Los Alamos Undergoes Clean Air Act Audits
First Independent Environmental Review of a U.S. Nuclear Weapons Plant
By Arjun Makhijani and Joni Arends1

In one respect, 1997 was an historic year for environmental issues in the U.S. nuclear weapons establishment. That year the first independent environmental audit of a nuclear weapons installation, conducted with public oversight and under court order, began. The installation was the world's best funded nuclear weapons laboratory, the Los Alamos National Laboratory (LANL) in New Mexico, owned by the U.S. Department of Energy (DOE) and operated for it by the University of California. Two audits of the compliance of LANL with the Clean Air Act have been completed. One more will take place in 2002, with the potential fourth audit in 2004 (as yet undecided). This is a brief summary of what led up to the audits and what has transpired to date. The reports of the audits are on the web at www.racteam.com/Experience/Projects/LANLAudit.htm. IEER's comments on those reports, made as part of the court-supervised process, are on the web at http://www.ieer.org/reports/lanl/audit1.html.

LANL is a complex operation, devoted primarily, but not exclusively, to nuclear weapons and associated scientific and technical experimentation, theoretical work, and computer modeling. Plutonium, uranium (in various enrichments), and tritium are stored and processed there (the last for both weapons and nuclear fusion power research). Large amounts of radioactive waste are also stored there. LANL has a small-scale capacity to manufacture nuclear weapons, mostly for prototyping purposes, as well as the associated chemical and physical processing facilities. It was the laboratory where the first nuclear weapons, tested in New Mexico and used on Hiroshima and Nagasaki in 1945, were made. It has also been designated as the site for future stockpile plutonium pit production and has an aggressive nuclear sub-critical hydrotesting program.

Experimentation at LANL involves many radionuclides, and results in a variety of discharges to the air and water. In 1991, with ink hardly dry on the regulations implementing the 1990 Clean Air Act,2 the Environmental Protection Agency (EPA) found LANL to be in violation of it. LANL had not done its dose calculations in the prescribed manner.3 The required calculation method would have estimated the dose to the hypothetical individual residing at the site boundary at more than the allowable maximum of 10 millirem per year.4 The use of a hypothetical individual to estimate a maximum dose is common practice in radiation regulations. If the location and circumstances of the maximum exposure are properly specified, such a regulatory procedure ensures that everyone else in the general public is also protected and will not be exposed to more than the allowable radiation dose limit.5

LANL's measurements of releases of radionuclides to the air, and its assessment of the impacts of unmeasured sources, also left much to be desired. A subsequent agreement between EPA and LANL, called the Federal Facilities Compliance Agreement, settled the issue for the two bureaucracies. But the public interest group Concerned Citizens for Nuclear Safety (CCNS) believed that the EPA had allowed LANL to remain in violation of the requirements of the Clean Air Act as regards making measurements and assessing radiation doses to the public.

Quite apart from the actual level of releases of doses, technical requirements regarding measurements and modeling are the heart of assurances to the public that a facility is in compliance. In 1994, CCNS filed a lawsuit alleging that LANL continued to violate the Clean Air Act and that its violations should stop.6Of legal importance in and of itself was that CCNS was granted the standing to sue, especially considering the Federal Facilities Compliance Agreement between DOE and EPA. IEER served as a technical consultant in the lawsuit.

Senior Judge Edwin Mechem issued a summary judgment that LANL was in violation of the Clean Air Act as alleged in CCNS's lawsuit. This was based largely on official documents issued by LANL itself. The judge ordered LANL to negotiate with CCNS to attempt to resolve the lawsuit. If not, LANL faced large fines and an order shutting down the facility until it could be brought into compliance. LANL chose to settle the case. It should be noted that LANL management consciously chose to keep operating while knowing that it was operating in violation of federal environmental law.

A Consent Decree embodying the settlement was filed in federal court in March 1997. It is a complex settlement, but its core provision relates to independent audits. CCNS and LANL agreed that the Risk Assessment Corporation, led by John Till, would set up an Independent Technical Audit Team, which would conduct up to four independent audits. CCNS would be able to retain its own consultant to monitor the audit ‹ that is, to check whether the audit was being done in a thorough and competent manner. CCNS chose IEER to provide the scientific personnel for this monitoring function.7

The federal government was to pay for the audit and the monitoring function, but the DOE did not control the funds. The Risk Assessment Corporation was paid directly by the Department of Justice (DOJ). DOJ also paid CCNS, which then paid IEER and its own expenses in monitoring the audit. Making the audits even more historic, Dr. Till decided to open the whole audit process, including site visits, to state and tribal government representatives and the public. LANL and DOE personnel went to extraordinary lengths to ensure access to the monitoring equipment and facilities to everyone who went on the site visits.8

LANL was to provide all relevant records and documents and make its air monitoring facilities as well as personnel available to the auditors and monitors. The issues raised during the audits were:

  • Were the environmental monitoring systems and the stack monitoring systems adequate for emissions monitoring and dose estimation?
  • Was the record keeping for unmonitored sources appropriate and were the procedures for estimating emissions from hundreds of these sources appropriate?
  • Were the computer models used by LANL appropriate for assessing doses given the complex terrain, consisting of canyons and mesas, and the patterns of emissions?
  • Did the assumptions about releases of radioactivity reflect the reality of operations in the various facilities?
  • Were the procedures for sampling and analysis, both at LANL and at the laboratories to which the stack samples were sent, appropriate and adequate?
  • Were quality control and assurance procedures adequate and were they being followed?
  • Was the location of the hypothetical maximally exposed individual appropriate or might it underestimate radiation doses under certain circumstances?

The main source of emissions (over 90%) at LANL during the 1990s was a proton accelerator, known as LANSCE (Los Alamos Neutron Science Center). After the EPA cited LANL in 1991 for emissions relating to this facility, measures were taken to reduce those emissions. Yet, the conclusion of the first audit, which examined LANL compliance for 1996, was that LANL was out of compliance with the Clean Air Act and, in addition, that there were scientific deficiencies in LANLšs compliance program. The main findings in this regard were:

1. "Lack of documentation of radionuclide inventory. An inventory for 1996 could not be determined from documentation that currently exists. The lack of inventory data in certain buildings was a primary deficiency that prevented the audit team from verifying what sources may have existed and, therefore, quantitatively verifying compliance."

2. "Absence of independent verification of calculations."

3. "Certain environmental sampling techniques and assumptions are not well described or documented."

4. "Sample losses. An evaluation of aerosol particle loss in the sample transport systems is required by the effluent sampling guidance in the regulation. However, LANL has failed to analyze losses in probes and transport lines for three sampling systems that do not employ shrouded probes."

The finding of non-compliance was issued in May 1998, midway through the first audit process, to allow LANL to fix the problems that were identified to that point.

The funds allocated for the first audit turned out to be insufficient and the audit was admittedly incomplete. IEER issued its assessment of the audit, agreeing with the finding of non-compliance. However, IEER disagreed with the audit team's "considered judgment" that the 10-millirem dose limit had not been exceeded. IEER found that, since a number of issues relating to measurements and models had not been resolved and no uncertainty analysis had been performed, such a conclusion by the audit team was unwarranted. IEER did not make any claim that doses were higher than the limit of 10 millirem, just that the analysis performed by the audit team did not permit such a statement of compliance with the dose limit. LANL disputed the auditšs non-compliance finding, but nonetheless began to implement many of its recommendations.

The main concerns that IEER raised early on about LANL's approach to compliance and EPA's agreement with LANL as to how LANL was to demonstrate compliance were:

  • The model used by LANL for making dose calculations, and approved by the EPA, is a "flat earth model" while the terrain at Los Alamos is anything but. The model is called CAP-88. Until the flat earth assumption could be shown to be uniformly conservative (that is, providing overestimates of doses under all prevailing conditions), the model was not scientifically appropriate.
  • The CAP-88 model estimates annual average doses, and assumes uniform releases throughout the year. However, some releases are not uniform and could result in doses to individuals higher than those estimated by the annual average model.
  • IEER also raised the question as to whether a passer-by, known as a transient receptor in technical jargon, such as a jogger passing by a radionuclide-emitting facility, might not under certain circumstances be exposed to a higher dose than the facility-designated hypothetical maximally exposed individual. This inquiry also led to the question of how the hypothetical maximally exposed individual was to be identified.

The second audit was conducted in the year 2000 for data relating to the year 1999. LANSCE, the main radiation source, was not in operation during 1999. As a result, the maximum estimated dose was 0.32 millirem. There are uncertainties associated with this number, which continue to be unquantified. But there was general agreement that LANL was in compliance with the Clean Air Act radiation regulations for 1999. This finding of compliance was expected and also not as significant as it would have been had LANSCE been in full operation. It should be noted that LANL has taken steps to reduce LANSCE emissions.

The audit team also did a check of the kinds of results that might be anticipated if a model that reflected the actual terrain at Los Alamos were to be used. It was found, based on admittedly limited modeling, that the flat earth model gave conservative results under most, but not all, circumstances. According to IEER, since the complex terrain model showed doses that were higher than the flat earth model under some circumstances, a more detailed and thorough investigation was needed before the continued use of the flat earth model could be justified.

Overall, the audit and monitoring process allowed sound science that the public could trust to emerge as regards one environmental aspect of the nuclear weapons complex. Of course, its limitation, understood by all parties from the start, was that the process could not be used one way or another to promote nuclear disarmament, a treaty-related obligation of the United States as well as a declared goal of some of the parties involved including IEER. On the other hand, the process has allowed us to raise some crucial national environmental questions in the context of an official process. In particular, CCNS and IEER have taken up the question of the inappropriateness of the flat terrain model under circumstances similar to LANL elsewhere, both at government-owned and privately-owned installations. We have also raised the question of the transient receptor. So far the EPA has agreed to look at the question of the flat terrain model and how it should be validated.

The third audit is scheduled to begin in June 2002 and will examine LANLšs 2001 compliance. An additional focus of the audit will be to develop a working model for independent audits and monitoring of audits for the public to use at government-owned and industrial sites regulated under the Clean Air Act.


1 Joni Arends is the Waste Programs Director at Concerned Citizens for Nuclear Safety. 2 The part relating to radionuclide air emissions is in the Code of Federal Regulations at 40 CFR 61 Subpart H.
3 LANL used an unapproved "building shielding" factor. EPA disallowed it after John Stroud of Concerned Citizens for Nuclear Safety formally questioned its use to the EPA.
4 The Clean Air Act does not limit all radiation doses via the air pathway, only those delivered by radionuclides. Specifically, neutron doses are excluded since neutrons are, technically speaking, not radionuclides in that they are not in the periodic table of elements.
5 See Section 2 of Setting Cleanup Standards to Protect Future Generations, by Arjun Makhijani and Sriram Gopal, IEER, December 2001, online at www.ieer.org/reports/rocky/2critgp.html. A dose less than the regulatory limit does not mean zero cancer risk. It means less than the risk specified in the regulations. The risk of a fatal cancer to an adult from exposure to radiation is considered to be proportional to exposure. The official risk factor is equal to about 4 in 10 million risk of a fatal cancer per millirem of exposure.
6 Jay Coghlan, Caron Balkany, Esq., and Carol Oppenheimer, Esq., played leading roles in the strategy and litigation. John Stroud was the originator of the lawsuit.
7 Bernd Franke (as a consultant to IEER) and Arjun Makhijani served as IEERšs monitoring team on the audit process.
8 Of course, access to classified areas was restricted to those with appropriate security clearances.

Back to Clean Air Act