Thursday, February 4, 2010

Spiraling Out Of Control

I have covered the troubles the ITER fusion project is having in ITER Gets Clipped which covered the American view of ITER's troubles. The The European Voice is taking a look at the problems from an European view.
ITER's projected costs have soared since the first estimates were made in 2001. Contributions will generally be made in kind (through provisions of construction materials, reactor components, labour and expertise). The EU's total in-kind contribution was estimated at €1.491 billion in 2001. By 2008, when the EU's Fusion for Energy agency, which was set up to manage the EU contribution to ITER, reviewed the costs, the estimate had risen to €3.5bn.
Rising costs

Concerns about the ballooning budget led the Commission last year to set up an expert group tasked with reviewing the construction costs. The group's report, released to member states last month and seen by European Voice, said that the construction costs alone could rise as high as €1.5bn (compared to a 2001 estimate of €598 million).

The report said that the increase was a result of “omissions or underestimates” in the original estimates, inflation in concrete and steel prices and “changes in specifications”.

The Commission has set up a task-force to identify sources of additional funding for ITER. One option being considered is a loan from the European Investment Bank.
The latest budget numbers I have seen have the project estimate at around $7 billion US (€5.1 billion).

Interesting that the budget was low balled to get things going and then things started spiraling out of control. Making up for missing resources in out years always costs a lot more than budgeting for them from the start. We see this in the space program all too often. The reason is that you have people you have to keep on board while changes are being made. What we in engineering like to refer to as "the burn rate" - the amount you have to spend to keep going while actual progress halts to make the changes. Every day's delay can cost millions of dollars. Then there is the problem of bringing new people up to speed. Adding people to a late project will often increase the delay over what making do with the people you have will cause. It is easy to get into a regenerative mode where you can never finish at an acceptable time with an acceptable budget. Another thing that happens when you add new people to a project is that the design suffers because the new people never know as much as the old hands.

Fredrick Brooks originally looked at this problem with respect to big software projects. He published his observations in a 1975 book called The Mythical Man-Month: Essays on Software Engineering.

It is probably the best book on big project management ever written so far. I have used his insights often in my engineering career. Management will hardly ever listen to these types of insights at the beginning. But occasionally you can get them to accept the insights provided once a project is in trouble.

Let me add that the much smaller Polywell Fusion project is not having any such difficulty. Physicist Rick Nebel said of his WB-7 experiment: it "runs like a top". Rick has been mum about WB-8 progress. Since he has the same team that did WB-7 working on WB-8, I expect he will deliver the knowledge required on time and within budget. Of course he has an advantage. It is easier to keep a small project ($ millions) on time than it is to do the same for a large project ($ billions). If the experiments look promising I expect that he will have a lot more trouble getting a real power plant operational. The logistics get harder.

You can look at recent list of the design problems ITER faces at Talk Polywell.

No comments: