This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


Energy economics – why system costs matter

Milton Caplan's picture
President MZConsulting Inc.

Milt has more than 40years experience in the nuclear industry advising utilities, governments and companies on new build nuclear projects and investments in uranium.

  • Member since 2018
  • 97 items added with 129,296 views
  • Feb 2, 2022

In our last post, we quoted from recent reports that clearly lay out the environmental benefits of nuclear power.  This month we want to start off the year by launching a short series addressing some of the issues that impact energy economics.  Today we will talk about the importance of system costs in understanding the relative costs of different generation technologies. 

Last year at this time we wrote about the IEA/NEA report, Projected Cost of Electricity 2020, that shows nuclear is competitive with alternatives in most jurisdictions using the traditional Levelized Cost of Electricity (LCOE) approach.  LCOE is a great way to compare costs of electricity as it is generated from two or more different options to be implemented at a single spot on the grid with similar system characteristics.  With intermittent variable renewables on the system, LCOE alone no longer provides a sufficient basis for direct comparison.  By their very nature, deploying these renewables add costs to the system to be able to deliver reliable electricity in the same way as more traditional dispatchable resources like nuclear, hydro and fossil generation.   


What are system costs?  In a report issued by the OECD Nuclear Energy Agency (NEA), system costs (see the report for a full definition) are basically the additional costs to maintain a reliable system as a result of intermittent variable renewables only producing electricity for a limited number of hours when the resource is available (e.g. daytime for solar), their uncertainty due to the potential for days with little resource (e.g. rainy or cloudy days), and the costs to the grid to be able to access them given their more distributed nature (e.g. good source of wind but far from demand).

A 2018 study undertaken by MIT “The Future of Nuclear Energy in a Carbon Constrained World” considers the impact of nuclear power on the cost of electricity systems when deep decarbonization is desired.  It looks at various jurisdictions around the world and the conclusion is always the same; the cost of electricity is lower with a larger nuclear share than trying to decarbonize with intermittent variable renewables (and storage) alone. 

The reason for this impact is fundamentally due to the relatively little time these resources produce electricity.  Solar and wind only generate when the sun shines and the wind blows, meaning they produce only some of the time and not always when needed.  The average capacity factors of these technologies vary by location with world average capacity factor of just below 20% for solar and about 30 – 35% for wind (capacity factor is the amount of time a resource produces compared to if it would produce 100% of the time).  Contrast this with the 24/7 availability of nuclear power, which can operate at capacity factors of more than 90%.

The impact on electricity systems is clear.  Given the limited duration of operation of intermittent variable renewables, there is a need to dramatically overbuild to capture all the electricity needed when the resource is available to cover periods when the sun is not shining, and the wind is not blowing (all assuming there is reasonable efficient storage available which is not yet the case).  The result is a system with much larger capacity than a system that includes nuclear (or any other dispatchable resource).  In the MIT study for example, the system in Texas would be 148 GW including nuclear but would require 556 GW of capacity with renewables alone.  In New England a system with nuclear would have a capacity of 47 GW but would require a capacity of 286 GW with renewables alone.   In the UK this would mean 77 GW with nuclear compared to 478 without.  And so on.  The costs of adjusting the system to accommodate these much larger capacities is significant.

Since that time study after study finds the same result.  This includes a study in Sweden in which 20 different scenarios for full decarbonization always come out the same; in every scenario the most cost-effective system has continued long-term operation of existing nuclear.  And more recently a study in France has shown that decarbonizing without nuclear means a system more than twice as large as one with nuclear and the more nuclear in the system, the lower the overall average cost of production.

So, what does this mean for planning?  The approach to implementing a reliable economic low carbon electricity grid must start with looking at the entire system.  A study should assess the total costs of deploying the system under a range of scenarios using different shares of available resources.  Different forms of generation have different capabilities and these need to be modelled.  Once an efficient mix is determined, a plan should be put in place to implement it (i.e., X% nuclear, Y% solar, Z% wind, A% storage, etc.).  When looking to deploy each technology, LCOE can be used to compare various options.  For example, when comparing one solar project to another or one nuclear project to another.  And of course, should the costs of any given technology vary too significantly from the assumptions in the system study that determined the efficient mix, then the system study should be updated.

Today’s energy markets are most often based on the assumption that all electricity generated is the same (to be discussed in a future post).  This is true at the moment of generation when yes, an electron is an electron.   Unfortunately, the ability of any given technology to actually be there to produce at the moment it is needed varies substantially.  Therefore, a direct comparison of the LCOE of one option vs another is only part of the story.

To fully understand the costs of electricity generated, the costs of integrating any given technology into a reliable system must also be considered.  After all, what really matters is how much we pay as customers for our electricity and the studies are clear, nuclear as part of a fully decarbonized system is always lower cost than a system based on renewables alone.

Milton Caplan's picture
Thank Milton for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member
Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.
Peter Farley's picture
Peter Farley on Feb 2, 2022

This study is full of unsubstantiated assumptions. Here is a real-world example.

In the 8 weeks ending 30th December France' nuclear output varied between 34.8 GW and 49.5 GW but the share of supply from nuclear varied from 58~77%. France has 63 GW of capacity and yet peak output was only 78% of nominal capacity. Peak demand over the period was 85.7 GW. Minimum demand was 44.2 GW and maximum exports were 11.3 GW and peak pumped storage demand was 4 GW. In other words, France could have easily maintained a minimum load on its nuclear of 44.2+4+11 = 48-59 GW even if exports fell to zero. So why did the nuclear capacity vary between 55% and 79% of nameplate if nuclear is so cheap and reliable.

In fact, given that French nuclear output can fall as low as 55% of nameplate, to be certain that a nuclear dominant system could meet peak demand over those 8 weeks, it would have needed a nominal 85.7/0.55 = 155 GW of capacity when average demand through the year is 51 GW. As France is well blessed with hydro and peak output is 15.7 GW then the peak nuclear demand would have been 70 GW so to meet that capacity means that France would have only needed a mere 127 GW of nuclear capacity to guarantee zero carbon supply. As hydro does supply about 70,000 GWh, nuclear would be supplying 380,000 GWh assuming imports and exports were balanced. So, the nuclear fleet would actually average 43.3 GW or 34% capacity. Of course, France could increase its pumped hydro capacity from 5 GW to 25 GW and then assuming 85% peak output from that and 55% from nuclear and 15 GW from other hydro, it would only need 49/.55 = 89 GW of nuclear to supply an average of 43.3 GW still less than 50% utilisation.

Now let's assume that we neglect construction finance costs and build a new nuclear and pumped hydro fleet at 70% of the cost per MW of Hinckley Point and new pumped hydro at $2m/MW, which is an investment of $625 bn. On the other hand, if France builds 100 GW of wind and 70 GW of tracking solar and 150 GW of rooftop solar and 50 GW 6 hours of batteries as well as the 20 GW of new pumped hydro, it will have capacity to generate 650,000 GWh from wind and solar and could meet peak demand with hydro, batteries and only 5% of wind.  That is an investment of $480m at current costs. Now if we allow that nuclear costs can be reduced by 30% it is only reasonable that we allow wind, solar to follow current trends and also fall by 30% so the real investment would be less than $360bn, but there would be enough spare electrical energy that combined with smart charging it could supply the entire 140 TWh/y needed to electrify France's entire ground transport system.

Now according to the US EIA, nuclear plants cost an average of US$40/MWh in operating costs, maintenance, insurance, and staff costs. Wind and solar vary between $10 and $20 with storage adding about $5/MWh because most of the wind and solar is supplied direct to the load with only 5~15% going via storage. So, to supply 450 million MWh per year the operating costs of the nuclear system are $18 bn/y. To operate the wind/solar/storage system $9~10bn.

On top of the nominal costs you must add construction finance, typically 10~15% of a renewable system but up to 40% of a nuclear system because of the long build time. You also need to account for the spinning reserves needed to be kept online in case of a plant trip or transmission failure. In Texas Ercot has demonstrated that due to the diversified nature of wind and solar and small generator sizes you need less spinning reserves than a thermal system. Nuclear systems in particular need large spinning reserves because of the large generator sizes and slow ramp rates of backup capacity. Then there is the question of water. A one GW nuclear plant evaporates enough water every year through its cooling towers to supply a town of 150,000 people. In the inland areas of the country where is that water going to come from in a drying world  

In conclusion a renewable system for France will be 20~40% cheaper to build and operate than a nuclear system. In the US due to better wind and solar resources and greater geographic diversity, the advantages of the renewable/storage system would be even more pronounced.




Milton Caplan's picture
Milton Caplan on Feb 2, 2022

You are suggesting that studies by reputable organizations such as MIT, the OECD NEA and RTE – France’s transmission system operator use bad inputs. This is not the case. We are showing that the studies we have seen are all consistent in their conclusions. This may or may not be the case for other jurisdictions. What we are promoting is that system costs matter and that in any jurisdiction a good study be done to optimize the system and then that be the technology mix that is implemented. If it includes nuclear as many studies are showing, great. If not, also OK. Each jurisdiction should do what is in their best interests.

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »