It's even worse. By his own admission he's just a useful idiot, who's not even smart enough to get paid for his efforts. It should be obvious from the fact that no matter how smart or sensible a given course of action or policy might appear to be on the surface, if it's good for certain for-profit companies or industries it's bad for the rest of us. I might have to cancel my subscription because this kind of analysis that makes too much sense obviously isn't digging deep enough to find out what's really going on.
For a guy who understands economics at a very sophisticated level I often wonder why he doesn't get the difficulty of implementing what is really the logical argument that I accept and agree with. If you want to electrify everything then you need vast amounts, even excessive amounts of generating capacity of every sort. But individual projects have their own economics. These are predicated on producing power the maximum amount of time.
No one is going to produce a generating facility like a nuclear power plant or a natural gas powered plant that is not permitted to operate enough of the time to make them economically viable. This applies to wind and solar too. Both of which are running into the economic wall that makes them not worth building.
I thought this was very interesting, but this open-space argument seems pretty overblown to me. Didn't you write a book about how America is really large?
Also, offshore windfarms are a thing, and roof-like solar panels for crop fields that double-serve as protection against hail etc. already exist. Plus, natural gas also takes up a lot of space because you need to build pipelines.
Think of space as a proxy for backlash. There is already have push back, and we need 10-20x+as much to be built + a simular increase in transmission. My worry is that buildout of renewables will stall out far short of the amount actually needed, and we will have burnt through our collective will and resources to continue the transition with other forms of energy.
Yeah, but that would require that there is a correlation between "space used" and "backlash generated" right? Put another way: I'm unconvinced that renewables have a higher or even equal ratio of "backlash per square feet" than a nuclear power plant, for instance.
Backlash per square foot less than 100x as high favours nuclear... renewables are very dispersed and if we put nuclear at old coal plants arguments around natural beauty etc are not there.
Sure, America is large, but who owns the land? And how easy is it to create right-of-ways for transmission lines?
Just as one example in my own area, we have a local coal plant. The state and utility operators want to shut it down sooner than its planned life and replace it with renewables potentially backed up by gas. But the site of the plant is pretty small. Locating enough land to build enough solar to replace the plant's capacity isn't easy, cheap or simple and would probably require imminent domain. And that's before considering tying that solar into the existing grid and having sufficient backup and peaking capacity.
It doesn't have to be quite as correlated with the location of people as the car infrastructure does though - if it were possible to locate every Walmart parking lot 10 miles outside the farthest suburb, they would be a *lot* less disruptive.
As a separate point, we shouldn't disregard the amount of technological innovation that has happened around wind and solar over the last decade. They both are on a constant trajectory to become cheaper and more powerful. You can build solar panels and wind turbines in places that would've been unthinkable ten years ago. Sorta like fracking technology, where it has become profitable to even go for the smallest specks of oil and gas in the most remote places
Maybe you’re right about technological progress of renewables, but predicting the future is hard. In the present are you for or against addressing the intermittency issues in the ways Matt suggests? I think just saying renewables will work someday isn’t really addressing the points made in the article.
If we tax the net CO2 emission, we get private incentives to find the lowest cost solutions on all the margins. CO2 capture and sequestration may turn out to be the marginal use of much of the intermittently surplus capacity, depending on how it scales.
That article doesn't have anything regarding cost in it that I could see. I'm still pretty doubtful that battery storage is going to scale with any current technology. Maybe there is a significant break through that enables it, but I haven't seen it so far.
*I'm also confused by the end of the article saying
"a recent analysis suggests that the state needs 37GW of batteries over the next 20 years"
while it describes
"The 256 solar-plus-storage projects representing 72GW of solar and 64GW of batteries make up the vast majority of hybrid projects in the CAISO queue..."
Why is there nearly twice the battery production in queue than they estimate they will need - or is this simply another example of projects under discussion that will likely never come to fruition?
There's a bit of a wild-west aspect to storage deployment right now because in competitive generation markets, project owners can use energy arbitrage to make a ton of money very quickly when grid prices spike due to supply constraints. Early storage adopters can under bid gas and diesel peaker plants that only run for a hundred or two hours in a whole year but that generate huge revenues when they run due to price spikes. Storage developers know that as more and more storage is deployed, lucrative price spikes will become rarer and that may make the economics of these projects tougher. OTOH, costs seems to still be on a steeply declining trendline, so it'll probably all work out in the end.
I'm very familiar with the Lazards report. They continue to show that storage is VERY expensive - to the tune that SV and storage are equivalent to nuclear which is also very expensive but has a much longer life than current solar or storage.*
My argument was less about the future and more about the present. I feel like Matt underestimates the current capabilities of renewables. I do agree that natural gas is a good complementary fit for renewables though, as long as it's produced domestically and hydrogen is not yet an option.
I'm not persuaded that manufacturing H2 as a substitute for C oxidation will be cheaper than C oxidation + capture and sequestration. But who knows? It depends on how technology evolves.
They still suck from an engineering point of view. And if you did want to overbuild to get some predictability on output to make the power they produce dispatchible to demand then you need a few other very expensive things like a continent spanning agile grid and massive amounts of energy storage that make wind and solar very expensive indeed. Advocates of wind and solar like to ignore this or pretend it is someone else's problem but it isn't. It will still turn up on your power bill.
Matt never said we would run out of space, but the land required to replace the MWs of a five acre coal plant is like 500 acres or something like that, and I think he is pointing out that there is a non-trivial environmental impact from developing that much unused land. It's not impossible, but it should be taken into account.
America is large, but transmitting power long distances is expensive and inefficient. The panels need to be in the general area of where the power is going to be used, which is more constrained. There are also ecological consequences to shading large areas of ground, once you run out rooftops / parking lots / etc that you can safely shade.
For the amount of solar panels needed to run the country at peak production, this probably isn't an issue. But if you want a 100% solar grid, you're talking about a 5-10x overproduction. ~20,000 square miles could be layered on top of existing infrastructure. 200,000 square miles (5% of the country) is less doable. And that's without taking into account the tremendous battery facilities required.
You miss one important dynamic: methane, the major component of "natural gas," is itself a potent greenhouse gas, 80 times more so than CO2, molecule for molecule, and accounts for about 10% of warming. Compared to CO2, at least some of methane's human-generated sources are easily addressable: leaks in extraction infrastructure and pipelines (which can be fixed) and landfills (which can be managed better). (Ruminants, especially cows, also produce methane, which is one of many reasons you should stop eating beef.) Tackling methane can also pay off quickly, as it decays from the atmosphere whine about a decade. So...any expansion of methane production needs to be accompanied by better regulation of leaks. Thankfully, leak detection has never been easier due to advances in satellite imaging and ground sensors, and the new International Methane Emissions Observatory will soon start to provide frequently-updated, granular methane emissions data. If you're interested in learning more, I recommend Cut Super Climate Pollutants Now! by Miller, Zaelke, and Anderson.
I do think the methane leak aspect was underappreciated when doing the initial calculus of replacing coal with natural gas, but I'm pretty sure that even taking it into account we are still much better off with wind+gas than coal.
Yes to wind, and you can't peak coal plants effectively. I think it's at least debatable that gas is better than coal on a standalone GHG basis when you take methane leakage into account. But you can solve leakage. You can't solve coal burning, which also emits a lot more non-GHG pollutants...including methane, which coal seams release when exposed to air. But it's not like we have a choice between coal and methane/natural gas. We passed the either/or stage a few decades ago. All fossil fuels need to die.
Yes, but when? [And there will always be some niche uses where it makes more sense to oxidize C atoms in one place and reduce (de-oxidize) them in another.]
I hear you, but I worked for a natural gas power plant and the only reason we would concern ourselves much with leaks was from a safety standpoint, we didn't worry much about the cost. Every year when we shut down the plant for maintenance we would vent whatever was left in the pipes to atmosphere prior to performing work. Could we have saved some money by trying to capture that gas and put it back in the pipes when we were done? Yes, but getting the equipment to do that would have cost money and the time and effort it would have taken would cost more money so we would never do it unless someone forced us. Companies don't want to lose money on gas leaks but fixing leaks costs money too so they don't always have enough incentives.
No one anywhere is taxing carbon. Doesn't matter how liberal the state, no one can do it. Even Washington state has voted it down twice. It's just not going to happen.
I think it will; it just need time for progressives to get serious about climate change. The places where it has not succeeded, progressives were not solidly behind it.
Um...you think the problem is not that average voters won't pay more to fix climate change, it's that climate change activists aren't fighting hard enough? The average voter won't spend $100 a year to fight climate change. Have you read anything MY has written? The mobilization myth ring any bells?
Matt's knowledge of clean energy is like the deployment of renewables on the grid: awesome, growing rapidly, and ready for even more valuable growth.
1. "awesome": this is an excellent writeup that will net-educate 95% of its readers. Kudos!
2. "growing rapidly": see previous. Also, he is shifting from nuclear-is-the-answer (false) to nuclear-is-a-tool (true).
3. "ready for even more valuable growth": see below
Broadly, what was true in the last decade (natural gas was the only viable complement to variable renewables) will not be true in the next decade. EV batteries, collectively, will provide enormous load shifting potential. Offshore wind is being built out at city-scale right now. As Matt has written elsewhere but mysteriously omits here, transmission is a valuable complement to variable renewables, and a dozen key additions would enable space-shifting at city-scale. Some of them will get built. Enormous investments in electrolysis and alternate-chemistry electric storage happening now promise economy-scale impact.
Shifting back to solar, wind, and lithium-ion storage: their exponential improvements in price/performance will continue. Exponential growth is hard to grasp, especially when two curves interact. As solar becomes "too cheap to meter", we'll find new ways to exploit it. As road transportation becomes a grid resource, it will soak up that cheap solar and displace some of the natural gas that is now turning on in the evening.
Resources for folks who would like to understand the next decade of clean energy growth:
1. @JesseJenkins, @TimMLatimer, and the rest of #energytwitter
I think these vastly overstate exponential cost reduction when we're talking about things out in the physical world. A solar panel and the battery system it connects to, at this point it's a much different point on the cost curve than at the early days people like to cite.
And demand is going to skyrocket for all of the minerals that are needed to build these things. On the one hand as prices rise, people will mine more of it, but the lags, mismatches, and volatility are really going to surprise a lot of people who just expect this nice, clean and continued reduction in price.
And if you make your entire plan contingent upon rapid continued price decreases, somebody ends up paying if you don't get it done... Whether it's the rate payer, the taxpayer, or the people suffering through reliability issues.
Thanks for the resources, I will have to take a look at those.
I don't understand your assumption that ev batteries will shift load but I can only assume that you think that we will eventually use people's cars to store energy and only charge them when the grid has excess power and then drain them or not charge them when the grid doesn't have enough. That just isn't going to happen. Range anxiety is already the number one obstacle to EV adoption, dictating when people can or can't charge their cars simply won't be acceptable to most people. Evs will change when electrical consumption happens, but most charging is going to happen when people get home from work, which is already one of the daily peaks in power consumption, making the problems of non-baseload power worse, not better.
"assumption ... eventually ... just isn't going to happen": we are doing this now. Dozens of efforts are under way, including my project and a second I know of at my Fortune-200 employer. You are correct to note the co-incidence of "peak plugin" with the evening peak, but that *incredible opportunity* is being exploited today and will scale to 10s of GW in the next few years; 100s soon thereafter.
I think these programs are a very good idea, since making consumers react to power prices to alter their consumption habits is good. But I'm a little skeptical of how much energy consumption can be altered this way and how many people will accept it. Very little of the energy I personally use can be scheduled for off-peak consumption, and the few things I could possibly change would be annoying enough that I wouldn't do it without a large financial incentive. My current energy bill wouldn't be high enough to change my consumption habits and if you try to tell a large group of people to change their consumption or face higher prices you will have the same pushback you get with a carbon tax. As far as charging goes, if I'm going to delay charging one of my vehicles I have to be pretty confident that I won't need it any time soon. I have a PHEV and the few times I have decided I didn't need to charge it right away when I got home I ended up forgetting I needed to run an errand later and always regret it.
Agree that many consumers won’t make many changes without significant incentives and the the fear of not having a charged car will limit uptake. A couple minor points though: 1) Peak pricing will be pushed by utilities, not the legislature like a carbon tax, so faces a much smaller implementation hurdle. 2) More and more devices will shift demand automatically, especially with thermostats and water heaters, as you can easily bank some of the temperature changes overnight.
No utility is going to change my thermostat temperature or decide when I get to take a shower. Pretty sure most people feel the same way. I would rather pay more for electricity than live in a cold house.
Alternatively, you use excess capacity from solar/wind to make hydrogen which you then run through fuel cells when capacity is needed. To make hydrogen, you need water and energy. No fossil fuel needed. Some storage issues, yes (hydrogen is highly explosive and difficult to store) but the technology exists already.
Methane synthesis is only modestly less efficient than hydrogen, and is much easier to store (liquifies at much higher temps, doesn't embrittle metals, etc.), so I think if you want to go that route, making synthetic natural gas from captured CO2 and then re-burning it in methane peaker plants is probably the smarter play.
I was wondering if aluminum refining might be the thing that uses all the summer daytime excess electricity, but hydrogen probably makes a lot more sense. But who knows, maybe it's superpowerful machine learning networks that need a huge amount of electricity to train, so that AI are only trained in the summer.
Aluminum would be worse than general grid use. These kinds of industrial process really do not like changing rates or worse turning off. Aluminum needs constant 24-7 power, so it pairs well with hydro, and would with nuclear, but there would be even more storage required for solar than for general grid.
It’s also why companies mined bauxite in various far-away places and shipped it to Iceland to take advantage of Iceland’s sweet, sweet geothermal capacity.
I think it's important to remember that making green hydrogen is an energy storage process, just like charging a battery. Battery storage is many magnitudes more efficient than turning electricity into hydrogen and then turning hydrogen back into electricity. But hydrogen is much easier and cheaper to store in enormous quantities. It is by no means certain at this point which side is going to make bigger technological advances in time to transform our energy grid.
Worth noting that while renewables are cheap GLOBALLY it is not really the case that they are cheap in the United States.
Solar still costs ~ double what it costs in Europe and nearly triple what it costs in China. Figures are comparable for wind, but not my area of expertise.
There is basically no market for “merchant” solar - selling into the market at market rates. New solar construction needs significant revenue subsidy (in addition to the 26% cost subsidy) to be built.
Activists need to reconcile themselves with the fact that consumer energy prices will have to rise significantly as part of a successful energy transition.
The implicit carbon savings of renewables more than makes up for any subsidies. If ratepayers were carrying the cost of their pollution, gas and coal would be much more expensive.
Sure, but people don't want to pay much higher electricity for carbon savings. Rising electricity costs in California aren't popular and they're only at 20% solar penetration.
Interesting omission of hydrogen. You can use electricity to create hydrogen, which can be stored and then used to create either heat or electricity in fuel cells. Getting to "cheap hydrogen" isn't easy, but it would seem to help you escape the problem with storing renewables.
He mentions energy storage in the article, but thankfully doesn't waste much time talking about it. Currently energy storage is so expensive and inefficient that you are probably better off building an oversupply of renewables than investing much in energy storage. Lots of companies are trying to tackle this problem so I'm confident that we will get there eventually, but it depends on some pretty major technological advancements to make the cost come down and then it will take a decade or more to build the facilities and infrastructure to store even a small percentage of our grids power. So if you want to wait decades to replace coal that's fine, but all the progress we have made so far and most of the progress we make in the next decade will still depend on natural gas.
I’m not sure how many more times it need be said, but the chances that nuclear turns out to be cheaper than massive redundancy and storage are essentially nil.
I am once again begging for a follow up post on this topic : ) Maybe a guest post by an expert or a few? Would love to pull on this thread further.
Milan, given that Matt is very pro-nuclear (as am I to some degree) it would be really interesting to dive into this more given that seemingly extremely well informed commenters keep making strong critiques and the importance of the topic.
As an example for Ontario (16GW average, 25GW summer peak) to use wind and solar only we would need about 5x overbuild (75GW wind, 50 GW solar) and 100 hours of storage. That does not include decrabonising heating or transport either. For nuclear we need about 27GW depending on the exact margin required. While this is a simplified case nuclear would have to be on the order of $15,000/kW, which is well above even current FOAK projects, let alone serial build. That being said, if the industry can't get serial production under control given an oppertuniy to do so, then it is an own goal.
And how is that a slant? Be specific. Real world will be neither all nuclear or all wind solar, and end out using the most valuable parts of both. My point is that if we don't get a clean firm base support, we are just not going to pay the cost to decarbonise. Not in the heating limited climates anyway.
You threw out a flat assertion with no backing at all. But fine. Source, own work using ontario hourly 2021 data. Wind and solar scaled from actual hourly production vs hourly total production data. Assumed $1500/kW for solar and $2500/kW for wind including transmission. Storage $100/kWh+$200/kW. Quantities of each optimized to give minimum install cost, while only having a few hours with shortfall to show it is a true minimum Storage. Certainly could be refined more, but this is not my day job...
My simulator got similar results for an all storage + solar grid. I used real-world GHI data from Albuquerque, New Mexico (one of the friendliest environments for solar) and wasn't able to get the simulated grid to 100% uptime without about 7x as many panels as would be required for their nominal peak production and 120 hours of storage.
Amortizing the battery cost over a 10 year lifespan, that's about $185 per MWH. And substantially worse pretty much everywhere else (the same sim run on data from Madison Wisconsin is not pretty).
Using gas helps drop it down, but I wasn't able to get it below $100 / MWH with > 50% solar contribution to the grid.
Anyone who thinks nuclear is more expensive than solar+storage has not seriously investigated what's required to maintain grid uptime under those conditions. The long tail of cloudy winter days sucks.
Yes the long low output times are the killer. Demand looks to be much less variable than renewable supply. I would much rather build in curtailable demand (hydrogen primarily) that can use the extra firm power when the other demand is not there, but can be reduced at critical times to cover peak cooling or heating loads. Doing the same with renewable gives very spiky extra energy availability, with frankly terrible capacity factors on the curtailable demand.
I want to try simulating natural gas synthesis for power storage (low efficiency but functionally infinite capacity) and nuclear reactors (high predictability but long spin-down times), but need to do some more in-depth work on modelling idle times, because that's a major shortcoming of my model right now when looking at things like gas and nuclear where the idle and production costs are very different.
I'm gonna need to see your work on this one, because there's very little I've seen that supports these numbers.
Wind has a capacity factor of around 30% in Ontario, there's significant hydro base load both in-province and imported from Quebec, and it's also worth noting that Ontario, like Illinois, is already running around 60% nuclear, and those plants absolutely should be kept alive.
As for the nuclear numbers, that sounds pretty much smack on for costs. "Serial build" is a fantasy par excellence, one that will *never* be achieved with Gen. III+ PWRs and most likely not with any of the Gen. IV SMR concepts either.
I did a pure wind and solar model based on ontario data because I kept seeing these endless discussions of just how much does renewable actually cost for 100% penetration. Ontario has convenient data for hourly production of each generator and for demand and is a northern climate. I was honestly expecting renwables to be a lot easier, but the winter low periods are killer. In reality yes, Ontario is a hydro + nuclear grid with a little supplemental wind and gas, and is likely to stay that way. Maybe Alberta would be a better test case for a grid that doesn't have much base that we want to keep already. Ontario is about 18%CF solar and 35% Wind, but last year had at least a week where both together averaged under 10%.
The gas component will go away because it's easy, and my guess is that enhanced geothermal will replace nuclear as those plants just fall the hell apart, but otherwise, yes, there's no reason to build out a purely solar/wind grid that far north.
If we had a nuclear package that worked worth a damn, I agree it'd be great. But at the end of the day, all the hope folks keep placing on nuclear, IMO, much better placed on enhanced geothermal, which uses a well-understood package of technologies in a new way and genuinely does face permitting and regulatory obstacles.
Not Gen. IV SMRs, which are almost all Powerpoint reactors and which will still require extensive shielding and site work.
You're right to note that they can be built at the scale of a sizable petrochemical plant, but the challenges are very different, so it's not a guarantee that they can be built as eaisly. Even if that pans out and we get middling-good at it, $1ish billion for 200 MW is still bloody expensive compared to the potential of "run fracking operation, pump water, set up turbine".
We will find out within ~6-8 years. OPG has a target date of 2028 for a 300MW SMR that they have started site prep for (so actually spending money = serious). If they can keep it within a year or two of that for a FOAK I will be much more optimistic. The project scale is in line with the refurbishments that they have been actually doing well on for the CANDUs so if any North American utility can prove it out, this is probably the chance to see.
If I were to place a bet, I'd go with "they get lapped by at least one, if not both, of low-cost gravity batteries and enhanced geothermal."
Possibly by multiple different start-ups in each space.
I have a fair bit of experience with off-site prefabrication and on-site assembly, and it's a bitch for structural systems with low tolerances and high margins of safety, let alone the opposite. With that in mind, I'd be surprised if they come in anything less than 2X their mooted timeline and 1.5X budget on the first project.
Enhanced geothermal, meanwhile, is really not going to require *anything* new at all. Just a marriage of fracking and a damned steam turbine.
Matt's persistent discussion of nuclear is weirdly detached from the empirical reality. He blames regulators, which may be right, but if the plan is "major regulatory reform in a highly politically sensitive area, then a massive building program", you're talking decades. It's not a near-term solution
The reality is that we have almost no data on how well these very old machines are going to function as they get older and older. Think about keeping a car with 300k miles on the road and having it reliably get you to work every day.
To make matters worse, as renewable penetration on a grid increases, nuclear plants have trouble competing because they can't economically adjust output. Grids with high renewable penetration often see several hours a day of near zero or even negative marginal energy prices. Although nuclear power has relatively low operating costs, they're not zero. More importantly, most of these costs don't drop as output drops (eg nuclear plants need 500-1000 employees compared to a dozen or two for a gas plant) so as nuclear power plant capacity factors fall, you get less and less MWhs of output to amortize operating costs over.
"Grids with high renewable penetration often see several hours a day of near zero or even negative marginal energy prices."
How does that play out in terms of ROI for utility solar? If the average cost of a MWh is $50 so that's what you plan for in terms of your capital repayment plan - but then you're selling when there is already surplus so its selling prices is much less, won't that significantly degrade the value of the project, increase the cost of capital, and make the solar power more expensive. Especially at higher penetrations?
"As a result, the standard model for solar projects is to have some sort of output agreement that either provides for the long-term sale to a utility of the energy output (and associated environmental attributes) at a specified price or that provides a hedge against the price volatility inherent in the spot market. The primary vehicle used in this regard is a long-term (generally 20 years) PPA with an offtaker under which the offtaker agrees to purchase, at a specified price, all energy and related environmental attributes as and when the same are produced by the solar project. That offtaker is often a load-serving utility, but in recent years large commercial and industrial customers have been significant players in the PPA arena in order to accomplish corporate renewable energy goals and/or hedge their own power costs."
I think these projects often leave part of their output unsold. I think a recent trend is to just presell enough output to make debt payments then "play the market" with the rest of the output. This is probably why many of these projects are adding storage to give them more flexibility for when they sell their extra output.
It's weird to hear people say this while 20% of their power comes from nuclear. Clearly it is perfectly possible to make cheap energy from nuclear power. We understand the costs and potential better than storage or large scale geothermal which are both somewhat unproven, so your calculus here is clearly flawed.
Get back to me when someone, somewhere, under any bloody regulatory regime, in any damned construction market, using any freaking technology, successfully builds a nuclear power plant for a cost that is equivalent to an LCOE less than US$100/MWh.
No one has, in decades, despite dozens of attempts. Which leads to the inevitable conclusion that no one *can*.
I don't have LCOE data, but a lot of countries have built nuclear power plants in the last several decades, presumably they all like money. Argentina, Belarus, Brazil, China, Czech Republic, India, Pakistan, South Korea, Romania, Russia, Ukraine, and the UAE. China built 20 nuclear reactors between 2016 and 2020 and they have something like 26 reactors currently under construction. So the inevitable conclusion is that lots of places find it cost effective to build nuclear power plants, and a lot more places would be building more of them...if they didn't last fifty years or more. Those batteries won't last that long.
The two plants have a combined output of 15.8 TWh per year. Assuming a 50 year lifespan (conservative), that would 15.8*1,000,000*50 = 790,000,000 MWh.
$4.58 billion USD / 0.79 billion MWh = $5.8 USD / MWh in upfront capital costs.
I'm aware this isn't the same thing as LCOE since LCOE needs to take into account cost of capital & also operating costs, but I couldn't find a good source for the latter and the former is dependent on the capital-discount rate. I'd be happy to run the numbers again if you know what capital-discount rate is used.
That makes some intuitive sense to me, but are there articles out there making that case in detail? I know the nuclear question tends to revolve around the degree of strict safety regulations and redundancy that drive up construction costs. But is that worldwide or just a feature of the US system? France is always touted as a nuclear power success story - is that replicable?
Even France can't economically build nuclear power plants anymore :
"A third reactor at the site, an EPR unit, began construction in 2007 with its commercial introduction scheduled for 2012. As of 2020 the project is more than five times over budget and years behind schedule. Various safety problems have been raised, including weakness in the steel used in the reactor.[1] In July 2019, further delays were announced, pushing back the commercial introduction date to the end of 2022.[2][3] In January 2022, more delays were announced, with fuel loading continuing until mid-2023"
No one has brought a nuclear power station across the finish line anything like on-budget and on-time in decades. Not the US or UK, not France or Russia, not even "dynamic" China or South Korea.
If a family of technologies cannot be successfully employed across two dozen national regulatory regimes, using 10 different firms' IP, hundreds of different construction firms, 4-5 different construction management practices, multiple different financing models... then the problem isn't the regulatory regime, it's the technology.
No matter what I'm building, I cannot simply have a reactor pressure vessel sitting around with no shielding structure, and the densely-reinforced concrete shield structure is the single biggest factor which makes these projects expensive, difficult to coordinate, fraught with schedule bottlenecks, and likely to go off the rails and require extensive rework on multiple occasions.
That will be true for small modular reactors too; I still cannot just leave them scattered hither and yon for anyone to access. The only real solution to avoiding a shield structure would be to bury them, which introduces a host of other problems in all but the driest, most geologically stable regions, and deep foundation structures/tunnels are the other two ruinously expensive civil engineering works that never end on-budget.
The technology is such that it must be protected everywhere it's deployed, and nothing will change that short of a fairly fundamental change in the human condition. Unless I can fix everything which causes megaproject failures (not likely), I can't come anywhere near rolling nuclear power out at the needed scale.
Whereas mass production and standardized installation processes (for example, PV solar farms or wind mills) are easy. The only thing that need be done is run over loud-mouthed opposition with a truck.
Don't get me wrong, plenty of folks will tell you I'm completely wrong.
But the best understanding I've been able to muster, and I've been involved in project delivery for a while now, is that nuclear is the epitome of everything about construction that we do badly.
As we ramp up solar production we might not be able to cover all of our needs when there is a shortfall of production but at the same time we'll have more electricity than we know what to do with when the sun is shining brightest. I wonder if it would make sense to use that extra electricity doing carbon capture from the air and synthesizing more natural gas from that and water.
Since it comes from atmospheric CO2 it would be carbon neutral on net and we could keep using all of our existing natural gas infrastructure instead of just abandoning it. Also, while electric flight might make sense for very short range flights I don't see anybody taking an electric plane across the pacific any time soon baring things like beamed microwave power from satellites. Designing planes that can use compresses methane seems doable, though, and it's a lot easier to work with than hydrogen.
We could also synthesize more complex fuels, but that's sort of complicated and I'd imagine that minimizing capital intensity would be a big factor in enabling this.
I think more likely we'd take already-existing electricity-hungry industries that don't need to run 24-7 and run them preferentially during the sunny times. I'm thinking aluminum smelting, but some have mentioned hydrogen generation, or perhaps certain kinds of cloud computing.
Most energy intensive industries don't change rates nicely. Hydrogen would work for some kinds of electolysers, but the utilization would be low so CAPEX high per produced quantity. There will be a utilization where even free power is not cheap enough to make it cost effective.
Low capital utilization means that your capital expenses will be higher than they would otherwise be. That doesn't necessarily imply that the savings from lower input costs won't overcome the loses from higher capital costs. And especially as electricity costs can get very low at certain times there's a lot of incentive to find ways to produce cheaper but less efficient plants. I couldn't find any numbers from a full Sabatier plant but this paper on hydrogen costs suggests that electricity price savings can cover a large decrease in duty cycle and have it come out a wash. https://www.hydrogen.energy.gov/pdfs/20004-cost-electrolytic-hydrogen-production.pdf
Based on the numbers they use, if you have free power and a $1000/kw electrolyzer you have $6/kg H2 at ~20% capacity factor. A big portion of the ‘excess’ power will be available less often than this.
> As a broader strategy, though, almost all of the really promising dams have already been built. But even if they hadn’t been, the ecological cost of damming rivers is non-trivial. And that’s the real issue here. There’s no such thing as a way to make electricity that doesn’t have some kind of ecological footprint, and that includes the renewables themselves. After all, the escalating cost of renewables once you run out of gas complementarities really just amounts to the fact that you need to build more stuff.
This is an extremely important point, and I think it should be made louder, more broadly, more generally. Hydro has environmental costs just like solar requires mining some somewhat rare elements and just like we're not really sure what to do with gigantic composite wind turbine blades when we decommission worn-out units. And just like "zero-emission" cars do not exist because lithium mining isn't cheap and because tire wear is (for circa-2020 vehicles) a major source of particulate emissions, no matter whether you have a tailpipe attached to your car or not.
This isn't meant as a bad-faith whataboutism, although unfortunately you'll read whataboutist arguments that look similar. I fully expect my next car to be at least a plug-in hybrid. It's instead an earnest plea that individuals make personal decisions, and policymakers make policy, based on real actual fact patterns rather than well-meaning slogans that sound nice but aren't true.
Still marginal at present for a one-car household that leaves the city regularly. If it's more than 4-5 years from now, then probably, yea. Need a feature set approaching a Model S or a Lucid Air to be pretty much standard before they're a full ICEV replacement.
Good post but Matt neglects discussing the role of housing which consumes lots of energy. We moved into a condo this January from a detached house and our utility bills for both electricity and gas dropped by 70%. Our living space decreased by only 20%. In our building all heating/cooling is done by individual heat pumps up on the roof. We get abundant sun during the day and the heating system hardly runs during the winter. Kitchen is all electric and the induction range uses less energy than our old gas cook top in the house. Town houses use less energy that a detached house of the same square footage as the exposure to the elements is on averaged reduced (end units would have more; middle units less). Moderate/high rise units have even less environmental exposure.
No doubt -- that's Matt's bid for the neo-lib shill prize -- but he has also talked in many places about the ecological advantages of greater density in housing, whether by urbanization or by building up.
Solar and A/C in southern climates pair well. Renewables and heating in Northern climates do not. Peak demand would be start of work but before the sun is up so no help from solar at all, and some days wind is very low. At the same time heat pumps are less efficient so peak demand is even higher.
I don't really understand the enthusiasm for heat pumps. They don't work well in really cold weather, so they can't be that efficient in northern states during the winter.
(1) Heat pump technology is continuing to improve and get more efficient in very cold weather (e.g. https://www.youtube.com/watch?v=_v8vizQXwss -- sponsored by the vendor, FYI).
(2) Lots of people live where it's cold enough to need heating, but not super cold very often, and it's easy to integrate a gas or electric heat strip for those unusually cold days..
(3) Because they function as AC units too, it can be an all-in-one solution.
Your experience must have been with very old tech. Nowadays temps of 15-20 degrees are fine for most heat pumps and more expensive hyper-heat units can put out heat down to -13 degrees.
"Leading products are now capable of performing well below -10°F and operating at more than double the efficiency of resistance or gas systems below zero. These aren’t just manufacturer claims: heat pumps have been successfully field tested in Minnesota (which has some of the coldest winters in the Continental United States) and as far north as the Arctic Circle!"
For a good cost comparisons between heating sources in a relatively cold climate, check out the second table here:
There are also geothermal heat pumps that are good for really cold climates. They are more expensive and not suited for a retrofit in, for example, a NYC apartment.
This is outdated. Modern heat pumps can work well down to 0F and below. Pair it with electric heat strips if you’re worried. We’ve loved getting off gas, despite being in the Rockies.
I have a heat pump at my 900 sq ft apartment in NYC. In the coldest month each winter, my electricity bill is anywhere from $650-$800. Hard to believe this technology is the Green utopia. On the other hand, I fill the gas tank of my car about four times/year, so a cost of about one week of heating my apartment. Don't have any idea how this all translates into my personal carbon footprint, but based on what I pay, the heat pump is a prime offender.
Do you happen to know what the temperature is at which a gas-powered furnace and an electric-powered heat pump use the same amount of energy per degree of heating delivered? I was trying to figure this out in mid-February 2021 when my heat pump here in Texas was struggling against the deep freeze to keep our house in the upper 50s. I assume that in a place where houses are designed to stay warm, and with slightly newer heat pumps and ones selected for a colder climate, there could be somewhat better performance. But it wasn't clear to me whether the technology had reached Minnesota winter needs yet. (It seems very likely to be able to meet Boston/New York needs, particularly in apartment buildings where you just have fewer exposed surfaces to lose heat through.)
Which confirms that the correct strategy for me economically (located in SE PA) is to eke out the life of my 80% efficient furnace as far as possible and then *probably* replace it with a heat pump with gas back-up in no less than 6-8 years once the technology has improved a bit.
Speaking of advanced geothermal and nuclear, i wonder if we could build enough of those to render hydro dams pointless. A free flowing Columbia river would be an amazing feat.
The four pink areas (the ones with the highest heat flow), which are Yellowstone, Clear Lake north of SF, the Imperial Valley in SoCal, and Valles in NM, are all volcanoes. They have an active magma body creating a large amount of heat. In fact, the geothermal field at Clear Lake is the largest in the US. The problem with geothermal in most of the rest of the US is that there is no active magma body, even in the redder parts of the map. You can drill and get some heat for a while, but you will eventually drain all the heat and the site will no longer produce any.
Drawing heat from Yellowstone won't make much a difference either way. The eruption depends on the amount of water in the magma and the amount of rock above putting downward pressure on the magma. Think of when you shake a pop bottle and then open the cap. Same concept.
In what ways? The basic calculations on watts per square meter from renewables seem pretty solid up to today. People mention that solar is more efficient than his expectations, but when I talk to people it seems to still be in the right ballpark.
Nuclear sucks as a complement to intermittent renewable energy.
Very high fixed costs, slow ramping, and low variable costs? That sounds like something you're going to want to run as much as it can.
But in a grid dominated by cheap intermittent renewables, you need two things to complement it - short-term dispatchable power (and/or demand maangement), and seasonal storage of some kind. Neither of these use cases are a good match for what nuclear can do.
At this point, somebody's going to say something silly like "oh, we could use the surplus energy to recharge <energy storage technology of choice>". Well, guess what, if you're going to use energy storage tech, you may as well use the cheapest source of energy available to fill the storage, and it ain't nuclear.
I agree totally that anybody prematurely shutting down a safely operating nuclear plant is nuts. But the track record of nuclear is so unpromising, and it's such a poor match with the rest of the energy landscape, that wasting political effort trying to streamline the NRC and establish an effective construction pipeline for the things seems like misdirected effort.
To repeat myself, the payoff from matching Australia's rooftop solar installation costs is a far juicier, and far more politically viable, than trying to get nuke plants built.
My concern is that this is both true and fully decarbonising with this approach is prohibitively expensive. Is building out this approach locking us into an incomplete path or stranded assets?
So I believe Matt's central contention -that electricity costs are going to spiral out of control once renewable penetration reaches levels a lot lower than what's required to get emissions down to ~0 - is wrong, at least for the US and Australia.
The real question:
Is Matt in the pay of Big Gas? Big Solar? Big Wind? Big Nuke?
Or is it....
[jarring chord]
All of the above?
It's even worse. By his own admission he's just a useful idiot, who's not even smart enough to get paid for his efforts. It should be obvious from the fact that no matter how smart or sensible a given course of action or policy might appear to be on the surface, if it's good for certain for-profit companies or industries it's bad for the rest of us. I might have to cancel my subscription because this kind of analysis that makes too much sense obviously isn't digging deep enough to find out what's really going on.
"...who's not even smart enough to get paid for his efforts."
Exactly. How can you even trust someone who isn't on the take?
It just screams naivete.... What does he expect us to do, evaluate the arguments on their merits? I thought we'd all outgrown that!
For a guy who understands economics at a very sophisticated level I often wonder why he doesn't get the difficulty of implementing what is really the logical argument that I accept and agree with. If you want to electrify everything then you need vast amounts, even excessive amounts of generating capacity of every sort. But individual projects have their own economics. These are predicated on producing power the maximum amount of time.
No one is going to produce a generating facility like a nuclear power plant or a natural gas powered plant that is not permitted to operate enough of the time to make them economically viable. This applies to wind and solar too. Both of which are running into the economic wall that makes them not worth building.
I thought this was very interesting, but this open-space argument seems pretty overblown to me. Didn't you write a book about how America is really large?
Also, offshore windfarms are a thing, and roof-like solar panels for crop fields that double-serve as protection against hail etc. already exist. Plus, natural gas also takes up a lot of space because you need to build pipelines.
Think of space as a proxy for backlash. There is already have push back, and we need 10-20x+as much to be built + a simular increase in transmission. My worry is that buildout of renewables will stall out far short of the amount actually needed, and we will have burnt through our collective will and resources to continue the transition with other forms of energy.
Yeah, but that would require that there is a correlation between "space used" and "backlash generated" right? Put another way: I'm unconvinced that renewables have a higher or even equal ratio of "backlash per square feet" than a nuclear power plant, for instance.
Backlash per square foot less than 100x as high favours nuclear... renewables are very dispersed and if we put nuclear at old coal plants arguments around natural beauty etc are not there.
Sure, America is large, but who owns the land? And how easy is it to create right-of-ways for transmission lines?
Just as one example in my own area, we have a local coal plant. The state and utility operators want to shut it down sooner than its planned life and replace it with renewables potentially backed up by gas. But the site of the plant is pretty small. Locating enough land to build enough solar to replace the plant's capacity isn't easy, cheap or simple and would probably require imminent domain. And that's before considering tying that solar into the existing grid and having sufficient backup and peaking capacity.
The amount of solar we're talking about for an all renewables grid is the built environment footprint of cars. Doable, but really really significant
It doesn't have to be quite as correlated with the location of people as the car infrastructure does though - if it were possible to locate every Walmart parking lot 10 miles outside the farthest suburb, they would be a *lot* less disruptive.
As a separate point, we shouldn't disregard the amount of technological innovation that has happened around wind and solar over the last decade. They both are on a constant trajectory to become cheaper and more powerful. You can build solar panels and wind turbines in places that would've been unthinkable ten years ago. Sorta like fracking technology, where it has become profitable to even go for the smallest specks of oil and gas in the most remote places
Maybe you’re right about technological progress of renewables, but predicting the future is hard. In the present are you for or against addressing the intermittency issues in the ways Matt suggests? I think just saying renewables will work someday isn’t really addressing the points made in the article.
If we tax the net CO2 emission, we get private incentives to find the lowest cost solutions on all the margins. CO2 capture and sequestration may turn out to be the marginal use of much of the intermittently surplus capacity, depending on how it scales.
The issues caused by the intermittency of renewables are basically solved thanks to steeply declining costs for utility scale battery storage.
Today, battery storage is being cost-effectively paired with renewables to provide flexible/dispatchable power.
https://www.pv-magazine.com/2022/03/14/californias-solar-market-is-now-a-battery-market/
That article doesn't have anything regarding cost in it that I could see. I'm still pretty doubtful that battery storage is going to scale with any current technology. Maybe there is a significant break through that enables it, but I haven't seen it so far.
*I'm also confused by the end of the article saying
"a recent analysis suggests that the state needs 37GW of batteries over the next 20 years"
while it describes
"The 256 solar-plus-storage projects representing 72GW of solar and 64GW of batteries make up the vast majority of hybrid projects in the CAISO queue..."
Why is there nearly twice the battery production in queue than they estimate they will need - or is this simply another example of projects under discussion that will likely never come to fruition?
Here's a link to a utility dive article that has some cost figures and also mentions that about 25% of the projects in the queue end up getting built.
https://www.utilitydive.com/news/developers-increasingly-pair-batteries-with-utility-scale-solar-to-combat-d/608117/
Also, here's a link to a widely respected summary of cost info from an investment banking firm that tracks this stuff from the finance side of things.
https://www.lazard.com/perspective/levelized-cost-of-energy-levelized-cost-of-storage-and-levelized-cost-of-hydrogen/
There's a bit of a wild-west aspect to storage deployment right now because in competitive generation markets, project owners can use energy arbitrage to make a ton of money very quickly when grid prices spike due to supply constraints. Early storage adopters can under bid gas and diesel peaker plants that only run for a hundred or two hours in a whole year but that generate huge revenues when they run due to price spikes. Storage developers know that as more and more storage is deployed, lucrative price spikes will become rarer and that may make the economics of these projects tougher. OTOH, costs seems to still be on a steeply declining trendline, so it'll probably all work out in the end.
I'm very familiar with the Lazards report. They continue to show that storage is VERY expensive - to the tune that SV and storage are equivalent to nuclear which is also very expensive but has a much longer life than current solar or storage.*
My argument was less about the future and more about the present. I feel like Matt underestimates the current capabilities of renewables. I do agree that natural gas is a good complementary fit for renewables though, as long as it's produced domestically and hydrogen is not yet an option.
I'm not persuaded that manufacturing H2 as a substitute for C oxidation will be cheaper than C oxidation + capture and sequestration. But who knows? It depends on how technology evolves.
They still suck from an engineering point of view. And if you did want to overbuild to get some predictability on output to make the power they produce dispatchible to demand then you need a few other very expensive things like a continent spanning agile grid and massive amounts of energy storage that make wind and solar very expensive indeed. Advocates of wind and solar like to ignore this or pretend it is someone else's problem but it isn't. It will still turn up on your power bill.
Matt never said we would run out of space, but the land required to replace the MWs of a five acre coal plant is like 500 acres or something like that, and I think he is pointing out that there is a non-trivial environmental impact from developing that much unused land. It's not impossible, but it should be taken into account.
America is large, but transmitting power long distances is expensive and inefficient. The panels need to be in the general area of where the power is going to be used, which is more constrained. There are also ecological consequences to shading large areas of ground, once you run out rooftops / parking lots / etc that you can safely shade.
For the amount of solar panels needed to run the country at peak production, this probably isn't an issue. But if you want a 100% solar grid, you're talking about a 5-10x overproduction. ~20,000 square miles could be layered on top of existing infrastructure. 200,000 square miles (5% of the country) is less doable. And that's without taking into account the tremendous battery facilities required.
You miss one important dynamic: methane, the major component of "natural gas," is itself a potent greenhouse gas, 80 times more so than CO2, molecule for molecule, and accounts for about 10% of warming. Compared to CO2, at least some of methane's human-generated sources are easily addressable: leaks in extraction infrastructure and pipelines (which can be fixed) and landfills (which can be managed better). (Ruminants, especially cows, also produce methane, which is one of many reasons you should stop eating beef.) Tackling methane can also pay off quickly, as it decays from the atmosphere whine about a decade. So...any expansion of methane production needs to be accompanied by better regulation of leaks. Thankfully, leak detection has never been easier due to advances in satellite imaging and ground sensors, and the new International Methane Emissions Observatory will soon start to provide frequently-updated, granular methane emissions data. If you're interested in learning more, I recommend Cut Super Climate Pollutants Now! by Miller, Zaelke, and Anderson.
I do think the methane leak aspect was underappreciated when doing the initial calculus of replacing coal with natural gas, but I'm pretty sure that even taking it into account we are still much better off with wind+gas than coal.
Yes to wind, and you can't peak coal plants effectively. I think it's at least debatable that gas is better than coal on a standalone GHG basis when you take methane leakage into account. But you can solve leakage. You can't solve coal burning, which also emits a lot more non-GHG pollutants...including methane, which coal seams release when exposed to air. But it's not like we have a choice between coal and methane/natural gas. We passed the either/or stage a few decades ago. All fossil fuels need to die.
Yes, but when? [And there will always be some niche uses where it makes more sense to oxidize C atoms in one place and reduce (de-oxidize) them in another.]
As soon as flipping possible.
* "whine" = "within"
I hear you, but I worked for a natural gas power plant and the only reason we would concern ourselves much with leaks was from a safety standpoint, we didn't worry much about the cost. Every year when we shut down the plant for maintenance we would vent whatever was left in the pipes to atmosphere prior to performing work. Could we have saved some money by trying to capture that gas and put it back in the pipes when we were done? Yes, but getting the equipment to do that would have cost money and the time and effort it would have taken would cost more money so we would never do it unless someone forced us. Companies don't want to lose money on gas leaks but fixing leaks costs money too so they don't always have enough incentives.
That's the point of taxing methane emissions just like net CO2 emissions except at higher rates.
No one anywhere is taxing carbon. Doesn't matter how liberal the state, no one can do it. Even Washington state has voted it down twice. It's just not going to happen.
I think it will; it just need time for progressives to get serious about climate change. The places where it has not succeeded, progressives were not solidly behind it.
Um...you think the problem is not that average voters won't pay more to fix climate change, it's that climate change activists aren't fighting hard enough? The average voter won't spend $100 a year to fight climate change. Have you read anything MY has written? The mobilization myth ring any bells?
Matt's knowledge of clean energy is like the deployment of renewables on the grid: awesome, growing rapidly, and ready for even more valuable growth.
1. "awesome": this is an excellent writeup that will net-educate 95% of its readers. Kudos!
2. "growing rapidly": see previous. Also, he is shifting from nuclear-is-the-answer (false) to nuclear-is-a-tool (true).
3. "ready for even more valuable growth": see below
Broadly, what was true in the last decade (natural gas was the only viable complement to variable renewables) will not be true in the next decade. EV batteries, collectively, will provide enormous load shifting potential. Offshore wind is being built out at city-scale right now. As Matt has written elsewhere but mysteriously omits here, transmission is a valuable complement to variable renewables, and a dozen key additions would enable space-shifting at city-scale. Some of them will get built. Enormous investments in electrolysis and alternate-chemistry electric storage happening now promise economy-scale impact.
Shifting back to solar, wind, and lithium-ion storage: their exponential improvements in price/performance will continue. Exponential growth is hard to grasp, especially when two curves interact. As solar becomes "too cheap to meter", we'll find new ways to exploit it. As road transportation becomes a grid resource, it will soak up that cheap solar and displace some of the natural gas that is now turning on in the evening.
Resources for folks who would like to understand the next decade of clean energy growth:
1. @JesseJenkins, @TimMLatimer, and the rest of #energytwitter
2. Chris Goodall's weekly roundups: https://www.carboncommentary.com/newsletter-archive
3. https://www.canarymedia.com
4. California generally: our grid is at solar-saturation now, and we're pushing the GW-scale alternatives to gas as fast as we can. https://www.utilitydive.com/news/california-lowers-electric-sector-ghg-target-directs-procurement-of-more-t/618733/
5. China: https://www.pv-magazine.com/2022/02/25/state-grid-of-china-unveils-plans-for-100gw-battery-fleet/ https://www.reuters.com/world/china/china-aims-build-450-gw-solar-wind-power-gobi-desert-2022-03-05/ @EnergyIceberg
6. 100% clean islanded grids via solar overbuild , li-ion, and electrolysis: https://www.pv-magazine.com/2022/02/21/barbados-to-host-50mw-128-mwh-solar-hydrogen-battery-facility/
For resources, reddit.com/r/energy can be useful as well.
There are some shilling posts there, but also several good commenters and the discussions can be pretty interesting.
I think these vastly overstate exponential cost reduction when we're talking about things out in the physical world. A solar panel and the battery system it connects to, at this point it's a much different point on the cost curve than at the early days people like to cite.
And demand is going to skyrocket for all of the minerals that are needed to build these things. On the one hand as prices rise, people will mine more of it, but the lags, mismatches, and volatility are really going to surprise a lot of people who just expect this nice, clean and continued reduction in price.
And if you make your entire plan contingent upon rapid continued price decreases, somebody ends up paying if you don't get it done... Whether it's the rate payer, the taxpayer, or the people suffering through reliability issues.
Thanks for the resources, I will have to take a look at those.
I don't understand your assumption that ev batteries will shift load but I can only assume that you think that we will eventually use people's cars to store energy and only charge them when the grid has excess power and then drain them or not charge them when the grid doesn't have enough. That just isn't going to happen. Range anxiety is already the number one obstacle to EV adoption, dictating when people can or can't charge their cars simply won't be acceptable to most people. Evs will change when electrical consumption happens, but most charging is going to happen when people get home from work, which is already one of the daily peaks in power consumption, making the problems of non-baseload power worse, not better.
When people charge will also depend on the cost of charging as different times of day.
Maybe this is how your power bill works, but my utility does not differentiate cost based on time of day.
So write whoever regulates your electric utility [or whatever form of political praxis you prax] and demand that they goad the utility to do so. :)
Yeah...no thanks. I'd rather just ask them to increase power from carbon neutral base-load sources.
Agreed but you have to be sure the carbon neutral base-load is lowest cost. That why you need at least a "shadow" carbon tax.
"assumption ... eventually ... just isn't going to happen": we are doing this now. Dozens of efforts are under way, including my project and a second I know of at my Fortune-200 employer. You are correct to note the co-incidence of "peak plugin" with the evening peak, but that *incredible opportunity* is being exploited today and will scale to 10s of GW in the next few years; 100s soon thereafter.
Higher electricity prices at peak rates will become more and more common and will definitely have an effect on when people charge.
https://www.pge.com/en_US/residential/rate-plans/rate-plan-options/time-of-use-base-plan/time-of-use-plan.page
I think these programs are a very good idea, since making consumers react to power prices to alter their consumption habits is good. But I'm a little skeptical of how much energy consumption can be altered this way and how many people will accept it. Very little of the energy I personally use can be scheduled for off-peak consumption, and the few things I could possibly change would be annoying enough that I wouldn't do it without a large financial incentive. My current energy bill wouldn't be high enough to change my consumption habits and if you try to tell a large group of people to change their consumption or face higher prices you will have the same pushback you get with a carbon tax. As far as charging goes, if I'm going to delay charging one of my vehicles I have to be pretty confident that I won't need it any time soon. I have a PHEV and the few times I have decided I didn't need to charge it right away when I got home I ended up forgetting I needed to run an errand later and always regret it.
Agree that many consumers won’t make many changes without significant incentives and the the fear of not having a charged car will limit uptake. A couple minor points though: 1) Peak pricing will be pushed by utilities, not the legislature like a carbon tax, so faces a much smaller implementation hurdle. 2) More and more devices will shift demand automatically, especially with thermostats and water heaters, as you can easily bank some of the temperature changes overnight.
No utility is going to change my thermostat temperature or decide when I get to take a shower. Pretty sure most people feel the same way. I would rather pay more for electricity than live in a cold house.
Matt... you're not even in the Neoliberal Shill bracket this year... there's no need to do this! :p
Alternatively, you use excess capacity from solar/wind to make hydrogen which you then run through fuel cells when capacity is needed. To make hydrogen, you need water and energy. No fossil fuel needed. Some storage issues, yes (hydrogen is highly explosive and difficult to store) but the technology exists already.
Methane synthesis is only modestly less efficient than hydrogen, and is much easier to store (liquifies at much higher temps, doesn't embrittle metals, etc.), so I think if you want to go that route, making synthetic natural gas from captured CO2 and then re-burning it in methane peaker plants is probably the smarter play.
It also lends itself to synthesis of other fuels like, for instance, aviation gas.
I was wondering if aluminum refining might be the thing that uses all the summer daytime excess electricity, but hydrogen probably makes a lot more sense. But who knows, maybe it's superpowerful machine learning networks that need a huge amount of electricity to train, so that AI are only trained in the summer.
Aluminum would be worse than general grid use. These kinds of industrial process really do not like changing rates or worse turning off. Aluminum needs constant 24-7 power, so it pairs well with hydro, and would with nuclear, but there would be even more storage required for solar than for general grid.
It’s also why companies mined bauxite in various far-away places and shipped it to Iceland to take advantage of Iceland’s sweet, sweet geothermal capacity.
That makes sense - thanks for the info.
I think it's important to remember that making green hydrogen is an energy storage process, just like charging a battery. Battery storage is many magnitudes more efficient than turning electricity into hydrogen and then turning hydrogen back into electricity. But hydrogen is much easier and cheaper to store in enormous quantities. It is by no means certain at this point which side is going to make bigger technological advances in time to transform our energy grid.
Worth noting that while renewables are cheap GLOBALLY it is not really the case that they are cheap in the United States.
Solar still costs ~ double what it costs in Europe and nearly triple what it costs in China. Figures are comparable for wind, but not my area of expertise.
There is basically no market for “merchant” solar - selling into the market at market rates. New solar construction needs significant revenue subsidy (in addition to the 26% cost subsidy) to be built.
Activists need to reconcile themselves with the fact that consumer energy prices will have to rise significantly as part of a successful energy transition.
The implicit carbon savings of renewables more than makes up for any subsidies. If ratepayers were carrying the cost of their pollution, gas and coal would be much more expensive.
Sure, but people don't want to pay much higher electricity for carbon savings. Rising electricity costs in California aren't popular and they're only at 20% solar penetration.
I've even heard that net CO2 taxation is not yet popular. :)
Interesting omission of hydrogen. You can use electricity to create hydrogen, which can be stored and then used to create either heat or electricity in fuel cells. Getting to "cheap hydrogen" isn't easy, but it would seem to help you escape the problem with storing renewables.
He mentions energy storage in the article, but thankfully doesn't waste much time talking about it. Currently energy storage is so expensive and inefficient that you are probably better off building an oversupply of renewables than investing much in energy storage. Lots of companies are trying to tackle this problem so I'm confident that we will get there eventually, but it depends on some pretty major technological advancements to make the cost come down and then it will take a decade or more to build the facilities and infrastructure to store even a small percentage of our grids power. So if you want to wait decades to replace coal that's fine, but all the progress we have made so far and most of the progress we make in the next decade will still depend on natural gas.
I’m not sure how many more times it need be said, but the chances that nuclear turns out to be cheaper than massive redundancy and storage are essentially nil.
I am once again begging for a follow up post on this topic : ) Maybe a guest post by an expert or a few? Would love to pull on this thread further.
Milan, given that Matt is very pro-nuclear (as am I to some degree) it would be really interesting to dive into this more given that seemingly extremely well informed commenters keep making strong critiques and the importance of the topic.
WARNING: Not an expert, do not attempt to Shanghai me into writing anything except long-ass comments.
I'm sure we can find an expert guest poster to write something that looks at this, though.
As an example for Ontario (16GW average, 25GW summer peak) to use wind and solar only we would need about 5x overbuild (75GW wind, 50 GW solar) and 100 hours of storage. That does not include decrabonising heating or transport either. For nuclear we need about 27GW depending on the exact margin required. While this is a simplified case nuclear would have to be on the order of $15,000/kW, which is well above even current FOAK projects, let alone serial build. That being said, if the industry can't get serial production under control given an oppertuniy to do so, then it is an own goal.
Vogtle 3 and 4 are currently estimated to cost $28.5b, with a capacity of 2.25GW, which is on the order of $15,000/kW.
https://en.wikipedia.org/wiki/Vogtle_Electric_Generating_Plant
These numbers seem like a master class in how to slant the hell out of your analysis while sounding reasonable.
And how is that a slant? Be specific. Real world will be neither all nuclear or all wind solar, and end out using the most valuable parts of both. My point is that if we don't get a clean firm base support, we are just not going to pay the cost to decarbonise. Not in the heating limited climates anyway.
You threw out a bunch of numbers with zero sourcing, you're not in the strongest of positions to demand footnotes from others...
You threw out a flat assertion with no backing at all. But fine. Source, own work using ontario hourly 2021 data. Wind and solar scaled from actual hourly production vs hourly total production data. Assumed $1500/kW for solar and $2500/kW for wind including transmission. Storage $100/kWh+$200/kW. Quantities of each optimized to give minimum install cost, while only having a few hours with shortfall to show it is a true minimum Storage. Certainly could be refined more, but this is not my day job...
Fair on Vogtle. If the industry can't do much better than that on serial production then there is little hope for us staying industrialized.
My simulator got similar results for an all storage + solar grid. I used real-world GHI data from Albuquerque, New Mexico (one of the friendliest environments for solar) and wasn't able to get the simulated grid to 100% uptime without about 7x as many panels as would be required for their nominal peak production and 120 hours of storage.
Amortizing the battery cost over a 10 year lifespan, that's about $185 per MWH. And substantially worse pretty much everywhere else (the same sim run on data from Madison Wisconsin is not pretty).
Using gas helps drop it down, but I wasn't able to get it below $100 / MWH with > 50% solar contribution to the grid.
Anyone who thinks nuclear is more expensive than solar+storage has not seriously investigated what's required to maintain grid uptime under those conditions. The long tail of cloudy winter days sucks.
Yes the long low output times are the killer. Demand looks to be much less variable than renewable supply. I would much rather build in curtailable demand (hydrogen primarily) that can use the extra firm power when the other demand is not there, but can be reduced at critical times to cover peak cooling or heating loads. Doing the same with renewable gives very spiky extra energy availability, with frankly terrible capacity factors on the curtailable demand.
I want to try simulating natural gas synthesis for power storage (low efficiency but functionally infinite capacity) and nuclear reactors (high predictability but long spin-down times), but need to do some more in-depth work on modelling idle times, because that's a major shortcoming of my model right now when looking at things like gas and nuclear where the idle and production costs are very different.
I'm gonna need to see your work on this one, because there's very little I've seen that supports these numbers.
Wind has a capacity factor of around 30% in Ontario, there's significant hydro base load both in-province and imported from Quebec, and it's also worth noting that Ontario, like Illinois, is already running around 60% nuclear, and those plants absolutely should be kept alive.
As for the nuclear numbers, that sounds pretty much smack on for costs. "Serial build" is a fantasy par excellence, one that will *never* be achieved with Gen. III+ PWRs and most likely not with any of the Gen. IV SMR concepts either.
I did a pure wind and solar model based on ontario data because I kept seeing these endless discussions of just how much does renewable actually cost for 100% penetration. Ontario has convenient data for hourly production of each generator and for demand and is a northern climate. I was honestly expecting renwables to be a lot easier, but the winter low periods are killer. In reality yes, Ontario is a hydro + nuclear grid with a little supplemental wind and gas, and is likely to stay that way. Maybe Alberta would be a better test case for a grid that doesn't have much base that we want to keep already. Ontario is about 18%CF solar and 35% Wind, but last year had at least a week where both together averaged under 10%.
The gas component will go away because it's easy, and my guess is that enhanced geothermal will replace nuclear as those plants just fall the hell apart, but otherwise, yes, there's no reason to build out a purely solar/wind grid that far north.
If we had a nuclear package that worked worth a damn, I agree it'd be great. But at the end of the day, all the hope folks keep placing on nuclear, IMO, much better placed on enhanced geothermal, which uses a well-understood package of technologies in a new way and genuinely does face permitting and regulatory obstacles.
Not Gen. IV SMRs, which are almost all Powerpoint reactors and which will still require extensive shielding and site work.
You're right to note that they can be built at the scale of a sizable petrochemical plant, but the challenges are very different, so it's not a guarantee that they can be built as eaisly. Even if that pans out and we get middling-good at it, $1ish billion for 200 MW is still bloody expensive compared to the potential of "run fracking operation, pump water, set up turbine".
And will require *decades* to scale.
We will find out within ~6-8 years. OPG has a target date of 2028 for a 300MW SMR that they have started site prep for (so actually spending money = serious). If they can keep it within a year or two of that for a FOAK I will be much more optimistic. The project scale is in line with the refurbishments that they have been actually doing well on for the CANDUs so if any North American utility can prove it out, this is probably the chance to see.
If I were to place a bet, I'd go with "they get lapped by at least one, if not both, of low-cost gravity batteries and enhanced geothermal."
Possibly by multiple different start-ups in each space.
I have a fair bit of experience with off-site prefabrication and on-site assembly, and it's a bitch for structural systems with low tolerances and high margins of safety, let alone the opposite. With that in mind, I'd be surprised if they come in anything less than 2X their mooted timeline and 1.5X budget on the first project.
Enhanced geothermal, meanwhile, is really not going to require *anything* new at all. Just a marriage of fracking and a damned steam turbine.
Matt's persistent discussion of nuclear is weirdly detached from the empirical reality. He blames regulators, which may be right, but if the plan is "major regulatory reform in a highly politically sensitive area, then a massive building program", you're talking decades. It's not a near-term solution
Even if building new nuclear plants is not a near-term solution, keeping the ones running we already have is.
This is only sort of true. The current nuclear fleet is very old and this is resulting in less reliability and higher maintenance costs.
https://www.reuters.com/markets/europe/edf-extend-civaux-nuclear-outage-shut-down-reactors-chooz-safety-measures-2021-12-15/
The reality is that we have almost no data on how well these very old machines are going to function as they get older and older. Think about keeping a car with 300k miles on the road and having it reliably get you to work every day.
To make matters worse, as renewable penetration on a grid increases, nuclear plants have trouble competing because they can't economically adjust output. Grids with high renewable penetration often see several hours a day of near zero or even negative marginal energy prices. Although nuclear power has relatively low operating costs, they're not zero. More importantly, most of these costs don't drop as output drops (eg nuclear plants need 500-1000 employees compared to a dozen or two for a gas plant) so as nuclear power plant capacity factors fall, you get less and less MWhs of output to amortize operating costs over.
"Grids with high renewable penetration often see several hours a day of near zero or even negative marginal energy prices."
How does that play out in terms of ROI for utility solar? If the average cost of a MWh is $50 so that's what you plan for in terms of your capital repayment plan - but then you're selling when there is already surplus so its selling prices is much less, won't that significantly degrade the value of the project, increase the cost of capital, and make the solar power more expensive. Especially at higher penetrations?
Subsidizing with a variety of tax credits, state renewable power certificates, selling of carbon credits, etc, etc.
There are many moving parts with these projects. A lot of the output from a project is pre-sold using power purchase agreements.
https://www.stoel.com/legal-insights/special-reports/the-law-of-solar/sections/power-purchase-agreements-utility-scale-projects
"As a result, the standard model for solar projects is to have some sort of output agreement that either provides for the long-term sale to a utility of the energy output (and associated environmental attributes) at a specified price or that provides a hedge against the price volatility inherent in the spot market. The primary vehicle used in this regard is a long-term (generally 20 years) PPA with an offtaker under which the offtaker agrees to purchase, at a specified price, all energy and related environmental attributes as and when the same are produced by the solar project. That offtaker is often a load-serving utility, but in recent years large commercial and industrial customers have been significant players in the PPA arena in order to accomplish corporate renewable energy goals and/or hedge their own power costs."
I think these projects often leave part of their output unsold. I think a recent trend is to just presell enough output to make debt payments then "play the market" with the rest of the output. This is probably why many of these projects are adding storage to give them more flexibility for when they sell their extra output.
It's weird to hear people say this while 20% of their power comes from nuclear. Clearly it is perfectly possible to make cheap energy from nuclear power. We understand the costs and potential better than storage or large scale geothermal which are both somewhat unproven, so your calculus here is clearly flawed.
Get back to me when someone, somewhere, under any bloody regulatory regime, in any damned construction market, using any freaking technology, successfully builds a nuclear power plant for a cost that is equivalent to an LCOE less than US$100/MWh.
No one has, in decades, despite dozens of attempts. Which leads to the inevitable conclusion that no one *can*.
So can we please stop pretending on the topic?
I don't have LCOE data, but a lot of countries have built nuclear power plants in the last several decades, presumably they all like money. Argentina, Belarus, Brazil, China, Czech Republic, India, Pakistan, South Korea, Romania, Russia, Ukraine, and the UAE. China built 20 nuclear reactors between 2016 and 2020 and they have something like 26 reactors currently under construction. So the inevitable conclusion is that lots of places find it cost effective to build nuclear power plants, and a lot more places would be building more of them...if they didn't last fifty years or more. Those batteries won't last that long.
South Korea built the Shin-Wolsong Nuclear Power Plant No. 1 and No. 2 for $4.58 billion USD in 7 years (http://www.koreaherald.com/view.php?ud=20151109001126, https://en.wikipedia.org/wiki/Wolseong_Nuclear_Power_Plant)
The two plants have a combined output of 15.8 TWh per year. Assuming a 50 year lifespan (conservative), that would 15.8*1,000,000*50 = 790,000,000 MWh.
$4.58 billion USD / 0.79 billion MWh = $5.8 USD / MWh in upfront capital costs.
I'm aware this isn't the same thing as LCOE since LCOE needs to take into account cost of capital & also operating costs, but I couldn't find a good source for the latter and the former is dependent on the capital-discount rate. I'd be happy to run the numbers again if you know what capital-discount rate is used.
That makes some intuitive sense to me, but are there articles out there making that case in detail? I know the nuclear question tends to revolve around the degree of strict safety regulations and redundancy that drive up construction costs. But is that worldwide or just a feature of the US system? France is always touted as a nuclear power success story - is that replicable?
Even France can't economically build nuclear power plants anymore :
"A third reactor at the site, an EPR unit, began construction in 2007 with its commercial introduction scheduled for 2012. As of 2020 the project is more than five times over budget and years behind schedule. Various safety problems have been raised, including weakness in the steel used in the reactor.[1] In July 2019, further delays were announced, pushing back the commercial introduction date to the end of 2022.[2][3] In January 2022, more delays were announced, with fuel loading continuing until mid-2023"
https://en.wikipedia.org/wiki/Flamanville_Nuclear_Power_Plant
To make matters worse, as their fleet ages, it's becoming less and less reliable:
https://www.reuters.com/markets/europe/edf-extend-civaux-nuclear-outage-shut-down-reactors-chooz-safety-measures-2021-12-15/
No one has brought a nuclear power station across the finish line anything like on-budget and on-time in decades. Not the US or UK, not France or Russia, not even "dynamic" China or South Korea.
If a family of technologies cannot be successfully employed across two dozen national regulatory regimes, using 10 different firms' IP, hundreds of different construction firms, 4-5 different construction management practices, multiple different financing models... then the problem isn't the regulatory regime, it's the technology.
No matter what I'm building, I cannot simply have a reactor pressure vessel sitting around with no shielding structure, and the densely-reinforced concrete shield structure is the single biggest factor which makes these projects expensive, difficult to coordinate, fraught with schedule bottlenecks, and likely to go off the rails and require extensive rework on multiple occasions.
That will be true for small modular reactors too; I still cannot just leave them scattered hither and yon for anyone to access. The only real solution to avoiding a shield structure would be to bury them, which introduces a host of other problems in all but the driest, most geologically stable regions, and deep foundation structures/tunnels are the other two ruinously expensive civil engineering works that never end on-budget.
The technology is such that it must be protected everywhere it's deployed, and nothing will change that short of a fairly fundamental change in the human condition. Unless I can fix everything which causes megaproject failures (not likely), I can't come anywhere near rolling nuclear power out at the needed scale.
Whereas mass production and standardized installation processes (for example, PV solar farms or wind mills) are easy. The only thing that need be done is run over loud-mouthed opposition with a truck.
Thanks for the explanation!
Of course.
Don't get me wrong, plenty of folks will tell you I'm completely wrong.
But the best understanding I've been able to muster, and I've been involved in project delivery for a while now, is that nuclear is the epitome of everything about construction that we do badly.
As we ramp up solar production we might not be able to cover all of our needs when there is a shortfall of production but at the same time we'll have more electricity than we know what to do with when the sun is shining brightest. I wonder if it would make sense to use that extra electricity doing carbon capture from the air and synthesizing more natural gas from that and water.
Since it comes from atmospheric CO2 it would be carbon neutral on net and we could keep using all of our existing natural gas infrastructure instead of just abandoning it. Also, while electric flight might make sense for very short range flights I don't see anybody taking an electric plane across the pacific any time soon baring things like beamed microwave power from satellites. Designing planes that can use compresses methane seems doable, though, and it's a lot easier to work with than hydrogen.
We could also synthesize more complex fuels, but that's sort of complicated and I'd imagine that minimizing capital intensity would be a big factor in enabling this.
I think more likely we'd take already-existing electricity-hungry industries that don't need to run 24-7 and run them preferentially during the sunny times. I'm thinking aluminum smelting, but some have mentioned hydrogen generation, or perhaps certain kinds of cloud computing.
Most energy intensive industries don't change rates nicely. Hydrogen would work for some kinds of electolysers, but the utilization would be low so CAPEX high per produced quantity. There will be a utilization where even free power is not cheap enough to make it cost effective.
The problem with that is the CAPEX is very high when the utilization is low. The way above demand times are a small portion of the time.
Low capital utilization means that your capital expenses will be higher than they would otherwise be. That doesn't necessarily imply that the savings from lower input costs won't overcome the loses from higher capital costs. And especially as electricity costs can get very low at certain times there's a lot of incentive to find ways to produce cheaper but less efficient plants. I couldn't find any numbers from a full Sabatier plant but this paper on hydrogen costs suggests that electricity price savings can cover a large decrease in duty cycle and have it come out a wash. https://www.hydrogen.energy.gov/pdfs/20004-cost-electrolytic-hydrogen-production.pdf
Based on the numbers they use, if you have free power and a $1000/kw electrolyzer you have $6/kg H2 at ~20% capacity factor. A big portion of the ‘excess’ power will be available less often than this.
> As a broader strategy, though, almost all of the really promising dams have already been built. But even if they hadn’t been, the ecological cost of damming rivers is non-trivial. And that’s the real issue here. There’s no such thing as a way to make electricity that doesn’t have some kind of ecological footprint, and that includes the renewables themselves. After all, the escalating cost of renewables once you run out of gas complementarities really just amounts to the fact that you need to build more stuff.
This is an extremely important point, and I think it should be made louder, more broadly, more generally. Hydro has environmental costs just like solar requires mining some somewhat rare elements and just like we're not really sure what to do with gigantic composite wind turbine blades when we decommission worn-out units. And just like "zero-emission" cars do not exist because lithium mining isn't cheap and because tire wear is (for circa-2020 vehicles) a major source of particulate emissions, no matter whether you have a tailpipe attached to your car or not.
This isn't meant as a bad-faith whataboutism, although unfortunately you'll read whataboutist arguments that look similar. I fully expect my next car to be at least a plug-in hybrid. It's instead an earnest plea that individuals make personal decisions, and policymakers make policy, based on real actual fact patterns rather than well-meaning slogans that sound nice but aren't true.
Make that next car a full electric
Still marginal at present for a one-car household that leaves the city regularly. If it's more than 4-5 years from now, then probably, yea. Need a feature set approaching a Model S or a Lucid Air to be pretty much standard before they're a full ICEV replacement.
Good post but Matt neglects discussing the role of housing which consumes lots of energy. We moved into a condo this January from a detached house and our utility bills for both electricity and gas dropped by 70%. Our living space decreased by only 20%. In our building all heating/cooling is done by individual heat pumps up on the roof. We get abundant sun during the day and the heating system hardly runs during the winter. Kitchen is all electric and the induction range uses less energy than our old gas cook top in the house. Town houses use less energy that a detached house of the same square footage as the exposure to the elements is on averaged reduced (end units would have more; middle units less). Moderate/high rise units have even less environmental exposure.
"...Matt neglects discussing the role of housing..."
You're absolutely right: town houses, apartments, and condos are all more energy efficient than single family dwellings.
But given how often MY trumpets the advantages of building upwards, I think we should welcome his giving the topic a rest on this occasion.
I thought the standard Matt/YIMBY position is "just let people build what they want".
A distinction without a difference. In effect letting people build what they want would have a ton of building upwards.
"just let people build what they want"
No doubt -- that's Matt's bid for the neo-lib shill prize -- but he has also talked in many places about the ecological advantages of greater density in housing, whether by urbanization or by building up.
Solar and A/C in southern climates pair well. Renewables and heating in Northern climates do not. Peak demand would be start of work but before the sun is up so no help from solar at all, and some days wind is very low. At the same time heat pumps are less efficient so peak demand is even higher.
I don't really understand the enthusiasm for heat pumps. They don't work well in really cold weather, so they can't be that efficient in northern states during the winter.
(1) Heat pump technology is continuing to improve and get more efficient in very cold weather (e.g. https://www.youtube.com/watch?v=_v8vizQXwss -- sponsored by the vendor, FYI).
(2) Lots of people live where it's cold enough to need heating, but not super cold very often, and it's easy to integrate a gas or electric heat strip for those unusually cold days..
(3) Because they function as AC units too, it can be an all-in-one solution.
Currently, how cold is too cold for a heat pump?
In my experience, anywhere under about 40 degrees.
Your experience must have been with very old tech. Nowadays temps of 15-20 degrees are fine for most heat pumps and more expensive hyper-heat units can put out heat down to -13 degrees.
https://rmi.org/heat-pumps-a-practical-solution-for-cold-climates/
"Leading products are now capable of performing well below -10°F and operating at more than double the efficiency of resistance or gas systems below zero. These aren’t just manufacturer claims: heat pumps have been successfully field tested in Minnesota (which has some of the coldest winters in the Continental United States) and as far north as the Arctic Circle!"
For a good cost comparisons between heating sources in a relatively cold climate, check out the second table here:
https://www.nh.gov/osi/energy/energy-nh/fuel-prices/
There are also geothermal heat pumps that are good for really cold climates. They are more expensive and not suited for a retrofit in, for example, a NYC apartment.
This is outdated. Modern heat pumps can work well down to 0F and below. Pair it with electric heat strips if you’re worried. We’ve loved getting off gas, despite being in the Rockies.
40 would be pretty bad. Even Atlanta consistently gets under 40 at night in winter.
I have a heat pump at my 900 sq ft apartment in NYC. In the coldest month each winter, my electricity bill is anywhere from $650-$800. Hard to believe this technology is the Green utopia. On the other hand, I fill the gas tank of my car about four times/year, so a cost of about one week of heating my apartment. Don't have any idea how this all translates into my personal carbon footprint, but based on what I pay, the heat pump is a prime offender.
There's something wrong with your heat pump, or it's bloody ancient.
My parents have a brand-new heat pump because their AC died and they only ever had electric baseboard heat before.
They're spending $120 a month to heat a 1970's-era 2200 sq. ft. freestanding home northwest of Philadelphia to 65 degrees.
7-8 years old. I agree it's junk. The one before it failed altogether during a cold snap after five years.
Do you happen to know what the temperature is at which a gas-powered furnace and an electric-powered heat pump use the same amount of energy per degree of heating delivered? I was trying to figure this out in mid-February 2021 when my heat pump here in Texas was struggling against the deep freeze to keep our house in the upper 50s. I assume that in a place where houses are designed to stay warm, and with slightly newer heat pumps and ones selected for a colder climate, there could be somewhat better performance. But it wasn't clear to me whether the technology had reached Minnesota winter needs yet. (It seems very likely to be able to meet Boston/New York needs, particularly in apartment buildings where you just have fewer exposed surfaces to lose heat through.)
I think this might give you something close to what you want:
https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.aceee.org/sites/default/files/publications/researchreports/a1602.pdf&ved=2ahUKEwiHzoOZyfD2AhVgmGoFHSScB-cQFnoECDIQAQ&usg=AOvVaw0_llJso_rmd26ZijD33jj-
Which confirms that the correct strategy for me economically (located in SE PA) is to eke out the life of my 80% efficient furnace as far as possible and then *probably* replace it with a heat pump with gas back-up in no less than 6-8 years once the technology has improved a bit.
Speaking of advanced geothermal and nuclear, i wonder if we could build enough of those to render hydro dams pointless. A free flowing Columbia river would be an amazing feat.
Maybe nuclear. Geothermal would be much more difficult because it isn't as "renewable" as it's made out to be.
Look at this surface heat flow map of the US: https://www.smu.edu/-/media/Site/Dedman/Academics/Programs/Geothermal-Lab/Graphics/SMUHeatFlowMap2011_CopyrightVA0001377160_jpg.jpg?la=en
The four pink areas (the ones with the highest heat flow), which are Yellowstone, Clear Lake north of SF, the Imperial Valley in SoCal, and Valles in NM, are all volcanoes. They have an active magma body creating a large amount of heat. In fact, the geothermal field at Clear Lake is the largest in the US. The problem with geothermal in most of the rest of the US is that there is no active magma body, even in the redder parts of the map. You can drill and get some heat for a while, but you will eventually drain all the heat and the site will no longer produce any.
If we draw some heat out of the Yellowstone basin, do we make the Supervolcano more or less likely?
I'd hate to have to rely on meteorites for the extinction of humankind -- they have a real intermittency problem.
Drawing heat from Yellowstone won't make much a difference either way. The eruption depends on the amount of water in the magma and the amount of rock above putting downward pressure on the magma. Think of when you shake a pop bottle and then open the cap. Same concept.
New geothermal is not subject to those limitations: https://thebreakthrough.org/issues/energy/take-geothermal-seriously
@matt if you haven’t already read you would love this online version if Sustainable Energy Without the Hot Air.
https://www.withouthotair.com
It’s from a UK perspective but it grounds the energy transition in “how much energy does the UK use va how much energy per meter do renewables give?”
It is still very relevant, and informs my thinking to this day on the importance of nuclear and gas.
It doesn’t touch as much on cost, but does imply similar conclusions about the important of non variable backups to solar and wind.
Sadly Dr McKay died too young, but you can find his lectures on YouTube.
https://youtu.be/GFosQtEqzSE
His work also led to the creation of government calculators for showing how you would achieve the energy transition https://my2050.beis.gov.uk/?levers=111111111111111
Was great; is badly outdated.
In what ways? The basic calculations on watts per square meter from renewables seem pretty solid up to today. People mention that solar is more efficient than his expectations, but when I talk to people it seems to still be in the right ballpark.
Nuclear sucks as a complement to intermittent renewable energy.
Very high fixed costs, slow ramping, and low variable costs? That sounds like something you're going to want to run as much as it can.
But in a grid dominated by cheap intermittent renewables, you need two things to complement it - short-term dispatchable power (and/or demand maangement), and seasonal storage of some kind. Neither of these use cases are a good match for what nuclear can do.
At this point, somebody's going to say something silly like "oh, we could use the surplus energy to recharge <energy storage technology of choice>". Well, guess what, if you're going to use energy storage tech, you may as well use the cheapest source of energy available to fill the storage, and it ain't nuclear.
I agree totally that anybody prematurely shutting down a safely operating nuclear plant is nuts. But the track record of nuclear is so unpromising, and it's such a poor match with the rest of the energy landscape, that wasting political effort trying to streamline the NRC and establish an effective construction pipeline for the things seems like misdirected effort.
To repeat myself, the payoff from matching Australia's rooftop solar installation costs is a far juicier, and far more politically viable, than trying to get nuke plants built.
My concern is that this is both true and fully decarbonising with this approach is prohibitively expensive. Is building out this approach locking us into an incomplete path or stranded assets?
Short version: no.
Longer version: there are multiple studies/plans out there suggesting that very high proportions of renewables won't break the bank.
There's this one for the US from a group of Stanford academics: https://web.stanford.edu/group/efmh/jacobson/Articles/I/21-USStates-PDFs/21-USStatesPaper.pdf
And the draft grid management plan for the Eastern Australian energy grid has as its central scenario ~95% renewables by about 2040: https://aemo.com.au/-/media/files/major-publications/isp/2022/draft-2022-integrated-system-plan.pdf?la=en
So I believe Matt's central contention -that electricity costs are going to spiral out of control once renewable penetration reaches levels a lot lower than what's required to get emissions down to ~0 - is wrong, at least for the US and Australia.
I have worked in electrical power generation for over ten years and this is the best article I have ever read summarizing our energy situation.