Hacker News Reader

Comment by dvh | Hacker News Reader

What would be equivalent solar/wind installed power?

Enough wind + solar + storage to deliver

> which gives Hinkley Point C a total output of over 3.2GW.

at least 3 GW 24 hours a day, 7 days a week, 50 weeks a year (allowing for some step down in output for maintainance, etc) for some 30+ (?) years.

The UK’s average nuclear load factor for 1970 to 2017 was 67.4 per cent

3 GW 24 hours a day, 7 days a week, 34 weeks a year.

3 GW 16 hours a day, 7 days a week, 52 weeks a year.

Interesting, thank you for that.

Do you have a source for that and any context (eg: "not power" time being used for "experiments" related to UK military nuclear uses, etc)?

No context sorry - there are enough experimental reactors in the UK for experimentation without interruption the commercial fleet.

What would be interesting is expected vs unexpected downtime, some sort of reliability factor.

https://assets.publishing.service.gov.uk/media/5c9a5d37ed915...

No drama, good link, thank you.

It appears approx ballpark with Europe reactors and as if they're designed with a 70% of max load operating target to allow for downtime inspections and statutory fettling, etc.

To be fair the European average (depending on which average that is) is skewed by France where the reactors load-follow.

IAEA's PRIS database is good for reactor information.

For instance https://pris.iaea.org/PRIS/CountryStatistics/ReactorDetails.... shows Sizewell B has a load factor of 83.1% which is better than the AGR fleet.

That is an excellent link thanks. Really sad reminder of how many sites were opened and how many reactors built in the 60's & 70's and how little has been done since (except for power down reactors).

It is possible for reactors to come back from shut-down. See Palisades and TMI 1.

https://www.world-nuclear-news.org/articles/palisades-on-sch...

https://www.world-nuclear-news.org/articles/constellation-to...

And China seems to be taking all forms of energy security seriously (intermittents as well as nuclear, combined with EVs).

60 years

It's a common mistake to price up renewables delivering flat nuclear shaped energy in an effort to inflate the price.

It's as silly as demanding that a nuclear plant must emulate the delivery profile of solarn and install batteries capable of shifting night time power to the day. With a similar effect on prices.

Demand is not flat. It's actually a bit more solar shaped on many markets which is why initial solar deployments have a high "capture rate" (i.e. they are delivering when the demand and price is high).

> It's a common mistake to price up renewables delivering flat nuclear shaped energy in an effort to inflate the price. > Demand is not flat.

The system as a whole has to deliver baseload plus variable demand. If the demand and supply do not match then there are problems. Load shifting/Demand Flexibility Service may be a promising concept, but there is only so much that will shift.

> It's actually a bit more solar shaped on many markets

That is not the case for the UK, the country being discussed, with peak electrical demand in winter. Electrification of heating (and not so great a need for cooling in the summer) indicates this will continue to be so.

> That is not the case for the UK, the country being discussed, with peak electrical demand in winter.

Wind also peaks in winter, while nuclear fails to rise to the occasion so maybe we should force nuclear to invest in seasonal storage so that we get a really fair comparison?

Wind also peaks in peaks, leading to gluts. Summarising something so irregular in a single number misses out essential properties. The peak/trough pattern is so good for heating without plenty of thermal storage (not something the UK’s buildings are renowned for).

> maybe we should force nuclear to invest in seasonal storage

Probably simpler to build a bit more nuclear, then use it in summer to drive DACCS (heat and electricity) or processes like raw material smelting. Better to have industry operating on fairly predictable schedules (summer on, winter off) than pre-empting them on a regular basis (as startup/shutdown could reduce useful work done/harm equipment).

Overbuilding is a good strategy, cheaper to do it with wind and solar though.

This has sometimes been referred to with the cheesy catchphrase of "SWB Superpower":

https://www.rethinkx.com/energy/in-depth/super-power

Dogger Bank Wind Farm

    Each phase will have an installed generation capacity of 1.2GW and represents a multi-billion pound investment. Combined, they will have an installed capacity of 3.6GW and will be capable of powering up to 6 million homes annually.

~ https://doggerbank.com/

Sounds about right in scale .. how many homes in the UK in total?

> how many homes in the UK in total?

About 25 million it seems.

https://www.statista.com/statistics/378391/uk-england-housin...

Thing is, "capacity" is the maximum generation in ideal conditions, so for wind the actual produced amount will be much lower than this.

The UK is a bit of a special case here. Although what you said is literally true, our wind energy has only 5% of the country's capacity, but provides nearly 30% of the consumed energy.

The Dogger Bank Wind site recognises this and bases the number of households serviced on proven reasonable wind conditions (apparently)

    6 million homes powered per annum based on Typical Domestic Consumption Values (Medium Electricity Profile Class 1, 2,900kWh per household; OFGEM, January 2021), typical 55% wind load factor, and projected installed capacity of 3.6GW.

Right, but doesn't that mean projected output of 55% of 3.6Gw? i.e., about 2GW

There's probably a more detailed report to be found if you go looking, and

the same qualifications apply to the nuclear reactors and turbines - it's unlikely they'll always run at peak all the time either.

The obligatory "what's the capacity factor" discussion should happen here; 40% to 50% might be reasonable, with highly correlated availability.

Additionally the wind farm requires backup (CCGT or OCGT), which cost money to sit there. https://assets.publishing.service.gov.uk/media/6556027d046ed... (take it with a pinch of salt) gives a figure of £80/kW/year for an H class CCGT; that's £250M/year (2015 prices).

Another talking point might be the lifetime of wind farms (20 or 25 years) versus 60 years design lifetime for the EPR. Note that there are lifetime extensions to 80 years being mooted in the US and Switzerland.

The capex per watt of gas backup is vastly lower than a nuclear plant. Like, by a factor of 10 to 20.

> The capex per watt of gas backup is vastly lower than a nuclear plant. Like, by a factor of 10 to 20.

https://www.iea.org/reports/projected-costs-of-generating-el... tables Table 3.2a and 3.4a show that Korean CCGT cost 838 USD/kWe and Korean PWRs were 2157 USD/kWhe (all 2018 prices).

Anyhow, gas backup fuel costs (and carbon taxes) are where the real big costs are.

The Korean nuclear numbers are unbelievable. They certainly didn't come anywhere close to that in UAE (more like $6/W, and possibly more if the concurrent military deal was being used as a slush fund to hide overruns.)

> the concurrent military deal was being used as a slush fund

Citation?

> The Korean nuclear numbers are unbelievable

See https://world-nuclear-news.org/Articles/Contract-awarded-for... and https://world-nuclear-news.org/Articles/Contract-for-Shin-Ha... for construction deals relating to their latest reactors in South Korea.

> Citation?

There's no hard evidence, but the presence of the side deal does raise suspicions. S. Korea's nuclear industry has been notably corrupt, with many years of prison sentences being handed out.

I believe there was a defense pact to sweeten the deal, but that seems like relationship building on top of a very competitive bid for nuclear power stations.

https://thediplomat.com/2018/03/risky-business-south-koreas-...

> Additionally the wind farm requires backup

So does a nuclear reactor.

Nuclear capacity should be around 90% when running baseload, with < 1 unplanned trip per year. The US have been very successful at getting the capacity factor higher; it is the cheapest way of producing more from the given assets. In addition the 10% unavailability is predictable and can be planned many months in advance (and can be delayed by a few weeks or so at lower power output). So multiple plants can plan their outages to not coincide ... one backup plant will serve as backup for more than one nuclear power stations.

However offshore wind has a capacity factor of 40% to 50%, and its lack of output is correlated with other wind farms. Therefore one backup power station can only really serve one wind farm.

Not only is it capacity factor but demand correlation that weights against the intermittents.

The UK’s average nuclear load factor for 1970 to 2017 was 67.4 per cent. Furthermore, France demonstrated the correlation of nuclear downtime in the summer of 2022 when it was too summery, compounded with overrunning maintenance. At its lowest point, France’s nuclear availability sat at around 40% of maximum capacity for about a month.

All single power generators need to be backed up for downtime.

> The UK’s average nuclear load factor for 1970 to 2017 was 67.4 per cent.

Finland's load factor over the last 3 years has been above 90%, as was the USA. Germany was up at 94% until ... stuff happened.

> At its lowest point, France’s nuclear availability sat at around 40% of maximum capacity for about a month.

Would they have shut down those reactors if alternatives were unavailable? I think they would have probably continued to run them, with permission of the regulators (at the behest of the government).

> All single power generators need to be backed up for downtime.

I'm not sure I follow what this statement means. What does "single" mean, as the preceeding comment talks about correlated downtime. Should we ensure that all methane plants are backed up too?

The French nuclear power stations are often built near rivers where the water was too hot for cooling or just not enough due to drought. The UK nuclear power stations are all on the coast drawing in much cooler sea water.

France has around 20GW on the coast (21,444MW including Flamanville 3).

Also France looks to have winter-dominated electrical demand (50TWh in January 2023, 40TWh in April, 37TWh in August, 46TWh in November). See https://www.iea.org/data-and-statistics/data-tools/monthly-e... and select France.

Indeed. Though, as I said there are others next to rivers.

The map here:

https://www.connexionfrance.com/news/map-where-are-frances-n...

shows 6 reactors on the coast, and 12 inland on rivers which are susceptible to summer river water temperatures and levels.

> susceptible to summer river water temperatures and levels

Yes, when the French grid needs them the least.

> Additionally the wind farm requires backup (CCGT or OCGT), which cost money to sit there.

Or batteries, but yes.

Battery manufacturing capacity is growing fast, but still — figures from recent years as reported by the IAE are [2022: 1720 GWh, 2023: 2500 GWh], but that's global and I'm not sure if demand is currently driven (no pun intended) by vehicles or by grids.

Still, if those batteries need replacement every decade, that global manufacturing supply can backup 150 GW of generation if they each need an independent 1 week backup and can't e.g. just use a broader grid to get e.g. Spanish sun for any of that.

>Or batteries, but yes.

Grid batteries make sense for those 30 second to 30 minute outages but it seems unlikely that they would be viable for those 2 week cloudy windless spells we get.

My understanding is the UK plans to fill that gap with peaker gas plants with CCS. They are being contracted for a maximum of two weeks in the year.

Currently sure, hence the numbers I gave. Theoretically possible already, but the UK isn't buying 30% of global supply for the next decade.

Forecasts suggest supply will rise to the level where this is doable… but over the next decade or so of increasing global manufacturing capacity.

Considering what happened to global and local trade in the last decade, I'm not confident beyond 5 years, and even then only somewhat confident.

There's a weakness in the way the National Grid works which means we don't fully utilise the battery capacity that's available

But in California batteries supply 20% of the power in the evening https://archive.is/IruIQ

I'm not saying we shouldn't use batteries, rather that they don't solve every problem. In particular they probably won't ever be suitable for week long outages.

> installed capacity of 3.6GW

I'm not aware what the capacity factor for wind is in the north sea, I think it's about 25% globally.

> how many homes in the UK in total?

About 25 million.

(But also: ongoing political issue is that this isn't enough and housing is too exaggerated vs. vested interests in housing being expensive).

25% would be for onshore wind. I've seen claims as high as 69% for very large offshore wind turbines.

I recall a previous exchange on hacker news https://news.ycombinator.com/item?id=40336108

  Offshore wind is at the extreme of renewables. It may or may not make sense, but it's not needed to kill nuclear.

Hornsea One (1.2GW from 174 × 7 MW turbines) generated 4,862,438 MWh in 2023 and 4,982,624 in 2022. This represents and 45.6% and 46.7% capacity respectively.

Triton Knoll (855MW from 90 × 9.5 MW turbines) generated 1,687,138 MWh in 2024H1, giving a capacity of 45.1% for this larger turbine.

Source :- https://dp.lowcarboncontracts.uk/dataset/actual-cfd-generati... and doing pivots on the table.

Thats a very good comparison-- to be fair though, 3.2GW nuclear are comparable to pretty much the whole Dogger Bank (including Sofia/South) at ~8GW (nameplate power), because offshore capacity factor is about ~40% in the UK.

Total cost for the wind farm is probably around 20-ish billion $, but the lifetime is likely shorter than the reactor (possibly less than half!).

It's surprising to me that the reactor at $35billion is still competitive, cost-wise, with the offshore wind farm.

Does the $35billion also include decommissioning the nuclear reactor and storing the nuclear waste for the duration at which it is considered harmful? If not, then what amount are we talking approximately if that was accounted for as well?

Rough estimate for decomissioning the reactor could be something like £5Bn.

Decomissioning the offshore wind turbines could get into the same range though, assuming £400M per installed GW (nameplate), this would be £3.2Bn

Numbers need to be taken with a large grain of salt.

Pretty good for a rough estimate!

The best I could find was a document by the National Audit Office from 2017 (using 2016 estimates); page 18, figure 2 - The expected costs of Hinkley Point C.

https://www.nao.org.uk/wp-content/uploads/2017/06/Hinkley-Po...

Back then the construction was estimated at £18.2 billion. Ongoing operation costs (including fuel, staffing, and grid charges) were priced at £29.3 billion and decommissioning (without the uplift) at £7.3 billion. Total project cost was put at £54.8 billion.

Again, it was from 2017 so things have changed.

The report states plant decommissioning is 2.7B GBP and ILW disposal is 0.3B GBP. That sort of agrees with https://www.eia.gov/todayinenergy/detail.php?id=33792 which puts PWR at perhaps 1B USD per plant.

Handling and disposing on the used fuel is 4.4B GBP (or 73M GBP per year of running, or 2.6 GBP per MWh of electricity). Cheap!

Do you mean it's competitive due to the longer lifetime?

Is the operation cost of the reactor over such a long duration not significant?

Also why are you comparing 8gw to 3.2gw? Is it due to peek output? if so won't this be solved by grid level storage?

> Do you mean it's competitive due to the longer lifetime? > Is the operation cost of the reactor over such a long duration not significant?

Yes. Also because maintenance costs are probably somewhat similar-- offshore turbines are still mechanical systems in a very unpleasant environment, after all.

> Also why are you comparing 8gw to 3.2gw? Is it due to peek output? if so won't this be solved by grid level storage?

Yes-- the turbines give you 8GW, but only 40% of the time. Reactor gives you its ~3GW basically all the time. This difference would NOT be erased by storage-- you would need the storage on top of the windpower, but it is not trivial to tell how much storage you really need (electrical grid connectivity can also somewhat compensate, or large consumers that only turn on when energy is cheap...).

Interesting. I've been reading about the much lower costs of grid level storage and the fact that it's now getting into economies of scale. I also assume there's a lot of room for cost reduction in wind over the next decade, while I'm guessing Nuclear won't see the same level of changes. Won't this change the situation significantly over the next decade.

> I also assume there's a lot of room for cost reduction in wind over the next decade

https://www.theguardian.com/business/2023/jul/20/giant-windf...

What are the reasons behind believing prices will reduce further?

> I'm guessing Nuclear won't see the same level of changes

  Parsons says the second reactor is being built 20% more quickly than the first one.
  Parsons says he expects that Sizewell will be built between 20% and 30% faster than the second Hinkley reactor.

In addition I would think reduced financing costs (risk reduction and sharing) will help bring prices down.

> What are the reasons behind believing prices will reduce further?

Economies of scale and room for technological/process innovation. The fact that solar reduced so significantly and battery costs too. This article seems a bit cherry picked about a specific project and not a global trend in costs. Offshore will naturally take time to become competitive since it requires complex installation and processes, but those seem like solvable problems for an admittedly laymen as myself.

> Economies of scale and room for technological/process innovation.

This applies when going from 1 to 2 to 4 reactors.

See https://www.oecd-nea.org/mdep/events/conf-2023/presentations... page 27 for the Chinese approach to HPR1000 (Hualong One) innovation.

Plus the linked article mentioned improvements to manufacturing outside the constraints of the containment.

Nuclear needs for more specialist disciplines than turbine maintenance. The systems are more complex. And you have to deal with a wider range of hazards. All whilst having the worlds most onerous rules.

Dogger Bank will when fully built-out at a cost of some £11bn.

Lifetime cost will be much lower - near 0 staff - no anti-terrorism policing operation - no fuel. If it were onshore, cost would be significantly cheaper and would be able to scale onto the grid and generate revenue (nearly) one turbine at a time as they are installed which makes a massive difference for financing and servicing a debt.

Cost wise certainly not orders of magnitude difference, and it is competitive. The sensible non partisan thinkers out there know it isn't a choice of building one or the other, you build both.

£11Bn is just Dogger Bank A, B, C (which is only 3.6GW total nameplate power, not 8GW)-- but this is in fact where my "~£20Bn total cost" estimate came from :P

Maintenance cost for offshore wind is NOT negligible. This is possibly as high as the reactor maintenance cost already by itself. Then, for a fair comparison, you would also have to add costs for improving grid connectivity and/or local battery storage...

I'm not arguing for more nuclear power btw, just honestly suprised that the expensive reactor is somewhat competitive, still...

Ahh fair comparison!

GE's 170 full-time turbine servicing jobs for Dogger Bank phases A, B and C will be based out of the Port of Tyne. That is more staff than I would have guessed, one worker for every 1.6 turbines!

Definitely not negligible.

https://www.irishtimes.com/ireland/2022/10/19/very-intense-r...

Yeah, that's one point that gets lost in the brouhaha over nuclear cost overruns:

Even the most catastrophic nuclear projects are competitive if not better than the best renewable projects.

Yeah, I know this sounds like frothing-at-the-mouth nuclear fanboiism, but run the numbers!

I did, and once I did I started sounding like a frothing-at-the-mouth nuclear fanboi.

¯\_(ツ)_/¯

Looking at the numbers for this, I don't see how it is better or even competitive at today's cost. Not taking into account the operating costs and decommissioning costs.

Off-shore wind is still scaling to become cheaper and cheaper. In 30 years it might cost much less. How can Nuclear compete?

> How can Nuclear compete?

A factory built plant or any economies of scale.

Start a company with a mobile workforce and go country to country building 5 or more large scale reactors in parallel in each country.

Small scale reactors (again factory built) are the other alternative. They would have to be installed in a 'farm' together for security reasons.

Even 4 stick-built reactors in a row is pretty good; look at Barakah in UAE, South Korea, China and Canada (Darlington, Pickering, Bruce).

> and decommissioning costs

https://www.eia.gov/todayinenergy/detail.php?id=33792

So perhaps 1B-2B USD a plant. 2,000,000,000/(60 * 1,600 * 365 * 24 * .9) = 2.6 USD/MWh

> Off-shore wind is still scaling to become cheaper and cheaper.

Why did the UK have to raise the maximum bid price for the Contract For Difference AR6? Wind farm costs have exploded, and turbine manufacturers are having reliability issues.

https://www.reuters.com/business/energy/what-are-issues-with...

https://www.reuters.com/business/energy/vineyard-wind-incide...

> Why did the UK have to raise the maximum bid price for the Contract For Difference AR6?

Bidders have to sell the government the electricity at the strike price. They pushed up the strike price to £54.23/MWh from a previous of £37.35/MWh with a current market price of £70-£80/MWh.

Nobody was going to offer to sell electricity for £37.35 for the next 5 years in a market where you can currently sell it for £70 - £100 (and where you could sell it for £300-£500 when the crisis occurred in 2021 / 2022).

> Nobody was going to offer to sell electricity for £37.35 for the next 5 years in a market where you can currently sell it for £70 - £100

The beauty of the CfD scheme is that you can choose when to start using it. But once you're in you're in, you're in until the end of the 15 years.

https://www.current-news.co.uk/significant-commercial-incent....

In truth I think it will be a mug's game to build offshore wind without a CfD. See https://www.squeaky.energy/blog/the-unintended-consequences-... for a view of what the day-ahead auction will turn into in a few years.

Reply on news.ycombinator.com