Positive Energy

Wednesday, August 25, 2004

What the Heck is a LUEC?

  • Levelized Unit Energy Cost (LUEC)
The LUEC is a means for calculating the cost of a unit of energy. It does this in a manner that allows one form of technology to be compared with another. It includes all the factors that affect the total cost, and views them over a long interval so that end game operations such as plant decommissioning can be factored in. Now there are problems with this approach - some industries do not associate any cost with their most harmful activities. Coal fired generators spray radioactive dust all over the countryside and do this for free. Water dams release methane, a very disruptive green house gas, but don't bother to measure and report this bad news. Other technologies are so different that cost comparisons are of little value. Wind mills that do not generate power when the air is calm cannot be compared with steadily operating CANDU reactors. Even if the power produced by wind mills is inexpensive, its unreliability is a drawback that is difficult to cost. So it is not easy to make a comprehensive comparison and numerically determine what is best. However, there are some good examples available where researchers have used this approach rigorously to aid actual decision processes. One such case is:

Brian Moore and Sylvana Guindon
Natural Resources Canada

So I have decided to give it a try, if only to learn through experience what the issues are. I wanted to prove to myself what the cost of nuclear power is, or at least what the range of this cost is.

What factors should be considered for a CANDU reactor LUEC? One factor I decided to ignore is inflation and related financing costs. These can be important since the plant operating periods extend over twenty, forty, or sixty year terms. However, for a first approximation I decided to concentrate on the direct technical costs. I used the following:
  • Startup - the design and construction of a new facility up to the point where it begins operating
  • Operation - the cost of staffing it during its operational period
  • Fuel - uranium oxide, how much is needed and at what price
  • Maintenance - paying for all the material and things needed for repairs
  • Administration - executives, lawyers, committees, etc.
  • Decommissioning - restoring the plant location to a state that allows future use
  • Disposal - dealing with the material produced by operating the plant
  • Term - How long does the plant operate? When do we have to replace it?
  • Output - How much energy is produced during the plant's operating term?

How much does it cost to build a new CANDU reactor? Apparently the Ontario Darlington station was constructed for a cost of $14.5 billion dollars. This amount has been reported as an alarming figure so I assume that we can use it as a high value, or a worst case cost.

Source [2] states that Darlington is a 3524 MWe(net) facility. Source [3] states that a reactor can be built for 1540 USD/kWe. Combining these figures, the cost of Darlington should have been 3524000 * 1540 * 1.33 = 7.22 billion CAD.

So I decided to use the 14.5 billion cost as a worst case amount.


Source [3] reports that reactor operation and maintenance costs are 0.009 USD/kWh. This can be split evenly between Operation and Maintenance for this calculation.

For one year the operation cost would be 365.25 * 24 * 3524000 * (0.009/2) * 1.33 = 0.185 billion CAD.

Looking at this another way, I would guess that 200 people would be working at Darlington at any given time. It has to run 24 hours per day, seven days per week, so the staff needed for three shifts including weekends would be 200 * 3 * 1.5 = 900. Assuming an average salary of $85000, the annual operation cost would be 900 * 85000 = 0.077 billion CAD.

So the 0.185 billion estimate is again a worst case value.


Various reports state that fuel costs for CANDU reactors are less than five percent of the LUEC.

Source [3] reports fuel costs as 0.004 USD/kWh. Therefore the annual fuel costs would be 365.25 * 24 * 3524000 * 0.004 * 1.33 = 0.164 billion CAD.

These two guidelines have been considered to calculate the fuel costs for this case.


Source [3] reports that a reactor operation and maintenance costs are 0.009 USD/kWh. This can be split evenly between Operation and Maintenance for this calculation.

For one year the maintenance cost would be 365.25 * 24 * 3524000 * (0.009/2) * 1.33 = 0.185 billion CAD.


I have assumed that this area involves executives, committees, some research activities, and various public relations initiatives. For a plant like Darlington I assume that 100 people are needed here, each costing $150000 per year. So the annual price tag would be 0.015 billion CAD.


I don't have any details about this cost. However, if we built the plant for 14.5 billion surely we can tear it down for 10.0.


CANDU plants are not disposing their used fuel at present, so this cost is zero. A reasonable way of looking at this would be to count the additional cost involved to reuse this fuel. That would make disposal costs about twice the original fuel costs, or 0.330 billion CAD annually. Note that this would eliminate the new fuel costs for that year, so an estimate for this cost should be the difference, namely 0.165 billion annually.

Alternatively, we could just bury this valuable "used-once" fuel for about 0.050 billion annually.

I decided to again use a worst case estimate of 0.165 billion CAD annually for this factor.


Once built a CANDU reactor is inexpensive to operate, and the operational costs are stable due to the plentiful fuel supply. So a longer term will show a lower LUEC value. CANDU reactors have established histories of twenty year operational cycles, with design objectives for forty year intervals. The newest designs are using sixty years as a design objective. Repeating the LUEC calculation for different operational terms is straightforward, so I decided to use twenty, forty, and sixty years.


Source [2] states that Darlington is a 3524 MWe(net) facility. So its annual output is 3524000 * 365.25 * 24 = 3.09E10 kWh.


Term Amounts in Billions
Fixed Cost
Annual Amounts
20 years
40 years
60 years
14.5 billion


0.185 billion

0.164 billion

0.185 billion

0.015 billion
10.0 billion


0.165 billion



3.09E10 kWh

0.063 $/kWh
0.043 $/kWh
0.036 $/kWh

Conclusions -

So if my assumptions are anywhere near the mark it appears that CANDU generated nuclear energy is no more expensive than other technologies, and probably less expensive since I have used several worst case values.

Source [5] below is a serious attempt to make this same calculation. My example above corresponds to their Candu 6 item. Their base case findings were:
  • Current technology CANDU (Candu 6) = 0.06344 $/kWh
  • New technology CANDU (ACR-700) = 0.05306 $/kWh
  • Gas = 0.07285 $/kWh
  • Coal = 0.04772 $/kWh
I therefore concluded that CANDU produced energy costs about the same as gas or coal produced energy, i.e. CANDU energy is not expensive.

And I concluded that attributing a cost of zero to carbon dioxide and soot release as is done for gas and coal makes these fuels appear inexpensive when they are not. If we do our cost accounting in a fair and accurate manner CANDU produced energy is the best deal available.


[1] C.1 How do the economic benefits of nuclear power compare to other sources in Canada?

[2] C.7 Why was the cost of Ontario's Darlington plant so high?

[3] Status Report on Nuclear Power Plant Life Cycle Management

[4] Reducing the Cost of the CANDU System

[5] Levelised Unit Electricity Cost Comparison of Alternate Technologies for Baseload Generation in OntarioLink