. If you value this article or any others please share and consider a subscription, or perhaps buying a virtual coffee. At the same time allow me to express my eternal gratitude to those who already support my work — without you this site could not exist.

1

The United Arab Emirates, along with it’s neighbors around the Gulf, receives an average of 3,568 hours of sunshine annually. That’s almost ten hours a day. For comparison Berlin, Germany, gets less than half of that (1730 hours). Yet, in order to deliver a steady stream of 1 GWh of power to the grid — even in such a sunny location as the Arabian Peninsula — one needs to install 19 GWh of battery storage in addition to overbuilding solar by 5 times. Isn’t that a bit excessively cautious? Well, in a perhaps oversimplified example, these numbers mean that the batteries would have supply the electric grid 19 hours a day on average, and then would have to be fully recharged in 5 hours, when the sun is at its highest. (Which by the way corresponds to the average peak sun hours in the UAE at 5 hours and 50 minutes a day). The solar panels would then need to play a dual role during these five peak hours: besides delivering power to the grid (at a rate of 1 GWh), they would have to charge the battery cells using the remaining 4 GW of their rated capacity. Based on this back of an envelope calculation the numbers seem to be just right, with only an hour or so of wiggle room left to cover day to day variances.

This little exercise, crude as it may be (1), shines a light on a rarely discussed metric when talking about power generation: the capacity factor; or the ratio of actual electrical energy output over a given period of time to the theoretical maximum electrical energy output over that same period. For example, if we were talking about a 5 GW gas turbine power plant, which could operate unimpededly 24/7 — day and night, rain or shine — for months, we could theoretically generate 120 GWh of electricity a day. Contrast that to the above solar panel array with a rated capacity of 5 GW, delivering around 29 GWh at best. This 29 out of the theoretically possible 120 GWh corresponds to a capacity factor of 24% — which equals the real life performance of the already built Mohammed bin Rashid Al Maktoum Solar Park in Dubai.

Lesson #1: Solar can only deliver less than a quarter of it’s rated capacity on an annual basis; even in the middle of the sunniest of deserts.

2

Contrast that performance to solar panels installed in Germany, and in other Central European states, where the length of day varies greatly across the year (from 8 hours in winter to 16 hours during the summer), and where cloud cover can persist for weeks, sometimes even months, between November and April. (And this is not to mention the lack of wind during the same time of the year, called the dunkelflaute or winter doldrums in German). Germany’s dismal geographic location thus yields a capacity factor of 11% for solar on an annual basis. Practically speaking you would need to buy PV panels with ten times the nameplate capacity — compared to what you actually need — and we haven’t even talked about battery storage…

Storing electricity across seasons would require an immense battery capacity, and a yet-to-be-developed technology which could prevent these batteries from losing their charge over months of storage. Pumped hydro, hydrogen etc., while theoretically being much better options than Li-ion for long term storage, are also not available at such a scale. For reference Germany produces 500,000 GWh of electricity a year, while battery storage is slated to reach 3 GWh by the end of this year. Call that a long way to go…

Low capacity factors for “renewables” in Germany (11% for solar, and 20–25% for wind) and a lack of storage explains why German electricity output keeps falling year over year — despite an exponential growth in installed “renewable” generation capacity. The shut-down of hard coal due to the depletion of Germany’s once rich reserves and the banning of cheap imports, combined with the dismantling of nuclear power plants, could not be compensated by wind and solar due to the simple arithmetic explained above. Without an adequate supply of cheap natural gas to take coal’s place (hello, blown up pipelines), Europe’s biggest economy has rapidly became a net importer of electricity from one of it’s biggest exporters.

Solar capacity up, coal down. Win-win? Source

Lesson #2: The proposed solar power plant in the U.A.E. is simply nowhere near technically feasible elsewhere — especially not in Europe.

3

The question poses itself: does a 24/7 solar plant with batteries even worth the investment? The answer, as usual, it depends. Solar projects generally have a much higher capital expenditure (capex) than operating expenditure (opex), which is a fancy way to say: it costs a whole lot more to build a solar park than to run it. In case of this project, we are talking about a budget of $6 billion, which somehow must be compensated on the revenue side. And this is where the offtake price (a legally binding, pre-agreed contractual price) comes into the picture.

As per a 2023 report from the same market research site linked above, solar power is priced in the country at around 1.35 cents per kilowatt-hour. This translates to $13,500 per GWh. Doing the math results in a mere 51 years till a return on investment is realized. Ahem, that cannot be right… Right!? OK, let’s be more generous here — we are talking about a 24/7 power supply after all — and take the price businesses pay for power (11 cents/kW), divided by two. (Grid maintenance fees, conversion and transmission losses, as well as taxes also have to be reckoned with.) Again, I’m not privy to the contract, I’m just applying my critical thinking skills here. This rough estimate of an offtake price, however, still translates to $55,000 for each GWh produced, resulting in a payback time of 12.5 years... That is just little higher than the expected lifespan of the grid scale batteries used in the project. Oh, and by the way, that also presumes that the power plant never goes down for maintenance, the batteries don’t lose an iota of capacity during use, and all employees work for free. Let’s be honest — at least to ourselves — the numbers simply don’t add up.

Lesson #3: Realizing a return on investment in a 24/7 solar plus battery project is highly unlikely without resorting to heavy government subsidies.

4

‘Then surely costs will climb down! Who knows, ten years from now we could pay half as much for the same set of batteries and panels.’ — While this might have been the case 10 years ago, the meteoric fall of battery and solar costs have slowed down considerably in recent years. As a recent Rystad Energy report found: “We see global solar costs stabilizing within a fairly narrow range over the next five years, while wind costs are likely to push higher and have more variation around the world.” So while solar capex intensity has indeed dropped from a global average of more than $5 per watt in 2010 to around $0.80 per watt in 2025, that was due to a slew of one-time efforts. Realizing economies of scale, utilizing cheap electricity from coal, resorting to government subsidies and preferential loans to start a solar business — all that in a single, special location… Yes, I’m talking about China. As the Rystad report found: “More than 90% of the global manufacturing capacity for ingots, wafers, cells, and modules is located in China, meaning that most of the solar PV projects developed across the world have Chinese components.”

Reducing costs for solar PV manufacturing has started with low hanging fruits (moving energy intensive manufacturing to places where energy is the cheapest), then progressed ever slower till all potential cost saving measures and government incentives were used up. Thanks to intense competition and a relentless cost pressure, Chinese manufacturers of both solar panels and batteries have been operating on thinner and thinner margins for years now, until they started to make a loss of -7% this year. This makes the solar miracle not only totally unrepeatable but wholly unsustainable. The levelised cost of electricity for solar PV (and onshore wind as well) is thus approaching its lower boundary determined by physics (around 50–55 EUR/MWh).

The LCOE is the discounted lifetime cost of building and operating a generation asset per MWh of electricity. It is calculated by dividing the present value of all relevant costs faced by the generator over its lifetime by its total electricity production. Source: Ember

Problem is that the energy needed to mine and deliver raw materials (diesel fuel) and to process those inputs (coal, natural gas) can only be expected to become scarcer after fossil fuel extraction peaks then begins to decline. The solar cost curve, as a result of this coming energy and raw material scarcity, will most likely turn around and begin to rise again… As it did temporarily in 2022, when EU sanctions on energy destroyed European wind and solar supply chains. Back then China stepped into the void, but what would happen when coal peaks there as well, on top of a worldwide oil shortage later this decade…?

Lesson #4: The cost of wind turbines, solar panels and battery cells now seem to have reached their minimum, and rely on one major supplier, China. As energy and raw materials become scarcer, and geopolitical tension grows, costs can be expected to rise in the years and decades ahead.

5

If PV panels and batteries cannot expected to become cheaper still, can electricity prices climb high enough to support an increase in investment costs? According to mainstream economic thinking, if there is a scarcity of anything, prices will rise to incentivize more production. Right? Well, not in the case of “renewable” electricity. Since solar panels and wind turbines do not generate power as needed, but according to the weather, there is often a huge production surge in the middle of the day, followed by a lull in the evening. (And the further north you go, where long summer days create periods of massive oversupply and a lack of electricity in the winter, the worse this predicament gets.) These wide swings in power generation result in similarly wide swings in electricity prices (sometimes going into negative territory), ruining any return on investment calculation and forcing producers to curtail electricity generation (2).

As more and more batteries and other storage get added to the mix to reduce intraday price swings somewhat, the price difference between noon and evening gets reduced as well. Consequently the economic incentive (the arbitrage) is also reduced to a point where the return on investment of a battery farm also disappears. This, on the other hand, disincentivizes the further deployment of storage, making the issue permanent. Again, a beautiful example of self-regulation, to the detriment of all participants. The solar market remains crowded during the day, ruining ROI calculations for new entrants, while battery storage will seem to be lagging forever behind.

Another unintended side effect of this fluctuating supply is the bankruptcy of baseload generators. Nuclear, coal, or natural gas power plants also need stable (and relatively high prices) to remain profitable. Now, if they are forced to shut off during the day then restart in the evening, that not only forces them to idle their workforce, but reduces plant fuel efficiency (as these generators need to warm up before they can start producing). Keeping these machines in a state of high readiness, on the other hand, also takes fuel. In the aftermath of the Spanish blackout earlier this year, and in order to prevent another such event, grid operators are now idling gas turbines, resulting in a 100% cost increase for grid services. Aging coal power plants do not enjoy this luxury and thus will have to be closed, making the grid even less stable and unable to cope with demand (again take a look at the case of Germany).

Lesson #5: Moving away from a grid supported and stabilized by fossil fuel use seems highly unlikely. As fossil fuel supply becomes scarcer still, grids relying heavily on “renewables” can be expected to experience frequent blackouts and power rationing.

6

Supporting grids by burning fossil fuels when needed, is of course no issue for the Gulf states. Or is it? According to the latest reports OPEC likely has not much spare oil production capacity left — i.e.: it cannot increase oil output even if it wanted to. Phrased differently: oil production in the Gulf is on a high plateau, and as their once prodigious fields age and deplete faster with time, their output can be expected to decline in the decades ahead. Saudi Arabia, already struggling to balance it’s budget, is freeing a million barrels a day for export by cutting oil from utilities and investing 8 billion in solar. The U.A.E. — another major oil producer — does the same, aiming to generate 75% of Dubai’s power from clean energy by 2050.

Despite all these efforts by wealthy Arab states, McKinsey projects that fossil fuels will still account for 41–55% of global energy use in 2050. They see energy affordability and security outweighing climate goals in shaping policy and investment; besides a lack of affordability holding back the widespread adoption of new low-carbon technologies. What will happen, when the fossil fuel supplies prove to be inadequate to fulfill their current role is, of course, another story.

Lesson #6: Gulf states already experience limitations in oil production and clearly see that demand might soon outstrip supply, as world oil output peaks then rolls over. Hence the recent multi-billion surge in solar investment to free up as much oil for export as possible.

7

Last but not least, what is the future of solar then? As we have seen from the reasons laid out above, solar power is not without its shortcomings. High variability (both within the day and across the year) and a low overall capacity factor are part of their nature and not something which can be solved by technical means. And while intraday variability can be offset by batteries, the economics only works in sunny deserts with ample government subsidies. Elsewhere, fossil fuels remain the only option to provide a stable electricity supply — especially in higher latitudes. (Nuclear simply does not scale and is not a tad bit more sustainable than fossil fuels.) The levelised cost of electricity cannot be expected to fall further, either, even as prices often reach negative territories. In fact, as fossil fuels needed to make solar panels become scarcer and geopolitical tension grows, costs can realistically expected to go higher.

This sheds some light on the biggest blind spot, when it comes to solar deployment: it’s complete inadequacy of replacing fossil fuels in electricity generation. We are approaching the problem of pollution and depletion from the wrong end. Solar is not there to replace this slowly failing system, but to aid the long descent ahead. Instead of building expensive solar farms with batteries, governments should support and, in fact, incentivize self-sufficiency aided by photovoltaics. Imagine using solar as a back-up for blackouts: not to replace everything we did with grid power, but to enable basic services to be run: like water and sewage treatment. Or building resilient homes with small solar DC (direct current) backup networks, instead of wasting 10% of the power generated in an inverter, which, by the way cannot operate during a blackout, and clogging the network with excess electricity fed in during noon. Households then could use their small solar plus battery DC grids to charge their communication devices, pump water, run a fridge, and keep their lights on — in case the main grid would become unavailable for shorter or longer periods of time.

Lesson #7: There are so many things which are possible using the power of the Sun, but running a 24/7 high consumption lifestyle, unfortunately, is not one of those.

Until next time,

B

Thank you for reading The Honest Sorcerer. If you value this article or any others please share and consider a subscription, or perhaps buying a virtual coffee. At the same time allow me to express my eternal gratitude to those who already support my work — without you this site could not exist.

Notes:

(1) In the crude example above I was not calculating with battery charging inefficiencies, nor with DC/DC and DC/AC conversion losses, let alone with safety margins preventing batteries from being totally drained or fully loaded (and thereby damaging battery life and capacity) — as none of this data was made available. Also, I considered changes in the length of the day negligible, or at least within the boundaries of the systems capacity. In Dubai, for example, daytime varies in length between 10 hours and 34 minutes in December and 13 hours and 43 minutes in June, which is a rather mild oscillation compared to other locations. As for seasonal variations in weather, mist can be a problem in “winter” blocking sunlight for a few hours in the morning, but further inland (away from the shores) that’s not an issue either. Persistent cloud cover, lasting for weeks, is also not a problem either — hence the massive amount of annual sunshine hours recorded.

(2) The Rystad report goes on to explain in detail: “Solar tends to have a lower capture rate than wind, as its generation is concentrated during the same hours, which compresses revenues through competition. This is often referred to as the cannibalization effect. Storage is expanding, but its growth has not kept pace with the rapid rise of solar and therefore does not yet provide sufficient flexibility. In New South Wales, this has contributed to solar capture prices falling to around 40%. Several high solar penetration markets across the world are also seeing yearly average capture rates in the range of 40–60% or are expected to reach this level in the coming years.”