Costs are a vital ingredient to evaluate fusion but are, unfortunately, very difficult to accurately evaluate at the present time. Fusion advocates, including the ITER website, suggest the electricity price from fusion will be competitive with fission, however given the complexity and the cost of special materials for fusion, this is difficult to accept. 

It is more likely that the actual cost of electricity from fusion would be several times that from fission.


Estimates based on ITER Construction Costs

The construction cost of ITER is not completely transparent because of the in-kind contributions which are costed in different ways by the different countries involved. Estimates exist in the range $22-65 billion. Michel Claessens estimates the total ITER cost based on the known EU contribution, which is 45.6% of the total, to be €41 billion (~$44 billion). The US Dept of Energy cost estimate is the high end, $65 billion. ITER is projected to generate 500 MW of thermal power ( zero electrical power) for short periods taking several hundred MW of electricity from the grid to operate.  

Notwithstanding the complexity of using these figures to predict the cost of commercial fusion power, a number of people have made such estimates. Reinders in his book, The Fairy Tale of Nuclear Fusion, suggests fusion would be $22000/KWh based on ITER costs, about 4 times more expensive than fission in the numbers he uses. Hirsch in an article,  Fusion: Ten Times More Expensive Than Nuclear Power , suggests fusion as 10 times more expensive than fission. The difference in these estimate is likely due to the different ITER estimates taken as the basis.


Published Fusion Reactor Cost Estimates

The paper by Entler et al. :Approximation of the economy of fusion energy makes an evaluation of the fusion costs, compared to other technologies, and these are illustrated in the following figure which shows investment cost.

This data gives the average investment cost for fusion $5999/ KWh about 1.4 times that of fission.

There seem to be no recent published cost estimates for current DEMO reactor designs.


Estimate  by comparing constituents

Crude estimates for the relative costs of fusion and fission can be obtained by comparing the scale of material and the different systems needed for the reactors. The figure below compares the scale of a fusion reactor and a fission reactor with similar, GW, electrical power output.

Fission reactors are very much smaller than Tokamaks. The former are of the order of 100 m2 , while a Tokamak, with similar power output, would be of the order of 10000 m2 . This leads to a crude estimate that material costs for a fusion reactor would be about a factor 100 more than fusion. Beyond the heat generation elements, the system to convert heat to electricity will be similar in scale and cost. For a fusion reactor, these costs are approximately 5% of the total and so will not affect relative fusion to fission costs much.

A slightly less crude estimate, can be based on the breakdown of projected costs for a fusion power plant from the PROCESS reactor design code, given in Nuclear fusion: What of the future? by Richard Kembleton . This breakdown is in the plot below.

Taking into account only component costs (i.e. ignoring construction, contingency and interest), an estimate of the relative cost, fusion to fission, can be made by dropping items which do not appear in fission. The biggest item here, is the "Magnets & power supplies" which represent almost half of the component cost of a fusion reactor. The "Replacement in-vessel costs" (the inner wall, diverter and breeder blanket) do not appear for fusion either, nor do "Heating and vacuum systems". The items "Heat extraction and power generation systems" as well as "Reactor systems and in-vessel components", are likely similar and so are included in the estimate unchanged. A guessed fraction of 50% for "Buildings and land" and for "Miscellaneous plant" is taken for fission. With this, fusion comes out 4.4 times more expensive, for component costs, than fission.