Hacker News new | past | comments | ask | show | jobs | submit login
Carbon emissions and costs of subsidizing New York nuclear vs renewables (sciencedirect.com)
54 points by toomuchtodo 59 days ago | hide | past | web | favorite | 69 comments

One of the authors is known for heavily cherry-picking his facts, making some pretty extreme and outrageous assumptions, and suing fellow academic critics who point this out. Caveat lector.

Oh, that guy. Suing PNAS and the author of a study critical of his work is honestly quite shocking. One would hope the work would stand for itself, or at least the response should’ve been another paper rather than a lawsuit.

You find it shocking that he might have a valid case against them, or that he has money to waste on legal harassment ?

It is thoroughly inappropriate to respond to legitimate criticism, made in good faith, published in a reputable journal, with a lawsuit.

It doesn't matter if you're right. If you're right; arguments can be made to show that your opponents are wrong (and he was in fact allowed a rebuttal letter in PNAS). Filing a lawsuit breaks any semblance of civility of discourse and dispassionate pursuit of truth; and results in everyone worse off, as the discussion has become thoroughly toxic.

It's burning down the entire house because someone disagreed with your analysis.

Whether it was appropriate or not depends on whether the criticism was legitimate or not. That is the legitimate purpose of legal proceedings to determine.

I guess you would be a heretic to be someone who complains that while the rich want to pay for renewable energy and can afford to live in such a Friedman-esque society in which money buys your civil liberties from the legal system--or even just protection from criticism... meanwhile cost of living, suicide rates, homelessness, civil liberties continue to deteriorate for the working class.

Edit: maybe the downvotes suggest my guess was right.

It is pretty easy to find him doing just that in this paper.

Here's some examples

> Nuclear energy is often seen as a fundamental or bridging technology for future low-carbon systems (International Energy Agency, 2015a; Echavarri, 2013). While it is true that electricity production from nuclear energy is characterized by very low CO2 emissions during the operation phase of the plant, its full life-cycle CO2 emissions, including all up- and downstream processes, are typically much more CO2 intensive.

While this is true, the paper does not make an equivalent comparison to their scenarios. They do not include the upstream and downsteam emissions from PV and wind. Which should include their energy storage requirements. When addressing storage requirements they hand wave, pointing to two papers that also hand wave the requirement. They all but flat out deny the duck curve. But regardless, they don't include these parts into their own scenarios.

Later when they present their 6 scenarios, they assume no change in nuclear. If you do want to make a good comparison moving forward, you should have a scenario 7 that would shut down the current reactors and replace them with new genIII or genIV reactors. You might say I'm being facetious, but the issue is that this is what the nuclear camp wants. The nuclear camp does not want scenario 1 (business as usual), and agree that it is a bad idea. So to make a claim about which direction we want to move forward to be the cheapest and most environmentally friendly you need to address the other positions. While I wouldn't be surprised if this was still a more expensive option, by not addressing it the authors are creating a strawman argument. In reality they do even worse than this, they create scenario 6 which is scenario 1 but with decreased efficiency (which they don't cite evidence for the number they use).

tldr: I'm not sold on the paper. There's merit to it, but the study was not rigorous enough and did not consider the arguments of the opposing scenarios that it is specifically countering.

Don't miss the assumptions they're relying on:

>Emissions are considered per kWh of produced electricity (kWhel), including emissions that occur over the complete life-cycle of a technology (cradle to grave). We use the following values (based on Sovacool (2008), Lenzen (2008) and updated values from Jacobson (2009); nuclear: 66 g-CO2/kWhel, onshore wind: 10 gCO2/kWhel, PV (no difference between utility-scale and rooftop): 30 g-CO2/kWhel.

That paper Jacobson (2009), the subject of the dispute mentioned above, factors in emissions from the burning of cities from nuclear war (at the high end, one city every 30 years), as well the "opportunity cost from planning-to-operation delays." The latter assumes 10-19 years to set up a nuclear plant, during which time emissions from a hypothetical incumbent coal plant are attributed to it. Somehow, PVs and wind have no opportunity costs from delays (though section 4b says they should take 2-5 years). Apparently PV production scaling will never bottleneck as they're deployed grid-wide.


Good to know. I posted it here because it'll get a critical review.


"His core defense: That massive, unexplained increase in hydropower capacity wasn’t a modeling error; it was a modeling assumption. He assumed that the U.S. could simply add turbines on existing dams such that they could deliver 15 times more instantaneous capacity, even though total flows for a year would remain constant"

This seems like a pretty big error. Enough for me to be unsure of the paper linked here.

Judge the math, not the man.

Refute the critics, not sue the critics.

Regardless of what your opinion of the author's character (and my polite rebuttal to look at the data instead), you might be biased on the topic [1]. I'm attempting to discuss from a place of intellectual honesty, and I expect the same from others.

"And it’s the 21st century — its embarrassing that we are talking about wind energy!"

"Simply put, we need nuclear power. We need to spend more time and energy trying to innovate on nuclear energy. That’s what is needed to support the power needs of 10 billion people who will have the energy demands of a modern western nation today. Solar and wind will just not cut it."

[1] https://news.ycombinator.com/item?id=18907686

Hang on. Didn't you just say, "Judge the math not the man." What are the ground rules here. Are we allowed to take into account authors' previous bad behavior or aren't we?

> What are the ground rules here.


"Be civil. Don't say things you wouldn't say face-to-face. Don't be snarky. Comments should get more civil and substantive, not less, as a topic gets more divisive."

"When disagreeing, please reply to the argument instead of calling names. "

"Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith."

"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."

Sure, I agree with all that. But you're dismissing the parent poster based on his comment history, while castigating him for dismissing your link based on the author's lawsuit history. That seems a little inconsistent to me, is all I'm saying.

Personally, I think a history of suing one's critics does open one's subsequent work up to significantly increased skepticism from the community, because it's a strong signal that something's off. There's some norm-breaking going on. I wish I could read this article but it seems like it's behind a paywall.

Posting rational thoughts in other threads does not mean you are biased in everything else. Unless you mean the author is biased against dogma?

>We compare the cost of maintaining a proposed subsidy for New York's three upstate nuclear power plants with the cost of replacing the plants with renewable technologies from 2016 to 2050

Is operating the plants until 2050 really something people consider? From a brief google search those nuclear power plants were all built in the 70s-80s. Power plants are usually designed for 30-40 years of operation. Operating them for 70-80 years seems ill advised from an economic and safety standpoint.

Back when those designs were licensed we didn't have experience under our belts telling how the plant would hold up longer term. Now we do, and thus many plants around the world have gotten extended operating licenses.

I imagine building a new plant -- even to replace an existing one -- would be a politically uphill battle.

There are plenty of folks who consider coal or natural gas to be "more renewable" or "more green" than a nuclear plant.

Wait who considers coal more green than nuclear? That's nonsense. Renewable is understandable as a sort of redefinition of terms, but I don't see how anyone explains away coal as "green."

There is such as thing as "green coal", a.k.a. "clean coal". Methinks it's an oxymoron, but some people seem to disagree :/

I don't think there's anyone who believes in "clean coal" that isn't a lobbyist or a dupe of one. Theoretically coal-with-carbon-capture could be clean, but I'm not sure if any of these have ever been built and they would have high running costs.

So, knowing the way technology works, one has to figure older legacy systems are painful to deal with. Not always, but usually.

Knowing that most code deployments for server side software cannot claim the benefits of total rewrite from scratch, many of us would love to get a chance at trying exactly that.

What would this assessment look like for a nuke-to-nuke replacement, instead of nuke-to-wind/solar?

In other words, if they abruptly terminated use of all the aging nuclear plants today, and immediately started building totally new, highly modern nuclear power facilities, how would that stack up against this research favoring renewables?

Nuclear power doesn't look that great from a cost standpoint once you factor in decomissioning of the plant after use, storing the waste from both fuel and decomissioning (some of it basically forever), etc. A lot of those have been externalties payed by the government, making it profitable for the operating company, but that shouldn't cause us to build more.

If you used modern cutting-edge designs things might look differently, but anything that is actually being built are updated versions of designs from the 70s. Nobody wants to spend lots of money on research and take the risks with building more radical new designs.

Nuclear plants are required by law to save up capital as they go for decommissioning themselves.

They used to be required to pay into a fund for long-term waste storage, and that fund has about $40 billion now. But after the Yucca Mountain project was shut down, the nuclear operators had to store the waste in concrete enclosures on their own properties, so they went to court saying they shouldn't also have to pay the government to do nothing. In 2013 the courts agreed.

There are actually a lot of companies that want to spend money on R&D and build radical new nuclear designs, including Terrapower, Moltex, Terrestrial Energy, Flibe, Thorcon, Elysium, and many others. In the U.S. the NRC makes this extremely difficult.

A couple years ago I got to sit in a meeting between representatives from about a dozen of these companies and a former head of the NRC. The reactor people said their biggest problem was that the NRC requires a near-complete design before they'll even look at it, and then they give a straight yes or no. If yes you still just have a paper reactor, and if no then you're done. The design process costs several hundred million dollars and you have no idea whether the NRC will let you do anything with it. It's a very difficult environment for investors.

The reactor people said just a more phased process would help a lot. The NRC person dismissed their concerns, said it wasn't the NRC's job to help develop new technology or promote nuclear power, and wasn't interested in climate change.

I can understand that this makes new nuclear development in the US a non-starter. Why is the outlook for nuclear just as abysmal in the UK and China though if their regulators are not as disincentivized to permit innovation?

I don't know about the UK, but China has been building new reactors. In fact, that's what's gotten some reactors to a TRL status where they can build them in America.

But from people that I've talked to in the UK, they seem to be more afraid of nuclear than Americans. This seems to be a common trend in Europe. Many people citing radioactive mushrooms from Chernobyl.

China has spent over a billion dollars on molten salt reactors, was partnering with Terrapower on a fast reactor project until the Trump administration stopped it, and has a bunch of other GenIV reactor projects in progress.

Another bright spot is Canada, whose regulators are much friendlier to reactor R&D. Terrestrial Energy has gotten through the hardest part of their licensing process, and expects to have a demo MSR operating by 2025 or so. Seeing their success, Moltex has moved into Canada as well, after making little progress with UK regulators.

There's probably an actual software rewrite in there somewhere, too. Like, for dual-redundant shutdown routines or something.

> In other words, if they abruptly terminated use of all the aging nuclear plants today, and immediately started building totally new, highly modern nuclear power facilities, how would that stack up against this research favoring renewables?

Probably not well.

https://www.technologyreview.com/s/612564/chinas-losing-its-... (China’s losing its taste for nuclear power. That’s bad news.)

https://www.theguardian.com/business/2019/jan/17/hitachi-set... (Hitachi scraps £16bn nuclear power station in Wales)

https://www.nytimes.com/2017/07/31/climate/nuclear-power-pro... (U.S. Nuclear Comeback Stalls as Two Reactors Are Abandoned)

https://www.independent.co.uk/voices/hinkley-point-managemen... (Hinkley Point management is the perfect illustration of how not to conduct a modern energy policy)

https://www.tampabay.com/news/business/energy/trigaux-abando... (Trigaux: Abandoning another nuclear project, Duke Energy mimics its failed Levy plant)

https://www.wfae.org/post/duke-cancels-florida-nuclear-proje... (Duke Cancels Florida Nuclear Project, Plans Solar Expansion)

Shutting existing nuclear without ramping renewables would be catastrophic. Best case scenario is extending their operating licenses 10-15 years in the future until generation capacity has been replaced. It only takes about a year to replace the generating capacity (including capacity factor) of a nuclear generator (which takes ~10 years to build) with renewables.

In the US, there are 65 pressurized water reactors (PWRs) with a combined capacity of about 65 GWe and 33 boiling water reactors (BWRs) with a combined capacity of about 34 GWe – for a total capacity of 99,221 MWe. ~50% more than this amount of generating capacity is coming online as utility scale solar over the next five years in the US. Five years! Replace the entire nuclear fleet with replacement nuclear in five years? Never.

https://pv-magazine-usa.com/2019/01/01/solar-tsunami/ (Developers have applied to build 139 GWac of large-scale solar projects in the territory of six grid operators – around five times what is currently online across the country – and that figure doesn’t even cover the entire United States. By any metric, we are looking at an unprecedented boom in solar development over the next five years.)

To really get a good idea why this is the case, check out the maps on the right side of https://www.eia.gov/electricity/monthly/#tabs_unit-4, and compare the maps between generating units coming online in the next 12 months to the map of generating unit retirements in the next 12 months. All new generation is solar, wind, and natural gas. All retired generation is primarily coal or nuclear.

To be blunt, no further commercial nuclear generation will be built. Maybe we subsidize renewables more? That'd be swank.

That didn’t really answer the question. It was just a blanket statement that it’d be a bad idea.

The links all just point to trends. The inventory just says the way things are. And then a map of actual plans? Who asked about actual plans?

That wasn’t the question at hand. The fundamental question at hand was the difference between old versus new nuclear technology, whether or not whatever the actual plans are.

But yeah, just drop links and say no. Really great. Top notch discussion.

Several of the links I posted explain that new reactor designs were used, and that estimated cost overruns exceeded almost $15 billion in some cases (South Carolina). Can you be more specific about what evidence you're looking for that nuclear, built today, isn't cost effective besides the various citations I've provided? I've provided more information below, and if that isn't sufficient, feel free to email me with additional questions. The TLDR is "Nuclear kept getting more expensive, renewables plummeted in price, and the growth of energy demand in first world countries plateaued".

From the Union of Concerned Scientists:


> The first generation of nuclear power plants proved so costly to build that half of them were abandoned during construction. Those that were completed saw huge cost overruns, which were passed on to utility customers in the form of rate increases. By 1985, Forbes had labeled U.S. nuclear power "the largest managerial disaster in business history.”

> The industry has failed to prove that things will be different this time around: soaring, uncertain costs continue to plague nuclear power in the 21st century. Between 2002 and 2008, for example, cost estimates for new nuclear plant construction rose from between $2 billion and $4 billion per unit to $9 billion per unit, according to a 2009 UCS report, while experience with new construction in Europe has seen costs continue to soar.

Feel free to read the entire blog post to understand exactly why nuclear is so expensive it can't be built. There is also a link to a report the UCS published in 2009 and a financial analysis published in 2011 that goes into great detail.

> Who asked about actual plans?

It's no use saying it would be great in theory if it turns out not to work or be cost effective in practice. This problem dogs a lot of technologies but nuclear has a particularly bad case of making promises that it doesn't meet in practice.

> Duke Cancels Florida Nuclear Project, Plans Solar Expansion

And what Floridians are going yo do at night? Fornicate?

Are they including the cost of storing power for peak demand and low generation times?

Not really. They hand wave it around the end. But their comparisons are not for full life cycles of PV and wind. At least as far as I can tell.

New York state has pumped hydro storage facilities. Any other gaps will be filled by natural gas until utility scale battery storage comes down in price.




> gaps will be filled by natural gas until utility scale battery storage comes down in price.

So emissions will actually rise, until Magic Future Tech (TM) arrives to save our bacon? Great plan. Just great.

Well, if not the climate, at least the fossil industry will be happy about this paper.

That is the same conclusions I reached. The gaps makes up such a significant portion of a year that the economical aspects of running fossil based power plants has not really changed while the overall power consumption moves towards renewable sources. A fossil based power plant will run when the renewable sources can supply the demand, dictating the cost by which ever of gas, coal and oil has cheapest combination of fuel price, investment and c02-costs.

> So emissions will actually rise, until Magic Future Tech (TM) arrives to save our bacon? Great plan. Just great.

That's exactly how we arrived with solar and wind so cheap between the 70s and now. Batteries will get cheaper, as demand for them for EVs brings more manufacturing capacity online.


> “At the end of July, Gigafactory 1 battery production reached an annualized run rate of roughly 20 GWh, making it the highest-volume battery plant in the world by a significant margin. Consequently, Tesla currently produces more batteries in terms of kWh than all other carmakers combined.”

> Panasonic recently commented that ‘production at Tesla is gaining momentum’ and they plan to add 3 battery cell production lines. It should result in an annual production rate of 35 GWh.

> At 35 GWh, they would be producing at the originally announced production rate, but Tesla has since increased the planned total capacity to 105 GWh of battery cells and 150 GWh of total battery pack output.

I agree with you. That demand will drive production up and economies of scale and research will drive costs down. But I still think the article should include all the facts. The storage / peak supply side of things is very expensive at present and it seems to have been left out of the equation. I'll be super happy when renewable are end to end the cheapest options. That's going to be a great day, but I don't think we're there yet.

We have 30GWh worth of battery production capacity annually worldwide. For the cost of a single nuclear power plant at $15 billion dollars you could buy all of this years production.

Pumped hydro is not a solution. It blows the whole cost effectiveness of renewables at any kind of scale.

It’s much simpler to say that NY will just have it’s future indefinitely wedded to fossil fuels.

Considering that literally every nuclear powerplant is a double digit billion dollar boondoggle I don't see why battery storage has to come down in price. We can afford boondoggle priced battery storage today.

Winder power is never just wind power. Looking at the power generation of example Germany we see a massive 75%+ use of wind power in good conditions, but then see it all replaced with coal + oil + gas when it is not.

Power generation is no longer based on demand. Weather sets the energy price, which dictate when it is economical to use fossil fuels. Wind is already at the point where the price per wat is fantastic, the environmental effect minimal, but if you include the need to infuse dirty energy when weather is not optimal the average wat start to look very different.

If we have to pick between wind + coal, or only nuclear, the carbon emission will favor nuclear. An other version is wind + fewer nuclear plants and lower energy usage during low wind conditions, but there is only so much that the energy usage drop when prices goes up. There is also a bunch of wind + "battery" research, but I have not heard of any national power grid fully investing in that.

I'm not sure why you're being downvoted. You make some reasonable points.

I would say though, in the short term, we could stand to increase renewables massively with out needing to solve the storage problem, virtually every watt of installed capacity we add is directly offsetting dirty producers.

Grid scale battery storage is just starting to be implemented, theres also hydro, and 'smart' grid solutions. Whether any of this turns out to be a silver bullet I don't know.

Yes. Short term renewables supplement existing grid and reduce the average pollution. It is good to reduce the problem and should be used so that on good conditions it will practically take over all production. However it can not replace the polluting producers on its own, who seem to have started to switch operation to mostly when there is not enough renewables to supply the demand (marginal costs like fuel + investment vs the so-called merit order).

Nuclear could operate just like the dirty producers and only operate when the combined wind and solar is too low for the demand, which would further reduce the environmental impact, but the massive investment needed in nuclear makes the economics of it a bit complex and I have not read if the marginal costs of nuclear operations make this a viable tactic. It would however reduce the nuclear pollution down to the gaps where wind and solar is not enough.

That leaves future tech solutions. Smart grids sounds great, through I do not know how big part of it is a political issue or technical problem. Just getting the steel industry to focus operation on energy producing peaks seems to be a major challenge, and they already have the incentive that the energy get cheaper if they do so. Converting hydro to create burst of power rather than continuously is also something very interesting but has also both political and technical issues to solve. Then we have different from of battery ideas, all from pumping water to different heat capturing concepts, through I think that is even further away form being implemented on a large scale.

The article talks about the proposed subsidy, and this is where I think we need to think about the grid as a whole. If the goal is to fill the gap when wind and solar can't supply the demand then the discussion should not be renewable vs nuclear. It is very possible that the specific subsidy is a bad tactic and they would be better served by trying to narrow the gap with even more wind and solar, but then we need data to show that existing dirty producers are preferred over nuclear in order to fill the remaining gap. That to me is a hard sell, but I am open to be convinced :).

"Nuclear could operate just like the dirty producers and only operate when the combined wind and solar is too low for the demand"

Not quite. Nuclear isn't that dispatchable.

In general I agree though, short term we should be using all tools at our disposal, long term depends on how energy storage etc works out.

Solar seems like a decent pair with nuclear. Use nuclear at the level of night-time load, solar covers the extra daytime demand, and any remaining discrepancy you cover with either storage or extra nuclear, whichever is cheaper.

A couple years ago MIT had a simulator online that let you try out different grid energy combinations to see their total cost, and this was the cheapest I came up with.

That's not it but it looks interesting!

The one I mentioned let you pick different proportions of energy sources for the entire U.S. grid, and told you how much it would cost, the carbon impact, etc.

>Looking at the power generation of example Germany we see a massive 75%+ use of wind power in good conditions, but then see it all replaced with coal + oil + gas when it is not.

The last time I looked they were importing power from abroad when it's not, which usually meant pulling in excess power from French nuclear plants.

Usage vs production is one of areas where I have had a hard time to get good data. I mostly base my own understand on https://www.electricitymap.org/?page=country&solar=false&rem... and https://www.energy-charts.de/power.htm?source=all-sources&ye...

That said, there is a lot of import/export in Europe on the energy grid.

Thing is nuclear can not be only nuclear. Or actually it can, but then you should compare it to other science fiction tech like solar in orbit, because the kind of reactors that can be steered to demand has just not been designed yet.

This study forget one important fact - reliability of the wind or solar. Unless there is a serious breakthrough in the storage technologies, nuclear is irreplaceable, be it for the good or bad.

Is nuclear good at scaling up or down production?

no. But thats not what its for. It provides solid dependable power. like most gigawatt power generation systems, it takes time to spin up/down.

In NY there are plenty of places where one can build on-demand hydro-electric dams, which can tide over short term peaks.

Everything else is mostly predictable. In new york state, in summer, peak power should correlate with solar power production during the day, but at night, not so much. So you'll need something to supply all those aircon units.


take spain, it has 35% renewable, but it costs 230g per kwh, france, which is almost entirely nuclear is 86g per kwh.

It's great. France literally does load follow based on demand: raises/lowers control rods throughout the day.

The US Nuclear Regulatory Commission does not permit automated load following of US nuclear generators. France and Germany nuclear regulators considered it earlier in the permitting process, hence the functionality exists there (load following incurs higher maintenance costs for components such as valves, but does not incur a higher fuel cost).

There is no need for a breakthrough because nuclear power is so ridiculously expensive that it can't even compete with battery storage at ridiculous boondoggle prices of 500$ per kWH.

> Abstract: "We compare the cost of maintaining a proposed subsidy for New York's three upstate nuclear power plants with the cost of replacing the plants with renewable technologies from 2016 to 2050. Keeping nuclear operating with subsidy until 2050 is the most expensive option, costing $32.4 billion (2014 USD) over that period in the base business as usual case. The least expensive option is to shut down nuclear today and replace it with onshore wind, saving $7.9 billion. All analyzed renewable scenarios lead to 20.1 to 27.4 Mt CO2 greater life-cycle emission reductions. In addition, re-investing the cost savings of the renewable scenarios into additional onshore wind increase CO2 savings up to 32.5 Mt."

This is the kind of junk science that ruins energy discussions. Seriously, hand waving away the energy storage problem of solar/wind by assuming a magic technology will be created?

Other than that Mrs. Lincoln, how was the play?

Does the "replacement" cost include all decommissioning and environmental remediation costs?

I haven't read the entire paper, so I can't answer if the paper addresses that. Note that work (decommissioning and environmental remediation) needs to be done regardless when those nuclear generators reach end of life. In similar fashion, renewables are less expensive than continuing to run existing coal generation facilities in several markets, and those coal plants will need to be decommissioned regardless of ongoing cost comparison.

The cost of remediating all of this legacy generation is going to be eye watering, every nuclear and coal power plant a Superfund site.

That's kind of baked in no matter what though, isn't it?

But you amortize it over the remaining lifetime of the plant. Those costs won't be materially greater after, say, 10 more years of use, but you get those additional years of output.

Fair enough!

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact