Health Canada
www.hc-sc.gc.ca
Common menu bar links
Institutional links
Priority Substances List Assessment Report for Releases from Primary and Secondary Copper Smelters and Copper Refineries - Releases from Primary and Secondary Zinc Smelters and Zinc Refineries
3.0 Assessment of "Toxic" under CEPA
None of the six risk quotients for the monitoring station at Noranda-CCR exceeded 1.0 individually. It should be noted, however, that the value for Cu (0.8) is close to 1, and also that the sum of the individual quotients is slightly above 1. If the additivity model described at the beginning of Section 3.1 is applicable, it is possible that sandy soils in the vicinity of Noranda-CCR are adversely impacted by the combined loadings of these metals. Furthermore, as indicated earlier, there is a significant probability that risk quotients based on TSP data - such as those for the CCR station - are low by a factor of 2-5. If the quotients for Noranda-CCR were increased by even a factor of 2, the value for Cu would exceed 1.
It may be concluded that there is potential for effects on aquatic and/or soil-dwelling organisms from exposure to steady-state concentrations of metals in the vicinity of copper smelters and refineries resulting from releases (especially of Cu) from these facilities. Impacted areas appear to extend up to about 14 km from the sources based on comparison to 25th percentile critical loads, which equates to an area of as much as 600 km2.
Exceedence radii were also estimated based on the generic deposition modelling described in Section 2.3.1.2.3 and detailed in SENES Consultants (2000).
Although many of the model input parameters are based on data for the facilities being assessed, it is important to recognize that the results of generic modelling do not correspond to any individual facility. In particular, results for a 95th percentile modelled deposition will draw from among the worst of each characteristic (emission rate, stack height, etc.) that has been used as input to the model, irrespective of facility.
The results shown in Table 23 are generally supportive of the deposition data based on monitoring. Maximum radii of exceedence of CL25s estimated from the empirical data are based on higher emitting facilities, and these radii generally fall between those determined from the 50th and 95th percentile modelled deposition values. In a number of cases, exceedence radii estimated by deposition modelling are less than those determined from monitoring data. This may be indicative of emissions from sources not being directly considered in these assessments (and therefore not used as model inputs), but which contribute to local background. Such sources may, for example, be due to the production and transport of concentrates or to metal containing dust blown from uncovered tailings areas. One exception is the estimate for Cu emitted from copper refineries. Based on the 50th percentile modelled deposition, the radius of exceedence of the CL25 for Cu is 7 km. This appears to be somewhat greater than the radii of exceedence indicated by monitoring-based data. This is in part because Cu emissions from the only stand-alone copper refinery considered in these assessments appear to be quite low. However, as was expressed above and in Section 2.3.1.2.2, values of total deposition calculated by applying a deposition velocity to TSP monitoring data are generally underestimates, especially close to the source. All of the empirically derived deposition estimates for Noranda CCR and the Sudbury region were based either fully or partially (i.e., "wet plus dry") on TSP data. The relatively large modelled radius of exceedence may reflect the fact that essentially all of the metals released from copper refineries are fugitives or are from low-elevation stacks. This suggests that Cu released from the Inco-Copper Cliff copper refinery may also contribute more to locally deposited Cu than the 13% expected based on source attribution (see Table 38). Thus, copper refineries may be more significant sources of local deposition than might be expected based on consideration of emission data alone.
Uncertainties: There are relatively minor uncertainties in monitored data and somewhat larger ones in use of the data to estimate annual depositions. In particular, estimation of total metal deposition from measurements of TSP appears to generally underestimate exposure. The relative reliability of results obtained using the different methods was discussed in Section 2.3.1.2.2.
Uncertainties are associated with selection of ENEVs, although care was taken in using high-quality studies focused on realistic (indigenous), most sensitive biological species. There was potential for significant uncertainty through use of the free-ion activity model in this work. While this approach is believed to be an improvement over standard assessment methods that consider exposure to total metal concentrations, it introduces uncertainties related to estimation of free-ion concentration as well as bioavailability. Parameters influencing metal uptake by organisms include free-ion concentration in solution, pH and hardness. The latter two were addressed by selecting effects studies that used pH and hardness conditions similar to those typically found in soft-water, circumneutral to acidic lakes on the Shield. Use of the free-ion activity model approach necessitated that effects studies that did not include sufficient data to allow estimation of free metal ion concentrations be ignored. As a result, some potential "most sensitive" studies may have been excluded from consideration. This approach also ignored ingestion as a pathway for uptake of metals, which may significantly underestimate exposure for some organisms.
There is significant uncertainty owing to assumptions made in using fate and transport models to estimate critical loads. These have been discussed in Section 2.4.1.1.3. One of the most significant ones is the assumption that there is no transfer of metal from catchment areas to lakes. It is believed that this could result in up to about a five-fold underestimation of metal exposure (equivalent to a five-fold overestimate of critical loads) in some realistic worst cases. The extent of underestimation could be even greater for a small percentage of watersheds. Critical load modelling was carried out to steady state in order to avoid complications owing to historic depositions, which have typically been much greater than current depositions. It is again pointed out that, in many cases, exposure levels must decrease over long periods of time to reach these steady-state concentrations. Hence, the potential for effects on organisms in the time period leading up to steady state may be underestimated.
It should be pointed out that risk quotients derived in this section have been based on comparison to 25th percentile critical loads, rather than on comparison to more protective levels, such as 10th percentiles. Further, the potential for additive effects of exposure of organisms to multiple metals has not been addressed in detail. This potential for additive effects should be kept in mind when considering risk quotients based on exposure to individual metals. Finally, for practical reasons, these assessments considered in detail a limited number of components released from these facilities. It is recognized that there is uncertainty introduced to the overall risk characterization associated with those compounds - such as Hg - that were not assessed.
The estimated range of impact is generally in keeping with results of aquatic and terrestrial field studies of the environmental effects of mostly historic accumulations of metals around copper smelters (see, for example, Freedman and Hutchinson, 1980; Couillard et al., 1993; Borgmann et al., 1998). Other evidence of the impacts of metals on organisms owing to past emissions from copper smelting facilities has been documented (see, for example, Sanderson, 1998, and references cited therein).
3.1.1.2 Releases to water (Noranda-CCR)
Risks: A site-specific screening-level risk assessment of releases from Noranda-CCR to surface waters was carried out. Release data were used to estimate the annual average exposure concentrations for organisms in waters receiving effluent from the MUC-WWTP, which receives releases from CCR (Table 26).
Risk quotients for aquatic releases from the MUC-WWTP are shown in Table 39. This table also contains estimates of the proportion of release components attributed to the Noranda-CCR operation. Risk quotients were determined by dividing the exposure values by the ENEVs that were discussed in Section 2.4.1.2.2 and summarized in Table 35. As shown in Table 39, there is a possibility of chronic effects on fish related to Cu and Ag (RQ=1.4 and 4.9, respectively) and of acute effects on zooplankton related to Ag (RQ=1.4), in the near-field plume in the St. Lawrence River. Toxicity testing of MUC-WWTP effluent (discussed in Section 2.4.1.2.1) also suggests a potential for aquatic toxicity in the plume, although there are questions about the representativeness of the samples tested. There are no environmental effects monitoring data. The CCR contribution to Cu and Ag in the MUC-WWTP effluent is very small. Thus, the potential for effects that has been identified is mainly attributable to sources other than the copper refinery.
Uncertainties: In the assessment of aquatic releases, some uncertainties have been accommodated by making conservative assumptions. These will tend to give high estimates for chemical exposure and low estimates for effect levels, so that potential effects are unlikely to be overlooked.
Conservative exposure assumptions include the assumption that fish with small home ranges are resident in the near-field plume, as well as the use of a two- to four-day exposure period in the acute toxicity benchmarks for zooplankton. In addition, the toxicity benchmarks are based on protection of sensitive species, which may not be present in the local receiving environments.
Alternatively, it should be noted that the discussion of risk considers potential effects of each element in isolation. As noted previously (see discussion of "Joint effect of metal emissions" - Section 3.1), there is the possibility that, due to simultaneous exposure to multiple elements, risks are greater than those predicted. Water/sediment distribution coefficients typically have order of magnitude uncertainty. There are also uncertainties in the appropriateness of the ability of the models applied to estimate the plume dispersion pattern accurately. Model validation was impeded by a shortage of near-field monitoring data.
- Risk quotients for zooplankton are based on acute effects. Risk quotients for fish and benthic organisms are based on chronic effects.
- Values in bold meet or exceed a risk quotient of 1.0
With regard to the MUC-WWTP specifically, data were not available to estimate short-term risk quotients (based on maximum monthly or four-day average loadings), which would be expected to be higher than those calculated using annual average loadings. Furthermore, since no data were identified on the chemical forms of releases, all metals were assumed to be in dissolved or adsorbed (i.e., bioavailable) forms.
3.1.2 Zinc plants
3.1.2.1 Releases to air
3.1.2.1.1 Sulphur dioxide
Data for the monitoring of ambient SO2 in the vicinity of zinc plants were provided by the companies. These represent exposure to ambient SO2 over time scales of the growing season as well as over 1 hour. Monitoring data were summarized in Table 10 and Table 11, respectively.
Table 36 shows risk information for ambient SO2. Derivation and interpretation of data in the table were described in Section 3.1.1.1.1.
As indicated in Table 36, four facilities have zinc plants. Release of SO2 from these plants is typically associated with roasting operations. Noranda-CEZinc is a stand-alone zinc processing facility. Eighty-five percent of SO2 emissions from Cominco-Trail are attributed to the zinc plant; the remainder are attributed to the lead plant. The zinc plant at Falconbridge-Kidd Creek is responsible for an estimated 15% of SO2 emissions from that facility. Finally, although HBM&S includes both a copper smelter and zinc plant, the pressure-leach technology used in its zinc plant does not result in the release of SO2.
The chronic "growing season average risk quotient" values in Table 36 for the three facilities with measurable SO2 releases show little similarity. This is in part due to the fact that data in the table have lost some spatial coherence, as sites are listed in increasing order of distance from the facility irrespective of direction. This obscures trends in the data somewhat, lacking consideration of geographical and meteorological factors.
The clearest trend is observed for Cominco-Trail due to the larger number of monitoring stations operated. Due to the valley location of this facility, most of the monitoring stations are downwind of the complex much of the time (i.e., either up or down the valley from the facility). Although growing season average risk quotients are not overly high (maximum of 3.4), they remain elevated over a significant area. Exceedences of a risk quotient of 2, indicative of "likely" effects on sensitive species, extend out to about 10 km within the confines of the valley.
At the Noranda-CEZinc facility, a relatively high chronic risk quotient (2.6) is observed at one monitoring station located 1.3 km from the plant in what is frequently a downwind direction. At another station located at a similar distance in a direction that is seldom downwind of the plant, the risk quotient is quite low. Insufficient data are available to determine the distance from the facility in the downwind directions that may be impacted.
The environmental impact of SO2 emitted by the zinc plant at Falconbridge-Kidd Creek is of lesser significance due to the generally lower risk quotients observed for monitoring stations near this facility and to the lower attribution of the zinc plant to total SO2 emitted by the facility. It should be recognized, however, that only 20% of the SO2 emitted is from a source not considered as part of these combined assessments, and that at one monitoring station a risk quotient exceeding 2 is observed.
Trends similar to those observed for chronic (growing season) exposure are seen for each facility when considering the risk information for acute (1-hour) exposure. At the Cominco-Trail facility, a moderate number of exceedences of the "possible" effect level (RQ=1) are observed with relatively few exceedences of the "likely" effect level for sensitive species (RQ=2). The observed outer limits for exceedence of these two levels are about 10 km and 4 km, respectively. Acute risk quotients above 1.0 occur relatively frequently at the monitoring station located downwind of the Noranda-CEZinc plant, reinforcing the expectation that the area impacted by ambient SO2 in the downwind directions extends somewhat further from the source. While there were a moderate number of 1-hour periods where the risk quotient was greater than 1 at some monitoring stations near Falconbridge-Kidd Creek, on only one occasion was RQ=2 exceeded.
The "maximum" risk quotient column shows the extreme values for acute (1-hour) exposure to SO2 for vegetation near the zinc plants. The highest quotient, 4.2, was calculated for two stations located within 4 km of the Cominco-Trail facility.
It may be concluded that there is the possibility for effects on vegetation from both acute (1-hour) and chronic (growing season) exposure to SO2 released from zinc plants.
Depending upon the facility type (SO2 releases are typically associated with roasting operations) and local meteorology and geography, the areas impacted may extend to about 10 km from the source.
Uncertainties: Uncertainties in the evaluation of risk to the environment due to ambient SO2 were discussed in association with releases to air from copper smelters and refineries (see Section 3.1.1.1.1).
3.1.2.1.2 Deposited sulphate
Sulphur dioxide emitted from zinc plants can be oxidized to sulphate in the atmosphere. Sulphur dioxide and sulphate can be transported long distances from the source, resulting in acidic deposition to soils and lakes over large areas.
Information needed for the evaluation of risk due to wet sulphate deposition is shown in Table 37. Derivation and interpretation of data in the table were described in Section 3.1.1.1.2. As was described, model parameters were scaled based on 1995 emission data for the facilities considered in these assessments, to estimate incremental contributions to deposition attributable to these sources. Of the four receiving areas shown in Table 37, the highest relative contribution to wet sulphate deposition from zinc plants was 0.2%, seen for the Montmorency location.
Uncertainties: Uncertainties in the evaluation of risk to the environment due to sulphate deposition were discussed in association with releases to air from copper smelters and refineries (see Section 3.1.1.1.2).
Although there are clearly detrimental acidification effects on lakes in eastern Canada owing to anthropogenic emissions of SO2, it may be concluded that current emissions from Canadian zinc plants contribute only a minor portion of the SO2 leading to lake acidification.
3.1.2.1.3 Deposited metals
Estimates of annual deposition of the metals Cu, Zn, Ni, Pb, Cd and As based on monitoring data in the vicinity of zinc plants are summarized in Tables 15, 17 and 18. Derivation of critical loads for these metals was discussed in Section 2.4.1.1.3, and annual critical loads were summarized in Tables 31 and 33.
Table 38 shows risk information for metals deposited in the vicinity of zinc plants. Derivation and interpretation of data in the table were described in Section 3.1.1.1.3.
Critical loads derived for sandy soils typical of those found on the Canadian Shield were used to calculate risk quotients for Cominco-Trail and Noranda-CEZinc. Although these are not located on the Shield, examination of local surface geology and soils maps (Mailloux, 1954; Fulton, 1984, 1996) indicates that sandy soils occur near each of these facilities, making use of soil critical loads suitable for application at these sites. For HBM&S and Falconbridge-Kidd Creek, the more sensitive of soil or surface water critical loads were used.
Emission-based source attribution information is also shown in Table 38. Noranda-CEZinc is a stand-alone facility, and all metal emissions may be attributed to zinc processing. The Cominco-Trail facility also includes a lead plant, which is responsible for large portions of the Pb, Cd and As emissions from that complex. The zinc pressure-leaching process used at HBM&S-Flin Flon results in insignificant emissions of the metals being assessed, excluding it from consideration in this section. The Falconbridge-Kidd Creek facility includes multiple operations. Emissions of Zn and As have significant proportions coming from the zinc plant. It should be noted, however, that because these attributions are based on only a partial inventory of sources (e.g., fugitive releases from tailings areas are not included), the relative contributions of zinc plants to metal deposition rates estimated from monitoring data may be somewhat overestimated.
It is apparent from the data in Table 38 that 25th percentile critical loads for Zn are often exceeded. At the Cominco facility, risk quotients significantly greater than 1 are seen at a monitoring station 10.5 km down the valley from the zinc operations. Considering the TSP-based risk quotient of 1.0 observed at 12.7 km, the impacted area likely extends beyond 13 km. Critical loads for Pb and Cd are also exceeded beyond 10 km, although only 4% and 36% of these metal emissions, respectively, are attributable to the zinc operations.
At the Falconbridge-Kidd Creek and Noranda-CEZinc facilities, the critical load for Zn is significantly exceeded out to a distance of at least 4 km - the locations of the furthest monitoring stations. At the Falconbridge site, however, only 24% is attributed to the zinc operations, the largest portion being attributed to the concentrator.
It may be concluded that there is potential for effects on aquatic and/or soil-dwelling organisms from exposure to steady-state concentrations of metals in the vicinity of zinc plants resulting from releases (especially of Zn) from these facilities. Depending upon the facility type (metal emissions from plants relying exclusively on pressure-leach technology may be negligible) and local meteorology and geography, the areas impacted may extend as far as about 13 km from the source.
Exceedence radii were also estimated for zinc plants based on the generic deposition modelling described in Section 2.3.1.2.3 and detailed in SENES Consultants (2000).
Table 23 shows the maximum distance from each facility type, or combination of facilities, where the 50th or 95th percentile estimates of total soluble deposition rates exceeded the benchmark 25th percentile critical load. As discussed in Section 3.1.1.1.3, these results do not correspond to any individual facility or combination of facilities. Again, the results shown in Table 23 are generally supportive of the deposition data based on monitoring, and estimates of exceedence radii based on modelling tend to be lower that those determined from monitoring data. Along with the possible explanation of other sources contributing to local background, it is pointed out that the metal emission values for the Cominco-Trail facility, which were one of the sets of input values used in the model, may have been underestimated (see footnote to Table 4). As was noted for copper refineries, essentially all of the metals emitted from zinc plants are from low-elevation sources, increasing the potential for these operations to contribute significantly to local deposition.
Uncertainties: Uncertainties in the evaluation of risk to the environment due to exposure to deposited metals were discussed in association with copper smelters and refineries (see Section 3.1.1.1.3).
3.1.2.2 Releases to water (Noranda-CEZinc and Cominco-Trail)
Risks: Site-specific screening-level risk assessments of releases to surface waters were carried out for two zinc facilities: Noranda-CEZinc and Cominco-Trail. Release and monitoring data were used to estimate the maximum monthly (chronic) and four-day (acute) exposure concentrations for the CEZinc and Trail facilities, shown in Tables 27 and 28, respectively.
Risk quotients for aquatic releases from the CEZinc and Trail facilities are shown in Tables 40 and 41, respectively. These tables also contain estimates of the proportion of release components attributed to the operations being assessed. Risk quotients were determined by dividing the exposure values by the ENEVs that were discussed in Section 2.4.1.2.2 and summarized in Table 35.
The potential effects of effluents from the CEZinc and Cominco-Trail facilities were evaluated by considering their local impacts on St. Lawrence River and Columbia River receiving waters. As indicated by the risk quotients for CEZinc shown in Table 40, there is a possibility of chronic effects on fish related to Se (RQ=4.9 based on 1995 release data) and Cu (RQ=1.5) under maximum loading conditions. Recent significant reductions in releases of Se, however, have likely reduced the quotient for that element. Toxicity testing of CEZinc effluent (discussed in Section 2.4.1.2.1) does not suggest a potential for aquatic toxicity in the plume given the pH control measures currently in effect. There are no EEM data.
For Cominco-Trail releases related to zinc operations (Table 41), there is a potential for effects on fish related to Cd and Tl (RQ=1.1 and 3.4, respectively) and benthos related to Zn, Cd, As and Hg (quotients up to 4.5). In the case of Cd and As, however, only a relatively small percentage of the exposure is attributable to Cominco's zinc operations. Toxicity testing of Cominco-Trail effluent (discussed in Section 2.4.1.2.1) does not suggest a significant potential for acute toxicity in the plume. Chronic toxicity testing has not been performed. EEM (discussed in Section 2.4.1.2.3) has found sediment toxicity and benthic/periphyton community effects in areas directly downstream of the outfalls, although these may have been related at least in part to historical slag deposits.
- Risk quotients for zooplankton are based on acute effects. Risk quotients for fish and benthic organisms are based on chronic effects.
- Insufficient data were available to evaluate maximum (1-month) exposure concentrations for Pb or ammonia. Of note is that the risk quotient determined for exposure of fish to ammonia was 0.58 (with 97% attributable to CEZinc) based on an annual average EEV (Beak International, 1999). A risk quotient based on a maximum short-term (1-month) EEV could be significantly higher.
- Values in bold meet or exceed a risk quotient of 1.0.
- Releases of Se from the Noranda-CEZinc facility are believed to have been significantly reduced recently; thus, this quotient likely overestimates risk.
- Risk quotients for zooplankton are based on acute effects. Risk quotients for fish and benthic organisms are based on chronic effects.
- Values in bold meet or exceed a risk quotient of 1.0.
Uncertainties: As noted previously (see Section 3.1.1.2), in the assessment of aquatic releases, some uncertainties have been accommodated by making conservative assumptions. As a consequence, the potential effects identified may not be realized. For example, it has been conservatively assumed that fish with small home ranges are resident in the near-field plume, and that organisms in local receiving environments are among the most sensitive identified in the literature.
Alternatively, it should be noted that risk quotients for several organism-element combinations were found to be only slightly below 1, and that if the additivity model discussed earlier (see "Joint effect of metal emissions" - Section 3.1) is applicable, it is possible that effects are greater due to simultaneous exposure to multiple elements. In addition, data were not available to calculate a maximum monthly exposure value for ammonia at the CEZinc facility. This omission could be significant given that the risk quotient determined for exposure to fish was 0.58, based on an annual average EEV. A risk quotient based on a maximum monthly EEV could be significantly higher.
Water/sediment distribution coefficients typically have order of magnitude uncertainty. There are also uncertainties related to the ability of the models applied to accurately estimate the plume dispersion pattern. Model validation was impeded by a shortage of near-field monitoring data.
3.2 CEPA 1999 64(b): Environment upon which life depends
As described in Section 2.4.2, based on the very small amounts of VOCs included in releases from Canadian copper smelters and refineries and zinc plants, these releases are not expected to contribute significantly to the creation of ground-level ozone. Similarly, because emissions of VOCs are low, and since sulphate aerosols formed from emitted SO2 are unlikely to migrate to the stratosphere, such releases are unlikely to contribute to stratospheric ozone depletion. Finally, based on the relatively small amounts of CO2 and other greenhouse gases included in releases from Canadian copper smelters and refineries and zinc plants, these releases are not expected to contribute significantly to global warming.
3.3 CEPA 1999 64(c): Human health
3.3.1 Exposure assessment
Based on the data summarized in Section 2.3.1, the airborne levels of metals, SO2 and PM are increased by releases from Canadian copper smelters and refineries and zinc plants. For those facilities where there is more than one monitoring site, the mean concentration of As, Cd, Cr, Ni, SO2 and PM is generally increased in relation to the proximity to the smelter. However, while the levels are elevated in this fashion at most of the copper smelters and refineries and zinc plants, particularly at those monitoring sites situated very close to the facility (i.e., less than 1 km), the mean concentration does not simply decline monotonically as a function of increasing distance. This is likely due to other factors that would influence dispersion of the emissions, including local meteorology and topography, as well as to the limited number of monitoring stations situated near all the facilities. It also appears that monitoring stations are often located in close proximity to local populations, which generally do not reside downwind of the copper smelters and refineries and zinc plants, rather than being situated where dispersion of the emissions can be tracked.
In addition, the airborne levels of each of these substances near Canadian copper smelters and refineries and zinc plants are consistently higher than regional background levels measured in areas removed from point sources. However, it is noted that there is considerable variation in the degree by which levels are increased between both substances and between facilities. Thus, concentrations of As, Cd and Pb are increased by up to approximately three orders of magnitude near some facilities, compared with more modest elevations in SO2 and in PM near all of the copper smelters and refineries and zinc plants. As well, levels of As, Cd and Pb are generally higher near those facilities where smelting is conducted compared with those where refining alone takes place, reflecting the lesser amounts of these metals emitted from refining (Table 4).
Hence, the results of monitoring near the Canadian copper smelters and refineries and zinc plants indicate that releases from these facilities result in increased potential for inhalation exposure (the route associated with the critical effect for these substances) to these and other substances.
The sizes and locations of local populations have not been characterized as part of this assessment, and the networks of monitoring stations are very limited for all of the facilities. Nonetheless, while the number of sites near each facility is very small, they are generally well situated with respect to local populations. Most of the monitoring stations are located in residential areas, and there is potential for exposure of the general population at the commercial and rural sites that comprise the bulk of the remainder. In addition, the available information indicates that, while the resident populations in many of these relatively isolated locales are not large, there are significant numbers of people (several thousand, and in some cases more than 100 000) residing within a few kilometres of virtually all of these facilities (SENES Consultants, 1996b; Fontana, 2000). In some instances, local communities are located within a few hundred metres of the smelters.
The focus of the health assessment is on evaluating the potential impacts of current releases of substances from copper smelters and refineries and zinc plants in Canada. To that end, the monitoring data from environmental media were restricted to those for ambient air, because levels in air were expected to reflect current releases much better than is the case for other media, which can be strongly influenced by high historical releases. The results of studies of environmental Pb near the Cominco lead smelter and zinc plant at Trail, B.C., provide support for this assumption (Hilts et al., 1998). In these studies, several lines of evidence indicated that air transport of re-entrained historical reservoirs of Pb was minimal compared with current emissions:
- The amount of Pb suspended in air was nearly four times higher when the wind blew predominantly from the smelter toward the sampling station compared with when it blew away.
- While total dustfall increased in summer months, when the ground is bare and the weather dry, the amount of Pb in dustfall was highest in winter months when emission dispersion conditions are poor.
- There were declines of up to 80-90% in airborne Pb and in dustfall Pb during a one-month shutdown of the smelter.
- Lead concentrations in dustfall were generally very high (>10 000 mg/kg), even higher than in the very fine fraction of soil.
Structural equations pathway modelling in this community explained 71% of the variation in blood lead in local children and indicated that the main direct contributor to blood lead was house dust lead loading and that environmental Pb passed from dust fall through street dust, soil and yard waste into house dust (Hilts et al., 1998). Following the introduction of a new lead smelter in 1997, which reduced emissions of Pb substantially, children's geometric mean blood lead concentrations declined by almost half, from 11.5 mg/dL in 1996 to 5.9 mg/dL in 1999 (Hilts et al., 1998; Hilts, 2000).
Thus, the results of studies near Trail confirm that both airborne levels and exposure of local populations to particulate metals are strongly influenced by current releases from the smelter. This would be expected to be even more pronounced near some of the other Canadian copper smelters and refineries and zinc plants, from which emissions of Pb and other metals are greater (in some instances, many times) than from the Trail smelter (Table 4).
3.3.2 Effects assessment
The epidemiological studies of human populations exposed to emissions from copper smelters and refineries and zinc plants in the environment are considered most relevant to the determination of "toxic" under Paragraph 64(c) of CEPA 1999, in terms of both the profile of substances to which they would have been exposed and the composition of the study populations (i.e., those exposed in the general environment, including the young, the elderly and compromised individuals).
However, with the exception of increased levels of lead in blood, the weight of evidence for health effects from epidemiological studies of populations in the vicinity of copper smelters and refineries and zinc plants is inadequate (Section 2.4.3). Even in the case of blood lead, while the most recent data from such populations in Canada indicate that roughly 10-20% of children surveyed had blood lead levels greater than or equal to the current intervention level of 10 mg/dL, such data are available only for a minority of the Canadian facilities, and most of these are at least several years old. In addition, children's current blood lead levels would reflect unknown contributions from both current and historical emissions of lead.
For these reasons, the results of the available epidemiological studies of populations resident near copper smelters and refineries and zinc plants are considered inadequate to characterize exposure response for both cancer and non-cancer effects.
For the substances for which recent data in ambient air near the Canadian facilities have been compiled in Section 2.3.1 (i.e., As, Cd, Cr, Ni, Pb, SO2 and PM), health assessments conducted under the PSL program and internationally are available. (These substances comprise the vast majority, on a mass basis, of those released to air from Canadian copper smelters and refineries and zinc plants [Tables 3, 4 and 5], as well as those considered a priori to be most relevant to health.) In selecting from among available health assessments for these substances, the criteria considered included whether the approach taken was consistent with the principles on which the PSL health assessments are based (e.g., whether the assessment was strictly health-based), whether the assessment was specific to the inhalation route of exposure, whether quantitative measures of exposure-response were developed, and how recently the assessment was conducted. On this basis, the assessments selected included those conducted for the PSL program for As (EC/HWC, 1993), Cd (EC/HC, 1994a), Cr (EC/HC, 1994c), Ni (EC/HC, 1994b) and respirable PM (EC/HC, 2000a), and in development of the WHO Air Quality Guidelines for Europe for Pb and SO2 (WHO, 2000).
In the next section, for each of the metals, SO2 and PM, a summary of an assessment of the substance under the PSL program or the WHO air quality guidelines program is presented. This includes a summary of the weight of evidence for the critical effect for each substance and the basis for the health-based measure of exposure-response or guidance value for the critical effect. It should be noted that the information provided is based entirely on the reports of these health assessments.14
For Priority Substances for which the weight of evidence of carcinogenicity is sufficient, where possible, estimated exposure is compared to quantitative estimates of carcinogenic potency to characterize risk and provide guidance for the establishment of priority for further action (i.e., analysis of options to reduce exposure). Potency is usually expressed as the dose or concentration that induces a 5% increase in the incidence or mortality due to relevant tumours (TD05 or TC05), based on data obtained in toxicological studies in experimental animals or epidemiological investigations in exposed human populations.
3.3.2.1 Exposure-response characterization for selected components of emissions from copper smelters and refineries and zinc plants
3.3.2.1.1 Arsenic
The following text, summarizing the PSL assessment for "arsenic and its compounds" (EC/HWC, 1993), has been taken from Hughes et al. (1994a):
An association between inhaled arsenic and increased mortality due to respiratory cancer has been consistently demonstrated in available epidemiological studies. In addition, ingestion of inorganic arsenic in drinking water has been consistently associated with an increased prevalence of skin cancer in exposed human populations with some indication of increases in mortality due to cancers of internal organs. Therefore, based on the weight of evidence of carcinogenicity in humans by more than one route of exposure, the group of inorganic arsenic compounds as a whole is considered to be carcinogenic to humans.
In the case of arsenic, potency estimates were developed for exposure by both inhalation and ingestion, based on epidemiological data. The TC05 for inhaled arsenic was based on data presented in the large studies of workers at the Tacoma smelter (Enterline et al., 1987), the Anaconda smelter (Higgins et al., 1986) and the Ronnskar smelter (Jarup et al., 1989), for which there was considerable information to serve as a basis for estimates of exposure. A negative exponential growth curve was used to describe the concave-downward relationship between concentrations of arsenic in air and mortality due to respiratory cancer (most of which were cancers of the lung) among workers for the Tacoma and Anaconda cohorts. This curve models the difference between a linear effect in exposure and a negative exponential term. A linear model was used to describe the relationship between exposure to arsenic and lung cancer mortality for the Ronnskar cohort. Excess risk of respiratory cancer was obtained using the predicted curves and age-adjusted lung cancer mortality rates for the Canadian population. Based on these data, the TC05s for inhaled arsenic were 7.8, 10 and 51 mg/m3 for the Anaconda, Tacoma and Ronnskar smelter workers, respectively.
3.3.2.1.2 Cadmium
The following text, summarizing the PSL assessment for "cadmium and its compounds" (EC/HC, 1994a), has been taken from Newhook et al. (1994):
Although an association between inhaled cadmium compounds and increased mortality due to lung cancer has been observed in some epidemiological studies, it is not possible to eliminate the potential influence of exposure to other heavy metals on these results. However, inhalation of cadmium chloride, oxide, sulphate or sulphide has induced lung cancers in several studies in rodents. Each of these compounds has also been carcinogenic in studies involving routes less relevant to environmental exposure i.e., subcutaneous or intramuscular injection, and cadmium chloride was carcinogenic in one of two adequate studies in which the compound was administered to rats in the diet. Concomitant exposure to zinc compounds reduced the carcinogenicity of inhaled cadmium oxide to rats (Glaser et al., 1990), and of cadmium chloride injected subcutaneously in rats and mice (IARC, 1976; Waalkes et al., 1989), indicating that it is most likely the cadmium ion itself which is carcinogenic.
On the basis principally of the results in inhalation studies in animals and supporting data on genotoxicity, inorganic cadmium compounds are considered to be probably carcinogenic to humans.
In the case of cadmium, the TC05 was derived from the data on lung cancers induced in rats by long-term inhalation of cadmium chloride aerosols (Takenaka et al., 1983); these data are considered to provide the most reliable estimate of the TC05, as a consequence of the clear dose-response relationship observed in this experiment for the incidence of total lung carcinomas (0 mg Cd/m3, 0/38; 13.4 mg Cd/m3, 6/39; 25.7 mg Cd/m3, 20/38; 50.8 mg Cd/m3, 25/35). The TC05, estimated by first fitting the multistage model to these data, and subsequently amortizing the exposure over the lifetime of the rat and adjusting to account for the duration of the experiment and the breathing volumes and body weights of rats and humans, is 5.1 mg Cd/m3. (TC05 values calculated from the total lung tumour incidences observed by Glaser et al. (1990) in rats inhaling cadmium chloride, cadmium oxide dust, cadmium sulphate, and cadmium sulphide are similar, ranging from 2.7 to 12.7 mg Cd/m3.)
3.3.2.1.3 Chromium
The following text, summarizing the PSL assessment for "chromium and its compounds" (EC/HC, 1994c), has been taken from Hughes et al. (1994b):
On the basis of its documented carcinogenicity in human populations exposed by inhalation in the occupational environment, the group of hexavalent chromium compounds as a whole is considered to be carcinogenic to humans. Available data are insufficient to support a hypothesized threshold for the carcinogenicity of hexavalent chromium, based on exceedence of the extracellular capacity to reduce hexavalent chromium to the trivalent species. Cellular uptake of trivalent chromium has been demonstrated (Alcedo and Wetterhahn, 1990) and the entry of hexavalent chromium into cells is rapid and extracellular reduction in the mucosal lining is incomplete (Witmer, 1991).
The TC05 was estimated on the basis of a study by Mancuso (1975), as this was the study in which the most information on exposure (inhalation) was provided. Although the cohort in this study was small, workers were classified into several categories of cumulative exposure to total chromium and soluble (principally hexavalent) or insoluble (principally trivalent) chromium. In addition, the period of follow-up was sufficiently long to account for the latency period of development of lung cancer. However, mortality by age group necessary for comparison with the general population was reported for total chromium only. Therefore, an estimate of the carcinogenic potency was derived based on exposure to total chromium.
The age-specific death rate for lung cancer was assumed to be a time-weighted quadratic function of exposure to chromium, which is additive to the death rate for the general population assumed not to be exposed to chromium. The increase in probability of death due to constant lifetime exposure to chromium was determined, based on the assumption that there are no competing causes of death and exposure is constant for a period equal to the median survival time of 75 years. The TC05 for inhaled chromium (total) was estimated to be 4.6 mg/m3.
An indirect estimate of the carcinogenic potency of hexavalent chromium may be derived from the study by Mancuso (1975). In an earlier study at the same chromate production plant, it was reported that the proportion of trivalent to hexavalent chromium present in most areas of the plant was about 6:1 or less (Bourne and Yee, 1950), although the number of workers in each area of the plant was not specified. Thus, the concentrations of hexavalent chromium may be estimated to be one seventh (1/7) of the reported concentrations of total chromium. Based on this assumption, the TC05 for hexavalent chromium has been estimated to be 0.66 mg/m3.
3.3.2.1.4 Nickel
The following text, summarizing the PSL assessment for "nickel and its compounds" (EC/HC, 1994b), has been taken from Hughes et al. (1994c):
There is sufficient and consistent evidence of the carcinogenicity of each of oxidic, sulphidic and soluble nickel in adequate epidemiological studies in different types of exposed workers and some weak evidence of genotoxicity in limited epidemiological studies. Although there may have been concomitant exposure to other compounds in these studies, the common predisposing factors in the various groups of workers examined appear to be these groups of nickel compounds. In addition, there is some supportive evidence of carcinogenicity and genotoxicity of these forms of nickel in principally limited studies in animal species. Therefore, each of oxidic, sulphidic and soluble nickel is considered to be carcinogenic to humans.
The epidemiological studies which provide sufficient information to serve as a basis for quantitative estimation of the carcinogenic potency of inhaled inorganic nickel are those of large cohorts (n = 3250 to 54 509) of exposed workers at two nickel refineries for whom the most extensive information on exposure is available: the Inco mining, smelting and refinery operations in Ontario and the Falconbridge refineries in Kristiansand, Norway (Doll et al., 1990). Estimates of the carcinogenic potency of oxidic, sulphidic and soluble nickel (combined), based on results at the Inco mining, smelting and refinery operations in Ontario, were considered the most relevant and reliable for several reasons: the cohorts were relatively large (e.g., total expected numbers of death of Copper Cliff sinter plant workers with 15 or more years since first exposure due to lung cancer was approximately 20); there was clear evidence of increased lung and nasal cancer mortality with increasing duration of exposure in the sinter workers and there was not exposure to metallic nickel (i.e., the estimates of total nickel concentrations did not include a form of nickel for which there is no convincing evidence of carcinogenicity). Although the potency of the various species may vary considerably, the TC05s estimated on the basis of the Inco cohort are based on oxidic, sulphidic and soluble nickel (combined) since available data do not permit separate estimates for each of the groups of compounds.
The Kristiansand cohort consisted of two clearly defined working groups (i.e., electrolysis workers with no employment in other high exposure departments and those employed in the roasting, smelting and calcining department). There was little exposure to metallic nickel in both groups. Based on the data presented for these workers, TC05s were developed for oxidic, sulphidic and soluble nickel (combined) and soluble nickel (specifically).
The age-specific death rate for lung cancer observed in the cohorts of the Copper Cliff sinter plant and Coniston sinter plant was assumed to be a linear function of the cumulative exposure to total nickel, whereas the age-specific death rate for lung cancer reported in the cohorts of the Port Colborne nickel refinery and Kristiansand nickel refinery was assumed to be an exponential function of the cumulative exposure to total nickel. The age-specific death rate was also assumed to be multiplicative to the death rate for the general population. The increase in probability of death due to a constant lifetime exposure to nickel has been determined, based on the assumption that there are no competing causes of death and a constant exposure for a period equal to the median survival time of 75 years. The estimates of the TC05 for inhaled oxidic, sulphidic and soluble nickel (combined) for lung cancer mortality ranged from 0.04 to 1.0 mg/m3. It should be noted that the TC05s based on data for workers in the Clydach refinery (although the numbers of workers in each occupational group were small) would not be substantially different. The TC05 for lung cancer mortality for soluble nickel was within this range of values (i.e., 0.07 mg/m3).
3.3.2.1.5 Lead
The following text is based on a recent review of lead produced for the "WHO Air Quality Guidelines for Europe" (WHO, 2000).
A variety of effects has been documented in humans exposed to lead, both occupationally and environmentally. In conditions of low-level long-term exposure, as for the general population, the most critical effects include those on heme biosynthesis, erythropoiesis and the central and peripheral nervous systems. The results of animal studies provide support for lead as the causative agent for these effects.
Children up to six years of age are considered to be more at risk for lead exposure and effects compared to adults for several reasons, including their lesser concern for personal hygiene and increased hand-to-mouth activity, substantially higher absorption in the gastrointestinal tract, a less developed blood-brain barrier, and lower thresholds for hematological and neurological effects of lead. In children, Lowest-Observed-Adverse-Effect Levels (LOAELs) for hematological and neurobehavioural endpoints have been summarized as follows. Reduced hemoglobin levels have been observed at blood lead concentrations around 40 mg/dL. Hematocrit values below 35% have not been reported at levels below 20 mg/dL; this is also true for several enzyme systems that may have clinical significance. Effects on the central nervous system occur at levels below 20 mg/dL; consistent effects have been reported for measures of cognitive functioning such as the psychometric IQ between 10 and 15 mg/dL, and in some studies below 10 mg/dL.
Based on the above information, the WHO (2000) identified a critical level of lead in blood of 10 mg/dL. This value was then used to derive an ambient air quality guideline, as follows. It was recommended that efforts be taken to ensure that at least 98% of the general population, including preschool children, should have blood lead levels that do not exceed 10 mg/dL. The corresponding median blood lead level was estimated at 5.4 mg/dL, compared with currently measured "baseline" blood lead levels of minimal anthropogenic origin of up to 3.0 mg/dL. The air quality guideline was calculated as the concentration of lead in air that was estimated to yield the difference (2.4 mg/dL). Based on regressions between levels of lead in ambient air and in blood, which indicate that 1 mg Pb/m3 directly contributes approximately 1.9 mg Pb/dL in blood, and calculating the indirect contribution through dust/soil, it was estimated that 1 mg Pb/m3 would contribute 5 mg Pb/dL blood (summarized in WHO, 1995). On this basis, an ambient air guideline of 0.5 mg Pb/m3 (annual average) was derived.
3.3.2.1.6 Sulphur dioxide
The following text is based on reviews of SO2 produced for the "WHO Air Quality Guidelines for Europe" (WHO, 1987, 2000).
Information on effects of exposure to SO2 averaged over a 24-hour period is based mainly on epidemiological studies in which the effects of ambient mixtures of SO2, PM and other associated pollutants are considered. Respiratory morbidity in patients with pre-existing conditions (asthmatics, bronchitics) was consistently observed when SO2 concentrations exceeded 250 m g/m3. This occurred in situations in which the air pollution arose principally from the inefficient burning of coal in domestic appliances. In several more recent studies involving the mixed industrial and vehicular sources that now dominate, increased mortality (total, cardiovascular and respiratory) and increased emergency department admissions for total respiratory causes and for chronic obstructive pulmonary disease (COPD) were observed at lower levels of exposure (mean annual levels below 50 mg/m3, and daily levels usually less than 125 mg/m3). The association with SO2 levels remained, in some instances, when Black Smoke and TSP were controlled for. There were also small effects on lung function at concentrations of SO2 below 300 m g/m3 in some studies, though it was difficult to separate out the effects of other pollutants.
With respect to the effects of longer-term exposures, in earlier studies during the coal-burning era, there were increased frequencies of respiratory symptoms and illnesses or effects on lung function associated with annual average concentrations of SO2 of 100 m g/m3 or more, in combination with other pollutants. The results of more recent studies have indicated adverse effects below this level, though it is not clear to what extent the findings may have been related to the different pollutant profile of earlier years. Cohort studies of differences in mortality between areas with contrasting pollution levels indicate that there is a closer association with particulate matter than with SO2.
Applying a two-fold uncertainty factor to the LOAEL reported above yielded ambient air quality guideline values of 125 mg/m3 and 50 mg/m3 for 24-hour and annual periods, respectively (WHO, 1987). In more recent studies, adverse effects were observed at lower levels of exposure. However, these values were retained in the recent revision of the guidelines (WHO, 2000), because of uncertainty as to whether SO2 was the responsible pollutant or merely a surrogate for some other correlated substance.
3.3.2.1.7 Respirable particulate matter (PM10)
The following text, summarizing the PSL assessment for "respirable particulate matter less than or equal to 10 microns," has been taken from EC/HC (2000a):
In numerous epidemiological studies from around the world, including Canada, positive associations have been observed between ambient levels of particulate matter (as PM10, PM2.5 or other particle metrics) and a range of health outcomes, including daily mortality, respiratory and cardiovascular hospitalizations, impaired lung function, adverse respiratory symptoms and medication use, restricted activity days and the frequency of reported chronic respiratory diseases. These associations could not be explained by the influence of weather, season, yearly trends, day-to-day variations or variations due to holidays, epidemics or other non-pollutant factors. While the populations studied were always exposed to other air pollutants in addition to particulate matter, associations of a similar magnitude were observed across numerous locations with differing air pollutant mixtures, and the association with particulate matter remained in analyses that adjusted for the effects of various other pollutants. These particulate matter-related health effects were observed at ambient concentrations that currently occur in Canada.
Therefore, the epidemiological evidence of mortality and morbidity in response to current levels of particulate air pollution meets a number of the criteria for causality, including consistency, dose-response relationship, coherence, temporal relationship and specificity of both outcome and agent. With respect to the biological plausibility of the association, the results of experimental studies in animals and humans provide some limited support for the epidemiological findings. However, both animal and human experimental work is constrained by the technological difficulties in reproducing environmentally relevant particulate matter, and this work has generally been conducted at high levels with artificial particles. Some of this work, and specifically the most recent work with concentrated ambient particles, has provided initial evidence of particulate matter-induced effects on the cardiorespiratory system, particularly in individuals with pre-existing respiratory and cardiovascular disease, and has provided preliminary indications of possible mechanisms. The database supports, therefore, a causal relation between current ambient PM10 and PM2.5 exposure and adverse health effects.
Table 42 summarizes the magnitude of the health effects associated with ambient particulate matter in the epidemiological studies, as the percentage increase in risk per 10 mg/m3 of PM10 for each endpoint. The average concentration of PM10 at which each PM-associated health effect has been observed in the studies reviewed in WGAQOG (1999) and in EC/HC (2000a) is also included.
3.3.3 Risk characterization
In this section, the risk posed to nearby populations by exposure to the various substances released from Canadian copper smelters and refineries and zinc plants has been characterized, by relating the concentrations in ambient air near these facilities to the health-based guidelines or measures of exposure-response for each substance. Based on the critical effects for each of the substances summarized in the previous section, the potential risks from As, Cd, Cr and Ni are considered together, followed separately by each of Pb, SO2 and PM.
Releases from copper smelters and refineries and zinc plants include complex mixtures of substances, including SO2 and numerous heavy metals. It is known that some of the components of these releases can interact in inducing toxic effects; for example, simultaneous exposure to Zn and a number of other elements is known to protect against the toxicity of Cd, and SO2 may enhance the respiratory carcinogenicity observed in workers at non-ferrous metal ore smelters (Krishnan and Brodeur, 1991). However, the available data are inadequate to characterize possible interactions among the numerous substances contained in releases from copper smelters and refineries and zinc plants, and in the following risk characterization, it is assumed that there is no interaction. In the case of those substances that are lung carcinogens, this amounts to assuming additivity.
3.3.3.1 Arsenic, cadmium, chromium and nickel
As summarized in Section 3.3.2, carcinogenicity is considered to be the critical effect for As, Cd, Cr and Ni, based on the sufficient weight of evidence for pulmonary carcinogenicity in occupational populations or experimental animals following inhalation of inorganic compounds of each of these metals. For such substances, estimates of exposure are compared with quantitative estimates of cancer potency to derive an Exposure Potency Index (EPI), in order to characterize risk and provide guidance in establishing priorities for further action (i.e., analysis of options to reduce exposure) under CEPA 1999 (Health Canada, 1994). The derivation of the relevant potencies (i.e., TC05s) for each of the metals is described in Section 3.3.2.
For each monitoring site at each Canadian copper smelter and refinery and zinc plant, a total EPI was developed as a measure of lung cancer risk (Table 43), as follows. A separate EPI was first calculated for each metal, as the ratio of the annual average concentration to the TC05 for lung cancer mortality/incidence, and then these values were summed for each site.
For those metals for which more than one TC05 was available (i.e., all but Cd), the values presented in the table are based on the lowest value; the impact of using these values on the total EPI was modest (i.e., most often four- to five-fold). The total EPI at a given site included only estimates for those metals for which data were available; based on data from those sites where monitoring was conducted for all four metals, the impact of the data for the individual metals that were most often missing for the other sites (i.e., Cr and Ni) was five-fold or less at most sites. For those facilities that are combined sources (i.e., the Inco nickel-copper smelter, copper refinery and nickel refinery in Sudbury, the Falconbridge-Kidd Creek copper smelter and refinery and zinc plant in Timmins, the Cominco lead smelter and zinc plant in Trail, and the HMB&S smelter in Flin Flon), that portion of the EPI that was attributable to the operation(s) that are the subject of this assessment was estimated, based on the source attribution presented in Table 17.
Table 43 Total Exposure Potency Index for lung cancer mortality at sites near Canadian copper smelters and refineries and zinc plants
The total EPI values and the corresponding margins between carcinogenic potency and estimated exposure for each monitoring site in the vicinity of Canadian copper smelters and refineries and zinc plants are presented in Table 43. Based on these margins, the priority for investigation of options to reduce exposure is considered to be in the high range for copper smelters, to range from low to high for copper refineries, and to range from low to high for zinc plants. In general, the margins are smallest near copper smelters, largest near copper refineries, and intermediate near zinc plants, although there is considerable variation among facilities of a given type (i.e., two orders of magnitude or more).
3.3.3.2 Lead
Increased exposure of children to Pb has been observed near copper smelters and zinc plants around the world, including some Canadian facilities (Section 2.4.3.2). The concentration of lead in blood is the most widely used and most generally accepted measure of dose, and there is extensive evidence linking blood lead levels to a variety of health effects. However, while there are data on blood lead levels in populations, mostly children, in the vicinity of some facilities in Canada (and some indication that the prevalence of excessive exposures in these populations studied has declined), these are not considered suitable as a basis for assessing risks to health from current releases of Pb. This is principally because the available data are inadequate to distinguish the contribution of current versus historical emissions of Pb to children's current blood lead levels.
Instead, the potential for health effects from exposure to current releases of Pb from Canadian copper smelters and refineries and zinc plants has been assessed by comparing recent data on levels of Pb in ambient air near these facilities to the WHO ambient air quality guideline of 0.5 mg/m3 (annual average). The annual mean concentrations of Pb in ambient air are elevated above regional background near all of the Canadian facilities (Table 24), although levels exceeding the WHO guideline occur near just a few of the facilities. These include two of the copper smelters (Noranda-Horne and HBM&S), where mean concentrations of Pb at some sites are elevated above the guideline, sometimes by a considerable margin. This appears to be a combined result of the proximity of these monitoring sites to the smelter and the substantial quantities of Pb emitted from these facilities (Table 4). The guideline is also slightly exceeded at one site near the Falconbridge-Kidd Creek smelting and refining complex. Although this facility also includes a copper refinery, zinc refinery and concentrator, virtually all of the Pb emitted from this facility is released from the copper smelter (Table 4). These results indicate the potential for lead-induced health effects, particularly neurodevelopmental and hematological effects, in populations in the vicinity of certain of the Canadian facilities involved in smelting copper.
3.3.3.3 Sulphur dioxide
The ambient 24-hour concentrations of SO2 in the vicinity of Canadian copper smelters and refineries and zinc plants are elevated. These increased levels are also reflected in exceedences of the 24-hour WHO Air Quality Guideline for Europe for SO2 of 125 m g/m3 (WHO, 2000), intended to protect sensitive individuals against health effects. While the guideline is exceeded occasionally near all of the facilities, this occurs most often in the vicinity of certain of those where copper is smelted. (The long-term WHO guideline of 50 mg/m3 was not exceeded near any of the Canadian copper smelters and refineries or zinc plants.) SO2 is also oxidized to sulphate particles in the environment, and as noted in the next section, an association between adverse health effects and airborne concentrations of respirable PM similar to those in the vicinity of Canadian copper smelters and refineries and zinc plants has been observed in numerous epidemiological studies. On this basis, there is some potential for SO2-induced cardiorespiratory health effects in individuals with pre-existing conditions (e.g., asthmatics) near these facilities. While some of the Canadian facilities include process streams that are not the subject of these assessments (e.g., the Cominco lead operations in Trail), the emissions of SO2 from the combined sources are estimated to be principally or entirely due to the copper smelters or zinc plants (i.e., between 80% and 100%; Table 10).
3.3.3.4 Particulate matter
Given the considerable uncertainties in such factors as the estimates of dose-response for the various health outcomes associated with exposure to PM and the background concentrations of PM in the regions of Canada where copper smelters and refineries and zinc plants are located, no attempt has been made to estimate the potential magnitude of health impacts of PM in the area near these facilities. However, it is noted that the range of annual mean concentrations of PM10 near the Canadian facilities overlaps the range of mean concentrations (most often averaged over a year or more) from epidemiological studies in which exposure to PM10 has been associated with a variety of adverse health effects (Table 25 and Table 42). For those facilities that are combined sources, most of the particulate emissions are expected to arise from the copper smelting or zinc plants, with the exception of the Cominco facility in Trail, where 1995 data indicate that approximately 90% was associated with lead processing (RDIS, 1995).
3.3.4 Uncertainties and degree of confidence in human health risk characterization
The exposure assessment was based on recent monitoring data for those substances that comprise the vast majority of current emissions and was specific to the environmental medium most relevant to the critical effects of exposure to these substances.
Nonetheless, there remains a fair degree of uncertainty in the exposure assessment for the health assessment of releases from copper smelters and refineries and zinc plants. The network of monitoring stations is very limited near all of the Canadian facilities, being small in number and apparently located near local populations, rather than being placed at points of impingement or located so as to track the dispersion of the emissions.
In addition, only a small number of substances was considered, limited to those that professional judgement indicated were most likely to be of concern and for which recent relevant assessments were known to be available. A large number of other substances not considered in these assessments are known to be released from these facilities, and risks to health may have been underestimated as a result.
On the other hand, some of the substances, most notably PM10, are not specific to the facilities that are the subject of these assessments, and some of the facilities, such as the Noranda-CCR copper refinery, are located near other major industrial operations; hence, other sources may have contributed substantially to the concentrations measured near some of the copper smelters and refineries and zinc plants.
Also with respect to PM10, it should be noted that most of the values used for this variable were estimated from the TSP data, rather than being measured directly; it is likely that the relative size distribution of airborne particles at a given location will vary depending on origin, composition and other factors affecting deposition rates, although the available limited data indicated that the long-term average concentration of PM10 estimated in this fashion was very similar to concomitant measurements of PM10. Moreover, this would not materially affect the assessment for this parameter, which is somewhat qualitative in any case.
There are no quantitative data on the species of metals present in ambient air near Canadian copper smelters and refineries and zinc plants. It is known that the various chemical species of a given metal can differ markedly in bioavailability and toxicity. Speciation was addressed to the extent possible in the PSL assessments for As, Cd, Cr and Ni and was clearly identified as an important information gap; however, these data do not appear to have been generated in the interim.
The overall degree of confidence in the exposure assessment is, therefore, moderate, owing principally to the limitations in the existing monitoring network near Canadian copper smelters and refineries and zinc plants.
There is also a fair degree of uncertainty in the characterization of effects for the health assessment of releases from these facilities. The principal uncertainty is the lack of meaningful direct data on effects on local populations of the mixture of substances released from copper smelters and refineries and zinc plants in Canada, which is the reason the scope of the health assessment is necessarily limited. This limited scope involved reliance on other assessments for information on exposure-response for the large number of components of releases that were considered; though these were not updated, the authors of this assessment are not aware of new data for these substances that would impact significantly on conclusions drawn under CEPA 1999 64(c).
With respect to those components of releases that affect the same endpoint (i.e., lung cancer), it has been assumed that there is no interaction among them, even though, for example, there is evidence that SO2 enhances the respiratory carcinogenicity of As. In addition, there is a lack of monitoring data for some carcinogenic metals (i.e., Cr and Ni) near some Canadian copper smelters and refineries and zinc plants. As a consequence of these factors, risks may have been underestimated. This is offset somewhat by the use of the most conservative TC05 values, though these have only a modest impact on the margin between potency and exposure.
Confidence in the effects assessment is increased by the fact that the critical effects for some substances (i.e., Pb, SO2 and PM10) have been determined based on epidemiological studies that were conducted at ambient levels of pollutants in the same range as those observed near the Canadian facilities considered in this assessment (though not necessarily the same mixtures of pollutants as for copper smelters and refineries and zinc plants) and on populations that included critical subpopulations in terms of exposure and sensitivity.
While releases from copper smelters and refineries and zinc plants can result in high blood lead levels, there is a lack of recent data on blood lead levels near all but one of the Canadian facilities, and there is inadequate information on the contribution of current emissions versus re-entrainment of historical deposits to these levels.
Overall, the degree of confidence in the effects assessment is considered to be low to moderate, owing principally to lack of data concerning the effects of environmental exposure to mixtures of substances emitted from copper smelters and refineries and zinc plants on local human populations.
3.4 Conclusions
3.4.1 Releases from copper smelters and refineries
CEPA 1999 64(a): Based on available data, it has been concluded that emissions from copper smelters and refineries of metals (largely in the form of particulates) and of sulphur dioxide are entering the environment in quantities or concentrations or under conditions that have or may have an immediate or long-term harmful effect on the environment or its biological diversity. Therefore, metals (largely in the form of particulates) contained in emissions from copper smelters and refineries and sulphur dioxide are considered "toxic" as defined under Paragraph 64(a) of CEPA 1999.
CEPA 1999 64(b): It has been concluded that emissions from copper smelters and refineries are not entering the enironment in quantities or concentrations or under conditions that constitute or may constitute a danger to the environment on which life depends.
Therefore, emissions from copper smelters and refineries are not considered "toxic" as defined under paragraph 64(b) of CEPA 1999.
CEPA 1999 64(c): Based on available data, concerning the effects of PM10, sulphur dioxide and compounds of arsenic, cadmium, chromium, lead and nickel, it has been concluded that emissions from copper smelters and refineries of PM10, of metals (largely in the form of particulates) and of sulphur dioxide are entering the environment in quantities or concentrations or under conditions that constitute or may constitute a danger in Canada to human life or health. Therefore, metals (largely in the form of particulates) contained in emissions from copper smelters and refineries, PM10 and sulphur dioxide are considered "toxic" as defined under Paragraph 64(c) of CEPA 1999.
Overall conclusion: Based on critical assessment of relevant information, metals (largely in the form of particulates) contained in emissions from copper smelters and refineries, PM10 and sulphur dioxide are considered "toxic" as defined in Section 64 of CEPA 1999.
3.4.2 Releases from zinc plants
CEPA 1999 64(a): Based on available data, it has been concluded that emissions from zinc plants of metals (largely in the form of particulates) and of sulphur dioxide are entering the environment in quantities or concentrations or under conditions that have or may have an immediate or long-term harmful effect on the environment or its biological diversity. Therefore, metals (largely in the form of particulates) contained in emissions from zinc plants and sulphur dioxide are considered "toxic" as defined under Paragraph 64(a) of CEPA 1999.
CEPA 1999 64(b): Based on available data, it has been concluded that emissions from zinc plants are not entering the environment in quantities or concentrations or under conditions that constitute or may constitute a danger to the environment on which life depends. Therefore, emissions from zinc plants are not considered "toxic" as defined under paragraph 64(b) of CEPA 1999.
CEPA 1999 64(c): Based on available data concerning the effects of PM10, sulphur dioxide and compounds of arsenic, cadmium, chromium, lead and nickel, it has been concluded that emissions from zinc plants of metals (largely in the form of particulates), of PM10 and of sulphur dioxide are entering the environment in quantities or concentrations or under conditions that constitute or may constitute a danger in Canada to human life or health. Therefore, metals (largely in the form of particulates) contained in emissions from zinc plants, PM10 and sulphur dioxide are considered "toxic" as defined under Paragraph 64(c) of CEPA 1999.
Overall conclusion: Based on critical assessment of relevant information, metals (largely in the form of particulates) contained in emissions from zinc plants, PM10 and sulphur dioxide are considered "toxic" as defined in Section 64 of CEPA 1999.
3.5 Considerations for follow-up (further action)
The assessment of risk to the environment was based on emissions to air of Cu, Zn, Ni, Pb, Cd and As (largely in the form of particulates) as well as SO2, while the health assessment included the same metals less Cu and Zn, plus Cr, SO2 and PM. These constituents were selected for evaluation from the complex combination of substances released from smelters and refineries, as these generally represent the substances released in the greatest quantity. This does not imply that other constituents do not pose a risk.
Thus, investigations of options for risk management should also take into consideration other substances of potential concern, some examples of which include Hg, Se, dioxins and furans. It should be noted in particular that, as a class, the facilities considered in these assessments are the largest source of Hg emissions in Canada. In 1995 (the most recent year for which comprehensive Canadian Hg emissions data are available), copper smelters and zinc plants emitted a total of about 3.8 tonnes of Hg. This represents about 35% of the 11 tonnes emitted by all anthropogenic sources in Canada in 1995 (data summarized in CED, 2000).
Risk to the environment due to aquatic releases was evaluated for only three of the facilities considered in these assessments. This was done because restrictions of time and resources precluded site-specific evaluation of aquatic releases from all facilities. In addition, aquatic releases from all six of the facilities not assessed are mixed with mining effluent prior to release to surface waters. As a result, their effluents fall under the Metal Mining Liquid Effluent Regulations and Guidelines (MMLER), passed in 1977 under the Fisheries Act.
Currently, these facilities are only subject to Guidelines under MMLER. However, all six facilities will have to conform to the revised Metal Mining Effluent Regulations (MMER), anticipated to come into effect in 2002. The MMER will include a requirement for EEM. It should be clearly stated that the exclusion of these facilities from site-specific risk assessment of aquatic releases does not imply that their effluents do not pose a risk to the environment. It is also of significance that aquatic releases from base metal smelting facilities are the subject of a number of other ongoing and planned risk management initiatives. Any investigations of options to reduce exposure as a result of the assessment of releases from copper smelters and refineries and zinc plants as Priority Substances under CEPA 1999 should also be integrated with these initiatives.
Screening-level risk assessment of aquatic releases from the three facilities evaluated (Noranda-CCR, Noranda-CEZinc and Cominco-Trail) indicated the potential for detrimental effects on the environment. The indicators of risk, based on the limited data available, were relatively low, especially given the slightly conservative nature of the screening assessment. Given existing controls on effluents put in place by the companies or imposed by Provincial governments or other authorities, Federal prevention or control actions under the Canadian Environmental Protection Act, 1999 (CEPA, 1999) are not recommended at this time. It is believed, however, that an increase in contaminant concentrations or loadings or changes in conditions affecting bioavailability (such as pH) have the potential to significantly increase risk to the environment. It is important that facility operators recognize that if information, such as monitoring data, shows a significant increase in contaminant concentrations or loadings or changes in conditions affecting bioavailability, such information may be subject to reporting under Section 70 of CEPA, 1999.
Comparison of estimated exposure to arsenic, cadmium, chromium and nickel in the vicinity of Canadian copper smelters/refineries and zinc plants with the tumorigenic potency indicates that the priority for investigation of options to reduce human exposure to releases from these facilities is considered to be in the high range for copper smelters, to range from low to high for copper refineries, and to range from low to high for zinc plants. Comparison of levels of lead, SO2 and PM10 in ambient air with health- based guidelines or with concentrations at which health effects have been observed also suggests that the priority for options analysis is high, especially for facilities where copper is smelted.
As a result of the Base Metals Smelting Sector Strategic Options Process, there are ongoing toxics initiatives designed to address air and water releases of inorganic As compounds, inorganic Cd compounds, dioxins and furans, Pb, Hg and oxidic, sulphidic and soluble inorganic Ni compounds from the base metal smelting sector. An assessment of options to reduce exposure as a result of the assessment of releases from copper smelters and refineries and zinc plants as Priority Substances under CEPA 1999 should be integrated with those for this ongoing initiative.
In addition, there are ongoing initiatives to control and reduce emissions of SO2 from major industrial sources in Canada. Any investigations of options to reduce exposure as a result of the assessment of releases from copper smelters and refineries and zinc plants as Priority Substances under CEPA 1999 should also be integrated with these initiatives.
Respirable PM less than or equal to 10 microns was the subject of a separate PSL risk assessment and was found to be "toxic" as defined in Section 64 of CEPA 1999. It was subsequently added to the list of toxic substances in Schedule 1 of CEPA 1999. As recognized in that assessment, SO2 is one of the major precursors in the secondary formation of PM2.5.
Risk from exposure to respirable PM is one consideration that has contributed to the proposal that releases from copper smelters and refineries and zinc plants be considered "toxic" under CEPA 1999. In determining any risk management measures to reduce exposure to respirable PM originating from these facilities, the fact that the facilities are major sources of SO2 must be recognized.
It should also be noted that since source attribution has been based on an incomplete inventory of emission sources that are not directly associated with copper smelters and refineries and zinc plants, the proportion of emissions estimated to have come from copper smelters and refineries or from zinc plants may be an overestimate. Sources not well represented in current inventories include, for example, emissions related to the production and transport of concentrates, as well as metal-laden dust blown from uncovered tailings piles. These sources augment exposure by contributing to local background metal concentrations. Any investigations of options to reduce exposure as a result of these assessments should take these other less well characterized sources into consideration. Further, inconsistencies between facilities in reporting of emissions have been a source of uncertainty in these assessments. More stringent standards for reporting, such as is suggested in the "Strategic Options Report for the Base Metal Smelting Sector" (Environment Canada, 1997b), and perhaps a greater level of industry accountability in future emissions reporting are warranted.
14 While there was no formal search strategy to identify recent data that may have impacted on the outcome of these assessments, the authors are not aware of new data that would impact significantly on the conclusions drawn under Paragraph 64(c).
