The Economics of Nuclear Energy

[Andrew Mason]

I am interested in comparing the cost of nuclear energy with the costs of fossil fuel and renewable fuels for electric power generation.

The figures I have for fuel consumption are:

1. a 1000 mW conventional LWR nuclear plant consumes 25 Tonnes of fuel (lightly enriched uranium) a year or about 1000 kg of ²³⁵U.

2. a 1000 mW coal plant consumes about 4,000,000 T of anthracite or bituminous coal in a year or about 10,000,000 T of Lignite coal

3. a 1000 mW biomass plant would consume about 8,000,000 T of biomass per year.

4. a 1000 mW natural gas plant would consume about 100,000,000 GigJoules of natural gas.

A tonne of bituminous coal is about $50 - $60 a tonne and about $30 for lignite coal (about $1.70/ gigajoule) so the yearly fuel cost for a coal plant is about $200 - $300 million.

A tonne of biomass is a little more variable. Wheat straw would be less than $30 a tonne or about $2 / gigajoule. Assuming you had a sufficient supply the annual fuel cost would be about $240 million.

Natural gas costs about $4-$5 a gigajoule so the annual cost would be $400 - $500 million.

I am having a little difficulty figuring out the total fuel cost for a nuclear reactor. The current spot market price for natural Uranium U₃O₈ is about $63 a pound or about $140 a kg, $140,000/T . But the conversion, enrichment and manufacturing costs add a fair amount to that (enrichment to 5% requires about 7 kg. of natural U for each kg of enriched U). The latest source that I can find is about 10 years old and put the price of enriched fuel at $1400 a kg, when uranium spot prices were much lower. So I will just assume it is about double that now: $3000 /kg or $75 million/year for 25 T.

Candu reactor fuel would be much cheaper because the Candu uses natural U, avoiding the enrichment costs. But it appears to use less of the U-235 so goes through more fuel.

So just looking at fuel costs: nuclear power is only 1/3 the cost of coal power and biomass power and 1/6 of the cost of natural gas. This seems high. I would have thought it would have been much less.

This ignores the much higher capital costs, the waste storage and disposal costs and the plant decommissioning costs, of course. So if you factor those in, the economic advantages of nuclear power are not very evident.

The potential economic advantage of nuclear could be there if the reactor technology could make better use of fuel to consume much less U and and to minimize waste.

Comments?

AM

 

[anorlunda]

ping. Can anyone answer this old question?

 

[BWV]

The price of natural gas in the US is about 1/3 the 2006 price and solar and wind are now competitive. Here is a current analysis (in $/MWH)

The cost of nuclear does not really matter, as it cannot be financed under the current political climate

https://www.lazard.com/media/450337/lazard-levelized-cost-of-energy-version-110.pdf

 

[DEvens]

@BWV What price factor does your data on wind costs include for backup? Considering that wind typically "enjoys" a capacity factor between 10% and 40%, depending on time of year and local expected weather conditions, and that it is usually not dispatchable, the cost of backup power should be included in the cost of the wind construction. And it should just about triple the cost, considering the low-end for gas is close to the high end for wind. And since gas is the typical backup for wind.

Also, what price factor does your data on wind include for the increased requirements for expanded and more complicated electrical grid? And what factor does it include for things such as environmental impact, neighbors complaining about noise, light flicker, and catapulted ice, cement production, production of the blades, dead birds, etc.?

Also, what does your at-the-wall-socket cost become if the conventional power sources are moved into backup-supply mode? If I have to operate a gas fired plant to supply the missing power for wind, I'm going to charge pretty close to the same amount per day as I would if I was operating at 90% capacity factor. I will still need to keep all my workers there, do all my maintenance, file all my paper work and do all my safety analysis. The only difference is my fuel costs are slightly lower because of the lower demand. So the per kWhr costs are going to be much higher.

In Ontario, wind is just about break-even at about 20 cents per kWhr, which is $200 Cnd per MWhr. That's something like $160US per MWhr. About three times the upper limit on your data for wind. If you make the land owners pay property taxes on the installation value then they need somewhere round about 25 cents to 30 cents per kWhr.

Meanwhile, Bruce nuclear and Darlington nuclear make substantial operating profits at 7 cents per kWhr. Or $70 Cdn per MWhr, or round about $55 US. That is, under the bottom end for your data on coal.

 

[gmax137]

UO2 is at $25 / pound. The $63 figure in the OP is way off today.

NEI has a page on fuel cost somewhere in their website; last time I looked the total cost for the fuel is around 0.75 cents ($0.0075) per KW-hr. That's the cost for fuel assemblies; not just the uranium cost.

The chart shown above (@BWV) shows higher costs than I would have expected -- my electric bill shows 9 cents/KW-hour (that's the per KW rate, not including the various taxes and $15/month service fee). So, $90 per MW-hour at my house (ie, includes both generation and transmission/distribution).

The cost of nuclear does not really matter, as it cannot be financed under the current political climate

There is some truth to this, certainly for building a new nuclear unit. On the other hand, the units can operate for a long time. Construction costs for the units built in the 1970s have been paid-off for decades; all of ones still operating have had their licenses extended to 60 years and many are now working to get the licenses extended again to 80 years. So, twelve billion dollars for a 1200 MWe plant seems outrageously expensive today (and it is) but it might look like a bargain from the perspective of the power company in the year 2199.

I know this doesn't hold up - the problem is, the construction cost is just too high relative to the entire net worth of the power company. And, failure to complete the project is a huge risk. The financiers have a long memory, and haven't forgotten the WUPPS default fiasco (not to mention the current mess in South Carolina).

 

[BWV]

The chart shown above (@BWV) shows higher costs than I would have expected -- my electric bill shows 9 cents/KW-hour (that's the per KW rate, not including the various taxes and $15/month service fee). So, $90 per MW-hour at my house (ie, includes both generation and transmission/distribution).

So $90/MW-hour vs some mix of combined cycle gas, wind and utility solar (which is the new electric generation triumvirate) works

Also $90/MW-hour vs marginal cost of coal and nuclear presumably works (as these plants are old and fully depreciated). These numbers include capital costs and do not reflect the marginal cost of continuing to operate an plant

 

[BWV]

@BWV What price factor does your data on wind costs include for backup? Considering that wind typically "enjoys" a capacity factor between 10% and 40%, depending on time of year and local expected weather conditions, and that it is usually not dispatchable, the cost of backup power should be included in the cost of the wind construction. And it should just about triple the cost, considering the low-end for gas is close to the high end for wind. And since gas is the typical backup for wind.

That is not relevant, the cost for wind is the cost for the electricity received. Its understood that wind and/or solar cannot be 100% of the generation capacity. Also, the cost for gas is combined cycle, which is in the same ballpark

In Ontario, wind is just about break-even at about 20 cents per kWhr, which is $200 Cnd per MWhr. That's something like $160US per MWhr. About three times the upper limit on your data for wind. If you make the land owners pay property taxes on the installation value then they need somewhere round about 25 cents to 30 cents per kWhr.

Not according to this:

Calgary, Alberta, December 13, 2017 – The Canadian Wind Energy Association (CanWEA) commends the Alberta Electric System Operator (AESO) and the Government of Alberta for the successful implementation of the first phase of its Renewable Electricity Program (REP), which has resulted in the lowest-ever price for wind energy generation in Canada.

As announced today, the AESO will secure power from four new wind energy projects representing approximately 600 megawatts (MW) of capacity at an average price of $37 per megawatt-hour. The three companies behind these four projects have each signed a 20-year Indexed Renewable Energy Credit (IREC) agreement with the AESO, providing predictable revenues while protecting Albertans against increases in the price of power. Under the IREC, when the market price is lower than the contracted price, the generator will be paid the difference and when the market price is higher, generators will be required to pay back the difference to the government.

https://canwea.ca/wind-energy/ontario-market-profile/

Meanwhile, Bruce nuclear and Darlington nuclear make substantial operating profits at 7 cents per kWhr. Or $70 Cdn per MWhr, or round about $55 US. That is, under the bottom end for your data on coal.

That is vs marginal operating costs, correct? Does not reflect the capital costs of creating a new nuclear plant

 

[anorlunda]

That is vs marginal operating costs, correct?

It is not a cost at all. It is a price determined by the energy auction. The text says that give a 7 cent profit, so I presume that includes interest and amortization of sunken costs.

 

[BWV]

It is not a cost at all. It is a price determined by the energy auction. The text says that give a 7 cent profit, so I presume that includes interest and amortization of sunken costs.

Dont see that, this is the breakdown in Darlingtons Annual Report, no depreciation is listed and the capital costs of the recent refurbishment are netted out of the $70 CAD number. If you look at the history, it cost $14B CAD to build and took a decade from breaking ground to beginning operations.

https://www.opg.com/document/2018-annual-report-2/

 

[BWV]

Here is an older EIA study (2014), the construction costs of nuclear still rendering it economic. However the out-of-control costs are largely a political and regulatory issue that could be solved

https://www.eia.gov/outlooks/aeo/assumptions/pdf/table_8.2.pdf

 

[BWV]

One last illustration on how Nuclear is dead in the US without a major regulatory overhaul:

https://www.vox.com/2016/2/29/11132930/nuclear-power-costs-us-france-korea

 

[DEvens]

That is not relevant, the cost for wind is the cost for the electricity received. Its understood that wind and/or solar cannot be 100% of the generation capacity. Also, the cost for gas is combined cycle, which is in the same ballpark

It's totally relevant. The average grid cost is what gets paid by the consumer. If wind messes up the grid then the grid passes that on to the consumer.

That is vs marginal operating costs, correct? Does not reflect the capital costs of creating a new nuclear plant

Considering wind turbines were originally costed out at a 20 year life span, but are typically not making that, then marginal cost is probably advantageous for wind.

Nukes in Canada have to pay a variety of charges for every kWhr generated. This includes, among other things, an amortized share of the debt the utility took on to build the station. And a fund for decommissioning and waste disposal.

 

[BWV]

The average grid cost is what gets paid by the consumer. If wind messes up the grid then the grid passes that on to the consumer.

No the average grid cost is the average of all the sources of power feeding the grid, no need to add additional costs to one source to reflect that it cannot generate 100% of the power needed

Considering wind turbines were originally costed out at a 20 year life span, but are typically not making that, then marginal cost is probably advantageous for wind.

Marginal cost for wind is essentially nothing, as it is all upfront capital cost. Agreed that if turbines degrade more quickly than assumption then the cost is higher. The Lazard piece uses 20 years, where first generation turbines are being refitted at 18 years

Nukes in Canada have to pay a variety of charges for every kWhr generated. This includes, among other things, an amortized share of the debt the utility took on to build the station. And a fund for decommissioning and waste disposal.

I am not against nukes, but look at the numbers I posted , the $72 CAD per Mwh from Darlington's annual report does not include any capital costs. The graph in #11 shows the last nuclear plants built in the US cost $8-9K per Kwh (in 2010$), so without regulatory reform, no nuclear plants are going to be built

 

[anorlunda]

Dont see that,

It is listed in the table as "Nuclear Generation Capital Expenditures". That is not the entire investment, it is interest plus this year's share of the investment which we call depreciation.

No the average grid cost is the average of all the sources of power feeding the grid, no need to add additional costs to one source to reflect that it cannot generate 100% of the power needed

Regardless of how you label them, the investor earns a profit (hopefully, maybe a loss). He has various fixed costs, variable costs. He gets various payments. Energy payments from the grid proportional to actual generation. Reserve service payments proportional to the capacity on standby. They can get payments for frequency control, and for voltage control. They get ICAP payments proportional to the capacity ready to bid to supply energy.

In the old days, generation was allocated so that ever source has equal incremental costs. In modern clearing price markets, that is no longer true. Some nuclear units have marginal costs far below the clearing price 24x7. Other times, the clearing price even goes negative although costs and marginal costs are never negative.

You can dig really deep and find dozens of costs, and dozens of revenues. The complexity is not very helpful for those trying to understand, or for those trying to compare one form of generation versus another. I do not recommend wallowing in that mud on public forums.

 

[BWV]

It is listed in the table as "Nuclear Generation Capital Expenditures". That is not the entire investment, it is interest plus this year's share of the investment which we call depreciation.

No "Nuclear Generation Capital Expenditures" is actual capex incurred during the year, not depreciation of past investment or the financing costs - just look at the income statement. Nothing in that number reflects the $14B CAD cost of building the plant.

 

[anorlunda]

No "Nuclear Generation Capital Expenditures" is actual capex incurred during the year, not depreciation of past investment or the financing costs - just look at the income statement. Nothing in that number reflects the $14B CAD cost of building the plant.

I'll take your word for that. Under US tax laws, we can use accelerated depreciation (such as ACRS) to deduct all the investment costs long before the end-of-life of the plant. So does that mean the value of the plant is zero after it is depreciated for tax purposes? Does the tax value relate to the real life value? Should the costs of new versus old plant of the same design be comparable?

Once again, those kinds of details are not helpful when comparing generation types, because each company's accountants use different definitions. One has to read the footnotes in the annual statements. The footnotes are where the real information is.

 

[BWV]

I'll take your word for that. Under US tax laws, we can use accelerated depreciation (such as ACRS) to deduct all the investment costs long before the end-of-life of the plant. So does that mean the value of the plant is zero after it is depreciated for tax purposes? Does the tax value relate to the real life value? Should the costs of new versus old plant of the same design be comparable?

Once again, those kinds of details are not helpful when comparing generation types, because each company's accountants use different definitions. One has to read the footnotes in the annual statements. The footnotes are where the real information is.

This is GAAP reporting, not income tax - income tax depreciation schedules are different from GAAP depreciation, which attempts to reflect economic reality. The value of the plant can be depreciated to zero on a GAAP or tax basis, but incremental capex (like the refurbishment listed in the schedule) adds to the capital value, so in practice this is unlikely. The depreciation schedule in the Annual report goes out as far as 74 years for nuclear stations (this is Ontario Power Generation's report, so it includes multiple stations)

 

[anorlunda]

I got lost about what we were debating about. I went back to #7

That is vs marginal operating costs, correct? Does not reflect the capital costs of creating a new nuclear plant

OK, I guess in the case of Bruce and Darlington I guess that's correct. But those numbers are prices, not costs. And you can't use that to presume that all nuclear power plants bury their capital costs in order to make a general statement about nuclear versus wind.

 

[BWV]

I got lost about what we were debating about. I went back to #7
OK, I guess in the case of Bruce and Darlington I guess that's correct. But those numbers are prices, not costs. And you can't use that to presume that all nuclear power plants bury their capital costs in order to make a general statement about nuclear versus wind.

Right, so the Darlington example does not work, then its back the levelized cost numbers in #3 which are an all-in number (see https://www.sciencedirect.com/topics/engineering/levelized-cost-of-electricity)

Can you find a source that has levelized cost for nuclear competitive with wind?

 

[anorlunda]

Can you find a source that has levelized cost for nuclear competitive with wind?

No. I don't have a source for that. It is a very elaborate equation. Even so, it's incomplete (No source of income other than selling energy.) It could be criticized for not including include the alternatives (including the alternative of no electricity; sitting in the dark.) It might be hard to find someone to put all that effort in for all the world's power plants unless there was a guarantee that nobody could claim, "The results are invalid because ... yada yada yada"