Canadian Coalition
for Nuclear

Regroupement pour
la surveillance
du nucléaire


Conformity Analysis of the
AECL Geological Burial Concept

An Analysis of the Lack of Conformity of

AECL'S Environmental Impact Statement
(AECL-10711 COG-93-1)


the FINAL GUIDELINES (March 1992) of the
Federal Environmental Assessment Review Office (FEARO)
for the Preparation of an Environmental Impact Statement
on the Nuclear Fuel Waste Management and Disposal Concept

Submitted on August 8 1995 by CCNR to FEARO


1. GAUGING THE DOCUMENT: EIS or Feasibility Study ?
2. CLARIFYING THE CONCEPT: Disposal or Storage ?
3. DEFINING THE DECISION: Approval Forever ?
4. NUCLEAR FUEL WASTE: The Reprocessing Option
5. REFERENCES: Independent and Accessible ?
6. NUCLEAR FUEL WASTE: Is It A Problem ?
8. HUMAN HEALTH: Unprofessional Treatment
11. MATHEMATICAL MODELS: Unproven Hypotheses ?
12. PARAMETERS: Grist for the Mill
13. MODELLING: The Heart of the Beast
14. VALIDATION: The Proof of the Pudding
15. TABOO SCENARIOS: Beyond the Pale ?


The most striking thing about the main AECL document is how much like an engineering feasibility study it is, and how peripheral the environment, or any considerations having to do with the environment, seem to be.

The biological effects of ionizing radiation on humans are covered, very poorly, in three pages (AECL, pp. 35-37). The biological effects of ionizing radiation on non-human biota are covered, very poorly, in another three pages (AECL, pp. 39-41). The biological effects of "chemically toxic elements" are covered in little more than a page (AECL, p. 42) -- yet more specific and useful information is hinted at in that one page, meager as it is, than in the six previous pages on the ionizing radiation emitted by "radioactively toxic elements" (which are, in fact, never identified or discussed as such).

The glossary is the poorest we have ever seen in a supposedly serious work. 'Health', 'mitigate', & 'safe' are included, but 'radiation', 'radioactivity', 'radioactive decay', 'radionuclide', 'actinide', 'fission', 'criticality', 'activation', 'model', 'validation', 'verification', 'serious genetic effect', 'alpha emitter', 'sievert', 'relative biological effectiveness' -- let alone 'spalling' or 'tortuosity' -- are nowhere to be found.

In Appendix A, the proponent announces (with no apology or justification) that "Because the structure . . . is different from the structure of the guidelines, we prepared this cross-reference to indicate where items in the guidelines are addressed in the EIS." A little checking soon shows, however, that a great many issues stipulated in the Guidelines are either totally absent or inadequately addressed. This, despite the fact that the Executive Summary of the Guidelines plainly states that these are "issues which the Panel has determined should be addressed in the EIS."

There is no index of subjects or of names. References to the R-documents are very imprecise, giving neither page numbers, nor section numbers or even chapter numbers. There is no master bibliography, but nine separate reference lists in the main AECL document and dozens more in the R-documents. And without a name index, it is impossible to determine where a given reference is cited in the text, or how often.

Our overall conclusion, after many long hours spent painstakingly chasing down labyrinthine "references within references", is that the main AECL document is not even close to being in conformity with the Guidelines.

Organizational defects compound the deficiencies in content, and serve to camouflage the many aspects of the document which are not in conformity.


  • The main AECL document is wholly unacceptable as an Environmental Impact Statement in both organization and content. So pervasive is the lack of conformity that it can only be rectified by writing a new EIS.
  • Throughout the document, the emphasis is on the engineering details of the project, while environmental impacts are relegated to the distant background. Health impacts, which are described entirely in terms of regulatory criteria (1), are treated as constraints to the project rather than as the raison d'être of the project.
  • ------------------------------------------------------------------------------------------
    (1) Safe is defined in the glossary as "Meeting criteria, guidelines and standards . . . " (AECL, p.495)
  • Moreover, the document is so poorly organized in relation to the Panel's Guidelines that a proper and comprehensive public review of conformity is seriously impeded. Because the document does not follow the organization of the Panels' Guidelines, is not indexed, and is not available in electronic form, it is quite difficult for public interest groups with moderate resources -- and even those with more extensive resources -- to verify the extent to which the document is or is not in conformity with the Panel's Guidelines.

  • In addition, the glossary is inadequate; cross-references to the R-documents are generally unhelpful because they give no page numbers or even section or chapter numbers; and there is no master bibliography.


  • AECL should be required to prepare and submit a new document -- a genuine Environmental Impact Statement or EIS; one which
    • highlights potential environmental impacts of the project throughout;
    • follows the structure of the FEARO guidelines sequentially;
    • is fully indexed not only in itself but also in relation to the R-docs;
    • contains a complete subject and name index for the EIS and R-docs;
    • contains a complete glossary for the EIS as well as the R-docs;
    • refers to R-documents by page and by chapter/section numbers (2);
    • gives a master bibliography of all refs cited in the EIS and R-docs;
    • is available electronically, on floppy disks and in CD-ROM format.
    (2) Giving chapter/section numbers as well as page numbers will facilitate communications in both official languages.
  • All other deficiencies identified by the FEARO Panel should be addressed by AECL in the appropriate sections of this new EIS.



We begin at the beginning. What is the mandate of the Panel? What is the AECL Geological Burial Concept? What does approval of that concept mean? In short, what is the Panel's task?

In September, 1988, AECL submitted for public review its concept for the deep geological burial of nuclear fuel waste in Canada, and subsequently a panel was appointed in October, 1989 ....

The Panel was given the mandate to undertake a review of the safety and acceptability of the AECL concept, along with a broad range of nuclear fuel waste management issues ... The Panel will take into consideration the various approaches to the long-term management of nuclear fuel waste, which is presently stored at reactor sites. This will include long-term storage with the capability for continuing intervention in the form of monitoring, retrieval and remedial action, and the transition from storage to permanent disposal . . . .

(Guidelines, Preamble ~ emphasis added)

It is quite clear from the proponent's documents that the AECL concept is submitted for approval solely as a "permanent disposal" option rather than a "perpetual monitored storage" option. In the Overview to the main AECL document (reproduced verbatim in each of the R-docs) the proponent states:

Current storage practices, while safe, require continuing institutional controls, such as security measures, monitoring, and maintenance. Thus storage is an effective interim measure for protection of human health and the natural environment, but not a permanent measure. As stated by the Atomic Energy Control Board, "For the long-term management of radioactive wastes, the preferred approach is disposal, a permanent method of management in which there is no intention of retrieval and which, ideally, uses techniques and designs that do not rely for their success on long-term institutional control beyond a reasonable period of time."

(AECL, p. i ~ emphasis added)

In the AECL Summary, on page 1, the proponent declares

"AECL proposes that the waste be disposed of in a vault, which would eventually be sealed, several hundred metres below the surface in rock of the Canadian shield." (AECL, p. 1, Summary ~ emphasis added)

Thus, if the Panel approves the AECL concept, such approval will be interpreted by the proponent and by others in Canada and around the world as proof that the permanent unmonitored disposal of high-level radioactive waste in geological formations is safe, and that the problem of disposing of high-level radioactive waste has been solved, at least in principle, once and for all.

Given the far-reaching and even global implications of any such approval, the Panel should require AECL in its new EIS [Deficiency #1] to define both the proposed concept [Deficiency #2] and the proposed approvals process [Deficiency #3] much more precisely than at present.


  • The essential, unalterable components of the AECL Geological Burial Concept are not sufficiently well delineated to allow for meaningful approval as a permanent disposal option.
  • Countless illustrative details of the concept are discussed in the main AECL document and the R-documents, any of which may arguably be regarded as "merely a detail" and "not really essential" to the concept. Without a much more precise definition of the unalterable, essential components of the AECL Geological Burial Concept, any approval of the concept by the Panel could be interpreted as a "carte blanche", allowing AECL (or the implementing organization) to change details at will at some later stage of development, possibly altering the concept dramatically, turning it into something quite different from what the panel thought it was approving.
  • Remedy

  • In the new EIS [Deficiency #1, above] the proponent should give a self-contained and complete description of the essential and unalterable characteristics of the AECL Geological Burial Concept which are considered necessary and sufficient for approval as a disposal option.
  • This definition should include technical specifications such as the type(s) of rock permitted, the range of depths considered, the maximum heat loading allowed, the minimum distance from emplaced waste to a major fracture zone, et cetera. All physical and engineeering aspects of the AECL Geological Burial Concept which are considered essential to the permanent unmonitored disposal of nuclear fuel waste -- all those aspects which are considered necessary and sufficient to ensure that such waste will remain isolated from the biosphere for the indefinite future -- should be included in the description.
  • The description should also include all essential sociopolitical or socioeconomic considerations which the proponent considers to be inviolable aspects of the AECL Geological Burial Concept.



In Section 1.7 of the main AECL document, entitled "Unusual Aspects of the Review and this EIS", the proponent makes no reference to the most unusual aspect of all - that the disposal concept has yet to be proven.

Indeed, nowhere in the document does the proponent explicitly acknowledge the unprecedented nature of this challenge, and the consequent necessity for the proponent to provide incontrovertible proof, acceptable to the Panel, that the AECL Geological Burial Concept will work reliably as a safe permanent disposal option.

Although the proponent draws a clear distinction between storage and disposal in the Overview to the main AECL document, as well as in the Summary, this distinction is blurred in the body of the main AECL document -- in Chapter 3 (sections 3.6.5 and 3.6.6), in Chapter 5 (sections 5.8.9 and 5.8.10), and elsewhere.

3.6.5 ~ Monitoring

Monitoring to determine whether the disposal system was performing as expected would be an essential activity during concept implementation. The AECB (1985) has stated that:

The performance of the waste repository in the postclosure period will be assessed on the basis of predictive modelling. Therefore measurements will be required during the preclosure period in order to ensure that input data for the models are sufficiently complete and representative of the repository environment.... The repository will only be allowed to close when sufficient evidence is available to lead to the conclusion, with a sufficient degree of certainty, that the facility could be abandoned without the need for postclosure monitoring.

Although the AECB views long-term postclosure monitoring as "a necessary element of protection-in-depth" (AECB 1991b), it requires that safety must not be compromised by any provisions that may be made for monitoring, either preclosure or postclosure (AECB 1985).

Participants in AECL's public consultation program and focus group discussions stressed the need for both preclosure and postclosure monitoring (Greber and Anderson 1989, pp.49-51; Pat Delbridge Associates Inc. 1989; Pieroni 1984, 1986). National surveys also indicate that monitoring would be very important to the public. For example, 51% of respondents to a recent survey considered that further monitoring beyond a 40-year operation stage would be required to make a decision about sealing the vault (Angus Reid Group 1990), and a survey conducted in 1986 indicated that a significant percentage of respondents (39% in Ontario and 51% in northern Ontario) considered that perpetual monitoring would be necessary (The Canadian Gallup Poll Ltd. 1986).

The proponent seems to take no note of the fact that these views run directly counter to the AECL concept of permanent, unmonitored disposal.

AECL should be required to indicate how the disposal option can be judged.


  • In the AECL Overview -- which is reproduced verbatim in each of the R-docs -- a clear distinction is drawn between storage and disposal.

    In the main AECL document, however, the distinction is blurred between

    1. long-term geologic storage of nuclear fuel waste, with continual, on-going monitoring and perpetual institutional control, versus
    2. permanent, unmonitored, walk-away disposal of nuclear fuel waste.

  • In our view, the Panel must be able to judge NOW whether or not the AECL Geological Burial Concept is acceptable as a PERMANENT DISPOSAL option.

    Otherwise, what is portrayed as a permanent disposal option might become instead an expensive, cumbersome never-ending, monitored storage option. If so, the stated rationale for the AECL Geological Storage Concept would evaporate, and many important considerations -- e.g. preferred geographic location, optimal depth of the repository, total funding needs, long-term institutional control, & desirability of geologic storage -- would alter significantly.

  • Yet AECL offers no clear decision framework to allow the Panel to choose definitively between these two quite different possibilities.
  • Remedy

  • In the new EIS [Deficiency #1] -- in addition to defining the disposal concept more precisely [Deficiency #2] -- the proponent should be required to outline a decision framework, comprised of a list of logical steps, each one capable of being verified and/or judged using specified decision criteria, which AECL believes will provide the Panel with the needed assurance that the repository can not only be built, operated and decommissioned safely, but that it can isolate nuclear fuel wastes from the biosphere for spans of time that transcend human experience.
  • If any significant doubt on this score remains, we believe that the Panel cannot, in conscience, approve the AECL Geological Burial Concept as a permanent disposal concept.



"The EIS should ... describe the origin and nature of nuclear fuel waste in order to provide a clear understanding of the requirements for its safe management."

Guidelines, Executive Summary, p. iv

Throughout the main AECL document (but not in the R-docs) the form of the nuclear fuel waste is described as either irradiated fuel, or solidified post-reprocessing waste.

"If used fuel were reprocessed, the most radioactive material that remained (the high-level waste) would be solidified. The term "nuclear fuel waste," as used in this document, refers to either the used fuel, if it is not to be reprocessed, or the solidified high-level waste from reprocessing." (AECL, Overview, p. i)

Consider the following points:

  • reprocessing is widely recognized as that portion of the nuclear fuel chain having the greatest detrimental impact on the environment;
  • reprocessing transforms a significant fraction of the solid irradiated fuel waste into liquid and gaseous forms which are potentially far more mobile in the environment than the irradiated fuel itself;
  • among the radionuclides liberated from the irradiated fuel during the dissolution phase of reprocessing are large fractions of the iodine-129 and carbon-14 inventories : these two nuclides, which figure prominently in the post-closure assessment scenarios carried out by the proponent, are an inextricable part of the nuclear fuel waste which the Panel must deal with;
  • reprocessing also yields strategic nuclear material which could be used in nuclear weapons, thereby posing extraordinary security problems;
  • there are strong indications that any future reprocessing plant may well be located on the same site as a future nuclear fuel waste repository, or nearby.

"If retrieval was intended to provide used fuel for future reprocessing and recycling, it would be desirable to select a site for centralized storage and disposal that was also suitable for a reprocessing facility." (AECL, p. 333)

Thus, failure to describe the environmental, social, and political impacts of reprocessing and the implications for a nearby nuclear fuel waste repository constitutes a glaring & unacceptable omission in the main AECL document.

If the Panel decides that it does not want to embark on an environmental assessment of the nuclear fuel reprocessing option at the present time, it should instruct AECL to drop all references to reprocessing and to post-reprocessing nuclear fuel waste from the AECL Geological Disposal Concept, except for purposes of information.


  • The proponent defines nuclear fuel waste to mean either used nuclear fuel or solidified high-level liquid waste from reprocessing.

    Accordingly, failure to describe the environmental, social, and political impacts of reprocessing, and the possible implications for a nearby nuclear fuel waste repository, constitutes a glaring and unacceptable omission in the main AECL document.

  • Remedy

  • In the new EIS [Deficiency #1, above] AECL should provide a thorough discussion of reprocessing technology, describing

    • all radionuclides released from irradiated fuel during the dissolution stage of the reprocessing operation;
    • all radionuclides released from high-level radioactive liquid waste during the solidification stage of the reprocessing operation;
    • environmental and safety records of reprocessing plants to date;
    • environmental and safety implications of siting a reprocessing plant on the same site as a geological nuclear fuel waste repository;
    • security and other socio-political implications of siting a reprocessing plant on the same site as a geological nuclear fuel waste repository;
    • potential criticality accidents in a reprocessing plant;
    • potential chemical explosions in a reprocessing plant;
    • accident analysis at a reprocessing plant;
    • aqueous and airborne radiological emissions from a reprocessing plant;
    • quantities and disposal of contaminated equipment from reprocessing;
    • handling of waste water from the reprocessing plant's laundry;
    • the characteristics and behaviour of solidified post-reprocessing nuclear fuel waste in comparison with the characteristics and behaviour of irradiated UO2 fuel in the vault and geosphere environments.



The Panel will receive the EIS submitted by AECL and will distribute it to review participants for comment. ...

... All sources of information and supporting data used in these analyses and assessments should be identified.

Guidelines, Preamble, p. ix

The nature of the concept, the approvals process, and the waste form having been considered, we now ask about the quality of the documentation provided.

Throughout the main AECL document and the R-documents, a large percentage of the citations are to internal AECL documents or other industry documents which are not readily available in libraries, and which have not been subjected to the independent scientific peer review process typically associated with articles published in refereed scientific journals.

It is exceedingly difficult to gauge the quality of a large body of work when the references are not readily available to those reviewing it. (Those wishing to review such citations must ask AECL to send them, sometimes waiting many weeks for the documents to arrive, and then receiving only a portion of what was requested.) It is particularly difficult to have confidence in the quality of references which have not been subjected to independent peer review.

As previously remarked, without a name index [see Deficiency #1] it is difficult to determine where or whether a given reference is cited, or how many times. Similarly, without a master bibliography [see Deficiency #1] it is difficult to assess the completeness or adequacy of the reference list.

At several critical junctures -- for example, the claim that accidental criticality is not possible in the vault (AECL, p. 280) -- there is only one reference given, and it is an internal AECL document not subject to the usual scientific standards of independent peer review. In some cases, no reference is given at all (e.g. "Those microbes present would likely form biofilms" -- p. 280). In a few cases, we learn that the sole reference for a given assertion is "in preparation".

Since the proponent admits there is no urgent need for geological disposal of nuclear fuel waste (3), it is difficult to understand why AECL can not wait, or should not be asked to wait, until most or all of the references have appeared in refereed journals.

(3) "There is no urgency to dispose of waste" -- Section 9.10, Recommendation 4, p. 346.


  • It is not readily apparent from the text of the main AECL document -- or the R-documents, for that matter -- what percentage of the citations therein are essentially in-house AECL publications.
  • Given the fact that AECL admits there is no urgent need for the geological disposal of nuclear fuel waste (4), it is difficult to understand why AECL can not wait until most or all of their references have appeared in refereed journals.

    (4) "There is no urgency to dispose of waste" -- Section 9.10, Recommendation 4, p. 346.

  • It is exceedingly difficult to gauge the quality of the work when the references are not readily available to those reviewing it.
  • It is also particularly difficult to have confidence in the quality of references which have not been subjected to independent peer review.
  • Remedy

  • In the new EIS [Deficiency #1] AECL should be required to clearly indicate, in the text itself, which references are internal AECL documents.
  • Thus, instead of (Brown, 1978) we would see (Brown, AECL-1978), making it easier to determine which references are in-house documents & which are not.

  • Moreover, AECL should be required to deposit, in selected libraries across Canada, a complete set of all AECL secondary documents cited in the new EIS and the R-documents for easy access by intervenors.
  • The Panel should consider requiring AECL to see that all or most of the in-house secondary references are published in peer-reviewed journals before approval of the Geological Storage Concept will be considered.
  • At the very least, and in any event, AECL should be required to publish separate volumes of all of the outside reviews of research work related to the Geological Burial Concept that have been commissioned to date by AECL -- so that the Panel and intervenors can gauge the extent to which the reviewer's comments have been taken into account by AECL.



We now move to a consideration of the requirements in Chapter 2 of the Guidelines.

"The EIS should define and explain the overall problem posed by nuclear fuel waste in Canada, and discuss the present magnitude and expected growth of this problem. The EIS should discuss the ethical and moral framework in which the problem posed by nuclear fuel waste should be evaluated, state the need for long-term management of nuclear fuel waste and discuss why this issue must be addressed now."

(Guidelines, Executive Summary and p. 3)

In science, as in life, one cannot evaluate a proposed solution without first understanding the problem. Very sensibly, the Guidelines ask AECL to define and explain

  1. the nature and magnitude of the problem of nuclear fuel waste;
  2. the ethical and moral framework for evaluating that problem; and
  3. the need and speed with which the problem must be addressed.
None of these is addressed in a satisfactory manner in the main AECL document.


Except in the most pedestrian sense (e.g. the number of fuel bundles) there is no analysis of the nature & magnitude of the nuclear waste problem. In Chapter 1 of the main AECL document we learn only that nuclear fuel waste

". . . is radioactive and contains some chemically toxic elements. Humans and other organisms are protected by isolating the used fuel from the natural environment, shielding humans and other organisms from its radiation, and cooling it to remove the heat produced by radioactive decay." (AECL, p. 2)

There is no explanation of why it should require -- or how one could justify -- the expenditure of several billions of dollars and many decades of effort to carry out the solution to a problem which has not been well or even adequately characterized.

References in Chapter 1 to the potential health and environmental problems associated with nuclear fuel waste are veiled and indirect, and can only be inferred from the use of such terms as: "primary safety feature" (AECL, p. 2); "institutional controls to maintain safety" (AECL, p. 2); "minimize any burden placed on future generations" (AECL, p. 3); "safe and permanent disposal" (AECL, p. 4); "human health and the natural environment must be protected" (AECL, p. 5); and the like. This is completely unsatisfactory.

A discerning reader cannot fail to notice the proponent's reluctance to delineate the basic problem with nuclear fuel waste in a frank & forthright way.


The proponent also fails to discuss "the ethical and moral framework in which the problem . . . should be evaluated." No mention is made of the most obvious moral question, namely, whether we should continue to produce such highly toxic wastes. Admittedly, it would be awkward for the proponent to raise this moral question, since the extraordinary toxicity of the nuclear fuel waste -- which is at the heart of the moral dilemma -- has not been adequately addressed either.

It seems that the only ethical point raised in Chapter 1 of the main AECL document is one relating to the distributional equity of risks and benefits:

"Disposal is also needed to minimize any burden placed on future generations resulting from the nuclear fuel waste produced by the present generation. The present generation, since it derives a significant benefit from the activities that result in the production of nuclear fuel waste, ought to assume, to the extent possible, the responsibilities associated with nuclear fuel waste disposal." (AECL, p. 3)

However, the ethical implications of asking less privileged members of a society to accommodate toxic wastes produced by those who are more privileged is not raised -- although it is an obvious extension of the stated principal that those who benefit should assume the responsibilities. The ethical acceptability of making irreversible decisions in the face of uncertainty, given the potential for harm to future generations, is also not raised. Nor is there any consideration of the morality of reassuring people that there is no serious risk associated with a given course of action, when in fact there may be, potentially, a very great risk (c.f. the Chernobyl accident). How many in the present generation, for instance, were ever fully informed at the outset by AECL or Ontario Hydro -- or even partially informed -- about the highly toxic nature of nuclear fuel wastes and the fact that they would remain a serious threat for periods of time longer than the span of human history?


When it comes to justifying the proposed disposal scheme, the proponent once again does a less-than-adequate job. The permanent disposal of any highly toxic material by perpetual isolation from the biosphere has never before been successfully carried out, and serious questions exist in many people's minds as to whether today's science and technology can accomplish this task. It is true that the notion of permanent disposal has received a great deal of support, but that support is based on the presupposition that permanent disposal is possible.

Thus the need for a disposal scheme is, in part, dependent on the proponent's ability to demonstrate that such a scheme would work -- ensuring the isolation of nuclear fuel waste from the biosphere forever. But there is more to it than that.

Desirability is one thing; feasibility is another; but need is something else. The need for a multibillion dollar permanent disposal option has not been demonstrated. Nor has any urgency been attached to this proposal.

Throughout the main AECL document, the proponent emphasizes how safe the current storage of nuclear fuel waste is:

"In our opinion, a decision to proceed with siting a nuclear fuel waste disposal facility would not be made unless it was considered that disposal was necessary and could be safely implemented . . . .

"[But] current storage methods have an excellent safety record. There is no urgency to dispose of waste because of a need to correct an unsafe situation.

"The question of the timing of disposal in relation to safety rests on the reliability of the institutional controls that are required to maintain the safety of storage facilities. If such controls cease, storage facilities could no longer be expected to protect human health and the natural environment." (AECL, pp. 75-76, emphasis added)

The last statement, which is unreferenced, is the closest the proponent comes to giving a possible justification for the AECL Geological Burial Concept. However, it remains an unsupported, unexplored, unquantified assertion.

We believe that the Panel should require the proponent to describe a scenario or several scenarios -- perhaps including a worst-case scenario -- in which existing storage facilities might fail due to lack of institutional controls, thereby endangering human health as well as the natural environment. The proponent should be required to describe in broad outline the extent of the damage that might result and the nature and scope of the harmful health and environmental effects that would be experienced in such a hypothetical case. The proponent should also be asked to quantify the results of such a scenario or scenarios to the extent that is possible.

Such a description would be useful not only in gauging the need or justification for the AECL Geological Burial Concept (5), but also in clarifying the nature of the problem of nuclear fuel waste vis-ą-vis safety, as required by the Panel in Chapters 1 and 2 of the Guidelines: "a discussion of . . . the risks of the wastes in general" (AECL, p. 1).

(5) provided the AECL Geological Burial Concept is proven to be a genuine disposal option.


  • In the Guidelines, the Panel has asked the proponent to "define and explain" the overall problem of nuclear fuel waste, to delineate the moral and ethical framework within which the problem may be situated, and to state the need for permanent disposal.

    In the main AECL document, these tasks have not been carried out adequately.

  • Remedy

  • The Panel should require the proponent

    • to provide a self-contained description of the problem of nuclear fuel wastes from the perspective of human health and the natural environment, in clear and easy-to-understand terms;
    • to quantify the toxicity of a typical irradiated fuel bundle by specifying the amount of water needed (hypothetically) to dilute all contaminants therein to the maximum concentration currently permitted in drinking water, and to present this as a function of the "age" of the bundle: (1) upon discharge; (2) within the first year, month by month; (3) within the first decade, year by year; (4) within the first century, decade by decade; (5) within the first millenium, century by century; (6) thereafter, millenium by millenium,
    • to justify the need for a permanent disposal option, by specifying the possible harmful consequences to human health and the natural environment that might follow from a postulated failure of current containment methods;
    • to specify possible mechanisms that might lead to a failure of current containment methods due to a lack of adequate institutional controls;
    • to explain the importance and difficulties of proving, beyond a reasonable doubt, that any proposed disposal option would, if implemented, ensure the isolation of nuclear fuel waste from the biosphere, essentially forever;
    • to record and assess the divergent views of those who do not support the concept of geological disposal of nuclear fuel waste at the present time;
    • to delineate, and discuss in a comprehensive manner, the ethical and moral dimensions of the each of the following: (i) nuclear fuel waste production; (ii) nuclear fuel waste storage; (iii) nuclear fuel waste transportation; (iv) nuclear fuel waste reprocessing; and (v) nuclear fuel waste disposal;
    • to discuss the moral and practical implications of our society possibly making a mistake by approving a disposal concept now, only to find, after closure, that the predictions were wrong and that significant amounts of radioactive contaminants are escaping from the repository into the biosphere.



The Panel's requirements on the subject of moral and ethical concerns go beyond the context of Chapter 2 of the Guidelines, or Chapter 2 of the main AECL document.

"Ethical and moral perspectives, along with various social issues, as evidenced by presentations to the Panel at the scoping meetings, are as important as scientific, technical and economic considerations. The proponent should investigate how relatively narrow and focused considerations of a scientific, technical or economic nature should be viewed in the much broader context of ethical, moral and social considerations."

(Guidelines, Executive Summary, p. viii ~ emphasis added)

"Discussions and analyses on the scientific, technical, ethical, social, and economic aspects must be considered with the same degree of attention and rigour throughout the EIS. Also, the proponent must present its analyses & assessments of the concept or components of the concept in quantitative terms wherever applicable and appropriate."

(Guidelines, Preamble, p. xii ~ emphasis added)

But very little explicit consideration is given to moral and ethical considerations in the main AECL document. What discussion there is, is concentrated in pages 64-69, where it is stated that "The ethical aspects of nuclear fuel waste disposal, including the redistribution of risk, are discussed in detail in R-Public". (Neither R-Public nor any of the other R-docs has been translated into French, so francophone Quebecers, Franco-Ontarians, and other French-speaking Canadians are out of luck.)

Morever, the quality of the work done is highly suspect [ see example below].

In Section 6.3.2 of R-Public, on Distributional Equity, we read that

"Some of the groups involved in the Public Consultation Program, as well as focus group participants, thought that the public had a moral responsibility to help in arriving at a solution, and suggested that acceptance of a waste disposal facility at a safe site is a contribution that a community could make (Greber and Anderson 1989, Pieroni 1984)"
(R-Public, pp. 145-146)

An examination of the references (AECL TR-M-21 and AECL TR-471) reveals that the participants were in fact strongly opposed to the idea of having a nuclear waste disposal site anywhere near them. It was only in response to a direct question (in Pieroni 1984), and after much discussion against the idea, that some people said they would accept siting decisions "if there were sufficient assurances of safety". Furthermore, a reading of the participants' comments reveals that only very powerful assurances would be accepted, and even then reluctantly. The overall tone of the participants was very negative toward having anything to do with nuclear waste.

If AECL can so misrepresent the contents of its own documents, dealing with ethical and moral principles, may not the same true of its scientific and technical documents?

Continuing on in the same paragraph from R-Public, we read:

It was proposed that society in turn has a responsibility to support and compensate the community for accepting the facility (Pieroni 1986). Kasperson (1983) and Shrader-Frechette (1991) also suggest that providing benefits to the host community can help offset inequities resulting from some people having to bear the risks and detriments for the benefits of others. (AECL, p. 146)

However, on consulting the cited Shrader-Frechette article, "Ethical Dilemmas and Radioactive Waste: A Survey of the Issues"(Environmental Ethics v. 13 n. 4) it appears that her views have been seriously misrepresented. What she actually suggested, in the context of an above-ground, monitored storage facility for nuclear wastes (6), was that the community be given funding to "control health and safety monitoring at the facility" and that "those who store radwaste actually set up a fund for compensation of future persons possibly harmed by the waste."

(6) not an underground disposal facility, to which she is strongly opposed -- see her 1993 book, Burying Uncertainty: Risk and the Case against Geological Disposal of Nuclear Waste.

She goes on to say (p. 340-341) that
"Providing equity, informed consent, and full compensation regarding the risks associated with managing radwaste requires that we forego use of nuclear power and the generation of additional radwastes,"
and then remarks (p. 343)
"A thousand years ago, the world's finest architectural and engineering talents were mobilized to build cathedrals. It is ironic that comparable talents & even more skills are today dedicated to devising foolproof nuclear garbage dumps."

The quality of scholarship in R-Public is deplorable, raising serious ethical questions that go beyond any the proponent has discussed.

If AECL had followed the Panel's guideline to treat scientific, technical, ethical, social, and economic aspects with the same degree of attention and rigour, we would have to conclude that AECL science is very slipshod, with data poorly recorded & misapplied.

In R-Public, 126 out of 263 references -- that is, 46 percent of the total -- are from the nuclear industry, of which 57 are in-house AECL documents (7).

(7) Only 26 of the references in R-Public are from academic journals, of which the Shrader-Frechette article is one. There are 58 Government Publications, 38 Books, 20 Conference Proceedings. The rest are difficult to identify.

CCNR requested many of these AECL documents, and received about half of those requested in time for the conformity analysis. Many of the AECL documents in question (including the two previously mentioned) bear the following inscription:

"The Technical Record covers ongoing work done as part of the Nuclear Fuel Waste Management Program, and is intended for quick dissemination of the information. The information it contains may not yet have been evaluated rigorously, and any conclusions may be subject to modification in the light of further information. This report is not to be listed in abstract journals..."

Nevertheless, these "quick", unevaluated and non-rigorous reports are cited and used as authoritative references by AECL. Dothe same standards of scholarship apply to the technical references? (They should, according to the Guidelines.)


  • Moral and ethical considerations are not treated adequately anywhere in the main AECL document. They are not even mentioned in Appendix A (purportedly showing its conformity with the Panel's Guidelines) nor are any related terms found in the glossary.
  • Remedy

  • The new EIS [Deficiency #1, above] should incorporate a thoughtful and comprehensive consideration of ethical, moral and social considerations "throughout the document", as suggested by the Guidelines. These should include, but not be limited to, a discussion of:

    • other ethical frameworks in addition to the "utilitarian" one, including a discussion of traditional moral and ethical views of aboriginal peoples;
    • the ethical implications of producing highly toxic waste materials of great longevity which cannot be neutralized or rendered harmless;
    • the moral implications of asking the less privileged members of a society to accommodate toxic wastes produced by those who are more privileged;
    • the moral implications of making irreversible decisions for future generations in the face of uncertainty, given the potential for harm;
    • the morality of continuing to produce nuclear fuel wastes in the absence of a proven safe solution or a socially or politically accepted disposal mechanism;
    • the ethical problems related to the export of nuclear technology to countries less well equipped to deal with long term waste management and disposal;
    • the ethical problems associated with conflict of interest arising from the fact that AECL and Ontario Hydro are both promoters of nuclear technology;
    • the ethical problems associated with couching guesses, opinions or unproven claims in scientific language, thereby giving them a veneer of certainty which may be seriously misleading to the uninitiated;
    • the ethical requirements for scientists to recognize and document the weaknesses & uncertainties of scientific approaches as well as the strengths;
    • moral, ethical and social problems related to the mystification & helplessness that many people feel when confronted with highly technical arguments;
    • the ethical implications of setting "acceptable" standards for public exposure to toxic materials without consulting the population(s) at risk;
    • the moral acceptability of reassuring people that they are not at risk when in fact they may be potentially at great risk due to circumstances beyond anyone's control.



"The EIS should discuss and provide background information about the risks to the health of humans and human communities ... that are associated with nuclear fuel waste. The sources used to obtain this information should be identified."

Guidelines, p. 3

As previously remarked, the discussion of radiological health risks to humans is very poor in the main AECL document, occupying only three pages (a bit more than half a percentage point of the whole 496-page document).

The discussion of radiation-induced health effects is not only inadequate, but inaccurate and misleading as well, showing clear signs of bias.

Example 1:

"Such [deterministic] effects do not occur below a certain level of dose (the threshold), and above the threshold their severity increases with dose and usually with dose rate." (AECL, p. 36)

This unreferenced statement sounds reassuring, but it is far too categorical to be considered true. For someone familiar with the literature, it is unsettling to witness the ease with which AECL rushes to generalize when the evidence does not warrant it.

Here, from BEIR V, is a more cautious and scientifically accurate statement about deterministic effects -- in children irradiated in utero:

"The effects of prenatal irradiation ... include gross structural malformations, growth retardation, embryo lethality, and central nervous system abnormalities. Major anatomical abnormalities have been produced in all mammalian species by irradiation of the embryo during early organogenesis; ... and the evidence suggests that there may be a threshold for many, if not most, major malformations.

"In humans, mental retardation is the best documented of the developmental abnormalities following radiation exposure.... In those irradiated between weeks 8 and 15, the prevalence of mental retardation appeared to increase with dose in a manner consistent with a linear, nonthreshold response, although the data do not exclude a threshold in the range of 0.2-0.4 gray."

BEIR V 1990 : pp. 354 and 362

Example 2:

"Genetic effects may occur in the descendants of the exposed person, although this has not been firmly established" (AECL, p. 36)

The statement misleadingly suggests to the uninitiated that there is substantial doubt about the mutagenic effects of radiation. Compare the following one from BEIR V:

"Radiation has been found to be mutagenic in all organisms studied so far, and there is no reason to suppose that humans are exempt from radiation's mutagenic effects. These mutagenic effects are expected to be harmful to future generations because, in experimental organisms, the majority of new mutations with detectable effects are harmful.... Indeed, the harmful effects of mutations that occur spontaneously in humans are well documented, because many of them result in genetic disease."

(BEIR V, p. 66)

Example 3:

"It should be noted that studies of populations residing in regions of high background radiation have not shown conclusive evidence for an associated increase in risk of cancer (BEIR 1990)." (AECL, p. 36)

Why is this observation from BEIR V worth mentioning twice on the same page, in two different paragraphs? And why is it not worth noting by AECL that the same report, BEIR V, announced a three-fold increase in the risk estimates for radiation-induced solid cancers, and a four-fold increase in the risk of radiation-induced leukemia, compared with earlier estimates (BEIR III, 1980)?

Example 4:

". . . there is some experimental evidence that low doses of radiation may be able to stimulate the repair of prior radiation damage or strengthen the body's natural defense mechanisms. Most of the experimental data on such effects, termed "hormesis", have been inconclusive, mainly because of statistical difficulties at low dose rates." (AECL, p. 36)

Outside the nuclear power industry, belief in beneficial effects of ionizing radiation at low doses is by no means widespread. There is simply no comparison between the mountains of evidence that support the mutagenic and carcinogenic effects of radiation, even down to very low doses, and the relatively sparse and flimsy experimental evidence for hormesis. Hormesis isn't even mentioned in the BEIR V index. Why is it given such a prominent place?

The inaccuracies in these examples are troubling; the evidence of bias even more so. In every case, the proponent seems to be bending the evidence in a pre-determined direction -- to minimize the perception of risk. If the proponent has no compunctions about slanting the health data -- for whatever reason -- may one not anticipate a similar slanting of the geological, hydrological, engineering and performance data?


  • The presentation and discussion of the biological effects of ionizing radiation on human beings in the main AECL document is inadequate, inaccurate, and biased.
  • Because potential radiological impacts on the environment are the main concern over the very long term, it is essential that the principal concepts associated with ionizing radiation and its biological effects be presented in a balanced, thorough and comprehensible way.
  • Remedy

  • The new EIS [Deficiency #1] should have a balanced, informative and thorough discussion of the entire range of radiation-induced effects in human beings and in non-human biota, with special attention to those of long-term significance.

    • an explanation of the difference between ionizing and non-ionizing radiation should be given, including the effects of the former at the cellular level;
    • a discussion of particulate versus non-particulate radiation should be given, with special emphasis on alpha emitters and their distribution in the body;
    • a short history of the medical, experimental and epidemiological evidence showing carcinogenic effects of ionizing radiation going back to 1900;
    • a history of the evidence showing genetic effects of ionizing radiation;
    • a history of the evidence showing teratogenic effects of ionizing radiation;
    • a discussion of the evidence showing mental retardation in children exposed in utero, and the possible long-term implications of this;
    • a discussion of other potentially significant effects of radiation, such as sterility, blood disorders, effects on the immune system, reproductive organs, litter size, and animal behaviour, particularly mating and nesting behaviour;
    • a discussion of effects on in terms of stillbirths and

    • a discussion of radiation-induced illnesses to which pregnant women and children are particularly susceptible, e.g. stillbirths & thyroid disorders;
    • a discussion of radiation dose and relative biological effectiveness (RBE) should be given, with special attention to (1) alpha emitters, (2) factors affecting RBE such as dose rate, critical organ, species, age and sex.
    • a careful discussion of threshold versus non-threshold effects of radiation, reflecting the uncertainties that exist in the scientific literature, and bearing in mind that our understanding of this has evolved & will continue to evolve.
    • a discussion of the stages of acute radiation sickness, with a clear statement that these symptoms are only manifested after massive exposure to radiation.


  • The treatment of radiation-induced "fatal cancers" and "serious genetic effects" (AECL, pp.36-37) in the main AECL document, is inadequate.
  • The limitation of radiological health effects to "fatal cancers" and "serious genetic effects" contradicts AECL's definition of health and is nowhere justified.
  • Remedy

  • The new EIS [#1, above] should include a thorough discussion of radiation-induced cancers and genetic effects, which should include but not be limited to

    • a table indicating which cancers are considered "fatal" and which are considered "non-fatal" [since any untreated cancer can be fatal];
    • a discussion of non-fatal cancers, which can cause physical pain, psychological torment, loss of employment, and costly treatment
    • a matching between cancer sites and various radionuclides implicated in producing cancers at those sites (e.g. strontium-90 and bone cancer; iodine-131 and thyroid cancer; plutonium-239 and lung cancer; etc.);
    • an explanation that radiation-induced cancers and genetic effects are (technically speaking) population health effects, which depend on the population dose, and not individual effects depending on individual dose;
    • an operational definition of genetic effects -- how do you identify them? -- and especially of "serious genetic effects", with examples;
    • the number of generations considered to be at risk by a "serious genetic effect" according to the AECB regulations, and the rationale for this;
    • the potential long term effects, both qualitative and quantitative, of damaged recessive genes, whose manifestations may not show up for generations;
    • the potential long term implications of increased impairment of human and non-human intelligence from chronic radiation exposure for generations;
    • a qualitative and quantitative discussion of the types of possible genetic effects of complex etiology and their potential significance [e.g. human ailments which have genetic linkages, such as Down's syndrome, hemophilia, muscular dystrophy, heart disease, diabetes mellitus, schizophrenia, etc];
    • a qualitative and quantitative discussion of the long-term consequences of radiation-induced genetic effects in birds, fish, mammals, insects, and plants;
    • an informed discussion of the possibility that microbes in the vault may adapt under mutagenesis to exploit the new materials introduced into the vault, including some of the radionuclides themselves, and the effects this might have on the dissolution of the waste form and the transport of contaminants;
    • an informed discussion of the possibility that radiation-induced mutagenesis in microbes or their vectors over very long time periods may lead to new strains of disease to adversely affect human and/or non-human populations.



"The discussion should include ... processes and mechanisms through which radionuclides (atoms which release radiation) and other contaminants may directly and indirectly impact on humans and various organisms in the natural environment."

Guidelines, Ch 2, p.3

The main AECL document discusses chemically toxic elements very slightly, and radioactively toxic elements almost not at all (except for Figure 2-10) .

About chemical toxics, the main AECL document states that

"These elements have a broad range of toxicities and environmental behaviours. ... The toxicity of an element is closely related to the quantity ingested or inhaled, to its chemical form or oxidation state, and to a host of environmental variables." (AECL, p. 42)

It goes on to mention examples of target organisms, critical pathways, accumulation in food chains, teratogenic birth defects (not mentioned at all in connection with ionizing radiation), "subtle changes in animal health", and effects at the population and ecosystem level. To read the main AECL document, you would never know that these considerations apply equally well, and with even greater force, to the large inventory of diverse radioactive poisons contained in the nuclear fuel waste.

Figure 2-10 (AECL, p. 37) shows the ingestion radiotoxicity of selected radionuclides in used CANDU fuel. While this information is interesting and relevant, it would be easier to understand if

  1. comparative toxicities of the chemically toxic elements and the radiotoxic elements were specified;
  2. the total toxicity of one unit of used CANDU fuel were given; and
  3. the toxicities were expressed per bundle instead of per gram.

Of course, it is inhalation and not ingestion that is the most dangerous pathway for some of the nuclides appearing in Figure 2-10 -- in particular the plutonium isotopes. AECL should illustrate their inhalation radiotoxicity in a companion figure.

Anther measure of ingestion radiotoxicity, easier for many people to grasp and more relevant to a saturated geological repository, is the amount of water required to dilute the contents of one bundle of used CANDU fuel to the maximum permissible concentration (of all contaminants combined) currently allowed in drinking water (8). [See Deficiency #6, Remedy #2] Again, a comparison between toxicities of the chemical toxics and radiotoxics, and a comparison between the toxicity of selected nuclides and the toxicity of the whole bundle would be appropriate.

(8) This is the measure that was used by the U.S. Geological Survey in its 1978 publication "Geologic Disposal of High Level Nuclear Wastre : Earth Science Perspectives" as well as by the Ontario Royal Commission on Electric Power Planning in its 1978 Report "A Race Against Time".
But this is only scratching the surface. The degree of hazard associated with a radionuclide depends in an important way on its chemical and physical properties (e.g. iodine is a vapour that can be inhaled; plutonium oxide is insoluble so it stays in the lung) as much as it does on its theoretical radiotoxicity.

Thus radioactive iodine is particularly dangerous to unborn babies and young children because it first concentrates in milk and then accumulates in the milk-drinker's thyroid gland, where it can trigger a host of thyroid disorders including stunted growth, mental retardation, non-malignant thyroid nodules, and -- of course -- thyroid cancer. None of this is mentioned in the main AECL document, even though iodine-129 is one of the most important radionuclides in the post-closure assessment scenarios (AECL, p. 309).

Carbon-14, another radionuclide of critical importance in the post-closure assessment scenarios (AECL, p.309) is another story. It does not concentrate in food chains or accumulate in specific organs, but -- because carbon is the basis of organic life -- it does get incorporated directly into organic molecules of all kinds, including DNA and RNA molecules. Although carbon-14 is a very weak beta emitter, whose radioactive emissions cannot penetrate through more than a few layers of cells, its long-term potential for doing genetic harm is still largely unknown. But it isn't discussed.

Radioactive decay chains can also play a significant role in determining the actual hazard associated with a given radionuclide. For example, radon gas -- which is an alpha emitter -- produces a number of solid radionuclides called radon progeny, some of which are also alpha emitters. In an enclosed space, the radon progeny attach themselves to tiny dust particles in the air, and when inhaled, these prove to be extremely potent carcinogens -- far more potent than anyone would believe just by looking at the inhalation toxicity of radon gas alone.

Strontium-90, which concentrates in bones, teeth and milk, offers another example. When strontium-90 undergoes radioactive decay it becomes yttrium-90, with a different chemistry altogether; it tends to go to the fatty tissues, including the gonads.

A particularly appropriate case study of radioecology at work, in the context of the Canadian North, is the well-known concentration of radioactive cesium in the lichen-to-caribou-to-man food chain, leading to unexpectedly high levels of radioactive fallout (comparatively) in the bodies of native caribou-eaters. The same food chain has resulted in northern meat-eaters having up to 80 times the levels of naturally-occurring polonium-210 in their bodies as those in the Canadian South.

The proponent has not adequately described the radioactive materials contained in the nuclear fuel waste from an environmental perspective.


  • The inventory of radionuclides in the irradiated fuel and the zirconium cladding is nowhere described in a biologically meaningful way.
  • Appendix E gives only incomplete physical information about each radionuclide (e.g. is it an alpha, beta, or gamma emitter? or some combination of these? is it fissionable? or fissile? One cannot tell from Appendix E.)
  • For a mix of toxic chemicals, we would need to know the physical, biological and ecological characteristics of each; the same applies to radiotoxic nuclides. Each has its own unique physical, biochemical and radioecological properties. The hazard potential depends on a combination of all these properties.

    An acceptable EIS should ensure that such information is not only available, but well presented with careful explanations, illuminating examples, and a discussion of gaps in current knowledge.

  • Remedy

  • A complete list of radionuclides in the irradiated fuel and the zirconium cladding should be given, preferably in the body of the new EIS [Deficiency #1, above] , together with the most significant physical, biological and ecological features of each one, such as

    • the type of radiation emitted together with energy levels and half-life;
    • whether it is a fissionable or fissile radionuclide;
    • its principal decay chain(s);
    • critical organs and retention times;
    • maximum permissible body burden;
    • maximum permissible concentration in drinking water;
    • relative importance of inhalation, ingestion in drink or in food, absorption through skin, groundshine, or other routes of exposure;
    • potential for bioconcentration and/or biomagnification in food chains, with examples where these are known and significant (e.g. the lichen-to-caribou-to-man food chain in the Canadian North);
    • relative degree of scientific knowledge of the specific biomedical and bioecological properties of each radionuclide, referenced, ranging from "thoroughly studied" to "little or no information available".
    • any particular biological concern(s) associated with a radionuclide other than cancer and genetic damage (e.g. thyroid disorders induced by radioiodine; fibrosis of the lungs induced by inhaled alpha emitters).



"The proponent should address the items and requests identified in these guidelines ...."

Guidelines, Ch. 1, p. 2

We are still at the beginning of Chapter 2 of the Guidelines. Evidently the main AECL document is not even close to being in conformity with the Guidelines. It is impossible to go through every point with the same degree of care and attention in the time available, but here are some other observations on Chapter 2 requirements.

We use a scoring system from 1 to 10, the highest score being complete conformity. In our scoring, we are trying to be generous to the proponent.

From Section 2.1 of the Guidelines

					 	    DEGREE OF
						   (OUT OF 10)

Definition of health, including 
community and social health (9).			4

Definition of risk and analysis of the viewpoints of various public groups, including aboriginal peoples, on risks (10) 2

Processes and mechanisms through which radionuclides and other contaminants may directly and indirectly impact on humans and various organisms in the natural environment 3

Explanations of effects of ionizing radiation on humans and the natural environment, including probability of exposure and quantification of risk (11) 2

Current and propopsed health regulations pertaining to ionizing radiation, past changes in these regulations, and reasons for those changes 1

Methodologies used in risk assessment, and in health assessment, their validity including a discussion of assumptions and the theoretical justification 1

Theoretical justification for the manner in which probability and magnitude of risk are calculated 1
Total 14
(Out of 70)

(9) AECL's definition of health does not include social and community aspects, and even AECL's own limited definition isn't followed in discussing radiological health effects.

(10) For AECL, risk is based on fatal cancers and serious genetic effects -- neither of which is adequately defined; the basis for the numbers given is unexplained; there is no real discussion of other views.

(11) See Deficiencies #8 and #9. There is no explanation of "population dose" in relation to stochastic effects, nor is there any discussion of non-conservative departures from linearity in cases of low-level exposure to alpha emitters.

From Section 2.2 of the Guidelines

					 	   DEGREE OF
						   (OUT OF 10)

Nature, magnitude, and origin 
of nuclear fuel waste problem (12)			5

Types, sources, quantities and locations of nuclear fuel waste, present and future, with uncertainties and reason for uncertainties, and various circumstance that might affect the situation (13) 7

Physical and chemical characteristics of nuclear fuel waste for all relevant time scales, including prominent radionuclides and their probable chemical form 2

Nature of physical and chemical changes that occur in nuclear fuel in the reactor, after removal, during storage or transportation (14) 6

Variations in characteristics of the components of nuclear fuel waste including heat, radiation intensity, radiation products and toxicity (15) 6

Possible changes in nature and characteristics of nuclear fuel wastes due to changes in nuclear power technology or energy policies (16) 2
Total 30
(Out of 60)

(12) There is no acknowledgement of any problem with nuclear fuel waste, except on page 389.

(13) There is no mention of the high level liquid radioactive waste buried at Chalk River; there is no discussion of uncertainties or reasons; there is no sufficient discussion of reprocessing although specifically requested by the Panel.

(14) Chemical changes are not discussed; physical changes are described only superficially. There is no adequate discussion of fission gasses and volatile radionuclides, physical and chemical defects in cladding, etc.

(15) Heat and radiation intensity are presented graphically, pp. 24-25, although more discussion would be appropriate; discussion of radiation products and overall toxicity not adequately done at all.

(16) There is no discussion of AECL proposals to 1) burn weapons grade plutonium in CANDU reactors; 2) import nuclear waste from other countries; there is no adequate discussion of advanced fuel cycles.


From Section 2.3 of the Guidelines

					 	    DEGREE OF
						   (OUT OF 10)

Current nuclear fuel waste management, 
objectives, ability to meetobjectives, 
environmental and ethical dimension 
of those practices					7

Performance of nuclear industry in managing nuclear fuel waste; containment failures; subsequent environmental impacts; history of experience with containment design and construction materials; comparison between Canada and other countries (17) 8

History of nuclear fuel waste management from beginning; preference for disposal versus storage; changes; other views (18) 2
Total 17
(Out of 30)

(17) Leakage of high level radioactive wastes from steel drums at Hanford, off-site contamination at West Valley, problems in Russia, the U.K, France etc. should be worth a mention. It could be pointed out that since Canada does not reprocess, we don't have such a mess here.

(18) No real history is offered, no indication is given that views have ever changed. Criticisms of AECL by (e.g.) Select Committee on Ontario Hydro Affairs (1980) are not reported or discussed.

From Section 2.4 of the Guidelines

					 	    DEGREE OF
						   (OUT OF 10)

Risk considerations on the basis of 
total population and individual (19)			4

Risk from ionizing radiation on the basis of physical and genetic effects 4

Risk criteria related to human health and environmental protectionand assumptions taken in development of these criteria 2

Distinction between risks due to current management of nuclear fuel waste and background radiation (20) 6

Risks resulting from social processes (e.g. accidental intrusion or criminal intervention), geological processes, and changes in the surface environment 6

Risks from possible accidents (e.g. unexpected major leakage) 2

Total 30
(Out of 60)

(19) Individuals are treated as entirely hypothetical -- i.e. "reference man" -- and risk discussions are based entirely on regulations. One seldom asks if "reference man" is pregnant -- or Inuit, for example. Total population impacts are never explicitly discussed except for non-humans (e.g. AECL, page 40).

(20) There is no discussion of the fact that the background level of most of the radioactive poisons in irradiated nuclear fuel was essentially zero before the advent of fission technology -- e.g. strontium-90 & cesium-135 were not problems prior to nuclear testing & nuclear reactors.)

From Section 2.1 of the Guidelines

					 	    DEGREE OF
						   (OUT OF 10)

Security for long-term management 
of nuclear fuel waste, in Canada and 
internationally, including loss of 
knowledge or nuclear expertise;
history and future of such provisions			1

Total 1
(Out of 10)

By this reckoning, which we truly believe to be generous to the proponent, we get a grand total score of 78 out of 160, or 48 percent. A dismal failure. We will be glad to review the marks and boost them if warranted.

How could AECL, a first rate nuclear research and development organization that has done so much fine work in the past, get such a miserable grade?




At this point,
we decided it was fruitless to try to follow the Guidelines
since the proponent has in large measure ignored them.

The remainder of this conformity analysis
deals with Mathematical Modelling
and the nature of the scientific and mathematical assumptions
made by the proponent
in attempting to justify the Geological Burial Concept
as a Permanent Disposal scheme.



"The EIS should discuss the selection and development of the mathematical models and other methods used in the performance assessment of the generic Multiple Barrier System."

Guidelines, 6.1, p. 23

"The EIS should discuss the procedures chosen to validate the models ..."

Guidelines, 6.3, p. 27

"The proponent is encouraged to present its EIS in the clearest terms possible, and to use language, where possible, that can be readily understood by the public. However, where the complexity of issues requires the use of technical or uncommon words, and words and phrases which may otherwise be interpreted in various ways, a glossary clearly defining terms should be included."

Guidelines, Introduction, p.2

Terms related to mathematical modelling are not included in the glossary in the main AECL document, even though these words can be extremely misleading if not properly defined and explained.

To the uninitiated, words like "Mathematical", "Verification" and "Validation" may suggest a degree of near-absolute certainty that is seriously misleading. Indeed, the very use of such language raises important ethical questions having to do with the proper role of science, mathematics and computers in our society.

In days of yore, people were often disenfranchised because of their gender, race or religion. Today, millions of people run the risk of being disenfranchised because of a lack of scientific aptitude or training.

Computers did not exist sixty years ago. Although most people have some experience with user-friendly applications, like video games or word processing, very few have any understanding of what computer programming is, what a mathematical model is, or -- for that matter -- what SCIENCE is.

The proponent has a moral and ethical obligation to explain both the strengths and weaknesses of computerized mathematical models, and to illuminate the technical meanings of the words validation & verification.


  • The principals of mathematical modelling are not adequately explained in simple language so that members of the public can understand the usefulness and limitations of this approach.
  • Remedy

  • In the new EIS [Deficiency #1] the proponent should be required to give a thorough discussion of the practical, moral and ethical aspects of mathematical modelling. Among other things, this treatment should include

    • a clear explanation of the scientific method -- observations are made, whence hypotheses are formulated, whence predictions are made, whence experiments are designed to invalidate -- or validate -- those predictions;
    • a clear explanation of the hypothetical nature of a mathematical model, and the consequent need to try to validate or invalidate a model in the same sense that we try to validate or invalidate any other scientific hypothesis;
    • simple examples of "good" (valid) and "bad" (invalid) mathematical models as illustrations, possibly including anecdotal examples from AECL's and/or Ontario Hydro's past experience where appropriate;
    • a discussion, with examples, of models that give incorrect results because of (1) incorrect data; (2) "bugs" in the code; (3) conceptual errors in the model;
    • an explanation of the vastly different behaviour that can be experienced by iterating linear models as opposed to non-linear models, with a brief explanation of so-called "chaotic behaviour";
    • an explanation of the special difficulties posed by modelling discontinuous processes as opposed to continuous processes;
    • an explanation of the dangers inherent in extrapolating a model's results beyond the range of parameter values within which the model has been experimentally validated;
    • an explanation of the fundamental difficulty of knowing whether or not a computer code has been fully de-bugged, with anecdotal examples of "bugs"that have lain dormant for years before showing up;
    • an explanation of the differences in predictive ability of mathematical models in
      • (a) the physical sciences;
      • (b) the biological sciences;
      • (c) the social sciences.



"The EIS should define all parameters and categories of data used in the generic mathematical models. As well, the EIS should discuss the methods used for determining these parameters and categories of data, and how they are used in the models."

Guidelines, 6.2, pp.25-26

Every computerized model starts with data -- numbers that have to be "fed" into the computer. These are the "parameters"; they are also called "input" values.

If the parameters are wrong, the results are worthless. It is very important to know just how good the parameter values are. One of the oldest saying in the computer business is: "GARBAGE IN EQUALS GARBAGE OUT"

The proponent does not claim to know the exact values of the parameters, because there is a considerable amount of uncertainty and variability in those numbers. For example, how fast is the water flowing through the rock? It depends on many things.

Instead, in some cases, the proponent assigns a RANGE OF VALUES to a parameter, by choosing two extremes -- the maximum and the minimum. Any of the numbers between these two extremes could be used as a value for that parameter; but any number outside that range would never be used. The question arises : how does the proponent know what the correct range of values is?

But that's not all. The proponent may also assign a PROBABILITY DISTRIBUTION to the parameter. The probability distribution of a parameter is very important, because it dictates which values of the parameter are considered most likely to be the right ones, and which values are the least likely. The question arises : how does the proponent know what the correct probability distribution is?

In actual fact, there is a lot of experimental data to draw upon -- careful measurements that have been made -- but there is also a lot of guesswork that has to be done.

The proponent has a moral and ethical responsibility to "put all the cards on the table" by revealing what assumptions have been made about the parameters, in terms of the range of values and the probability distributions.


  • The various assumptions that have been made about the parameters in each of the mathematical models used for post-closure assessment have not been explicitly stated in the main AECL document .


  • In the new EIS [Deficiency #1] the proponent should be required to provide a map of all the computerized models that play a role in the post-closure assessment. The map should indicate how the models are interconnected.
  • For each submodel, the proponent should be asked to list all of the input parameters in a table, with a brief explanation of what each one means. In the table, the probability distribution or range of values for each parameter should be specified, with an indication of how it was arrived at :

    • is the range of values known with a very high degree of certainty, based on extensive experimental or field data (please specify which)?
    • is the range of values based on a rough, orders-of-magnitude estimate, utilizing sparse or highly variable experimental or field data?
    • or is the range of values an informed guess, based on little or no data?

  • Similar questions apply to the SHAPE of the probability distribution (in other words, where the "most probable" values are assumed to be):

    • is the shape of the probability distribution known exactly, or nearly so, based on extensive experimental and/or field data (please specify which)?

    • is the shape of the probabiliy distribution known only very roughly, based on limited experimental or field data
    • is the shape of the probability distribution an informed guess, based on theoretical considerations unsupported by experimental or field data?
    • or is the shape of the probability distribution a simple guess based on no specific theoretical consideration or data from the lab or the field?



"The assumptions made in the model development and the resulting limitations inherent in the models should be discussed."

Guidelines, 6.12, p. 25

"The discussion should include ... the coupling among physical, hydrogeological, chemical, biochemical, and geomechanical processes and mechanisms."

Guidelines, 6.11, p. 24

"The discussion should include ... the representation of fracture systems in the rock mass, the associated groundwater flow systems, and the channelling of groundwater flow within the fracture systems."

Guidelines, 6.12, p. 25

For each computerized model that plays a role in the post-closure assessment, the proponent should be required to identify all major assumptions of either a mathematical or scientific nature.

Of particular importance are assumptions where linear models are used to represent or approximate processes which are probably not linear; these should all be identified.

Additionally, the proponent should identify where and how discontinuities are represented in the mathematical models, and where they are simply ignored; for example, the interface between grout and rock; the welds or joints in the containers; the existence of joints and fractures in the rock;

The proponent should discuss the over-riding assumption, which may not be true, that geological events and processes are random and therefore can be represented mathematically by probability distributions.

In addition, the proponent should identify in a table all potentially relevant couplings between different processes, and indicate whether or not -- and if so, how -- that coupling has been represented in the mathematical models.

A coupling refers to a linkage between different phenomena; for example, swelling caused by the heat of the nuclear waste could cause new fractures to form, or heat from the waste could change the flow characteristics of the groundwater. Thermomechanical, thermohydrological, thermohydromechanical, and thermochemical couplings are thought by some to be more important than the others; however a geological panel report in 1984 discussed 11 potential couplings and attached levels of significance to subprocesses within each.


  • In the main AECL document, the scientific and mathematical assumptions -- that are built into the various computerized mathematical models used in the post-closure assessment -- are not made clear.
  • Remedy

  • All mathematical & scientific assumptions that have been made about processes and mechanisms in constructing various mathematical models should be laid out, preferably in an organized tabular format.
  • There should also be an accompanying explanatory text to discuss the justification for & possible significance of those assumptions.
  • In particular, the proponent should identify all examples of, and discuss in an organized and comprehensive way, each of the following :

    • the use of linear models to model non-linear behaviour;
    • the way in which discontinuities are incorporated ( or not) into the models;
    • the way in which couplings are incorporated (or not) into the models;
    • the use of probabilistic assumptions in dealing with parameters and processes which may not be random;
    • over-riding assumptions (e.g. is it assumed that the imbalance in the temperature field due to the emplaced waste, combined with the imbalance in the stress field of the rock due to excavation, do not, either separately or in combination, cause any significant new fractures to form?)



"The discussion should include ... the validity and reliability of long-term performance modelling in general, in view of the limited experience and the limited validation that is possible."

Guidelines, 6.52, p. 30

When a mathematical model is used to predict changes over a 10,000 year period, one can be forgiven for being somewhat skeptical.

If we cannot predict the weather tomorrow, or the energy needs ten years from now, or the cost of an engineering project that is being built today, how can we hope to predict 10,000 years into the future?

If we will not live long enough to check out the accuracy of our prediction, what are the moral and ethical implications of acting on that prediction? Will future generations marvel at our forethought or curse our stupidity and our pride?

"Model validation might be better understood as model evaluation (Dormuth 1993), because direct observation of the behaviour of a disposal vault for thousands of years is not possible. Model evaluation would consist of comparing hypotheses and predictions of models of system components with observations of real systems and comparing model predictions with predictions of independently developed models." (AECL, p. 292)

The proponent is playing with words instead of facing the problem . Checking one model's predictions against another's accomplishes little except to confirm that "great minds think alike", or that "fool's minds run in the same channel". Trouble is, we don't know which.

We need to check our prediction against nature, not against each other.


  • The proponent should explain clearly why it is so difficult to validate all of the mathematical models that are used in the post-closure assessment. The proponent should also attempt to define what level of certitude we can hope to accomplish, and suggest an objective criterion to be used for deciding to accept the prediction or to reject it.

    In this way we have come full circle, back to Deficiency #2.



"The evaluation and screening of the potentially important factors indicated that many factors need not be explicitly included in the significant scenarios. In this section, we discuss what we consider to be the most important of these factors: criticality, microbial action, gas generation, glaciation, topographic change, biosphere evolution and climate change, open exploration boreholes, earthquakes, and meteorites." (AECL, p. 279)

The factors mentioned in this list from the main AECL document are disruptive developments that could seriously challenge the proponent's contention that the Geological Burial Concept will be safe for at least 10,000, and probably 100,000 years.

AECL has looked briefly at each of these possibilities, and has decided that none merits inclusion in scenario studies. In a few brief pages -- from page 279 to page 284 -- we can see the proponent working in the absence of computerized models.


The first item on the list is accidental criticality -- that is, a spontaneous fission reaction, turning the underground waste repository into a kind of nuclear reactor.

A self-sustaining nuclear fission reaction requires a "fissile element". In the case of irradiated nuclear fuel, the most important fissile element is plutonium-239. (The other one is uranium-235, but it has been depleted by fissioning in a CANDU reactor.)

"... an analysis of used fuel in a disposal container showed that criticality would not occur, whether the used fuel remained intact or was uniformly distributed throughout the container (McCamis 1992). The low porosity of the material surrounding the container would help to ensure that fissile material could not accumulate outside the container in a critical configuration. Moreover, the dissolution rate of the fuel is expected to be so low that the 239Pu present in the fuel would have decayed long before any such accumulation would be possible (R-Vault). Thus criticality would not occur because non-fissile materials in the fuel and its surroundings would prevent it." (AECL, p. 280)

The proponent does not explain what criticality is or what it would mean to the analysis. Would it add a large and unexpected amount of heat to the repository? Would it create steam and possibly melt some of the components of the irradiated fuel? Would it create new unanticipated stresses that might cause a deterioration in the quality of the vault and/or the surrounding rock mass?

The only reference provided for the proponent's assertion is an internal AECL document, which -- judging from the description -- does not prove that accidental criticality is impossible. (It deals only with two particular configurations.) Before criticality is discarded as a possibility, based on what may be spurious and incorrect arguments, should there not be a serious study of the consequences of being wrong?

Since plutonium is chemically different from uranium, there is a theoretical possibility that chemical processes could concentrate plutonium preferentially, eventually leading to a critical configuration. Morerover, when plutonium-239 decays it becomes uranium-235 -- another fissile material -- so that the proponent's reasoning at the end of this short passage is quite misleading.


"... even deep subsurface environments may contain microbial populations. However, the geosphere is nutrient-poor; thus microbial activity would be limited. Those microbes present would likely form biofilms in open fractures, which would tend to sorb radionuclides and thereby retard movement. Biofilms may also compete with colloids for the sorption of radionuclides, which would further reduce the effect that colloids are likely to have." (AECL, p. 280)

There are no references given to support these statements, which sound entirely conjectural. It is difficult to escape the impression that AECL doesn't know how to study this problem, and instead of saying so, is using what may be spurious and incorrect arguments to justify not doing the work.

Living things are incredibly adaptable. Microbes may be attracted to the untypical materials of the vault and shaft of the repository, not to mention the nuclear fuel waste itself. It is possible that our disruption of the subsurface environment may have unwittingly produced conditions more favourable to microbial populations.

There may also be mechanisms by which microbes can transport radionuclides within their bodies -- radionuclides that would have otherwise been sorbed -- or alter the chemical form of the radionuclides so that predictions made in the absence of microbes will not be borne out. In any event, such possibilities should not be rejected out of hand based on little more than inadequate scientific knowledge.

"... microbes are an integral part of the nutrient-rich biosphere, and their effects on radionuclide movement and dispersion in the biosphere have been implicitly included in the system model. The model is based on field and laboratory studies performed under natural conditions where microbes are allowed to thrive ...

"... methylation is expected to be of significance only in the biosphere. This process has been implicitly included in the system model (R-Biosphere)" (AECL, p. 281)

Is there any scientific meaning to the expression "implicitly included"? The Panel should ask AECL to refrain from using such language or to provide references.

If the repository leaks in an unpredicted manner, the supply of radionuclides to the biosphere will, for thousands of years, be quite untypical of the world we live in now.


Here again, as in the previous two examples, the proponent seems more interested in removing potential objections to the Geological Burial Proposal than in dealing with the concern in a straightforward manner.

This is consistent with our conclusion, in Deficiency #1, that the main AECL document is more of an Engineering Feasibility Study than an Environmental Impact Assessment.

Environmental concerns are treated more as obstacles to be overcome in implementing an Engineering Project rather than questions to be explored seriously in an objective search for the best possible answers.

"The disposal container specified for the postclosure assessment case study is the same as the titanium-shell packed-particulate container specified for the preclosure assessment case study ... except that the basket is made of a material that does not generate significant quantities of hydrogen gas upon corrosion. It would be feasible to make such a container from a ceramic material, for example." (AECL, p. 277)

"Note that the basket ... is made of a material that does not generate significant quantities of hydrogen gas upon corrosion." (AECL, p. 281)

This example also illustrates the need for a more precise definition of the concept, as in Deficiency #2. It isn't acceptable to keep changing the details of the concept so as to avoid the need to answer one objection or another.

If the Panel approves the concept, and implementation proceeds, the basket could end up being made of anything. We would like to know, and we believe the Panel should demand to know, what might be the consequences of "significant quantities of hydrogen gas generation" ?


The proponent informs us, giving no reference for the statements, that

"... extensive glaciation of the Canadian Shield is not expected within the next 10,000 years ... [and] major changes in topography are not expected within the next 10,000 years." (AECL, p. 281)

A generous-minded individual might be inclined to believe that these statements are based on solid scientific information that is not being divulged. However, the following two paragraphs in the main AECL document might shake that belief....


"... changes in climate, surface-water flow patterns, soils and vegetation types within the next 10,000 years are expected to be within the ranges currently observed....

"To the extent that acid rain, the destruction of the ozone layer, and global warming resulting from greenhouse gases have already affected the biosphere, they are implicitly reflected in the system model....

"Future interactions of humans with the biosphere are unpredictable. We have assumed they will not alter the biosphere in any fundamental way over long periods of time...." (AECL, p. 282)

None of this inspires confidence. So many camels are being swallowed in one mouthful that the credibility of the entire exercise is called into question. In addition, it seems unprofessional, unscientific, and ethically wrong to say that factors which are explicitly not included in a mathematical model are "implicitly included" just because the real world is experiencing these influences and therefore the data used as input parameter values for the model have been conditioned by those factors.

Such statements not only reflect adversely on the proponent's credibility, but call into question the level of professionalism and intellectual honesty which is brought to the task..


"When a borehole was no longer to be used for monitoring, it would be sealed immediately. Drilling and sealing records would be compared to ensure that no exploration boreholes were left unsealed.... If a borehole were left open 5 metres from a disposal room, the resulting dose would be less than 10 percent of the dose associated with the risk criterion specified by the AECB in R-104." (AECL, p. 282)

There are thousands of unlogged boreholes in the Canadian Shield. The proponent proposes to fill only those that it knows of. This is hardly a sufficient reason to exclude the possibility of unplugged boreholes from the scenarios.


"Earthquakes of significant magnitude are associated with faults, which are linear fracture systems along which shear movement takes place.... Moreover, almost all earthquakes on the Shield of magnitude greater than four occur near rifted areas, near Shield margins, or along very large faults.... Atkinson and McGuire (1993) estimated the probability that fractures caused by earthquakes would reach the disposal vault.... The estimates indicate that this probability is less than 10-- 8 per year provided the vault is * not within one kilometre of an active fault more than two kilometers long; * not within 200 metres of an active fault more than 500 metres long; * not within 50 metres of any active fault." (AECL, p. 283-4)

Contrast that statement with this one from Allison L. Bent, writing in the Bulletin of the Seismological Society of America (v. 84, n. 4):

"On 25 December 1989 the largest earthquake in northern Quebec in at least 60 years occured in the Ungava peninsula. This earthquake was unique in that it marked the first time surface faulting could be confirmed for a historical eastern North American earthquake.... The Ungava earthquake and many interplate events from several continents share a number of characteristics, including shallow hypocenters, source complexity ... and perhaps most disturbingly, epicenters in unexpected locations on faults that even in retrospect could not have been recognized as active faults prior to the earthquakes." (from the Abstract and Conclusion)

This earthquake in the Canadian Shield was totally unexpected and unpredictable. One should not overlook the Saguenay 1988 earthquake in this context either.

It should be borne in mind that earthquakes are not truly random events, and therefore probabilistic arguments are highly suspect. The strength in the rock of the Canadian Shield permits high levels of unrelieved stresses to exist for very long periods of time. In the event of a nearby earthquake, is it not conceivable that the weakening of the rock caused by excavation of the repository might make earthquake damage more extensive than it might otherwise have been?

More to the point, do any of us really understand what we are talking about?


"We estimated the impact of meteorite impacts for which these effects could disrupt a closed disposal vault with an area of 4 km2 at a depth of 500 m, We used the relationship between probability of impact, crater diameter, and depth of the effect given by Grieve and Robertson (1984). The estimated probabilities were less than 10-- 8 per year for each of these effects." (AECL, p. 284)

Although meteorite impacts are arguably more probabilistic in nature than earthquakes, it is also true that no probability argument can be legitimately used to prove that something that can happen is not going to happen.

It would be worth exploring what the effects of a meteorite impact could be; it may provide added insights into the Geological Burial Concept.

AECL seems overly fearful of analyzing scenarios in which something really bad goes wrong.


  • In the main AECL document, a number of important factors are excluded from the significant scenarios without adequately referenced or reasoned arguments to justify those exclusions.
  • Remedy

  • The proponent should be required to deal with the consequences of factors which could have significant implications for the Geological Burial Concept, regardless of the proponent's opinion that they are unlikely to come into play in the next 10,000 years.
  • The possible consequences of each of the following factors, over the next few millenia, should be explored in sufficient detail to understand the implications

    1. for the buried nuclear fuel waste, and

    2. for the assumptions on which each of the computerized mathematical models used in post-closure assessment is based
    • accidental criticality in the vault;
    • microbial action;
    • gas generation;
    • glaciation;
    • topographic change;
    • biosphere evolution;
    • climate change;
    • open boreholes;
    • earthquakes;
    • meteorites.


[ Deficiencies Identified by the Scientific Review Group ]

[ Radioactive Waste Sub-Directory ] [ COMPLETE DIRECTORY ]

Since March 27th 1996, there have been over
100,000 outside visitors to the CCNR web site, plus

(counter reset July 2nd 1998 at midnight)