Nuclear vs. Solar - CAPEX & OPEX
A comparison of using nuclear vs. solar to deliver 1.4GW of baseload power
Fundamentally in the discussion of using nuclear vs. solar power we need to look at the costs of each. They’re both zero carbon. They both run fine when a storm or other event shuts down distribution. With our present technology stack, this is the choice for green energy.1
Providing power during multiple days of overcast skies, a blizzard, etc. is an issue where we need additional solar generation and storage, the below assumes that does not happen. How long we might have degraded solar generation is a complex question. And if we’re pure solar, we can have gas backup for that situation, which is additional CAPEX and OPEX.
This analysis compares the total costs of delivering 1.4GW of reliable power year-round in Colorado using either a nuclear plant (APR-1400) or solar farms paired with three energy storage models. We assume no federal subsidies and use 2024 technology costs.
Key Assumptions
I found numbers all over the place, from reputable sources such as NREL, Lazard, etc. I think the following are what is being paid now.
Solar Generation : Colorado’s shortest winter day provides 4.5 peak sun hours.2
Solar Panel Generation : 400W3
Solar Panel Cost : $0.80/W (installed)4
APR-1400 Cost : $4.6 billion - $9.5 billion5
Solar Farm Design
To generate 33.6GWh/day in winter, the solar farm must produce 7.47GW DC capacity (33.6GWh ÷ 4.5h).
Solar panels needed : 18.7 million (400W each)
Land area (panels only) : 37.3 km² (14.4 mi²)
Total land required : ~181 km² (70 mi²)6
Solar CAPEX : $5.98B ($0.80/W * 7.47GW)
Storage Model 1: Batteries for Duck Curve + Gas
This model, which has significant CO2 emissions, is composed of batteries for the duck curve and uses gas turbines for the rest of the day. For this case we can remove ⅓ of the solar CAPEX/OPEX as we don’t need additional generation for overnight, just for the duck curve charging.
Design :
Batteries : Cover 4-hour evening "duck curve" ramp (5.6GWh).
Gas Plant : Provides 1.4GW for remaining 15.5 hours.7
Costs :
Batteries
CAPEX : $840M ($150/kWh)
OPEX : $112M ($20/kWh/year)
Gas Plant
CAPEX : $1.4B ($1,000/kW)
OPEX : $42M ($30/kW/year)
Transmission
CAPEX: $100M
Total
CAPEX : $8.32B
OPEX : $303M/year
Storage Model 2: 24-Hour Batteries
This model uses sufficient batteries to provide a continuous 1.4GW outside of the times the solar can directly provide it. This is the all renewables approach. This model adds 10% CAPEX/OPEX to the solar because the batteries are only 90% efficient..
Design :
Batteries : Store 33.6GWh (accounting for 90% efficiency).
Costs :
Solar Farm
CAPEX : $6.64B (8.3GW DC)
OPEX : $166M ($20/kW/year)
Batteries
CAPEX : $5.6B ($150/kWh)
OPEX : $739M ($20/kWh/year)
Transmission
CAPEX : $100M
Total
CAPEX : $12.34B
OPEX : $905M/year
Storage Model 3: Batteries + Pumped Hydro
This model uses pumped hydro as the backup. So mid-day the solar is both providing power and pumping up the water from the lower lake to the upper lake. It then uses that hydro over the rest of the day to provide a continuous 1.4GW. This model requires an additional 20% solar CAPEX/OPEX because pumped hydro is only 80% efficient.
Design :
Batteries : 4-hour duck curve (5.6GWh).
Pumped Hydro : Stores 21.7GWh (80% efficiency).
Costs :
Solar Farm
CAPEX : $7.04B (8.8GW DC)
OPEX : $176M ($20/kW/year)
Batteries
CAPEX : $840M ($150/kWh)
OPEX : $112M ($20/kWh/year)
Pumped Hydro
CAPEX : $3B ($2,000/kW)
OPEX : $70M ($50/kW/year)
Transmission
CAPEX : $100M
Total
CAPEX : $10.98B
OPEX : $358M/year
Nuclear Option: APR-1400
We compare each of the above models to the nuclear model.
Nuclear Plant
CAPEX : $4.6B - $9.5B
OPEX (including refueling) : $287M
Cost Comparison
Conclusion
Nuclear takes longer to build but is otherwise cheaper.
Solar + Gas is competitive over 20 years but relies on fossil fuels.
Solar + Batteries is prohibitively expensive due to storage costs.
Solar + Pumped Hydro balances CAPEX and OPEX but requires suitable geography and the hydro takes longer to build.
The bottom line is nuclear, even without taking into account the additional batteries or gas needed to handle overcast days, blizzards, etc. when solar generation drops precipitously, is cheaper.
It is fair to say that solar panel and battery efficiency will keep rising and costs will keep falling. But by the same measure, if we build 100 APR-1400 nuclear plants, the cost of that 100th plant will be a lot lower than the present $6 billion because we’ll learn a lot with each build that can be applied to the next.
So why are we building more solar farms instead of nuclear?
Total land includes spacing, access roads, and maintenance.
A bit less than that time because the solar is generating over about 9 hours of the day in winter. It’s just at under 100% for a fair amount of that.
Thanks for the analysis. If you want to amuse yourself, compare nuclear with residential solar sometime. The latter obviously will never fly as baseload, but even without backup, it’s absurdly expensive, and it pushes ongoing costs onto the ratepayers who can’t afford solar.
I’m comparing these numbers to NYS’s—they’re quite similar, as are the options. You state that nuclear takes longer to build, but it’s not always as different as you might expect here. We have solar projects that started the permitting and interconnection processes in 2017 and still haven’t finished—or sometimes started—construction. Meanwhile, costs have gone up considerably.
You forget one added fact - the solar plant will last 20 years, at which time it must be replaced.
The nuclear plant will last 60 years.
Hence the solar costs are actually triple what you are showing over 60 years. To put it another way, nuclear gets cheaper the longer time frame you look at.