SNS & Los Alamos: response to allegations
Frank Munger's Atomic City UndergroundA column I wrote earlier this week, "Was Los Alamos weak link on SNS?" attracted a bit of attention -- including a posting on a blog known as LANL: The Rest of the Story.
An anonymous commenter on the blog said the SNS wasn't the success story it's been built up to be and raised some technical issues about the neutron-producing facilities. I thought they were interesting enough to pursue, and so I asked Ian Anderson -- associate lab director at ORNL and the SNS chief -- to respond to them.
Anderson is currently in Europe and didn't have immediate internet access, but he provided responses via his Blackberry.
Here they are:
Allegation: "The linear accelerator has serious beam loss problems that lead to activation of the items in the accelerator tunnel."
Answer: Linac beam losses at the recent 340 kW are within a factor of 2 of design expectations, which is a roaring success by any measure.
Allegation: "The spallation neutron target can only operate at the specified power for a few weeks before it has to be replaced, a two-month effort."
Answer: Per the SNS design, the system has been designed to rapidly replace targets. The operation should take 1 week, which is built in to the operations schedule.
Allegation: "The superconducting cryomodules provided by the Jefferson Laboratory will ALL have to be totally reworked in order to meet their performance specifications."
Answer: Half the cryomodules exceed their performance specifications by approximately 30%. The other half are lower than their performance specifications by approximately 15%. The performance will be improved by in-situ methods. There is no plan to "totally rework" all cryomodules.
Allegation: "The entire system was designed and built with no consideration of the reliability..."
Truth: Our present availability of 86% is world-class. ISIS 10-year average availability is 88%, Lujan is 80%, PSI is 85-90%.
The SNS target uses liquid mercury. The beam pulses cause a shock wave which degrades (erodes) the target housing. The target was tested at LANL (LANSCE?) and problems found; this problem has been known but SNS continued to (proudly) plan to use such a device. Ask what they do with the highly activated mercury, every time they have to replace it? The solid metal LANL target at LANSCE lasts for years and has no mercury, although at lower beam power. SNS project had a "not invented here" mentality for this and other designs for their accelerator.
ReplyDeleteAlso ask about the sagging foundation for the beamline building, where one floor has drifted vertically inches since it was poured. All this, and the accelerator must be perfectly aligned.
Also, ask about cost-cutting design changes which forced LANL-designed (and others) systems into buildings with inadequate space to remove components easily for repairs. Or the late change to superconducting accelerator structures, after Los Alamos aleady began design of RF for normal temperature systems. Not a well-planned scheme for a big project. SNS project was a series of start/stop/change/halts, as most from the participating labs can attest to. In the end, they couldn't afford to have the labs do much of the installation of their systems.
The variability of the performance of the superconducting RF cavities from the other lab was unplanned for and has been a disaster. SNS management has been covering up this aspect of their startup. The have not reached design power for long periods.
It is a shame that they chose to blast LANL when talking to the GAO auditors, adding to the negative rhetoric that the report was designed to elict for political purposes. Since LANL is now their only competition (with LANSCE, with an already running SNS having a thousand users annually), it would be expected that they would make this move. Petty, but its dog eat dog when the dollars are slim.
Ask them about the power upgrade they have already asked DOE to fund, when they haven't met nameplate performance even.
There are many lessons to be learned from SNS, a collaboration of efforts from many laboratories. They (SNS) have found fault with systems from all the partners, yet, are the first to say that they have no problems and everything was peachy.
DOE Office of Science isn't about to say ANYTHING negative about them, as they bought in & sold the project to Congress, hook, line and sinker. Instead they chose to cut funding from other accelerator laboratories nationwide, including early shutdown of IPNS, one of the original neutron spallation sources in the country, that had a considerable backlog of experiements planned this year. Expect to see additional attacks on the remaining second Spallation Neutron Source, at LANSCE/LANL.
Regarding the statement about the cryomodules exceeding specifications:
ReplyDeletethat is normal design margins. You would not put an accelerator into service
that only "met specifications." In that case, the most minor abnormality would cause the system to not meet specifications.
The statement "Linac beam losses at the recent 340 kW are within a factor of 2 of design expectations, which is a roaring success by any measure." The specified beam power is 2 MWatts. SO, if we have 2X the design beam loss at 340 kWatts, then one would expect 12X the design beam loss at 2 Mwatts.
No comment on the cost overrun on the buildings.
Also, no mention that the floor of the linac tunnel has sunk about a foot.
I think that Anderson's comments are disingenuous at best.
ReplyDeleteA number of the cryomodules have een taken off-line due to failure of individual cavities. The radiation from them far exceeds the design levels and the insulation of internal electronic cables is failing. ORNL/SNS is in the process of building a complete cryomodule refurbishment facility.
And, he did not mention the sagging foundation in the tunnel.
The availablity of LAMPF exceeds 90% of the scheduled run time.
Yet another case of the usual irresponsible half-assed journlism. Why didn't Munger ask somebody at LANL for their side of the story before he went to press?
ReplyDeleteJournalists are scum. If we built nuclear weapons with the kind of care that journalists exercise there would have been a very large hole in the ground at Los Alamos a long time ago.
For those still awake their reaction to this thread is baaaaaaa....
ReplyDeleteFor the rest its zzzzzzzzz.....