Post

More from @pretentiouswhat

David Fishman

@pretentiouswhat

Sep 5

Oh no...🤦‍♂️

This op-ed on Chinese cleantech overcapacity and competition was in The Wire China a few days ago. Unfortunately it contains many huge errors about Chinese cleantech sectors I can't ignore.

Paywalled. I'll provide screenshots and comment. 🧵
thewirechina.com/2025/08/31/bei…

[Oh, and this will be another long thread. It probably should have been a long-form essay instead, but I already wrote more than half of it before I realized how long it had gotten. Sorry in advance.]

This piece has problems immediately in the second paragraph, starting with:

"China's domestic demand for green tech has also peaked given the massive frontloading of installed capacity during the last few years, fueled by subsidies."

This has two big errors:

1. Chinese demand for green tech has not peaked, as evidenced by the steadily rising annual installed capacity figures for wind and solar. In fact, the installed capacity isn't just rising each year, but even the volume of new installs in a single year has grown every year from 2020-2024. Last year saw 277 GW of solar PV and 80 GW of wind.

Even now in 2025, with the offtake policy reforms starting from 1 June, it looks like solar is going to at least match the capacity growth from last year, while wind is actually going to EXCEED the capacity figures from last year. Domestic demand is strong. As for next year, we'll see what the market reforms bring.

2. Chinese newbuild solar and wind farms have not been subsidised for several years already now (since 2021). Over the past few years (until 1 June 2025) they were built on a feed-in-tariff (FiT) basis, which means they earn a fixed on-grid price from the gridco, independent of what's happening in power markets.

If market prices are high, the FiT may be less than the market rate. If market prices are low, the FiT may be more than the market rate. In a power market context, this is very different from a subsidy (although it could be construed as/look like a subsidy if market prices end up lower than the FiT rate for long periods).

Same second paragraph, continued:

"plummeting external and domestic demand have forced Chinese tech companies to compete aggressively to gain market share by cutting prices"

This is wrong on both the domestic and international counts. We already know from the last post domestic demand for wind and solar installs is still rising.

Meanwhile, in the international space, Chinese solar panel exports totaled 236 GW in 2024, rising 13% YoY. Wind turbine exports were 5.2 GW, rising 42% YoY.

In 2025 to date, completed *panel* exports have fallen 5%, but cell and wafer exports are rising dramatically, up 73% and 26% YTD, respectively. Exports to some countries are down, but they have been more than offset by rising exports to other countries and regions. Unfortunately, I don't have a source on YTD wind turbine exports for 2025 so can't comment there.

The point I'm trying to make here is that while there's oversupply relative to demand, it's not reasonable to attribute much - if any - of this supply-demand mismatch to the demand side. Demand is fine. The primary driver of the supply-demand mismatch is coming from the supply side.

Read 15 tweets

David Fishman

@pretentiouswhat

Sep 3

https://twitter.com/SecretaryWright/status/1962861731174097195

This is funny and sad. Secretary Wright's post is so silly and unsophisticated, and yet Twitter's Community Noters managed to find a way to miss the argument entirely and "rebut" with an even more unsophisticated response. It's just all so tiresome. 😞

But let's break it down🧵

https://twitter.com/SecretaryWright/status/1962861731174097195

First, this this broader idea about not conflating energy with electricity is fine, even good and necessary. 💯

Electricity is what's called "secondary energy", a specific kind of energy that has been transformed from "primary energy" sources like coal, oil, sunlight, etc.

In 2024, about 21% of global energy consumption was electricity. This is called "share of electricity in final energy consumption".

China's electricity share is ~28% and rising rapidly, among the top 10 worldwide.

#1 in the world is Norway, at ~47%.

yearbook.enerdata.net/electricity/sh…

Read 15 tweets

David Fishman

@pretentiouswhat

Aug 30

Okay. Here we have a boldly stated series of ideas about Chinese solar. They are unsupported and wrong.

But it's a good opportunity to talk through some issues re: how we integrate solar, and some of the important and hotly-debated considerations. So I'll try to make this educational, not merely critical.

Let's dig in...

1.
"local solar prodution numbers are based on models, not measurements"

No. Models are typically forward-looking, used for forecasting, not describing the past. No matter whether the solar generator is offtaking to a power user, a power retailer, or the gridco itself, there's a business transaction going on there. The generator is being compensated based on how much electricity they generate, so you need measurement (i.e. metering). Without metering each kWh of power, how could you run a power business? 🤨

In China, the grid companies (e.g. State Grid and Southern Grid) and their subsidiaries have a monopoly over installation, maintenance, and reading of metering infrastructure. They report monthly, quarterly, and annual data to the NEA/NBS (or to the China Electricity Council, which compiles data on behalf of the NEA and NBS). That's the source of the generation stats. You could find always find a way to criticize the measuring and reporting methodology I suppose, but it's definitely a measurement approach, not a modeling approach.

2.
"Industrial solar production numbers aren’t backed by enough batteries to match consumption curves..."

The claim I think being made here is China's daily power consumption demand curves are not matched to the generation profile of solar, thus solar production would need to be time-shifted to a different time of the day with batteries for those production numbers to be realistic. But since China doesn't have "enough batteries" to do this, then the solar production data must be fraudulent.

This is speculative nonsense. I know OP pulled this argument out of thin air, because it's a fairly complex argument that would necessitate validation via datasets, models, or market access he doesn't have. Are China's 95 GW/222GWh of batteries installed nationwide "enough"? How would anyone besides a Chinese power trader, regulator, or market dispatch modeler know? You need to have a robust, data-backed argument here, because you're seeking to overturn/debunk reported data from State Grid.

But even without a quant-based argument, it's just a weird argument for solar.

Solar produces power during the day, aligning with human activity, as we are generally diurnal creatures. So solar already produces when humans tend to consume power. At low penetration levels for solar (say, <10%) you hardly even need batteries to time-shift solar to another part of the day because solar's production can be fully absorbed by the typical daytime rise in power usage.

At higher penetration rates for solar, you actually will get to the point where production can't be absorbed by the daytime consumption spike and thus storage becomes not just "nice-to-have" but necessary to avoid wastage - unless you have very flexible generators that can easily ramp up and down (more on this later). This issue can arrive earlier for wind than for solar, because wind tends to produce more when humans are asleep and power loads are lower, but we're talking about solar here.

There are a few provinces where daytime overproduction of solar is a real problem, like Shandong...and they are indeed installing storage rapidly, as you would expect from what I have just described. But Shandong is a frontrunner. Nationwide, solar provided just 8% of China's generation mix last year, so its ability to be disuptive is muted, paradoxically both massive, but also mostly absorbed with only a few ripples by the overall vastness of the Chinese power sector.

3.
"...but solar producers don’t pay for increased costs and waste induced in competing sources."

In China, this is broadly true, but not true everywhere, as it depends entirely on local market design. China's design has evolved over time but remains reasonable for China.

Non-flexible power sources are typically forced to try to "get out of the way" in spot markets when variable generators are producing, to their disadvantage. They incur losses caused by the existence of the variable generator. Coal generators in particular suffer from very low annual operating hours in China, although this was historically mostly because the coal generation sector is overbuilt and cannibalizes itself, with the "shouldering-out" effect from renewables starting to play a material role only since ~2021.

Generally speaking, generators being pushed out by renewables will usually have the opportunity to make up the losses on power sales by serving as load-following generators or suppliers of capacity, earning revenues in the ancillary services market or capacity market, respectively. Gas-fired power plants and battery farms are very well suited to play this role. Coal-fired power and nuclear can do it as well, although they are less well-suited and may need retrofits to perform in this way. This is how we keep these generators whole and happy and prevent them from exiting the market - assuming we felt we needed them to stick around.

Of course, these ancillary services fees or capacity fees have to come from somewhere, which brings us back to the original issue - whether solar producers are paying for it or not. Broadly, there are two competing philosophies here: "causer pays" and "beneficiary pays". Sometimes it makes sense to force the renewables generators to pay, as their existence necessitated the load-following service ("causer pays"), and sometimes it makes more sense to socialize those costs to everyone who benefitted from the service provided ("beneficiary pays").

The question of how to allocate these costs is a major point of contention in power market and power system design all around the world right now, and there is no "right answer"; it really depends on your country's economic and energy priorities and the ability of each of the stakeholders to pay. Someone who tries to force a one-size-fits-all to this question is not going to make a very good analyst or advisor!

China deals with these costs in a China-specific way that makes sense for its needs in this moment: Capacity charges for backup coal, batteries, gas, are mostly socialized to end-users (i.e., beneficiary pays). Ancillary services are broadly socialized too. Wind and solar generators are mostly exempt and hardly held financially responsible for any of the losses they cause for other generators in the market. They command a low-carbon privilege and the "victims" are primarily coal-fired generators, as China is trying to peak emissions, so this shouldn't be surprising.

Passing charges to end users is a more socially acceptable approach when you've already been highly effective in keeping power rates low. It also helps to have a huge base of power customers to share the burden. Another benefit of China's scale.

Nuclear generators are basically shielded from this mess, since they sell almost all their power via annual contracts to either the grid or large end-users, guaranteeing their dispatch and sidestepping the gladiator arena of the short-term markets. Even if they have to compete in spot markets someday though, they'll likely be fine, since they should be able to handily out-compete coal-fired generators and consistently earn dispatch. Chinese nuclear is cheap.

Read 5 tweets

David Fishman

@pretentiouswhat

Aug 26

Translation of key portions of the public statement from China's National Energy Administration (NEA) July press conference re: power supply and demand during the 2025 summer peak.

I'll add comments inline [like this] and after.
Translation next post. 🧵
nea.gov.cn/20250731/d34b8…

These comments were from Deputy Director Liu Mingyang of the NEA's Electric Power Department:

"Friends from the media, good morning. Next, I will introduce the power supply situation during this summer's peak demand period.

First, power loads during the summer peak repeatedly set new record highs. In July, peak temperatures were seen, with most provinces experiencing average temperatures 1-2 degrees C higher than the same period in the past. Together with the end of the rainy season in the south, the weather has been hot and humid.

GDP grew by 5.3% YoY in the first half of the year, rapidly driving power loads higher, increasing by over 200 GW versus the end of June. The national peak power load successively saw new records on July 4th, 7th, 16th, and 17th, exceeding 1500 GW and finally reaching a peak of 1508 GW, which is 57 GW higher than the peak load record last year. To date, 19 provinces have seen record-high peak loads, including Jiangsu, Shandong, and Guangdong, breaking records 46 times."

"Second, power supply across the country has been stable overall. In the first half of the year, over 200 GW of new generating capacity was brought online, including 30 GW of supporting and regulating power sources [DF: here, this means dispatchable capacity] including hydropower, gas-fired power, and coal-fired power. Three new cross-region transmission corridors were brought online, increasing cross-region transmission capacity by 16 GW. [DF: this means new UHV lines]

In July, another 10 GW of additional dispatchable capacity was connected to the grid, further strengthening the power supply guarantee. Since summer's start, all kinds of supporting and regulating power have been fully operating. Primary fuels including thermal coal and natural gas have been amply supplied, ensuring stable and orderly national power supply, with only Sichuan needing to implement demand response measures on the evening of the 17th. This demonstrates the system has withstood its first round of high-temperature, high-load challenges this summer."

[DF: This is the first I hear of this, but if Sichuan was truly the only province to use its demand response mechanism during the July peak, then it's quite impressive. I'm quite curious how August has been so far...I have heard some buzz that we might have set ANOTHER new peak a few days ago. Guess we'll find out in the August press conference later this week.]

Read 7 tweets

David Fishman

@pretentiouswhat

Aug 25

⚡️China Power Consumption Update July 2025 ⚡️

Wow.

Chinese power consumption soared to a staggering all-time high of 1,023 TWh (or over 1 petawatt hour) in July 2025.

This blows the previous one-month consumption record out of the water. 🧵

Industrial power consumption was up 4.7% YoY, rising to almost 600 TWh in the monthly of July.

Industrial sector performance continues to strengthen after weaker growth in previous months due to the tariff threat, but still a touch sluggish. Movement in the right direction tho.

Meanwhile, power consumption in the services sector was up a very strong 10.7% YoY in July to 208 TWh. I had to revise my chart's y-axis, as one-month power consumption cracked 200 TWh for the first time.

That's more than 2x the services power consumption of July 2019.

Read 10 tweets

David Fishman

@pretentiouswhat

Aug 24

https://twitter.com/michaeljmcnair/status/1959134225447182439

Ok, I was asked by 5+ people to weigh in, so here we go.

Quick answer: This is wrong. China currently has much more dispatchable capacity (MW) than peak load AND much more capability to generate power (MWh) than is consumed by customers, both by large margins.

Longer Answer:🧵

https://twitter.com/michaeljmcnair/status/1959134225447182439

Installed Capacity vs. Peak Load:

At the end of 1H 2025, China had roughly 2071 GW of dispatchable capacity (thermal, hydro, nuclear, batteries).

Meanwhile, national peak load on 17 July 2025 hit a new high of 1508 GW.

Thus, reserve capacity margin during the recent annual demand peak was ~37%. That's dispatchable capacity...not counting intermittent generators. Of course, in reality, the demand peak arrived at a period when solar generators WERE available (afternoon on a sunny, hot, midsummer day). China now has over 1000 GW of solar (and 650 GW of wind). So some of them were performing at this time, but it's a good conservative practice to not include them in a reserve margin estimation.

Most of the time, load is WELL below 1508 GW and effective reserve margins are higher. In mild evenings of shoulder seasons with no particular heating or cooling demand, reserve margins are at their highest, probably over 100%(?) But in the context of considering a data center that needs to operate 24/7/365, we might as well look at the annual peak as our limiter.

Generation vs. Consumption and Surplus:

Generation and consumption will be balanced, yes (after accounting for plent self-use and line losses). But if the generators COULD generate more, then that means you have a lot of unused potential. This is the case in China.

In 2024, the average annual operating numbers for China's thermal fleet of 1444 GW were just 3442 hours. That works out to an annual capacity factor of ~39% (the average is higher for coal and lower for gas). There's nothing stopping them from generating more, except the power consumption needs of the Chinese economy were already met by other generation sources.

Frankly, these thermal generators could easily generate 2x their current operating hours and it wouldn't be a problem at all (except for the carbon peaking target of course!). If the annual capacity factor of the thermal fleet doubled to a hardly-unreasonable 78%, it would yield an additional 4,970 TWh annually. That's more than the entire annual consumption of the USA.

Yes, China is squatting on an entire USA's worth of untapped power generation solely within the unused capacity of its current thermal fleet.

Additionally the curtailment of wind and solar means there are times throughout the year when these generators are producing power that the grid can't accept - thus it is wasted. National curtailment rates were around 3% last year, but much higher in individual provinces (and they've been rising in 2025).

Read 7 tweets

Share this page!

Enter URL or ID to Unroll

David Fishman

Try unrolling a thread yourself!

More from @pretentiouswhat

David Fishman

David Fishman

David Fishman

David Fishman

David Fishman

David Fishman

Did Thread Reader help you today?

Don't want to be a Premium member but still want to support us?

Send Email!