Does Big Solar pose a threat to Big Gas? [Gas in Transition]
Over the past generation gas has carved out a major role in the power market, with 23% of global electricity production (6.8 trillion kWh out of 28.4 trillion kWh) flowing from gas in 2021. That large industry consumed approximately 1.5 trillion m3 out of world gas production of just over 4 trillion m3 in 2021. Gas-to-power therefore accounts for 40% of the global gas market.
This remarkable market share was born out of availability and low prices, but today it has another key driver – climate concerns. Gas more or less drove oil out of the power market two decades ago (not completely, oil-to-power still accounts for about 5% of oil production). Now gas is slowly squeezing coal out of the market too, though again not completely as coal is still the cheap and easy go-to choice for a set of economies in the Global South as well as in China. Those minor competitive threats aside, in today’s power market gas is the emerging king, with nuclear and hydro at its left and right hands.
The king’s throne may not be so secure
Greek mythology turned on the concept of hubris and nemesis. The hubris of success is often followed by a painful nemesis. Might large-scale solar be a coming nemesis for gas? Most observers would quickly dismiss the idea as risible – solar is too small, too expensive and too time-constrained to threaten gas’ position in global power. But recent successful Big Solar projects are giving the lie to that lazy consensus.
First, I want to think about cost, and I’ll focus on best-of-class numbers because those are the numbers that will drive future decision-making.
Gas best-of-class costs are already minimised
The best-of-class output cost per kWh for gas is a function of a handful of key variables. One is the thermal conversion rate achieved by a new, best-in-class, combined cycle gas turbine. Large units in ideal conditions are now delivering 60%+ thermal conversion. Adding a little pessimism and accounting for smaller units 55% looks like a more realistic base number. There is little the industry can do to get that number up.
The second important number is, of course, the price of gas. Today’s Henry Hub is about $0.10/metre. Power of Siberia gas is about double that. These probably represent minima for the market going forward.
At 55% efficiency those minimised gas costs convert to a resource cost of $0.018-3.3/kWh. Add another $0.025/kWh for plant operating costs and depreciation and we get a “middle” cost of $0.05/kWh. Add a profit margin of 30% of the end price and we end up at $0.07/kWh.
That is not far from Siemens’ enormous Beni Suef plant in Egypt, which is an excellent example of “best-in-class” equipment and cheap resource, where the power contract reportedly comes out at around $0.06/kWh.
Big Solar costs have been much higher, but need to be much lower to become a threat
For Big Solar to be a structural threat to gas it has to come in at or below that number. There are some non-price factors to be considered. Solar wins on the carbon front, giving it some pricing power, but loses on the intermittency front since it operates for only half the time. It wins on the strategic front (you can’t sanction sunlight), and on the balance of payments front (sunlight is intrinsically domestic). Intermittency is a big issue because it forces a producer into backing solar with other sources, preferably ones with high flexibility and short spin-up times. Taking these factors together Big Solar needs to undercut gas by a material margin before it can become a strategic threat.
Big Solar prices are plummeting, and deserts are in good supply
So, what is happening to Big Solar output prices? Here a quiet revolution is in progress. A large-scale solar PV array now comes in at around $1/watt of nameplate capacity. Energy yields depend on insolation, which is largely a factor of latitude and to a smaller degree a factor of atmospheric water content (water=clouds which cuts insolation). We have three very large benchmarks to show where the best-of-class sits today – the 1,650-MW Benban project near Aswan in Egypt, the 1,000-MW Junma (“horse”) project 20 km southwest of Dalad Banner in China’s Kubuqi Desert, and the 800-MW Al Qudra project in Dubai.
Immediately obvious is that all three are located in deserts located within 40 degrees of the equator. Here they offer excellent insolation, plus low atmospheric water content and land that is effectively free. Benban has 2,300 kwh/m2, Junma and Al Qudra are somewhere around 1,800 kwh/m2.
High insolation and low latitudes drive high yields
The market has grown used to average solar yields in the high teens. Deserts go higher than this. Benban (Egypt) has a power yield of somewhere in the low 20s (I’ve seen numbers as high as 26% but I think those are more theoretical than everyday). Junma is claiming 25%, which might be true with newer and better panels. I’ve used an average of 23% for this analysis.
Big Solar’s big cost component comes in the form of depreciation, with a relatively small addition for maintenance, management and regular cleaning of PV panels (deserts are dusty). Junma’s control room is reportedly manned by just a handful of people, so running costs other than depreciation are small.
At the top line level, then, a 1,000 MW/$1bn Big Solar plant with a 23% yield and a 25-year life generates 2bn kWh/yr, with a depreciation charge of $0.02/kWh. Not a bad start.
Capital hurdle rates drive debt into the project
But to be a threat Big Solar has to generate a decent return on capital too. Most corporates will target a hurdle rate of around 18% pre-tax return on equity. In a nil-debt structure our 1,000-MW Big Solar project would need a pre-tax return of $180mn (18% of $1/watt capital invested). With production at about 2bn kWh/yr the hurdle rate margin would work out at $0.09/kWh, obviously a deal-killer.
But when we gear up the project to 80% at, say, a 5% coupon the hurdle rate power price falls to around $0.06/kWh. Benban’s actual feed-in tariffs fell in 2017 to $0.084/kWh – reflecting the fact that Big Solar was less proven back then, and that the project was divided into 41 less capital efficient separate lots to be built out by separate contractors and therefore at higher cost. In 2018 Egypt moved to open tenders, and ACWA tendered an amazingly low $0.027/kWh (that was probably a loss-leader, but indicative of where the market is heading).
Big Solar output prices are now on a par with Big (cheap) Gas
The ACWA deal apart, Big Solar is now clearly on a par with Big Gas for costs, with the added benefit of zero carbon, zero gas import risks and zero balance of payments penalties. That just leaves intermittency as the last barrier to competitive advantage.
Here Big Gas’ very flexibility is a possible breach in its own defensive wall. Unlike coal and nuclear, which are most efficient when run at full load 24 hours/day, CCGTs can spin up and down in minutes with no great energy penalty. Big Gas and Big Solar can therefore be seen as complementary, working a day/night shift together. Plant producers will be happy with that – economies will still need CCGT plants to partner with Big Solar – but gas producers will not, as they see a growing chunk of that 1.5 trillion m3 demand coming under pressure from Solar/Gas partnership structures.
We’ve got deserts, plenty of deserts
The uncomfortable truth for Big Gas is that China, Egypt, the UAE, Pakistan, India and many other economies with fast-growing power demands are also well-furnished with deserts or near-deserts within that 0-40 degree latitude. In Argentina, for example, a 500-MW Big Solar plant is under construction at Cauchari, 25 degrees south of the equator. South Africa has the desert potential (in the Karoo) but lacks market and financial depth. But even here there are the first seeds of a Big Solar switch with a 175 MW plant at De Aar (30 degrees south).
Big Gas might need to start worrying, but it probably does not need to panic quite yet. Even the best new PV panels still deliver yields in the low 20s. That means a 1,000-MW plant can only produce about 2bn kWh/yr. With a gas-to-power market of 6.8 trillion kWh/yr even galloping solar PV investment is not going to displace gas any time soon.
The core gas-to-power market might be safe, but the growth market is not
Big Solar’s effects on gas demand growth are likely to be more painful. Power demand in the Global South plus China is growing at about 6% per year – making for up-steps today of 400bn kWh/yr. Large chunks of that are being soaked up by nuclear (UAE has just commissioned its third reactor at Barakah, for example, with a fourth due any year now), a trend that will be accelerated when China gets its first commercial Molten Salt Reactor design into the market. From the other direction Big Wind is coming in fast too. Coal still lurks in the undergrowth as the cheap easy go-to solution for the short-of-cash.
If Big Solar starts adding a hundred 1,000 MW plants/yr globally (at $100bn worldwide, not a large swing in capital allocations) they could take half of the power market demand up-steps in the Global South and China between them. China’s latest 5-year plan alone foresees about half of that investment, and China is not the only Big Solar investor.
The core might be safe, but margins are not
If gas loses the edge in new power production it cannot rely on its established position to protect gas producer margins. Gas lives in an equilibrium market – its market price is only loosely connected to its production cost. That means a small marginal imbalance between supply and demand can spark substantial price changes (look at what happened in 2022 if you want evidence for that proposition).
With world gas production rising, if gas loses its hold on power demand up-steps a negative imbalance might set in, driving gas prices to painfully low levels. There is not much the gas market can do to change the ballpark. CCGT conversion efficiencies are about as high as they can go after two generations of technical development. CCGT plant costs too are not going to see a substantial change downwards. The cost of power from gas is about as cheap as it’s going to get, while the cost of Big Solar is going to head south.
It looks like gas has a painful power market fight on its hands.