Voisey's Bay Mine and Mill Environmental Assessment Panel Report

To establish baseline conditions, VBNC investigated existing levels of metals in water, sediments, soils and the tissue of selected organisms from representative sites adjacent to the Project. In a few cases, measured baseline values for some metals in freshwater, seawater and their sediments in the Project area, as well as for some benthos and fish, exceeded current guidelines (although the nature and extent of these exceedances were not always clearly stated in material presented to the Panel). These exceedances are considered to be a natural condition because of the presence of the measured metals in country rock. There is also an excess of mercury in caribou, attributed to the fact that lichen has absorbed mercury transported through the atmosphere from southern sources.

Using its proprietary IMPACT model, Beak International predicted that metals that VBNC would release into the environment would pose hazards to living organisms. IMPACT is a probabilistic model that accounts for the pathways and fates of released metals; the behaviour and properties of these metals in environmental media, including bioaccumulation in organisms; and the resulting risk to biological receptors such as fish, wildlife and humans.

Beak considered dust and other atmospheric emissions as a potential contaminant source; it also considered aquatic releases of treated effluent, including releases from the open pit, the mine and mill and the concentrate handling facility, as well as seepage and post-closure surface releases from the two tailings basins. Beak modelled effects for various phases of the Project, including post-decommissioning, for up to 140 years.

The ultimate source of virtually all metals that may be released is the mine rock itself. The model therefore relied on analysis of mine rock from the Ovoid, the Western Extension and the Eastern Deeps, and of predicted tailings composition and tailings water chemistry. As Beak noted, model results and interpretation depended strongly on the accuracy of VBNC's source terms and environmental transport forecasts. Beak characterized its modelling as a "screening level assessment" and suggested the results be used to identify requirements for more detailed assessments, and to identify the need for additional mitigating and monitoring measures where incremental exposure increases warrant.

Beak initially analyzed eight metals (copper, nickel, cobalt, lead, zinc, cadmium, aluminum and arsenic), based on their known environmental toxicity and on biological sensitivity to these metals. Although mercury is a known potential contaminant of country foods (particularly fish and marine mammals) and elicits widespread concern in northern Canada, it was not included in the original modelling. So the Panel requested and received additional information on mercury contaminant potential from VBNC.

The modelling exercise predicted incremental changes in water and sediment chemistry over time, and levels of metals uptake for a variety of aquatic and terrestrial receptors, including freshwater and marine invertebrates, fish, waterfowl, marine mammals, and large and small terrestrial mammals, selected for their ecological and cultural importance. Metal dosages were calculated for each receptor by pathway, and compared to the benchmark dosage at which chronic adverse effects might occur, in order to establish a hazard quotient.

Modelling accounted for the time that any particular species would be present in a potentially contaminated area. VBNC used high-end or worst case values for inputs and uptakes, based on site-specific information where available, or established literature values. Where metals were below detection limits, they were assumed to be at the detection limit for the purposes of modelling. Such conservative approaches overestimate potential environmental effects.

The Canadian Council of Ministers of the Environment has established water quality objectives. The model predicts that, as a result of the Project, contamination will exceed these objectives in freshwater and freshwater sediments, but not in marine water or marine water sediments. Exceedances are expected to result from tailings pond seepage and from periodic releases of water from the flooded pit into Camp Pond. As a result, contaminant levels would exceed United States Environmental Protection Agency chronic effects guidelines for freshwater snails and land-locked char at these locations. There would also be excess levels of nickel and aluminum for arctic char in the Throat Bay watershed and at the head of Kangeklualuk Bay during the post-operational period. No exceedances are predicted for terrestrial or marine mammals or for birds, since the incremental effects of the Project would fall below guidelines for all species of these animals modelled.

All samples of mine rock and tailings had levels of mercury lower than detection limits. VBNC predicts that, based on a balance of factors, the potential for mercury mobilization in water will not increase. Mercury will not become more bioavailable and there will be no significant increment available for bioaccumulation or biomagnification in the food chain.

Therefore, VBNC predicted that the environmental effects of the metals that the Project would release would be largely indistinguishable from localized minor effects already occurring due to existing natural levels of metals. It also predicted that contaminants would have no consequences for the country foods that local residents eat.

VBNC stated that it would monitor water, sediment and some biota in receiving watersheds for both modelled and other metals as appropriate, for the life of the Project.

Government and Public Concerns

The Department of Fisheries and Oceans (DFO) believed that VBNC did not substantiate its position that mercury mobilization would not be an issue. DFO did not assert that mercury mobilization would definitely be a problem, but considered that it could be. DFO's chief concerns included the uncertainty surrounding predictions about mercury mobilization and about the behaviour of metals in sediments and salt water. With respect to predictions based on modelling, DFO questioned whether they adequately addressed metal speciation and whether the IMPACT model properly addressed the complexity of mercury's effect on aquatic systems; DFO also questioned VBNC's choice of certain macrobenthos as indicators, as well as the model's failure to quantify probability ratings. However, DFO noted that the baseline chemical analyses for metals that VBNC reported were compatible with and comparable to its own data.

The Canadian Wildlife Service (CWS) of Environment Canada also considered VBNC's predictions about mercury to be optimistic, in view of the potential problem of acidification, which could mobilize even small quantities of available mercury. CWS considered that there were insufficient baseline data with respect to birds and mammals. It also identified several technical problems with VBNC's contaminant modelling, although some of these turned out to be errors of presentation. Overall, CWS considered that VBNC had underestimated contaminant hazards.

DFO and CWS made broadly similar recommendations. For example, both recommended that VBNC

• do more baseline sampling;
• further evaluate, strengthen and test the model, in cooperation with DFO and CWS; and
• monitor a broad range of species for contaminants throughout the life of the project, and establish a protocol for interpreting results and taking remedial action.

The provincial Forestry and Wildlife Branch recommended monitoring small mammals for contaminants.

An expert, appearing on behalf of the Innu Nation, also suggested several methodological flaws in the selection and application of the IMPACT model. These focused chiefly on the use of average values rather than a range, particularly with respect to stream flows and to bioconcentration in organisms; the failure to include extreme events; and the lack of sensitivity analysis. Thus, a worst case scenario was not modelled. She suggested that probabilistic modelling was required, and also noted that more baseline data were needed to provide an adequate range of input values.

LIA experts recommended a review of all existing data, further sampling to fill gaps, cooperative identification of monitoring targets and a monitoring program that, among other things, would include areas used for harvesting. LIA emphasized the need for a cumulative effects approach to contaminants, and provided a spatial framework for modelling sources, pathways and receptors in the Project area.

VBNC responded that it had presented a deterministic, not a probabilistic, analysis, using conservative values throughout to predict the outcome of a worse case scenario. It provided a comparison of predicted and measured values of metals in certain plants and aquatic organisms to confirm that the model tended to overestimate metals accumulation. VBNC noted that the model was not intended to predict the occurrence or severity of acute toxicity due to accidental events, but rather the effects of chronic low-level exposure on organisms over long periods of time. VBNC did not consider that more baseline research was required before construction. However, it was willing to discuss modelling issues with all parties, to consider monitoring the effects of contaminants on wildlife and to review the protocol for interpreting results, in consultation with DFO and CWS. VBNC reaffirmed its view that mercury is not an issue in the assessment of the Project, because Project activities would not increase existing levels of mercury in organisms.

Conclusions and Recommendations

The Panel considers that VBNC provided adequate baseline data on contaminants in water, sediments and biota adjacent to the proposed Project, for the purposes of the Environmental Assessment. The Panel also agrees in principle with VBNC's approach to modelling, in particular that

• it was appropriate to begin with deterministic modelling using conservative values; and
• the screening approach was appropriate for the purposes of the Environmental Assessment.

The Panel considers that the source terms were appropriate, and that the values and assumptions used within the modelling exercise tended to overestimate, rather than underestimate, metals hazards.

It would appear, on the basis of assessment done to date, that Project activities are unlikely to release metals into the environment at levels that would constitute a significant hazard to fish, wildlife or humans. The Panel was not presented with any clear hypotheses (as opposed to concerns) that released metals would significantly threaten ecosystem or human health, based on knowledge of the way these metals would be released and mobilized, the way they would become bioavailable, and the potential for bioaccumulation and biomagnification in organisms and food chains in the Landscape Region.

The Panel notes that while levels of some metals, particularly nickel, copper and aluminum, are predicted to exceed guidelines near the Project site, these metals do not significantly bioaccumulate or biomagnify in the food chain. Even if they accumulate at levels hazardous to aquatic organisms at a few specific sites, which is considered unlikely, they would not become hazardous to predator species because those species are not resident at those sites, and hence would not become hazardous to humans.

Nonetheless, the Panel believes that contaminants, as they may affect country foods or its consumption, are a significant issue in relation to the Project. The Panel therefore believes that two distinct monitoring programs are required. One, which should be VBNC's responsibility, is an effects monitoring program that is hypothesis driven and tightly focused on metals. The other is a more general contaminants monitoring program focused primarily on country foods and the health of the food chain. This should be a cooperative program under EAB direction.

Recommendation 14

The Panel recommends that VBNC develop an appropriate effects monitoring program for metals and other contaminants, in cooperation with DFO, Environment Canada, LIA and the Innu Nation. The program should include a protocol for interpreting results and for taking remedial action. The program should be in place before construction starts and should be subject to ongoing modification, as appropriate.

In view of the concerns expressed by various participants, and in keeping with Beak's warning, further consideration should be given to both the technical aspects of and appropriate targets for modelling for the purposes of monitoring.

Recommendation 15

The Panel recommends that a program be established to monitor contaminant levels in country foods on a continuing basis in northern Labrador. This general program should be a cooperative one involving primarily governments, LIA, and Innu Nation, although VBNC should contribute some technical and material support. The lead agency for this program should be designated by DFO, in its capacity as the Responsible Authority. This lead agency should be the primary funder of the program, and provide scientific resources to it, but the program should be under the direction of the Environmental Advisory Board (EAB). The objective of the program should be to address public concerns, and to minimize misunderstandings about the actual effects of the Project on the regional environment. The program should address the cumulative and synergistic effects of contaminants from all sources, and should include provisions for interpreting and communicating the results to the regional public on a continuing basis. It should fully incorporate the knowledge and experience of the federal Northern Contaminants Program and also develop cooperative links with it. The program should, at the outset, ensure that adequate baseline data are obtained on contaminant levels (not restricted to metals) in a broad spectrum of biota and locations in the region. It should assemble all existing contaminants data for the region from all relevant public and private agencies, and then add to them as required. These baseline data should be available prior to construction, subject to review and recommendations of the EAB.

Two issues appear uncertain, and require further examination. One is the potential for mercury mobilization and in particular under what circumstances acidification might occur at a level and scale which could increase it, and if it does, whether other factors might counteract this tendency. The other is the behaviour of metals in the marine environment and sediments, in particular whether they might become more bioavailable to marine organisms than VBNC has predicted. These matters should be considered on a continuing basis as part of the effects monitoring program, but they also require dedicated research.

Recommendation 16

The Panel recommends that DFO and Environment Canada jointly develop a problem statement and research design to identify the means by which mercury could become mobilized in the environment, within the parameters of this Project. If this exercise results in a clear hypothesis linking the Project to mercury mobilization at levels potentially hazardous to fish, wildlife, or humans, then DFO, Environment Canada, and VBNC should develop and fund a cooperative research program leading to prevention or mitigation.

The issue of the behaviour of metals in the marine environment is addressed in Recommendation 27.