Panel Report

10.1 General Principles and Comments

A major concern of northerners is that mining activity might interfere with their traditional use of the land. They want to know that it is safe to drink the water, to fish, to trap, to hunt and to harvest plants, both now and in the future when the mines have ceased operating. To ensure that the environment is safe for these activities, an objective assessment of mining impacts is essential. This requires collecting baseline data to determine the state of the environment before mining begins, observing changes that occur during mining, and monitoring the recovery or restoration of the environment after decommissioning. Conceptually, these procedures are easy to design, but in practice they are often difficult to implement.

People are very concerned about their land and their lakes and they want to protect them because that is where they make their living from and they want to live there the rest of their lives.

Emile Hansen, Chief of the Hatchet Lake Band, Transcript of Public Hearings, La Ronge, Saskatchewan, October 1, 1996, p. 31.

10.1.1 Location of Sampling Sites

Collection sites should be arranged along the predicted concentration gradients of emissions from a point where contaminant concentrations are highest to a point where they approach background levels. Three sites per gradient, representing possible high, medium and low impacts, would be sufficient. In addition, another set of collection sites should be placed where predictions indicate that no impacts would occur. These control or reference sites are necessary because ecological systems may change through time. Impacts are assessed by comparing the control and potentially impacted sites with one another and to their respective baseline (i.e. pre-impact) states. Replicate samples should be collected at all sites.

Although the spatial arrangement of the collection sites might seem simple to establish, practical difficulties frequently interfere with the ideal statistical design. Matching of sample sites is important, but undoubtedly there will be variation in site factors such as size and depth of lake, the substrate type, and the communities of organisms that live in an area. For example, a lake receiving mill effluent might be shallower and smaller than lakes further down the watershed, making precise matching of sites or the biota to be monitored impossible. Inevitably, the design of a study will be compromised to some extent, and judgment will be required to determine the best location of sample sites.

10.1.2 Components to be Monitored

The movement of contaminants through the environment is monitored by measuring their concentrations in different components of the ecosystem. The assessment of impacts on the biota is done by looking at changes in population size, biomass, species composition, or other measures of population or ecosystem health.

Mine operators have emphasized the monitoring of chemical contaminants in air, water, soil and sediments but have done relatively little monitoring of biological effects. The current and proposed monitoring programs sample fewer valued ecosystem components (VECs) than do the government agencies which monitor cumulative effects at points distant from the mines (see Section 10.4). This should not be so. To create an integrated monitoring program, the mine operators and government agencies must monitor the same components. An exception would be made in the case of caribou; because they are seldom present at mine sites, it would be appropriate to substitute another terrestrial mammal such as the vole or squirrel. Thus, the existing or proposed monitoring programs at the mine sites should be expanded to include a terrestrial mammal, spruce grouse, spruce needles, and aquatic macrophytes. Phytoplankton and zooplankton should also be monitored because they are among the best indicators of trends in relation to effluent concentration gradients at Key Lake.⁶¹ The inclusion of some of these components would require the collection of additional baseline information.

The range of contaminants to be monitored should be reviewed periodically. At the public hearings, a presenter suggested that biota should not be analyzed for thorium-230, because it is not very mobile in the environment, because it is present in only very small amounts in organisms, because it contributes a very small radiation dose to the biota, and because it is so expensive to analyze.⁶² The money saved in eliminating unnecessary chemical analyses would allow improvement elsewhere in the monitoring program.

10.1.3 Monitoring Methodology

The method of monitoring sediments should be changed at some sample sites. Typically, sediment cores are collected every three years and divided into two strata: 0-5 cm and 5-15 cm. However, in areas where sediments accumulate, contaminants enter at the sediment surface by adsorption and by the deposition of new sediment. Because the deposition rates of new sediments are low in northern Saskatchewan, the impacts from most mines would only be observed in the top few centimetres of sediment. Although there is some debate on this issue, evidence received from the Department of Fisheries and Oceans indicates that sediment cores provide a history of events for many contaminants of interest at uranium mine sites.⁶³ By dating the different strata in the core, it is possible to calculate the rate at which contaminants are accumulated in the sediments. This would provide an independent check of the predictions of some impact models. For this technique to be useful, it would be necessary to subdivide cores into 1 cm strata and to lower the detection limits of contaminants to ensure that there would be no “less than” detected values.

This improved methodology for sediment sampling would be more expensive; however, the increase in cost could be offset by taking samples less frequently, a justifiable approach because the accumulation of contaminants in sediments is a slow process. In addition, sampling at this level of detail would be required at only a few sediment sampling sites.

So it is one thing to go out and take a dredge of mud and come back and measure the various things in it. It is another to go out and take a core and do the same thing, and it is a third level, I think, of sophistication to date that core so you can use it to calculate fluxes.

Dr. L. Lockhart, Transcript of McArthur River Public Hearings, Saskatoon, Saskatchewan, September 18, 1996, p. 72.

Pathways analysis is used to analyze the movement of contaminants through ecosystems (see Section 10.4). The analysis involves compartment models which use transfer coefficients to estimate the proportions of contaminants moving from one compartment to another.⁶⁴ Frequently, transfer coefficients are derived from the literature and their applicability has been questioned.⁶⁵ For example, is the forage-to-meat transfer coefficient of radium-226 in an Ontario cow a reliable estimate for the forage-to-meat transfer of that element in moose in northern Saskatchewan? The calculation of site-specific transfer coefficients would alleviate questionable assumptions of this type. This could be done at no extra cost, providing samples of the various components were collected at the same time and place.

The biological availability of contaminants and their effects on fish, or on other biota, need careful assessment. The uptake of metals by fish should be assessed by measuring concentrations in kidney, liver, gills, and possibly scales. On the other hand, the effects of increasing metal concentrations on fish are less easily assessed. Various indicators have been suggested, including lipid peroxide, metallothionein, and histopathology, but the significance of these indicators is still debatable.⁶⁶ This aspect of ecosystem health should be studied further to facilitate the development of a definitive monitoring system.

Although contaminant analyses of fish muscle is appropriate to assess the suitability for consumption by humans, the use of muscle to I determine the biological availabiliv of metals isnot appropriate with the exception of mercury.

J.F. Klaverkamp, C.L. Baron, H.M. Cooley and R.V. Hunt, The Use of Fish in Environmental Effects Monitoring for Uranium Mines, Submission to the MeArthur River and Cigar Lake Public Hearings, Saskatoon, Saskatchewan, September 16, 1966, p. 2.

At public hearings, the panel was alerted to a concern that the amounts of calcium sulphate in mill effluent could adversely impact the downstream watershed. This compound, often thought to be largely inert, could, in fact, change the chemistry of the lakes, producing effects that might only become apparent well after decommissioning.⁶⁷ Exploratory research is desirable to identify a simple indicator that could assess the probability of potential negative impacts from calcium sulphate.

10.1.4 Assessment of Impacts

There are several approaches to impact assessment. The following discussion is not exhaustive but does consider examples of all the main methods of assessing impacts.

In one approach, models are used, prior to operations, to predict contaminant concentrations at different locations. The predictions are then compared to accepted benchmarks that characterize the risk, enabling a qualitative risk assessment to be made. This approach is used in pathways analysis (see Section 10.4) where predicted doses to humans are compared to ICRP standards. Another example is Cameco’s use of an Integrated Risk Management Approach (IRMA) to compare predicted contaminant concentrations to toxicity values in the literature (LC₂₀, or LC₅₀, values⁶⁸) for various species, or to regulatory standards such as the Saskatchewan Surface Water Quality Objectives (SSWQO) or the proposed Canadian Sediment Quality Guidelines.

This type of risk assessment can be useful, but it does not eliminate the need for doing biological effects monitoring to check the validity of the predictions. This is because IRMA, which uses toxicity values from the literature, screens only a limited number of species, but bases its results on the assumption that these species are representative of all the species in the ecosystem. For example, IRMA might screen only a single species of phytoplankton, although there could be more than 100 species living in the system. If the species selected were tolerant to metal pollution, all phytoplankton would then be assumed to be tolerant, although we know this is not true. Similarly, since regulatory standards like the SSWQO have no limits for contaminants such as uranium or total dissolved solids, conforming to the standard does not automatically confer protection for all species.

A second impact assessment approach uses standard toxicity testing of effluent and possibly of sediment. This may be done either by acute toxicity testing, or by chronic or subchronic testing. Acute toxicity tests measure the percentage of the population killed by different concentrations of a chemical in a relatively short period of time. Chronic or subchronic tests screen sublethal effects, such as the causing of non-terminal cancers, malformations, mutations, etc. These toxicity tests are useful to assess risk but can only screen a limited number of species and life stages. In this way the approach su.ffers from some of the same drawbacks as IRMA; however, it may be very helpful for detecting upset conditions and should be used to screen mill effluent when there is a change in the mill process.

A third approach uses a variety of statistical models to analyze changes in the number, density, or biomass of a species, or to examine changes in the species composition of communities. This approach might also be used to analyze changes in contaminant concentrations, either in the biota or in the physical environment. The ability to detect an impact will increase as the magnitude of the change increases, as the sample variation decreases, and as the number of samples increases. On the other hand, the chance of detecting impacts may be reduced by inappropriate choice of sample collection sites, unsuitable sampling techniques, or by poorly matched samples. For example, if sediments were sampled using an Ekman dredge, the top 10-20 cm would be mixed, and dilution would reduce the ability to detect an increase of a contaminant in the top 1-2 cm.

If the data are analyzed and no change detected, this might indicate that no impacts are present; however, it might also indicate that the impacts (changes) present could not be detected by the tests because of low statistical power.⁶⁹ Thus, a test done might not be able to detect a change in a population or contaminant concentration because of small sample size and a large variation between samples. Baseline and monitoring data should be subjected to power analysis at each sample site. As a general rule, it should be possible to detect a 50% reduction (or increase) in a population at each site with a high degree of power or probability (>95%). If there is little chance of detecting such a magnitude of change, the sampling program should be modified. This may require taking a larger number of samples at each sample site, or matching sites more carefully to reduce sample error. To compensate for increased costs, it might be possible to increase the time interval between samples, or to decrease the number of sample sites, as long as the overall range of potential impacts is not decreased. It is better to sample a few things well and obtain definitive results, than to sample many things poorly.

10.1.5 Northern Participation

Even if the monitoring program were carefully designed and executed, its results might fail to convince the people of the region of its validity. Residents of the north must be involved, especially in the implementation of the program, before they will trust the results. This involvement can be accomplished, in part, through the Environmental Quality Committees (see Section 11.2). They could fulfil a valuable liaison role by connecting their communities to the monitoring program. In addition, it is important that some mechanism be found to involve northerners directly in the activities of monitoring.

...the more involved they become in monitoring, the more they are going to trust the industry.

Dr. Pat Thomas, Transcript of McArthur River Public Hearings, Saskatoon, Saskatchewan, September 16, 1996, p. 189.

10.1.6 Design of Monitoring Programs

The monitoring programs are specified by the regulatory agencies in the various operating licences for each mine. The specification of programs is not a one-way process, but rather the result of discussions between the operators and their consultants, and the regulatory agencies and their advisory government departments. This consultation process, necessary because of the expertise required to develop a well-designed monitoring program, might lead to concerns about the independence of the regulatory agencies from the industry. The panel suggests broadening the technical group that advises on monitoring protocols to include appropriate expertise from universities and from government institutes. In addition, representatives of the Environmental Quality Committees should be included. The technical group could hold a workshop every five years to evaluate the monitoring programs at all mines, enabling adjustments to be made in protocols in a timely manner.

...But companies are stuck with regulations, and regulations are built from the information available at the time they were made. And so part of the problem may be that the regulations have not necessarily kept pace with the science.

Dr. L. Lockhart, Transcript of McArthur River Public Hearings, Saskatoon, Saskatchewan, September 18, 1996, p. 86.

10.2 Monitoring at the McArthur River Site

The EIS predicts that the mine would leave a remarkably small footprint on the surrounding environment considering the size of the ore body. This is because of the way in which the mine has been designed, the fact that the ore will be milled at Key Lake and the proposal to dispose of problematic wastes either in the Deilmann Tailings Management Facility (DTMF) or underground in the McArthur River mine.

Cameco predicts very low radon emissions for the McArthur River mine, compared to most other uranium mines, because minewater inflow would be controlled by ground freezing or grouting, and the ore stream would be contained from extraction to transportation. Consequently, the impacts of aerial emissions are likely to be small.

The environmental impacts likely to be greatest at McArthur River would result from the release of treated effluent into the muskeg that drains into the east end of Boomerang Lake, and onward into Read Creek, which supports a grayling fishery. The lakes downstream are also productive, making them an important resource to protect. Minewater inflow would increase as the mine is extended along the ore body and two more shafts are constructed. There would be a corresponding increase in the volume of treated effluent, to approximately 4670 m³/day in the year 2012. This would be only about 10% of the average flow rate immediately downstream of Boomerang Lake, but almost 30% of the mean low flow rate. However, the effluent would be considerably lower in dissolved salts than would that released from milling operations, and it is predicted that contaminant concentrations in Lucy Lake would be less than levels deemed acceptable by the Canadian Water Quality Guidelines and Saskatchewan Surface Water Quality Objectives.

The effluent would seep through the muskeg and enter Boomerang Lake or Read Creek in a very diffuse manner. It is also likely that many contaminants would be adsorbed by the muskeg. This fact has not been taken into account in predicting water concentrations in Lucy Lake, located about three-quarters of a kilometre downstream of Boomerang Lake; therefore, the predictions for Lucy Lake are likely to be conservative. The panel concludes that the risk to fish and other aquatic organisms is acceptable.

In general, the proposed monitoring program is acceptable, with the provision that biological effects monitoring should be extended to include a terrestrial vertebrate, aquatic macrophytes, and plankton (see Section 10.1.2). The method of sampling sediments should be improved (see Section 10.1.3) to include a detailed profile of the sediments in Little Yalowega Lake.

The panel heard concern that four rare or uncommon plants (Carex pauciflora, C. trisperma, Pinguicula villosa and Scheuchzeria palustris), found in the muskeg that would receive the treated effluent, might be adversely affected. Given that this habitat is commonly found throughout the region, it is not likely that these species are at risk; however, Cameco should monitor to determine if these plants suffer adverse impacts.

10.3 Monitoring at the Key Lake Site

Milling of the McArthur River ore at Key Lake would impact two watersheds: the David Creek and the McDonald Lake drainage systems. Mill effluent would be combined with treated water from the DTMF and released into Wolf Lake which drains via Yak Creek into David Creek. The impacts to this watershed would likely increase marginally above current levels. Contaminant loadings would increase because of milling a higher grade ore and an increase in the volume of treated water from the DTMF.

The relatively clean water intercepted by perimeter wells around the DTMF would be released into the McDonald Lake drainage system. The impacts from the release of this water would likely decrease below current levels. Concentrations of nickel and other contaminants would be reduced by the recently commissioned reverse osmosis plant, and the need to pump the perimeter wells would decrease as the water level in the DTMF is restored.

Unfortunately, the ability to detect aquatic impacts, particularly those resulting from the release of mill effluent, has been compromised by the inadequacy of baseline information and a poorly designed monitoring program during the operation of the Key Lake mine. Cameco commissioned a set of studies by Terrestrial and Aquatic Environmental Managers Ltd. to correct this deficiency,⁷⁰ and collected additional operational baseline information to facilitate the assessment of incremental impacts. However, because the ability to assess impacts depends on the magnitude of the change, the incremental effects would have to be large before their detection would be possible.

Aerial emissions from the mill will also impact surrounding ecosystems. However, the results of the aerial monitoring program indicate that sulphur dioxide emissions are not sufficiently large to acidify the surrounding lakes, and that contamination by radioactive dust and radon progeny is largely confined to the Key Lake site.

The proposed monitoring program could be improved in many ways by applying the principles discussed in Section 10.1. First, biological effects monitoring should be implemented for the water body into which effluent is released; namely, Wolf Lake, for mill effluent, and Horsefly Lake, for intercepted groundwater around the Deilmann and Gaertner pits. It is disturbing to note that Delta Lake, approximately 10 km downstream from the point of release of mill effluent, was selected as a nearfield sample station; this would actually be a low-impact station. At this distance, sodium and sulphate ions would be diluted to approximately 20% and uranium, to approximately 2 %, of effluent concentrations. Second, the biotic components to be monitored should be expanded to include aquatic macrophytes, a terrestrial mammal, spruce grouse, and plankton. Third, the sampling of sediments should be improved, as outlined in Section 10.1.3.

The operating licence for the DTMF would prescribe the monitoring method for the tailings, groundwater, the water in the pond overlying the tailings, and the water collected in the drainage sump beneath the tailings. Subsequently, the licence for the decommissioning phase would require monitoring of the final consolidation of the tailings and the restoration of the groundwater. Such monitoring is one of the responsibilities of the regulatory agencies; the panel is confident of their ability and determination to enforce an acceptable monitoring program for the facility.

The main outstanding issue to be addressed is the length of the monitoring period after the pumps are shut off and the water table is restored. Local people deserve to be assured that contaminants are being contained within the facility and that, in the long term, any leakage of contaminants will be sufficiently small as to not harm organisms. The panel believes that the only way in which the people of the region can be assured of environmental protection is to monitor the facility indefinitely; many impacts may become apparent only in the long term and it is, therefore, not possible to guarantee a walk-away, zero-risk tailings storage facility. The industry and its regulators must recognize that tailings management facilities will require monitoring, and possible mitigation, in perpetuity.⁷¹ The details of the monitoring program, possible maintenance of the facility, and ability to respond to any contingencies would require careful thought ,and appropriate funding (see Section 12.5).

10.4 Pathways Modelling and Cumulative Effects

Pathways modelling has been used to predict the movement of contaminants in the environment and the dose to human receptors at different locations (see Sections 10.1.3 and 10.1.4). The dose estimates obtained from pathways models are subject to criticism, which is not surprising in view of the complexity of the models and the many source terms and factors on which they are based. Although it is important to make models as realistic and accurate as possible, it is impractical to expect precise predictions, because of the many uncertainties in the source terms and factors. It should be recognized that the main purpose of modelling exercises is to assess potential risks, not produce exact predictions. Monitoring is necessary because it is the only acceptable way to assess actual risks. Furthermore, the model must be updated whenever a significant amount of new information becomes available, as a result of monitoring.

...that such a program incorporate more food chain pathways bioaccumulation studies and further that northern stakeholders be involved in all aspects of monitoring.

Bill Layman, South Central EQC, Transcript of McArthur River Public Hearings, La Ronge, Saskatchewan, October 1, 1996, p. 48.

The proponent used environmental pathways analysis to assess the radiological dose from mine and mill emissions to hypothetical humans. The analysis was done for humans living at the McArthur River and Key Lake mine sites and nearby areas, and for residents of Wollaston Lake, Hatchet Lake and Black Lake. The analysis of the latter group involved an assessment of the cumulative effects of all the mines in the eastern part of the Athabasca basin. Cameco estimated the doses to members of the public to be well below regulatory limits and a small fraction (approximately 1% or less) of the natural background dose. A person at the Key Lake camp would receive the highest incremental dose, amounting to approximately 10% of the natural background dose. The Atomic Energy Control Board did an independent assessment⁷² and obtained comparable estimates. Thus, it may be concluded that the potential radiological impact of the McArthur River development would be acceptably low.

Even when it is anticipated that levels of impact would be acceptable, it is necessary to confirm all predictions by using a well-defined monitoring program. Saskatchewan Environment and Resource Management (SERM) and the AECB agreed to establish such a program in response to an earlier recommendation of the panel (see Section 13.1). They formed a Cumulative Effects Monitoring Working Group (CEMWG) in 1994 with technical advice from representatives of Saskatchewan Health, Environment Canada, the Department of Fisheries and Oceans Canada, the Saskatchewan Research Council, the University of Saskatchewan Toxicology Centre, the Saskatchewan Northern Mines Monitoring Secretariat, and Terrestrial and Aquatic Environmental Managers Ltd. This working group continues to improve the IMPACT/AECB model, which evaluates cumulative environmental effects.

The CEMWG has also established a cumulative effects monitoring program to test the reliability of the model’s predictions using field observations. A total of 63 sample stations has been established. Valued ecosystem components (VECs) are monitored on a 3-year cycle, and include air, soil, lichen, blueberry, spruce needles, caribou, spruce grouse, water, depositional sediments, macrophytes, benthos, and fish. Each VEC is measured for concentrations of radionuclides and metals, together with other physical and chemical parameters.

The way in which sediments are monitored also needs careful consideration (see Section 10.1.3). A detailed profile of sediments at a few localities would provide an excellent assessment of the spread and flux of contaminants for the past few decades.

. . . on a regional basis for cumulative effects monitoring, a lot of northern residents are hunting, fishing, trapping all the time, and it would be much more cost-effective and tess of an impact on the environment, if they were the ones supplying those types of samples, rather than consultants or scientists flying up there, spending all this money to collect the samples, when there are people there that need the work, and could do it anyway. All they need is training.

Dr. Pat Thomas, Transcript of I&Arthur River Public Hearings, Saskatoon, Saskatchewan, September 16, 1996, p. 189.

The panel endorses the CEMWG initiatives and notes that the existence of this diverse and highly competent team of scientists should reassure Athabasca residents and other northerners about the safety of country foods. Attempts should, however, be made to give northern residents a sense of ownership of this project and its results. This could possibly be achieved if the residents were involved in the collection of information and had representation on the monitoring committees (see Sections 11.1 and 11.2).

10.5 Conclusions and Recommendations

The available information indicates that the environmental impacts of this project will likely be within acceptable limits; however, careful monitoring will be required to ensure protection of the environment and human health.

The proposed monitoring program should be revised as follows:

• it should include the monitoring of the water bodies which will receive effluent;
• biological effects monitoring should be expanded to include all VECs that are monitored for cumulative effects at more distant sites;
• sampling of sediments should be modified to measure detailed profiles of contaminants at some sample sites; and,
• statistical power analysis should be conducted on baseline and monitoring data. If the results indicate little probability of detecting even a large change or impact, the monitoring program should be revised accordingly.

Monitoring will be required over a much longer time span for the Deilmann Tailings Management Facility than that proposed. Arrangements should be made that will permit perpetual monitoring.

Local residents should be involved in all monitoring activity.

Regular workshops should be held to review and advise on the biological effects monitoring for uranium mines. Participants could include technically qualified people from the uranium companies, the regulatory agencies and advisory government departments, other government institutes, and relevant departments of universities. Representatives of the Environmental Quality Committees should also be included.

---

⁶¹ Terrestrial & Aquatic Environmental Managers Ltd., Operational impact assessment to support nickel water quality objective setting for the McDonald Drainage system, prepared for Cameco Corporation, December, 1993.

⁶² P. Thomas, Submission to McArthur River and Cigar Lake Public Hearings, Saskatoon, Saskatchewan, September 16, 1996, p. 7.

⁶³ J.F. Klaverkamp, C.L. Baron, H.M. Cooley and R.V. Hunt, The Use of Fish in Environmental Effects Monitoring for Uranium Mines, Submission to the McArthur River and Cigar Lake Public Hearings, Saskatoon, Saskatchewan, September 16, 1996, p. 2. W.L. Lockhart and P. Wilkinson, Lake Sediment Cores as Archives to Detect and Measure Environmental Changes, Submission to the McArthur River and Cigar Lake Public Hearings, Saskatoon, Saskatchewan, September 18, 1996.

⁶⁴ The environment is viewed as being made up of a number of compartments such as the air, water, soil or sediment, or organisms at a particular location. An organism compartment may be specified to the species level (e.g. caribou), or may include several species (e.g. plants that the caribou eats), or may include only a part of an organism (e.g. muscle or liver).

⁶⁵ P. Thomas, Submission to McArthur Mver and Cigar Lake Public Hearings, Saskatoon, Saskatchewan, September 16, 1996, p. 2.

⁶⁶ McArthur River Environmental Impact Statement, Addendum, Cameco Corooration. June 1996.

⁶⁷ R.H. Hesslein, A genenic proposal for research on the effects of greatly elevated levels of calcium sulfate in lakes down stream of ore processing operations, Submission to McArthur River and Cigar Lake Public Hearings, Saskatoon, Saskatchewan, September 10, 1996.

⁶⁸ LC₂₀, or LC₅₀ values refer to the concentrations of a substance that will kill 20% or 50% of a population.

⁶⁹ Statistical power refers to the ability of a particular statistical test to detect a given change in a population.

⁷⁰ McArthur River Environmental Impact Statement, Appendix 8, Cameco Corporation, December, 1995, Sections 4.6.3.3 and 4.6.3.4. McArthur River Environmental Impact Statement, Appendix 12A, Cameco Corporation, December, 1995.

⁷¹ D. Kirkwood, T. Peters and D. McCreath, Decommissioning of Uranium Mine Tailings Management Areas in the Eliot Lake Area, June, 1996, Supply and Services Canada.

⁷² L. Chamney, Atomic Energy Control Board, Regulatory Review of the Radiological Impact of Cigar Lake and McArthur River Uranium Projects, Submission to the McArthur River and Cigar Lake Public Hearings, Saskatoon, Saskatchewan, September 16, 1996.