4 ms·

Turkey is set to get 10% of its grid electricity from the Akkuyu Nuclear Power Plant [1] The build cost has been 22B (excluding loans) and it will have a capac

Turkey is set to get 10% of its grid electricity from the Akkuyu Nuclear Power Plant [1] The build cost has been 22B (excluding loans) and it will have a capacity of 4456MW meaning it's $4.9/W. On average the cost per Watt for solar is 27 cents (excluding loans and costs for storage) [2] [1] https://en.wikipedia.org/wiki/Akkuyu_Nuclear_Power_Plant https://en.wikipedia.org/wiki/Akkuyu_Nuclear_Power_Plant [2] https://ourworldindata.org/grapher/solar-pv-prices https://ourworldindata.org/grapher/solar-pv-prices EDIT: incorrect calculation

Retric 3y ago
That project isn’t finished they are targeting 2026 and hope it’s only ~10% over budget. Your solar numbers are wildly inflated because that’s per watt not per watt hour. Over 2/3 of nuclear power plants costs occur after you build the things. It would be comparable if 22B was the total cost and if nuclear didn’t need ~1,000 full time employees, fuel, decommissioning etc. Those costs delay how quickly you can pay back loans which then drives up how much interest nuclear pays over the plants lifetime. Companies have signed solar power purchase contracts at 2c/kWh. Adding solar redundancy and batteries for 24/7 power is actually more flexible and cheaper than nuclear because you can cheaply follow the demand curve. Meanwhile nuclear power plants need to go offline for multiple weeks every ~2 years and really want to sit at 100% the rest of the time. PS: Even just looking at construction costs is misleading because solar starts producing power so much sooner. You’re paying interest from day 1 of construction not day 1 of operations. So if you take 6 years to build that’s 5 years of interest on the first years construction costs without any revenue. Meanwhile solar’s been producing power for ~5 years and paying down debt.
- kragen 3y ago
  1.5 cents per kilowatt hour even https://pv-magazine-usa.com/2020/05/28/record-low-solar-ppas-in-the-southwest-means-carbon-capture-is-not-going-to-save-coal-plants/ https://pv-magazine-usa.com/2020/05/28/record-low-solar-ppas... note that the coal plant mentioned in that article has since been decommissioned https://en.wikipedia.org/wiki/San_Juan_Generating_Station https://en.wikipedia.org/wiki/San_Juan_Generating_Station
kragen 3y ago
these are great numbers, thanks keep in mind that almost all the cost of pv plants is construction (capex), while most of the cost of nuclear is during operation and decommissioning (opex) so 20 cents per watt capex is close to 60 cents per watt total, twice the price of pv as emmaengineer pointed out, though, you did the division backwards: it's 0.20 watts per dollar, not 0.20 dollars per watt. the correct quotient is 490¢ per watt, or maybe closer to 15 dollars per watt including opex
EmmaEngineer 3y ago
$22B / 4.456B = $4.9/W, about 20x more expensive than your number for PV.
- kragen 3y ago
  thank you for posting this correction; i had missed that obvious error
- apatheticonion 3y ago
  Oh you're right! Did it backwards. Correcting my post
kragen 3y ago
i just noticed that you're linking a page about panel prices two problems one, pv modules are typically about a third of the cost of a utility-scale solar farm; things like installation labor and power electronics are also part of the cost. so roughly we should expect solar projects to cost about 80 cents per peak watt based on that number two, that page ends in 02021, but pv module prices have fallen by about a third to half since the beginning of this year https://www.solarserver.de/photovoltaik-preis-pv-modul-preisindex/ https://www.solarserver.de/photovoltaik-preis-pv-modul-preis... in particular solarserver's 'mainstream' category was about 25 euro-cents throughout 02021 and is now at 19, 24% lower, and the lowest it's ever been it seems like the price-fixing cartel announced at the beginning of 02019 https://www.reuters.com/article/us-davos-meeting-solar-gcl-idUSKCN1PI2OQ https://www.reuters.com/article/us-davos-meeting-solar-gcl-i... has finally collapsed solarserver's 'low-cost' category is at the staggeringly low price of 11 euro-cents per peak watt; if the other costs didn't change, that would lower the solar plant cost from 80 cents to 65 cents or so per peak watt, though in fact they tend to increase somewhat because of the larger area of solar cells required to reach the same wattage (that's why the mainstream pv modules are more expensive) a third problem is the capacity factor, tho; nuke plants typically run at about 90% capacity, while solar farms run at closer to 20%, because of problems such as night. in very polar countries like germany it's 10%. so 70 cents per nameplate watt works out to more like 3.5 dollars per delivered watt which is still cheaper than 4.9 (or 5.4 dollars per delivered watt assuming that 90% capacity factor) but not implausibly cheaper anymore if solar farm builders can find ways to reduce the costs of solar farms proportionately to the vertiginous drop in module prices, which sounds implausible but has always been done successfully before, we should expect the price to drop to half that
- Retric 3y ago
  German solar policies are insane. Nation wide capacity factor was 11.6% in 2018. The numbers are so low because they keep building solar in the northern parts of the country which are absolutely terrible for solar. Brandenburg has a stupidly large amount of solar compared to a terrible amount of sunlight. https://en.wikipedia.org/wiki/Solar_power_in_Germany#/media/File:Germany_GHI_Solar-resource-map_GlobalSolarAtlas_World-Bank-Esmap-Solargis.png https://en.wikipedia.org/wiki/Solar_power_in_Germany#/media/... Meanwhile the best location in Germany the south southern tip of Baden-Württemberg only has a relatively small fraction of their total installed solar. I haven’t worked out the exact numbers but they are something like 20+% less efficient than they could be. Granted they would need to build more power transmission lines but it’s a small county so such projects would be fairly cheap compared to the efficiency gain. PS: May solar power plants on the US are over 30% capacity factor and the average is ~25%. We’re a long way from the equator. In a perfect location solar tracking PV should get around 40% capacity factor. You beat 1/pi because over the day an angled panel casts a wider shadow than the width of the panel.

Retric 3y ago

That project isn’t finished they are targeting 2026 and hope it’s only ~10% over budget. Your solar numbers are wildly inflated because that’s per watt not per watt hour. Over 2/3 of nuclear power plants costs occur after you build the things. It would be comparable if 22B was the total cost and if nuclear didn’t need ~1,000 full time employees, fuel, decommissioning etc. Those costs delay how quickly you can pay back loans which then drives up how much interest nuclear pays over the plants lifetime. Companies have signed solar power purchase contracts at 2c/kWh. Adding solar redundancy and batteries for 24/7 power is actually more flexible and cheaper than nuclear because you can cheaply follow the demand curve. Meanwhile nuclear power plants need to go offline for multiple weeks every ~2 years and really want to sit at 100% the rest of the time. PS: Even just looking at construction costs is misleading because solar starts producing power so much sooner. You’re paying interest from day 1 of construction not day 1 of operations. So if you take 6 years to build that’s 5 years of interest on the first years construction costs without any revenue. Meanwhile solar’s been producing power for ~5 years and paying down debt.

kragen 3y ago
1.5 cents per kilowatt hour even https://pv-magazine-usa.com/2020/05/28/record-low-solar-ppas-in-the-southwest-means-carbon-capture-is-not-going-to-save-coal-plants/ https://pv-magazine-usa.com/2020/05/28/record-low-solar-ppas... note that the coal plant mentioned in that article has since been decommissioned https://en.wikipedia.org/wiki/San_Juan_Generating_Station https://en.wikipedia.org/wiki/San_Juan_Generating_Station

kragen 3y ago

these are great numbers, thanks keep in mind that almost all the cost of pv plants is construction (capex), while most of the cost of nuclear is during operation and decommissioning (opex) so 20 cents per watt capex is close to 60 cents per watt total, twice the price of pv as emmaengineer pointed out, though, you did the division backwards: it's 0.20 watts per dollar, not 0.20 dollars per watt. the correct quotient is 490¢ per watt, or maybe closer to 15 dollars per watt including opex

EmmaEngineer 3y ago

$22B / 4.456B = $4.9/W, about 20x more expensive than your number for PV.

kragen 3y ago
thank you for posting this correction; i had missed that obvious error
apatheticonion 3y ago
Oh you're right! Did it backwards. Correcting my post

kragen 3y ago

i just noticed that you're linking a page about panel prices two problems one, pv modules are typically about a third of the cost of a utility-scale solar farm; things like installation labor and power electronics are also part of the cost. so roughly we should expect solar projects to cost about 80 cents per peak watt based on that number two, that page ends in 02021, but pv module prices have fallen by about a third to half since the beginning of this year https://www.solarserver.de/photovoltaik-preis-pv-modul-preisindex/ https://www.solarserver.de/photovoltaik-preis-pv-modul-preis... in particular solarserver's 'mainstream' category was about 25 euro-cents throughout 02021 and is now at 19, 24% lower, and the lowest it's ever been it seems like the price-fixing cartel announced at the beginning of 02019 https://www.reuters.com/article/us-davos-meeting-solar-gcl-idUSKCN1PI2OQ https://www.reuters.com/article/us-davos-meeting-solar-gcl-i... has finally collapsed solarserver's 'low-cost' category is at the staggeringly low price of 11 euro-cents per peak watt; if the other costs didn't change, that would lower the solar plant cost from 80 cents to 65 cents or so per peak watt, though in fact they tend to increase somewhat because of the larger area of solar cells required to reach the same wattage (that's why the mainstream pv modules are more expensive) a third problem is the capacity factor, tho; nuke plants typically run at about 90% capacity, while solar farms run at closer to 20%, because of problems such as night. in very polar countries like germany it's 10%. so 70 cents per nameplate watt works out to more like 3.5 dollars per delivered watt which is still cheaper than 4.9 (or 5.4 dollars per delivered watt assuming that 90% capacity factor) but not implausibly cheaper anymore if solar farm builders can find ways to reduce the costs of solar farms proportionately to the vertiginous drop in module prices, which sounds implausible but has always been done successfully before, we should expect the price to drop to half that

Retric 3y ago
German solar policies are insane. Nation wide capacity factor was 11.6% in 2018. The numbers are so low because they keep building solar in the northern parts of the country which are absolutely terrible for solar. Brandenburg has a stupidly large amount of solar compared to a terrible amount of sunlight. https://en.wikipedia.org/wiki/Solar_power_in_Germany#/media/File:Germany_GHI_Solar-resource-map_GlobalSolarAtlas_World-Bank-Esmap-Solargis.png https://en.wikipedia.org/wiki/Solar_power_in_Germany#/media/... Meanwhile the best location in Germany the south southern tip of Baden-Württemberg only has a relatively small fraction of their total installed solar. I haven’t worked out the exact numbers but they are something like 20+% less efficient than they could be. Granted they would need to build more power transmission lines but it’s a small county so such projects would be fairly cheap compared to the efficiency gain. PS: May solar power plants on the US are over 30% capacity factor and the average is ~25%. We’re a long way from the equator. In a perfect location solar tracking PV should get around 40% capacity factor. You beat 1/pi because over the day an angled panel casts a wider shadow than the width of the panel.