Thanks for your great explanations. Unfortunately, many of our state elected officials are eager to believe the disinformation coming from the likes of TPPF..... Any suggestions on how we can further diminish the influence of TPPF on energy policy in Texas?
This is a political/communication issue that has dogged us for decades. As a general rule, scientists are not likely to have the mindset and temperament of going up against political machinations. Find a professional lobbyist or try to bribe members of the TPPF to STFU.
Limiting your analysis to wholesale electricity prices is profoundly misleading.
I'll take your figures at face value: average wholesale price in August 2024 was $39 per mWh. In some sense, this would be the average marginal generation cost, which was probably not set by renewable facilities but by marginal fossil fuel facilities. In addition to this, renewable generators got a minimum federal production tax credit of $33.50 per mWh. The subsidy could be as high as $75 per mWh if the project met the Treasury Department's "Labor Requirements", or if the project was under 1mW size. (See page 3 of https://www.energy.gov/sites/default/files/2024-02/508%20Federal%20Solar%20Tax%20Credits%20for%20Businesses_Feb24.pdf). In addition to this, renewable facilities benefited from some state subsidies and tax exemptions - I don't have time to research this in depth, so let's ignore it for the sake of discussion.
Your counterfactual cost analysis recognizes that demand for electricity has increased since 1998. You say that, without new renewable capacity, retail prices would have increased. But this implicitly assumes that no new fossil-fueled capacity would have been added if there had been no new renewable capacity, because the actual history is essentially no net new dispatchable power installation in that time (see page 8 of https://www.ercot.com/files/docs/2025/03/14/ERCOT-2024-State-of-the-Grid.pdf). Is this a reasonable assumption? I realize this is somewhat outside your area of expertise, but we should at least recognize that some new facilities would have been built, meaning there would have been more capacity, which would imply lower average prices. At the very least, extreme price spikes would have been moderated because there would have been new capacity relative to 2018. Possibly, the spikes would have been much lower because the alternative generation facilities would have been dispatchable.
Natural gas prices were certainly a factor in electricity prices in Texas. Comparing 2018 prices to 2024, gas prices decreased by about 25%, (see https://www.eia.gov/dnav/ng/hist/rngwhhdm.htm) which would have had a significant impact on the cost of operating natural gas prices. Some of the decrease from 2018 to 2024 simply reflects lower natural gas prices.
You mention subsidies only to dismiss them, but they matter. The source you linked shows a large volume of subsidies for fossil fuels in 2024, broken down by explicit subsidies ($616 billion in 2023) and "implicit subsidies" (externalities). This source links to a study by the International Monetary Fund, which does not give specific estimates of the externalities, not does it explain how it makes its calculations, which leaves their validity open to question. None of the explicit subsidies are in the US. The IMF implies that the externalities in the US are also quite low.
So, as you've shown before, renewables are having some role in decreasing retail electric prices in Texas. In return, Texans pay part of the cost of their electricity through their state and federal taxes. You have not shown that renewables result in less expensive energy.
"I realize this is somewhat outside your area of expertise, but we should at least recognize that some new facilities would have been built, meaning there would have been more capacity, which would imply lower average prices."
This is definitely outside my area of expertise, but the construction time for new thermal is much longer than for PV, land-based wind and grid battery storage. Construction *costs* for thermal—lots of expensive industrial plumbing, and establishing a fuel feed—is also higher. (And ROI calculations for new thermal had been using over-optimistic and outdated expectations for income from selling power on a grid that has growing amounts of wind, pv and storage.)
Thanks for the feedback. It's been over 20 years since I worked in the power generation industry. At that time (the height of the new construction boom in the US) leadtimes for new gas turbine powerplants ranged from 2 years or more (for GE) to less than a year for some smaller OEMs. What construction times are you referencing?
What is your basis for claiming construction costs for thermal plants (I assume you mean gas turbines) is higher than for solar PV? I recognize the installed cost per kw is in the same range (with ranges for both). But if you include capacity factors, the capital cost per delivered kWh is much higher for the PV.
Construction times for wind/pv/storage are easily less than two years (grid connect overhead seems to be the limiting factor, at least on MISO). Solar farms construct racks of ever-cheaper photovoltaic panels which can be put up with low-skill or quickly-trained labor. Grid battery banks are plug-n-play, too, with pre-fab components delivered from a factory. Land-based wind farms are more involved, of course, and do require small dedicated teams of assembly experts (including really big cranes).
AIUI, nowadays most wind and solar farms have battery front ends, which automatically make it easier to stably interact with the grid.
The big problem for therm right now are the tariff roller coaster and the ongoing major hit to the long-established labor supply (deportations). A shortage of gas turbines foretells construction time for new gas plants as up to five years or more. Videographer Peter Sinclair has compiled a summary of those issues:
The graph of price vs net load is very interesting.
I assume the high net load, and therefore high prices, are at that awkward hour after sundown when everyone is still awake and turning their lights and TVs on, but the solar power isn't there to supply it.
You can see that the moments of highest price are indeed just after sundown, and your graph shows, as might be expected, that these are also the moments of highest net load = load - renewables.
The price at these times depends I suppose on the bids placed by natural gas peakers and battery facilities. Hopefully as battery prices decline, more people will buy some Megapacks and submit lower bids at these awkward hours.
This is a great comment and I’ve looked at this in some detail. Indeed, in the last few years, highest August prices occur around 8 PM as solar drops off and they need to fire up the gas peaker plants. However, in 2018, before there was a lot of solar online, The peak price was 5 PM, when total demand peaked. This is another interesting aspect of adding a lot of solar to the grid.
So far, the existing grid battery setup has worked like a dream for ERCOT. Or, rather, the profitable energy arbitrage (time-shifting cheap energy to times of high demand), takes full advantage of the ERCOT marketplace model, to both's benefit.
Renewable energy infrastructure isn't clean, as you assumed near the start. This is an oft repeated inaccuracy. I suppose the goal is to make people actually believe it but it is scientifically irresponsible.
You also, as many others do, try to show that there is no financial cost to "renewables," and, indeed, there is a financial gain. If this is true then renewables would just continue to accelerate the destruction of the environment. Those savings will get spent on something else, growing the economy and emitting more greenhouse gases.
I'm not sure why you think that AI would provide a good fact-check of my statements but at least it showed that renewables aren't clean and whilst they may be cleaner than fossil fuelled electricity, they are not clean. Unclean without the explainer may be misleading, as AI said, but so is referring to renewables as clean. Unclean is actually factual, clean is not.
AI talked about Jevons Paradox, I didn't. AI just talked about how nothing can be said about how people and companies will spend the savings but it didn't suggest that the savings will be burned. The savings will be spent. Spend on goods and services requires those goods and services to be made or provided and this takes resources and energy, as well as producing waste.
I don't think it's ultimately helpful to omit all of the downsides because it is misleading and, even if it slows the one aspect of environmental damage, it won't stop it. It won't do anything about most of the damage that modernity is doing to the biosphere and other natural systems.
"Renewable energy infrastructure isn't clean, as you assumed near the start."
Which aspect of renewable infrastructure do you consider not clean, especially in the face of the power sources it is displacing? The extraction footprints for wind and solar farms is minuscule compared to the fossil fuel plants and massive dams. Managers of aquifer recharge zones much *prefer* solar arrays with natural ground cover rather than row crops, and farmers that lease out part of their lands for solar arrays have the added benefit of soil rejuvenation in the interim.
And where is anybody saying there is no financial cost to renewables? Of course they have up-front costs. It's just that, in purely investment terms, the ROI of wind/solar/battery is typically much faster than for traditional thermal power generation. And the carbon payback period is typically within a few years.
I'm not talking about relative cleanness, only about the notion that renewables are clean. It takes mining, refining and fabrication to make renewables infrastructure. They all cause damage to the environment. The continued use of "clean" instead of "cleaner" is grossly misleading, perhaps intentionally so.
Of course there is a cost to doing anything. The idea I was referring to is the savings, over fossil fuel energy. It seems that renewables will save people money but we never hear that people will spend those savings, and so cause more damage that way. Whether the total environmental damage may end up being less than with more costly fossil fuels, is not my point. It's that renewables are not a panacea for all our ills. Environmental damage will continue if we ever transition to renewables. They may continue at a lower level but there will still be a predicament and the world will be in a poorer state, environmentally, than it is now, even if the state may be better than it would be if there had been no transition.
This is a great article. However, I wonder how applicable it is elsewhere? Elsewhere in the US? And elsewhere more generally, like where I live in NZ.?
Thanks, Kevin. This should be applicable to any grid that uses local marginal pricing to set the price of electricity. There are lots of places that do this in the US and Europe, but I don’t know about NZ.
Thanks for your great explanations. Unfortunately, many of our state elected officials are eager to believe the disinformation coming from the likes of TPPF..... Any suggestions on how we can further diminish the influence of TPPF on energy policy in Texas?
This is a political/communication issue that has dogged us for decades. As a general rule, scientists are not likely to have the mindset and temperament of going up against political machinations. Find a professional lobbyist or try to bribe members of the TPPF to STFU.
Limiting your analysis to wholesale electricity prices is profoundly misleading.
I'll take your figures at face value: average wholesale price in August 2024 was $39 per mWh. In some sense, this would be the average marginal generation cost, which was probably not set by renewable facilities but by marginal fossil fuel facilities. In addition to this, renewable generators got a minimum federal production tax credit of $33.50 per mWh. The subsidy could be as high as $75 per mWh if the project met the Treasury Department's "Labor Requirements", or if the project was under 1mW size. (See page 3 of https://www.energy.gov/sites/default/files/2024-02/508%20Federal%20Solar%20Tax%20Credits%20for%20Businesses_Feb24.pdf). In addition to this, renewable facilities benefited from some state subsidies and tax exemptions - I don't have time to research this in depth, so let's ignore it for the sake of discussion.
Your counterfactual cost analysis recognizes that demand for electricity has increased since 1998. You say that, without new renewable capacity, retail prices would have increased. But this implicitly assumes that no new fossil-fueled capacity would have been added if there had been no new renewable capacity, because the actual history is essentially no net new dispatchable power installation in that time (see page 8 of https://www.ercot.com/files/docs/2025/03/14/ERCOT-2024-State-of-the-Grid.pdf). Is this a reasonable assumption? I realize this is somewhat outside your area of expertise, but we should at least recognize that some new facilities would have been built, meaning there would have been more capacity, which would imply lower average prices. At the very least, extreme price spikes would have been moderated because there would have been new capacity relative to 2018. Possibly, the spikes would have been much lower because the alternative generation facilities would have been dispatchable.
Natural gas prices were certainly a factor in electricity prices in Texas. Comparing 2018 prices to 2024, gas prices decreased by about 25%, (see https://www.eia.gov/dnav/ng/hist/rngwhhdm.htm) which would have had a significant impact on the cost of operating natural gas prices. Some of the decrease from 2018 to 2024 simply reflects lower natural gas prices.
You mention subsidies only to dismiss them, but they matter. The source you linked shows a large volume of subsidies for fossil fuels in 2024, broken down by explicit subsidies ($616 billion in 2023) and "implicit subsidies" (externalities). This source links to a study by the International Monetary Fund, which does not give specific estimates of the externalities, not does it explain how it makes its calculations, which leaves their validity open to question. None of the explicit subsidies are in the US. The IMF implies that the externalities in the US are also quite low.
So, as you've shown before, renewables are having some role in decreasing retail electric prices in Texas. In return, Texans pay part of the cost of their electricity through their state and federal taxes. You have not shown that renewables result in less expensive energy.
https://chatgpt.com/share/685b0499-f568-8004-8ba4-a77b460d4b0a
Wow, auto-fisking!
I would never do this for an important response but it’s absolutely perfect for this.
"I realize this is somewhat outside your area of expertise, but we should at least recognize that some new facilities would have been built, meaning there would have been more capacity, which would imply lower average prices."
This is definitely outside my area of expertise, but the construction time for new thermal is much longer than for PV, land-based wind and grid battery storage. Construction *costs* for thermal—lots of expensive industrial plumbing, and establishing a fuel feed—is also higher. (And ROI calculations for new thermal had been using over-optimistic and outdated expectations for income from selling power on a grid that has growing amounts of wind, pv and storage.)
Thanks for the feedback. It's been over 20 years since I worked in the power generation industry. At that time (the height of the new construction boom in the US) leadtimes for new gas turbine powerplants ranged from 2 years or more (for GE) to less than a year for some smaller OEMs. What construction times are you referencing?
What is your basis for claiming construction costs for thermal plants (I assume you mean gas turbines) is higher than for solar PV? I recognize the installed cost per kw is in the same range (with ranges for both). But if you include capacity factors, the capital cost per delivered kWh is much higher for the PV.
Construction times for wind/pv/storage are easily less than two years (grid connect overhead seems to be the limiting factor, at least on MISO). Solar farms construct racks of ever-cheaper photovoltaic panels which can be put up with low-skill or quickly-trained labor. Grid battery banks are plug-n-play, too, with pre-fab components delivered from a factory. Land-based wind farms are more involved, of course, and do require small dedicated teams of assembly experts (including really big cranes).
AIUI, nowadays most wind and solar farms have battery front ends, which automatically make it easier to stably interact with the grid.
The big problem for therm right now are the tariff roller coaster and the ongoing major hit to the long-established labor supply (deportations). A shortage of gas turbines foretells construction time for new gas plants as up to five years or more. Videographer Peter Sinclair has compiled a summary of those issues:
https://www.youtube.com/watch?v=MHfwTFEKgg0
The graph of price vs net load is very interesting.
I assume the high net load, and therefore high prices, are at that awkward hour after sundown when everyone is still awake and turning their lights and TVs on, but the solar power isn't there to supply it.
Actually, I plotted Texas 2024 electricity prices as a function of time of day for my battery post: https://www.lesswrong.com/posts/mnaEgW9JgiochnES2/2024-was-the-year-of-the-big-battery-and-what-that-means-for#The_niche_pays_better
You can see that the moments of highest price are indeed just after sundown, and your graph shows, as might be expected, that these are also the moments of highest net load = load - renewables.
The price at these times depends I suppose on the bids placed by natural gas peakers and battery facilities. Hopefully as battery prices decline, more people will buy some Megapacks and submit lower bids at these awkward hours.
This is a great comment and I’ve looked at this in some detail. Indeed, in the last few years, highest August prices occur around 8 PM as solar drops off and they need to fire up the gas peaker plants. However, in 2018, before there was a lot of solar online, The peak price was 5 PM, when total demand peaked. This is another interesting aspect of adding a lot of solar to the grid.
So far, the existing grid battery setup has worked like a dream for ERCOT. Or, rather, the profitable energy arbitrage (time-shifting cheap energy to times of high demand), takes full advantage of the ERCOT marketplace model, to both's benefit.
Full dashboard set: https://www.ercot.com/gridmktinfo/dashboards
Combined wind and solar: https://www.ercot.com/gridmktinfo/dashboards/combinedwindandsolar
Energy storage: https://www.ercot.com/gridmktinfo/dashboards/energystorageresources
Renewable energy infrastructure isn't clean, as you assumed near the start. This is an oft repeated inaccuracy. I suppose the goal is to make people actually believe it but it is scientifically irresponsible.
You also, as many others do, try to show that there is no financial cost to "renewables," and, indeed, there is a financial gain. If this is true then renewables would just continue to accelerate the destruction of the environment. Those savings will get spent on something else, growing the economy and emitting more greenhouse gases.
https://chatgpt.com/share/685b5c39-3da0-8004-89c6-0b81e4e718c9
I'm not sure why you think that AI would provide a good fact-check of my statements but at least it showed that renewables aren't clean and whilst they may be cleaner than fossil fuelled electricity, they are not clean. Unclean without the explainer may be misleading, as AI said, but so is referring to renewables as clean. Unclean is actually factual, clean is not.
AI talked about Jevons Paradox, I didn't. AI just talked about how nothing can be said about how people and companies will spend the savings but it didn't suggest that the savings will be burned. The savings will be spent. Spend on goods and services requires those goods and services to be made or provided and this takes resources and energy, as well as producing waste.
I don't think it's ultimately helpful to omit all of the downsides because it is misleading and, even if it slows the one aspect of environmental damage, it won't stop it. It won't do anything about most of the damage that modernity is doing to the biosphere and other natural systems.
"Renewable energy infrastructure isn't clean, as you assumed near the start."
Which aspect of renewable infrastructure do you consider not clean, especially in the face of the power sources it is displacing? The extraction footprints for wind and solar farms is minuscule compared to the fossil fuel plants and massive dams. Managers of aquifer recharge zones much *prefer* solar arrays with natural ground cover rather than row crops, and farmers that lease out part of their lands for solar arrays have the added benefit of soil rejuvenation in the interim.
And where is anybody saying there is no financial cost to renewables? Of course they have up-front costs. It's just that, in purely investment terms, the ROI of wind/solar/battery is typically much faster than for traditional thermal power generation. And the carbon payback period is typically within a few years.
I'm not talking about relative cleanness, only about the notion that renewables are clean. It takes mining, refining and fabrication to make renewables infrastructure. They all cause damage to the environment. The continued use of "clean" instead of "cleaner" is grossly misleading, perhaps intentionally so.
Of course there is a cost to doing anything. The idea I was referring to is the savings, over fossil fuel energy. It seems that renewables will save people money but we never hear that people will spend those savings, and so cause more damage that way. Whether the total environmental damage may end up being less than with more costly fossil fuels, is not my point. It's that renewables are not a panacea for all our ills. Environmental damage will continue if we ever transition to renewables. They may continue at a lower level but there will still be a predicament and the world will be in a poorer state, environmentally, than it is now, even if the state may be better than it would be if there had been no transition.
This is a great article. However, I wonder how applicable it is elsewhere? Elsewhere in the US? And elsewhere more generally, like where I live in NZ.?
Thanks, Kevin. This should be applicable to any grid that uses local marginal pricing to set the price of electricity. There are lots of places that do this in the US and Europe, but I don’t know about NZ.
"For hours with total load above 69 GW, we need to make an assumption for the price when demand exceeds the highest demand in 2018."
The "extrapolated price" is clearly theoretical, because at that point we would be in residential or industrial brownout/blackout territory. 😉
I’m imagining a counter factual world where we’ve built enough thermal to handle the higher load.
I was thinking instead of ERCOT's "Forced Dispatchable Outages" I see on a regular basis, and presume they're happening to legacy thermal plants.