Wednesday, July 21, 2010

Show Your Work!

If I wanted to check the numbers on how much plutonium might still be in the environment, I’d go to the records on production, check to see if there were any updated reports, and I’d check the records of what has happened to the plutonium since. The biggest changes will be in the plutonium that has been recovered from decommissioned plants and from the environment in contrast to earlier estimates, along with reclassifications resulting from plutonium removed from weapons as a result of treaties.

I would set up a table with a column showing the earlier numbers and another showing the changed numbers, with references for all the numbers. Or maybe two tables: one with categories of how the plutonium was used (Nagasaki, weapons tests, research, lost to environment, etc.) and one with where the plutonium is. This sort of accounting for how much plutonium was produced against how much we know we have is called a mass balance.

Unfortunately, that is not what Robert Alvarez chose to do in his report claiming that the US plutonium inventory has gone up. I can’t make the numbers from Alvarez’s report add up to a mass balance, because he doesn’t supply all of them, and the ones he does supply are not all comparable. It’s possible he’s got a mass balance somewhere, but, if he does, he’s reporting only selected numbers.

Page van der Linden has compared Alvarez’s report and Matthew Wald’s New York Times article that seems to be based on it. I say “seems to” because Wald, in the MSM tradition, doesn’t make his sources available. It is possible that Wald interviewed Alvarez and based his article on that interview. I’ll concentrate in this post on problems I’ve found in the report.

Alvarez comes close to part of a mass balance in his Table 2, which gives plutonium in the waste inventories of the various DOE sites, comparing earlier and more recent numbers. The more recent numbers add up to his 12.7 metric tons, the basis for his claims and Wald’s article. What he says about how he arrived at his numbers is
Based on more recent waste characterization data (see bibliography), approximately 12.7 tons, more than 11 percent of the total amount of Pu-239 produced and acquired has gone into waste streams (Table 2). [p. 4 of report]
“[S]ee bibliography,” however, is not an explanation of method. Which numbers did he take from the sources in the bibliography? How did he treat uncertainties? Why did he choose these sources? Further, two of the references in Table 2 are not in the bibliography, and there are eleven references in the bibliography that are not referred to in the table. The report also has footnotes that are not in the bibliography. Very confusing.

Alvarez throws other numbers into the report, and it’s not clear which are included in his 12.7 tons.
This paper does not address about 7.6 tons of plutonium contained in DOE spent reactor fuel, and 61.5 tons of plutonium declared excess for weapons purposes with the exception of 3.5 tons discarded at the Rocky Flats Plant which is included in the 61.5 tons “excess” declaration. About 41.8 metric tons of the U.S. excess plutonium is expected to be processed so it can be mixed with uranium for fabrication into mixed oxide fuel for use in commercial nuclear power plants and subsequently disposed. Disposition options for 5 tons of “non-pit” plutonium include mixing with defense high-level wastes to be vitrified or direct disposal in WIPP. [p. 2 of report]
So there is a 3.5-ton exception to the exception, except it is included in the 61.5 tons to which it is an exception? Is it included in the 12.7 tons? Or is it somehow double-counted? A full mass balance would clarify this.

There are other points that are unclear and potentially misleading.
The dramatic increase from the DOE’s 1996 waste estimate appears to be due to: reclassification as waste of process residues originally set aside for plutonium recovery for weapons; underestimates of production losses; and improvements in waste characterization data. [p. 1 of report]
If material was not considered waste in the earlier estimate and it is now, that doesn’t imply errors in the earlier estimate.

“Production losses” have been a question for as long as I can remember. Material held up in glove-box ventilation systems and other production piping has been an explanation for a lack of agreement between production figures and current inventories. Now that Rocky Flats and other production buildings have been torn down, better estimates must be available for those holdup amounts. I have asked some questions on this, and the numbers are not publicly available. It would seem that releasing these numbers, if they improve the mass balance, would be good for the Department of Energy, showing that they knew what they were talking about. Does Alvarez have access to those numbers? He doesn’t say.

As van der Linden notes, Alvarez suddenly brings in the idea of having the IAEA monitor WIPP. It’s only one of several off-topic insertions, another of the features of this report that doesn’t inspire confidence. But it’s actually a good point, in my opinion. Less so is his suggestion that better knowledge of the mass balance of plutonium produced and lost over the years will improve treaty verification.

Neither the United States nor Russia have adequate records of plutonium production. I’ve tried to work with original shipping records. Secrecy, poor analytical methods, and just not seeing those records as important at the time all are factors. We simply don’t know, and won’t ever know, the amounts of plutonium produced down to tolerances of several kilograms. What will happen in treaty negotiations is that numbers will be agreed upon for the purposes of the treaty.

Something that the arms control community might make more of is IAEA monitoring within the United States. As Alvarez comments, it would be a way of indicating our transparency to the international community. I don’t understand why the US allows relatively little, and why the list of monitored sites is so closely held. I need to do more research on this to write in a well-informed manner, but maybe I’ll get around to that in another post.

4 comments:

Gwyneth Cravens said...

Cheryl, I appreciate your thoughtful analysis.

Is the object of Alvarez's paper to achieve something in the political arena--possibly having to do with IAEA inspections in the U.S. ? Could he also be channeling his longstanding opposition to commercial nuclear power?

Since the paper is being published by a policy institute rather than in, say, a nuclear waste management journal, he can avoid scrutiny by scientists prior to publication.

It seems to me that sound political policy in the arena of nuclear matters must first of all be based on scientific facts. As you point out, these facts can be known (at least up to a point). He evidently avoids addressing them. Unfortunately the Congress members and their aides who'll be sent this report are not likely to question its scientific basis or look to its sources.

In her post on armspolicywonk.com, Page van der Linden has pointed out that Alvarez avoids citing specific sources.

The collaboration of Wald and Alvarez is relevant. It shows how unscientific claims become accepted as science and how these are then distributed throughout the media and become "fact". As a result many wrong decisions are made that can have serious long-term consequences.

Kudos to you and Page for trying to clarify matters.

Cheryl Rofer said...

I'll let Alvarez speak for himself, if he cares to, on his purposes.

The Wald article says that the paper is to be published in a peer-reviewed journal. Usually such journals disallow other sorts of publication before theirs. At first, I wasn't sure that this report was in fact the paper to be published - Wald doesn't link to the report, and it took a bit of looking to find it. But the report link above says that the report is "featured" in the NYT article, so this must be it.

I doubt that there's actually a collaboration between Wald and Alvarez. Wald wants readers. Alvarez's claims will draw readers.

But it seems to me that what I've done above is just the sort of thing a reporter should be doing - no special scientific knowledge needed.

Gwyneth Cravens said...

I didn't mean a literal collaboration. But Matt Wald should have provided more background about Alvarez and made it clear that his background is not technical and that he is a controversial figure.

"Peer review" in the article's context implies that people with science credentials like yours and Page's would be fact-checking the article Alvarez decided put into the public media stream prior to print publication.

Rod Adams said...

Cheryl - excellent summary. This is an example of what I hope could become a trend that helps to correct some of the weaknesses of the current "peer review" system. The term has been used and abused to the point where it has little useful meaning.

For example, the journal where Alvarez is apparently planning to publish his paper is a journal published by a school that teaches policy wonks interested in public affairs. It is not a science journal, despite its name. I would guess that the peer review process for that journal does not include any of the detailed technical questions and review that you have included in your post.

As qualified people like you continue to publish detailed reviews like this, we might have another way to evaluate the credibility of a document other than resorting to a description like "published in a peer-reviewed journal".