September 25, 1997

Dr. Denny Erickson, Director
Environment, Safety, and Health Division
Los Alamos National Laboratory
Los Alamos, NM 87544

Dear Denny,

Thank you for the well-organized, well-moderated, and mildly entertaining public meeting of last night, concerning (among other topics) the rather esoteric question of where the trace amounts of plutonium (Pu) in Cochiti reservoir came from.  Surely it is clear to all parties involved that these levels have no public health significance whatsoever.  And, just as surely, this lack of public health salience has already been known for at least a decade, making the subject a safe one for detailed discussion, whether in public or in private with the tribes.  The percentage of Pu coming directly from LANL to Cochiti is too integrated a number to be meaningful for containment or cleanup decisions.

The plutonium in the canyons is more interesting, since a) it is present at higher concentrations, b) it is mobile, and c) some of it is uphill from tribal lands.  More could have been presented on this subject.  I gathered that after some hundreds of millions of dollars of environmental restoration expenses at LANL, there is still no clear idea of what (if anything) should be done about Pu in canyon sediments. 

Apparently mercury is also a salient contaminant, given its mobility and its ability to bioaccumulate in fish.  Perhaps, in hindsight, a risk-based analysis of all contaminants at LANL would have been an appropriate subject for presentation.  Has such an analysis been done?  If so, could it be provided?  If yours is not the appropriate office to answer this question, could you forward this letter to the appropriate person?

Could someone on your staff mail back to me the official screening action level LANL is using for plutonium?  Bruce on your staff said that the screening level used was 24 picocuries (2.4 X 10^-11) per gram of soil.  In Transuranium Elements: Vol. 2: Technical Basis for Remedial Actions (EPA 520/1-90-016), a 1990 EPA guidance publication, a screening action level of 0.1 microcuries per square meter is suggested for new contamination, and a level twice that for old plutonium contamination.  These translate, using a soil density of 1.5 g/cm³ and an assumed mixing/sampling depth of 5 cm, to 1.33 and 2.66 picocuries (i.e. 1.33 and 2.66 x 10^-12 curies) per gram of soil -- about ten times more stringent than the LANL screening level. 

I think it is appropriate to have an open public process, involving tribal leaders as well, to set screening levels as well as cleanup levels for these and other contaminants, since DOE and LANL are now highly motivated to increase the number of potential release sites (PRSs) that are placed in the "no further action" category.  Am I correct to assume that screening levels, not cleanup levels, will determine this status?

You may not know that DOE has an obligation to conduct a public site-planning process under DOE Order 430.1 ("Life Cycle Asset Management").  The University of California has, prior to the new contract at least, accepted the applicability of this order to LANL.  It is hard to do site planning without having agreed-upon screening levels that can be compared to the levels found at the PRSs.

The emphasis of the presentations last night was largely on contaminants already emplaced in the environment -- that is, a retrospective emphasis -- as opposed to those that might be emplaced in the future due to a) routine operations or b) one-time events, e.g. accidents.  I think LANL is doing a good job, after considerable prodding, in decreasing the environmental footprint of its routine operations.  Much more can, should, and I think will, be done.

But the risk of accidents with environmental consequences is hardly trivial, even as recent experience suggests.  Just one such hypothetical accident, the accidental detonation of a plutonium hydrotesting assembly prior to closure of the containment vessel, could put more plutonium in the open environment than all LANL operations to date.  Special facilities ("gravel gerties") are present at Pantex and the Nevada Test Site for this kind of dangerous operation.  LANL does not have one, raising the question of whether LANL's safety standards are comparable to these other sites.  No one will answer my questions about these activities, and indeed the very existence of some programs is highly classified.  How can a "black" program like "Apaloosa" (hydrotesting with "Cider", i.e. Pu-242) receive the independent safety oversight as well as the public acceptance implied by our democratic ideals?  Please respond to me on this as well as screening levels.

Returning to the historical perspective, it is (as I said) illogical to neatly divide the plutonium in Cochiti Reservoir into that due to LANL and that due to global fallout, since LANL was heavily involved in creating global fallout -- especially at the nearest upwind test site. 

But the effects of these tests are not really in the past.  "Fallout" -- a partial misnomer because much of it does not fall out -- continues to kill.  Some 430,000 cancer deaths worldwide are expected by the end of the century alone (Radioactive Heaven and Earth, International Physicians for the Prevention of Nuclear War, 1991, Zed Books). 

In the long run, perhaps 85% of the doses due to nuclear testing will come from carbon-14 (IPPNW, cited above).  Each megaton of atmospheric testing created about 2.25X10²⁶ neutrons, 95% of which were captured by atmospheric nitrogen to create radioactive carbon-14 (see Sakharov, translated and reprinted with commentary by Frank von Hippel in Science and Global Security, 1990, Vol. 1, pp. 175-187) (available in LANL library).  The expected deaths due to carbon-14 number in the low millions.  These impacts were in general known to LANL scientists at the time the tests were planned and conducted, just as they were to Sakharov in the Soviet Union.  Why were there no scientists raising ethical concerns at Los Alamos, in a country where it could be freely done?

Millions of early deaths are hardly of mere academic interest.  The femtocuries per gram of Pu in reservoir sediments, from fallout or whatever source, appear especially trivial in this somber light.  I bring it up to make a point: the effects of LANL, environmental and otherwise, have always been largely indirect and -- this last point being beyond the scope of this letter and therefore will be left mysterious -- are becoming increasingly unmeasurable.  Measuring femtocuries in mud misses the point.

The important book I hawked to you after the meeting is called Normal Accidents:  Living with High-Risk Technologies, Charles Perrow, Basic Books, 1984.  Perrow's work at Yale has spawned a large literature on the institutional analysis of safety.  You can see some of this literature in Scot Sagan's book The Limits of Safety:  Organizations, Accidents, and Nuclear Weapons (Princeton Univ. Press, 1993) (available at Otowi, I think).  LANL implicitly uses "high-reliability" theory in accident prevention, without really looking at its underlying assumptions, its limitations, or recognizing that it is only one school of thought.

Thank you for your attention and for your work.

Sincerely,
Greg Mello

cc:
Governors of the Cochiti, San Ildefonso, Santa Clara, and Jemez nations
Jay Coghlan, CCNS
Hank Daneman, CAB
John Parker, NMED
Tom Todd, DOE LAAO