Several US utilities back out of deal to build novel nuclear power plant (2020) | Hacker News Reader

Several US utilities back out of deal to build novel nuclear power plant (2020)

sciencemag.org

52 pointszeristor48 comments

> It also estimates the cost would climb from $4.2 billion to $6.1 billion.

The history of every new nuclear project includes promises of cost reduction followed by massive 50% overruns, sometimes multiple times.

> The deal protects UAMPS customers by specifying a maximum cost for electricity from the plant of $55 per MWh

This is good, though. Cheaper than Hinkley Point C.

> Bradford questions how reliable that reassurance can be. He notes that in the 1980s, Washington Public Power Supply System agreed to build several nuclear reactors in Washington that ran far overbudget and were never completed, leading to the biggest default on municipal bonds in U.S. history.

Possibly the only nuclear reactor to have caused a city-wrecking disaster without even being built.

From linked-linked article [1]: "For now, NuScale's reactors exist mostly as computer models [...] the company has built a full-size mock-up of the upper portion of a reactor"

I'd place the bet: never going to happend. Reductions in price for wind and solar power is going to kill any black bottom line for this project.

1: https://www.sciencemag.org/news/2019/02/smaller-safer-cheape...

Wind and solar do not work on demand. It's a huge difference. It's not always sunny nor windy. Only coal, natural gas or hydro in some cases can be a viable baseline energy sources.

I believe we have reached the point where solar/wind plus storage is cheaper than nuclear, and without the massive capital outlay, political problems, or risk. Even solar or wind plus a gas turbine is cheaper.

That's a pity because nuclear could be competitive if economies of scale and innovation occur in that space.

Edit: I'm getting down voted, maybe because I didn't provide a source. But neither do the people who disagree. I stand by it.

I honestly don't care if it's cheaper. I just don't want to keep running gas and coal plants, and continuing to let the planet warm up. The cost when we've burned this planet up will be much higher.

We have a solution to end coal, oil and gas fired power plants, and end a tremendous amount of pollution. But it's somewhat more expensive to run a nuke plant... Is money the only thing we are looking at here?

We keep fiddling around with solar and wind, and while it's great, we're still running goddamn fossil-fuel based plants for a huge percentage of our power. If solar and wind are cheaper, and price is what dictates this market, then we should have much more transitioned away by now. Solar and wind will get better and better, but we need to dump all this carbon-based power right now. We should be building nukes right now to replace the fossil-fuel based plants and then phase them out once wind and solar, and other, better tech gets developed.

We have a huge problem solved right now. It's nothing but politics and bullshit getting in the way.

Yes, the money matters when it's private industry putting it up.

If the government could do the smart thing and enact a carbon tax so the true cost of carbon based fuels could be accounted for, this problem would naturally go away with no need to convince anybody of anything else. Nuclear would maybe be viable, but at the very least building new solar and wind would become cheaper than just operating existing coal plants.

Maybe it's not politically popular? I don't know. Taxes automatically reduce the thing being taxed, but we tax good things like income, rather than bad things like illegal drugs, gambling, pollution, garbage, etc. It makes no sense, but that's why I'm a programmer and not a politician.

You can ask what CO2 tax would be needed to make a particular non-fossil source feasible.

For nuclear in the US, CO2 taxes would have to be $300-400/ton for nuclear to compete with natural gas combined cycle. This number is from the president of Exelon, which operates 20 or so power reactors in the US. Exelon has given up even planning for new nuclear power plants; they are too far out of the running.

I think it should be set at the first to sequester a ton of CO2. Which would be in that neighborhood.

You can't do that overnight, that'd be a huge shock to the economy. But you can implement it over ten to twenty years, ramping it up gradually.

>We have a solution to end coal, oil and gas fired power plants, and end a tremendous amount of pollution. But it's somewhat more expensive to run a nuke plant... Is money the only thing we are looking at here?

I see these types of comments over and over and over.

Wind doesn't always blow. Sun doesn't always shine. We don't have economical grid-scale electrical energy storage technology. Therefore, we can't actually shut down the fossil generation plants, and we end up paying for both.

What is so hard to understand about this?

Show me a working grid scale electrical energy storage system that enables us to spin down fossil generation.

Because I will invest all my money into it.

But it doesn't exist yet.

Um, pumped hydro? Teslas battery facility in Australia?

It certainly can be done. The economics aren't there against natural gas, without subsidies or carbon taxes.

People are working on it; no doubt. But look up the total capacity of these systems and compare to the size of the grid and you'll see that were at the beginning of a very long road.

Nuclear can be on demand, but if you are turning nuclear on/off, you have huge capex but less and less electricity.

Nuclear is so expensive to build, that it doesn't make sense to 'turn off' a nuclear plant. We can do it with control rods, but the economic fallout of leaving nuclear off is bad. (Nuclear will need to be retired after 50 years. Every minute it is turned off is a minute wasted of it's limited lifespan)

-----------

Natural gas is the opposite. High ongoing costs but very low capex. So it makes economic sense to cycle natural gas on and off.

> We can do it with control rods,

Also note that there is 'turning off/down' thermally and electrically.

Once the steam is created, all of it is usually sent to the turbines. But if a nuclear plant need to dial back output, it is possible to route the steam to some place else where it's dissipated and not used to spin a turbine.

You can see this in some of the generator numbers for the Ontario grid, specifically for today (2021-04-24), DARLINGTON-G1 had reduced output:

* https://www.sygration.com/gendata/today.html

> I'd place the bet: never going to happend.

Tend to agree.

Wind and (especially) solar were heavily subsidized by government over the last few years, but that allowed for economies of scale to kick in and the price to drop. There would need to be a few governments willing to 'subsidize' building nukes to kick-start the industry.

It would also be helpful if there weren't a bajillion different designs: if (say) ≤3 nuclear designs could be fixed upon (to not have a complete monoculture), then that would help things as well.

Right now it's basically bespoke construction every time.

This is why Small Modular Reactors have me excited for the future. Small reactors of a single design can be centrally manufactured and shipped to what's basically a concrete hole in the ground, easing regulation requirements and lessening the burden of manufacturing skill onsite.

Why can't we just use candu reactors and pebble bed reactors. They are proven technology made by American allies and work really well.

The cost overruns always come from demanding we have to come up with a novel way to do it.

We could if Canada would build them at the border and sell the power to US grids. The Northeast US already imports clean hydro and nuclear from Ontario, for example, and Canada could use the revenue to replace the oil and gas revenue from tar sands wind downs.

Most of the US south has enough wind and solar potential to avoid needing such a solution. Texas could be a net exporter to other US grids and Mexico if ERCOT gets its act together and upgrades the interconnectors between other grids, not to mention the vast solar potential of the US Southwest.

60% of Ontario capacity is already nuclear and they sell it to the US. So in a way they are already doing this. Just not to the degree they probably should.

If anyone is curious about the actual (real-time) numbers, go to this page† and click on the "Supply" tab:

* https://www.ieso.ca/power-data

Nuclear chugs along with the base load at 9000 to 9500 MW, next biggest is hydro with between 3000 and 5000MW (with some deep troughs at 2500). Wind is the third-biggest, but is very variable: it tends to peak where hydro troughs (2500). Natgas comes along every so often

Ontario retired all of its coal plants several years ago.

If I had a magic want in Ontario I'd say we should build another 2000-2500 nuclear plant, which would take care of some of the base load that hydro is handling. Then hydro would be more free to handle the variable load, and we could reduce/remove natgas.

Various CSV and XML files available with historical data:

* https://www.ieso.ca/en/Power-Data/Data-Directory

†Site run by the Independent Electricity System Operator (Ontario's grid).

Yeah I found out about these numbers when I was reviewing the policy book for the Alberta chamber of commerce.

If anyone really wants to get into the numbers, you can see the output on a per generator basis:

* https://www.sygration.com/gendata/today.html

> The Northeast US already imports clean hydro and nuclear from Ontario

Hydro probably more from Quebec, but Ontario exports as well. (Source: I live in Ontario.)

Ontario put out for bids for new nuclear plants a decade ago. The newest CANDU reactor came in 3x what they were looking for. It's simply not competitive.

https://www.thestar.com/business/2009/07/14/26b_cost_killed_...

Going forward, Canada will likely go with some combination of wind and hydro. There is enormous hydro capacity in Canada, and it pairs very naturally with wind. There are good wind resources in many places east of the Rockies in Canada, especially Quebec (near large hydro resources) and around Hudson Bay.

https://aws-dewi.ul.com/assets/Wind-Resource-Map-CANADA-11x1...

The problem with PBMRs was sposta be the possibility of a graphite fire. I got the impression that there is fear that nobody actually knows how to put one out. I do not know whether this is still a concern.

This article doesn't mention which "utilities" backed out. Probably because the readers would then laugh at the subsequent ridiculousness of the article's premise:

Here are a few: The city of Beaver, Utah. The city of Bountiful, Utah. The cities of Lehi, Logan, and Kaysville Utah.

Is the cost overrun due to engineering - we haven’t built these (NuScale) yet so we don’t really know how hard it is - or due to politics?

Regarding Nuclear power in general, it’s amazing to me that we seem to have lost the engineering ability to actually build technology that’s about 70 years old at this point.

But all the safety measures are not 70 years old. Of course we could technically replicate an old plant.

We have the engineering. In a real sense that is the problem. For any particular complication we can come up with designs and systems that handle and control that. But each and every one such complication whether it is a coating or a backup cooling system or whatever adds up. Ideally nuclear power could be cheap, and it is conceivable that with enough design effort invested we could get there. Unfortunately even the modern simplified designs are not cost competitive with solar, wind, and hydro.

Not to mention zoning laws, neighborhood petitions against it, and the myriad of laws and regulations that make it impossible to even want to attempt it.

The places it was attempted recently in the US (GA, SC) had great local and state support. That didn't stop the efforts from becoming expensive fiascos.

I explained why in https://news.ycombinator.com/item?id=26925722. The cost increase was due to redesigning the project to be larger.

Utah has excellent energy storage possibilities. In addition to lots of topographical relief for pumped hydro, there is a salt formation near Delta that could store enough hydrogen (in solution mined caverns) to power the entire US grid for 30 hours. It would make more sense for them to go with rapidly cheapening renewables + storage than to try one more time to make nuclear work.

Right now, nuclear power plants generate approximately the same amount of power as all renewables combined (in the US). Don’t see why we wouldn’t invest in all proven and environmentally-viable energy solutions. Not to mention the reasons in the article.

We wouldn't invest in it because it's now massively more expensive than the renewable alternatives, and also much more risky. The people handling the billions of dollars know their stuff, and they don't like nuclear for this reason.

That there are existing nuclear plants still running does not imply building new ones would make any sense.

> The people handling the billions of dollars know their stuff, and they don't like nuclear for this reason.

What do you mean by this? I’d be curious to read about that. As far as I knew, nuclear is one of the main investments of the current us government (pages like https://www.energy.gov/science-innovation/energy-sources/nuc... confirm)

Governments (and quasi-governmental entities) invest in things for reasons other than direct return on investment. Essentially all nuclear construction is of this kind.

Private investors spurn nuclear. That's because the cold financial details don't work for it. The first wave of nuclear investment in the US came to a crashing halt in 1978 not because of TMI, but because of the passage of PURPA which opened up the grid to merchant power producers (ostensibly cogeneration, but many were cogen in name only.)

If nuclear were really a good deal, we'd be seeing merchant nuclear plants, financed and built privately. But that money went to fossil fuels, and now increasingly to renewables.

I agree nuclear is generally govt funded rather than commercial. Although TFA is about a commercial nuclear plant (granted, about a minor setback for said commercial plant).... so there’s at least some commercial interest.

Should just paint the development and construction of new nuclear plant as a vital part of "Great Power Competition". Congress will pour money and resources in no time.

There is something to that frame. Based on one unreliable Wikipeida article [0] it looks like nuclear power is becoming something of an Asian technology.

[0] https://en.wikipedia.org/wiki/List_of_nuclear_power_stations...

Asia has 4.6 billion people, of 8 billion in total. If technology is distributed equally among all humans, most technology will be predominantly used by Asians.

Exactly. It's laughable to think that only ~900m people have the right to access new technology.

A much more informative link: https://www.powermag.com/shakeup-for-720-mw-nuclear-smr-proj...

Let us recall that France is 100% nuclear. If we have 20 years to save ourselves from climate change it's hard to understand why environmentalists didn't embrace this proven solution decades ago.

The planned cost was US$4.2 billion or US$4.3 billion; https://www.powermag.com/shakeup-for-720-mw-nuclear-smr-proj... says it was originally 600 megawatts, electric. That's US$7.17 per peak watt. That is not a good price, and the planned expansion to 720 MW and US$6.1 billion doesn't rescue it (US$8.47/W, somehow resulting in a lower LCOE estimate).

Optimistically, if you finance the original US$7.17 per peak watt at 3% per year, that capital is costing you 21.5¢ per year per peak watt. If you can run the reactor at an optimistic 95% capacity factor, that's 22.6¢ per average watt. There are 8765 hours in a year, so that's 2.6¢ per kilowatt hour or US$26 per megawatt hour. That's just the cost of building the plant; operating costs are added on top of that.

https://pv-magazine-usa.com/2020/05/28/record-low-solar-ppas... describes a recent photovoltaic PPA in Texas (not in Utah, but not far away) that's selling solar energy for US$15 per megawatt hour, and a battery-storage PPA that's selling at US$30/MWh, and notes that prices that year at the Palo Verde energy trading hub (in Arizona, a bit nearer to Utah) averaged US$26.58/MWh. With new generation capacity coming online and selling at $15 in the daytime and $30 at night, it's going to be hard to sell nuclear energy in the Southwest at $26/MWh plus opex and transmission costs. (The LCOE estimate in the POWER article, based on the original US$4.3 billion estimate for 600 MW capacity, was US$65/MWh). You can't turn your nuclear plant off in the daytime just because PV is driving LMPs negative. (And turning it off wouldn't reduce your capex even if you could, just your opex—maybe.)

Part of the problem is the quote from the former chair of the Nuclear Regulatory Commission in https://www.sciencemag.org/news/2019/02/smaller-safer-cheape...: "Nuclear does not do anything quickly." And that makes it expensive.

As I said three weeks ago, nuclear energy is the Amiga of energy sources: https://news.ycombinator.com/item?id=26674832

If this is what "Smaller, safer, cheaper" looks like, it's a definition of "cheaper" I'm not familiar with. How did this project get off the ground in the first place? Under what analysis did it look like a good deal?

Why do they need to put these containers together? Why not create a 30 MW box and drop it in suburbia?

That's kinda the design behind russian SVBR-150 (from mid 1990s, still waiting on money for finishing). Self-contained "cells" that can be delivered by railcar ready to use as power/heating/cogeneration plant. You literally need to attach normal COTS steam turbine or heat exchanger for municipal heating, and that's it as far as pure generation is involved.

For refueling you disconnect the pipes and cables, place it on the railcar, and send it to manufacturer.

Initial design for such cell was 150MW(e) each, thus the name.

Firstly it might be easier. One site, one grid connection, one access road for heavy loads, one civil engineering operation, one security fence. One construction compound, fewer land owners and nimby neighbors. All the same reasons you might build larger factories or distribution centers. But more so. Pursuading local residents in suburbia is not a fight worth having.

Also, a project designed to export 1GW can connect to a regional or national grid. Then it can sell power far more widely. Connected to a local grid you have the same capacity issues as solar. Why bother?

Reply on news.ycombinator.com