The details of arrangements for implementing ITER are in the process of being finalised, but the main features are illustrated below.

The items to be purchased can be divided into three categories.

1. Those that can only sensibly be purchased by the host country.
2. Those which are of minor technical interest or size, and whose cost burden must therefore be shared by all Parties. For these items a centrally administered fund can be established.
3. Items of interest to all the Parties due primarily to their high technology content. To ensure each Party obtains its fair share of these items, the Parties must agree beforehand which ones each will contribute. To do this, all Parties must agree on their value to the project. This requires an agreed valuation of items, as described below. Each Party will contribute its agreed items "in kind", using the purchasing procedures and funding arrangements it prefers. Thus the actual costs to each Party may not correspond to the project valuation - it may differ due to competitive tendering as well as different unit costs.

To arrive at the "evaluated cost" or valuation of ITER, construction has been broken down into about 85 "procurement packages" each defined at the level of detail appropriate for an actual procurement contract. Half of these are for the tokamak hardware in the "pit", the remainder for peripheral equipment. Industrial companies or large laboratories with relevant experience were invited by the Home Teams to analyse the manufacturing processes and to generate from these procurement packages their estimates of the physical quantities to be supplied, including manpower, materials, and tooling, and other costs pertinent to their normal industrial practices, to fit a given delivery schedule.

These physical quantities (e.g. man-hours, material quantities) were then assessed starting from the industrial estimates, and multiplied by a single set of labour rates (depending on the specialty and type of labour) and material unit costs across all packages, independent of the Party. Such an approach gives the Parties, jointly, an appreciation of the relative contributions each might make to building ITER and, individually, allows them to estimate the absolute costs in their own currency each might actually incur in providing a given overall percentage contribution.

To eliminate variations arising from fluctuating exchange rates or the various rates of domestic inflation, actual and assumed, cost estimates for ITER and its operation and decommissioning are all expressed in relation to a specific reference date and currency (US$ at January 1989 values, appropriate for the CDA of ITER).

The percentage breakdown of direct construction costs for ITER, in total estimated at $2,755 M, which is 49% of the estimated direct capital cost of the 1998 ITER design, is then:

As part of ITER negotiations, these procurements have now been shared out among the future ITER Parties.

The "ITER Project Team" referred to in the above diagram will incorporate both a Central Team at the ITER site, which will have overall responsibility to meet the project objectives and to ensure design continuity and coherence of technical specifications, and Field Teams located in the Party's territory, which will technically monitor the procurements contributed by that Party, and liaise with the domestic procurement agency of each Party. In the above model, the professional effort in the ITER ProjectTeam will total about 1800 ppy, with about 2800 ppy in support staff. Taking professional and support staff costs at $150k and $75k per annum, leads to staff costs for construction of $477 M. In addition, some manufacturing R&D will be needed. The present estimate for this R&D is ~$70 M. As is the case for the procurement of "contribution-in-kind" hardware, the actual costs to each Party of contributing these staff to the project may differ from the Party's nominal share of the above figure.

Totalling the above costs, and averaging over a nominal 9 year phase of commitments for construction, leads to an average annual expenditure ~$360 M. Notwithstanding the distinction between "value" and "cost" drawn attention to in the above, this, for instance, is less than the current annual expenditure on fusion in just one of the potential ITER Parties to construction (Europe).

The annual costs of operating ITER are estimated to be, on average, on the order of $188 M (54% of the 1998 ITER design operating costs), to be shared among the participating Parties, and totalling $3760 M over 20 years. These estimated costs include personnel costs (~32%), energy and tritium fuel costs (~20%), and capital improvements, spare parts and materials, and waste management operations (48%).

Using the decommissioning strategy, excluding costs of radwaste disposal, salvage credit for reusable materials, site restoration, and financing costs/benefits for deferred expenditure, the cost of decommissioning ITER would be $335 M.