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
2.0 Summary of Information Critical to Assessment of "Toxic" under CEPA 1999 (Continued)
2.3.1.1.3 Deposition of sulphate from air
As wet deposition monitoring is continuous, all results obtained over the period of one year were summed. The totals were normalized to a full year to account for any missing data. Values were converted to annual deposition rates (mg/m2/a), summed with the dry deposition values for the corresponding year, and converted to total soluble annual deposition rates using the appropriate solubility factor for each metal.
OME data for the Sudbury region determined as the sum of wet and dry deposition are shown in Table 18. IADN data were used only for calculation of regional background metal deposition (discussed later in this section).
Snowpack: Monitoring of metals in snowpack samples near the Noranda-Horne smelter was conducted by the Geological Survey of Canada (GSC) in the winter of 1997-98. Snow cores were collected over a three-day period at 82 locations, mostly at 3-km intervals along three transects, extending 50 km south, northeast and northwest from the smelter. Samples were thawed at low temperature (4° C) and filtered through 0.45-mm membranes, separating dissolved and particulate fractions. Each fraction was analysed for metal content, from which deposition rates were calculated. A detailed description of sample collection and analysis is provided in Kliza et al. (2000).
It should be noted that due to the method of handling the snowpack samples (filtration at 4° C a short time after thawing), the "dissolved" concentrations determined, and subsequent estimates of soluble deposition, are likely lower than those that would be determined after longer periods of time at higher temperatures.
Table 18 Deposition of soluble metals in the vicinity of copper and zinc production facilities - based on snowpack and combined ("wet plus dry") deposition sampling
The relation between dissolved deposition rate and distance from the facility for each metal was characterized by the GSC (personal communication, G. Bonham-Carter) by fitting an exponential curve to the empirical data. Results for the "dissolved" fraction based on this study are shown in Table 18. All results shown are based on the fitted curves. Deposition rates have been calculated at each of several arbitrarily chosen radii. Dissolved deposition rates determined from the equations were converted from ng/cm2/winter (the units of deposition used in Kliza et al., 2000) to mg/m2/a, assuming that the snowpack represents deposition over three months of the year.
Relative reliability of empirical deposition data: It is recognized that estimation of total deposition based in whole or in part on TSP data has significant uncertainty. There is uncertainty both in the deposition velocity assumed and in the fraction of total deposition due to dry deposition. It is further recognized that both of these can vary considerably as a function of distance from emission sources. Comparison of annual deposition estimated from TSP data with the more reliable annual deposition measured in dustfall samples indicates that TSP-based estimations are generally low by factors of 2-5.
There are also uncertainties associated with other methods. For example, dustfall monitors are known to slightly underestimate deposition due to their poor collection efficiency for very small particles. Estimation of annual deposition from snowpack monitoring assumes that deposition during the winter is representative of the entire year and that sampled cores contained one-quarter of the annual deposition. Based on these factors, the relative reliability of total deposition estimates considered in this report is likely of the order dustfall ≥snowpack >dry +wet >TSP.
Source attribution: The "Combined sources" sections of Tables 15, 17 and 18 include information on source attribution. Source attribution is the percent contribution of separate operations to total emissions from the facility. It will be used as an estimation of how much each emission source may be contributing to monitored ambient or deposited metal concentrations. Source attributions of metal emissions for the combined facilities were determined as follows:
- The HBM&S facility in Flin Flon includes a zinc plant and a copper smelter. Due to the pressure leach process used, metal emissions from the zinc plant are negligible. Therefore, all metal emissions are attributed to the copper smelter.
- The Cominco-Trail facility includes a lead smelter and a zinc plant. Attribution was determined based on 1998 emission data provided by Cominco (personal communication with facility operators). These data reflect the significant process changes that took place at the facility in 1997.
- The Inco-Copper Cliff facility includes a nickel/copper smelter, a copper refinery and a nickel refinery. Source attribution was determined using 1995 and 1996 NPRI (1995, 1996) emission data. No major process changes are believed to have occurred at this facility since 1995, and the differences in apportionment calculated for 1995 and 1996 were considered to reflect normal year-to-year variability. The average of these two years was used.
- The Sudbury region includes the Inco facility described above, as well as the Falconbridge facility. As these facilities share an airshed, results for monitoring in this region are influenced by the presence of the two. Attribution was estimated in the same way as for the Inco facility but with the inclusion of emissions from the nickel/copper smelter at Falconbridge.
- The Falconbridge-Kidd Creek facility includes a copper smelter, copper refinery, zinc plant and concentrator. Source attribution was determined based on 1995 emission data provided by Falconbridge facility operators. No major process changes are believed to have occurred at this facility since 1995. Emissions indicated as relating to storage and handling were equally distributed between the copper smelter and zinc plant. Emission of As as arsine from the zinc plant was included in the calculations.
It should be noted that these attributions ignore background (natural, regional, local) contributions to estimated deposition rates. This omission may be significant at some facilities that include major emission sources that are not subject to these assessments (e.g., wind-blown material from uncovered mill tailings). Details of the source attribution calculations are provided in CED (2000).
Estimation of soluble fraction: Environmental exposure in this assessment focuses on bioavailable metals that have been deposited to soils or surface waters in particulate or dissolved form. The assumption is made that the bioavailable portion of the metal is the free metal ion, which may be estimated from the fraction of deposited metal that is water-soluble.9 As monitoring methods generally provide information on total deposition only, some means of estimating the soluble fraction is required. A limited number of data sources allowing calculation of the water-soluble fraction of deposited metals were identified. Each is described below, and results are summarized in Table 19.
Monthly monitoring of dustfall at Butler Park, located 1.3 km east of the Cominco-Trail facility, has been conducted by the B.C. Ministry of Environment (MOE) since 1971. The procedure includes filtration of the samples through 0.45-mm membrane filters followed by analysis of both the dissolved and particulate fractions. The water-soluble fraction was calculated using data from the years 1989-1997.
When analysing TSP samples, the OME extracts the collection filters first with water, then with acid. Both fractions are analysed for metals. Data were obtained from the OME for 141 samples collected from 20 different monitoring stations in Ontario in 1995 and 1996. Samples collected at sites located within 100 km of Sudbury were treated as "near-field," while those collected further than 100 km were treated as regional background for the Canadian Shield.10
During the winter of 1997-98, the GSC collected snowpack samples from 82 sites within 50 km of the Noranda-Horne smelter. The samples were melted and filtered through 0.45-mm filters, followed by total metals analysis of each fraction. Again, it is pointed out that the method used may underestimate the soluble fraction somewhat, as the samples were filtered before dissolution equilibrium had been established.
The values for Trail shown in Table 19 were used to calculate soluble deposition from total deposition related to the Cominco facility. Those shown for Sudbury were applied to Inco-Copper Cliff and Falconbridge-Sudbury. Those for Rouyn-Noranda were applied to the Horne facility. Owing to a lack of other smelter-specific data, the averages of the Trail, Sudbury and Rouyn-Noranda values were applied to all other facilities. The values for regional background were used only in deposition modelling (see Section 2.3.1.2.3). The consistency in the solubility values between the different facilities shown in Table 19 is worthy of note.
Estimation of regional background deposition: Data used to estimate regional background deposition of metals on the Canadian Shield are summarized in Table 20. "Regional background," as used in this report, means levels of deposition that might typically be expected to be found in areas not locally influenced by copper smelters and refineries, zinc plants or other associated operations. These values, however, may include deposition originating from other anthropogenic sources. Indeed, the data used in estimation of regional background deposition for these assessments are based on monitoring both in relatively remote areas and in areas influenced by other distant industrial sources.
All data used to estimate regional background deposition are based on the sum of dry deposition calculated from TSP monitoring and directly measured wet deposition. Handling of these data types was discussed earlier in this section. Data for six of the sites considered were obtained from the OME. Data collected for use in the IADN (1997) were also used in estimation of regional background. The IADN produces estimates of spatially averaged deposition over the entire area of each of the Great Lakes. This is based on monitoring at a number of stations around each of the lakes. The IADN site located on Burnt Island in Georgian Bay was also selected to use on its own, owing to its significant location - upwind of Sudbury. Average regional background deposition rates were adjusted for solubility using the regional background soluble metal fractions shown in Table 19. The average values of regional background soluble metal deposition rates have been used in dispersion modelling (Section 2.3.1.2.3) and as benchmarks for empirical deposition data.
- Data for Trail were provided by the B.C. MOE (courtesy of E. Tradewell, Air Resources). Soluble fractions are based on the analysis of 58, 70, 44, 35 and 6 dustfall samples for Cu, Zn, Pb, Cd and As, respectively.
- Data for the Sudbury region were provided by the OME (courtesy of R. McVicars and D. Toner, Laboratory Services Branch). Soluble fractions for Sudbury near-field represent the average of results for 54 TSP samples (51 for Cd), and those for regional background represent the average of results for 82 TSP samples (58 for Cd).
- Data for Rouyn-Noranda provided by Natural Resources Canada (courtesy of G. Bonham-Carter, D. Kliza and K. Telmer, GSC). Soluble fractions represent the average of results for 82 snowpack samples for Cu, Zn, Pb and As, 40 for Ni and 77 for Cd.
2.3.1.2.3 Deposition of metals from air -modelled
To complement empirical data, dispersion modelling was also used to estimate metal deposition rates near "generic" facilities. The following is a summary of the approach used to estimate deposition rates. The method is described in detail in SENES Consultants (2000).
Existing copper smelters, copper refineries and zinc plants differ in the raw materials processed by each facility, the method used to process the raw materials, the types of control equipment used to limit atmospheric releases of air pollutants (and therefore total amounts of releases), the numbers and types of stacks from which the pollutants are released, and the geographic location in which each facility is located (which determines the dispersion meteorology). Consequently, in one sense, each facility may be considered to represent a relatively unique operation whose impact on the environment will be somewhat different from that of a similar facility in another location. On the other hand, these facilities also represent a limited set of industrial operations whose emissions of trace elements fall within a reasonably well-defined range of emission rates, and whose stack characteristics also fall within a known range of heights, diameters, exhaust gas temperatures and velocities. Using these known ranges of emission rates and operating characteristics, it is possible to define both upper and lower bounds for air-quality impacts due to these releases. Moreover, by assigning probability density functions to variables that differ between facilities, it is also possible to statistically determine the probability with which a given deposition rate is likely to occur for the given range of emission variables.
Table 20 Estimation of annual regional background soluble metal deposition for the Canadian Shield
In this analysis, differences in emission rates and source release characteristics between facilities were addressed through statistical dispersion modelling for a set of "generic" facilities using a four-step process. First, a representative range of trace element emission rates, particle size distributions and release characteristics were developed to define source characteristics based on data provided by industry representatives for individual facilities. Second, probability distributions were assigned to the ranges of emission rates and release characteristics as suggested by the data provided by industry, and/or using professional judgement where data were unavailable. Third, a dispersion model (CALPUFF) was used to determine unit deposition rates for a discrete number of particle sizes and for a range of source characteristics. Finally, deposition rates were estimated at each receptor grid point downwind of the generic facility by multiplying the unit factors by the set of trace element emission rates and release characteristics. By repeating the last step many times for a randomly selected set of emission rates and source characteristics (referred to as trials), a range of possible deposition rates was calculated at each grid point, and a full set of trials was used to define a probability distribution for deposition rates (i.e., 25th percentile, 50th percentile, etc.).
The dispersion modelling analysis was conducted to a distance of 200 km from the generic facility, using a nested fine grid spacing of 1 km to a 10-km distance from the facility, and a 10-km grid resolution for distances 10 -200 km from the facility. The analysis used meteorological data from North Bay, Ontario, averaged for the period 1989-1993. This period included one year (1990) with the highest total precipitation recorded over the Great Lakes region in the 50-year period 1948-1997. Therefore, potential wet deposition would be maximized using the 1990 meteorology. From a climatological perspective, the long-term trend in precipitation over the last 100 years in this region has been toward increasing precipitation levels. The probability of precipitation on any given day has increased for all categories of daily precipitation amounts. Therefore, it was considered appropriate that the extreme above-normal precipitation for 1990 should be included in the analysis.
Three generic types of facilities were considered in this analysis, specifically:
- a copper smelter,
- a copper refinery, and
- a zinc plant.
In addition to modelling the emissions from these three types of facilities individually, the analysis also considered the impact of combinations of facility types that may be located in close proximity, namely:
- a copper smelter and zinc plant,
- two copper smelters and a copper refinery, and
- a copper smelter, a copper refinery and a zinc plant.
Therefore, the dispersion modelling analysis was conducted for a total of six facility scenarios, including three scenarios where only one type of facility is located at a site, and three scenarios where two or more facilities are located at the same site. The results are not intended to be representative of any existing single facility or combination of facilities. Instead, the results of the analysis represent a statistical merging of various ranges in operating conditions and emission rates to provide ensemble probability frequency distributions of trace metal deposition rates that would be expected to occur for the set of operating conditions and release rates reported for these facilities.
1 Emission rates, as derived from Table 4, are based largely on NPRI data (NPRI, 1995) with additional information provided by facility operators. Further detail is provided in the text.
Trace metal emission rates provided by copper smelter operators were generally reported separately for process stacks, low-level sources and total emissions. However, only total emissions were reported for some facilities. Lognormal probability distributions were fit to the total emissions data for those smelter facilities where the process stack and low-level emissions were also reported. Typically, there was substantial variation in emission rates between the facilities.
The emission rates listed in Table 21 were derived from Table 4. Values listed as not determined (ND) or negligible (neg.) in Table 4 were assigned a nominal value of 0.001 tonnes per year. There were insufficient data on trace metal emissions from copper refineries and zinc plants to compute meaningful statistics for median, 5th percentile and 95th percentile values.
The total trace metal emissions were partitioned between high-elevation releases (i.e., stacks over 30 m high), low-elevation releases (stacks less than 30 m high) and fugitive releases, based on the information received from industry representatives. For copper smelters, the total trace metal emissions were increased by 5% to account for fugitive emissions that were not considered in the emission data reported by facility operators. The ranges of reported release rates are listed in Table 22.
Total deposition rates attributable to the facilities were determined at each location based on the simulated emission rates, partitioning between releases, particle size distribution and the modelled atmospheric dispersion. The total deposition rates were summarized across the probabilistic trials. The total soluble deposition rate for each metal was determined by multiplying the deposition from the facility by the average near-field soluble metal fraction (Table 19) and adding this value to the regional background soluble deposition rate (Table 20).
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 deposition rate. The benchmark considered was the 25th percentile for critical load (deposition rate), discussed in Section 2.4.1.1.3. For some facilities, the benchmark levels for some trace metals were not exceeded at any distance considered. Note that the benchmark level is not exceeded at all locations closer than the maximum distance reported in the table, since the atmospheric dispersion has directional effects. The total area over which the critical load is exceeded will be less than the area calculated using the maximum distance. Isopleths for soluble Cu deposition in the region of a copper smelter, as estimated by dispersion modelling at the 50th and 95th percentiles, are shown in Figures 2 and 3 respectively.
2.3.1.2.4 Concentrations of metals in ambient air
Data on the concentrations of As, Cd, Cr, Ni and Pb in ambient air were available for a small number of monitoring sites near Canadian copper smelters and refineries and zinc plants. In most cases, these were based on TSP collected using high-volume samplers, usually over a 24-hour period once or twice per week, and analysed for some or all of these metals (discussed in Section 2.3.1.2.2). A summary of the data, which were obtained from the companies or from provincial governments, is presented in Table 24. For each combination of site and metal for which data were available, the table includes the arithmetic mean concentration for the most recent representative year, as well as the identity, location and type of site (e.g., residential) and the number of samples. A relatively long averaging period was selected, because the critical effects for each of these metals are associated with long-term exposure. In those cases where there is more than one monitoring site, the mean concentration of the various metals is generally increased, sometimes quite markedly (i.e., by two or three orders of magnitude), at those sites nearest the facility. Further, the mean airborne concentration of each of the metals near the copper smelters and refineries and zinc plants is consistently and substantially higher than regional background levels measured in areas of the Canadian Shield and the Great Lakes removed from point sources, although there is considerable variation among the facilities in the degree to which concentrations are increased.
n.e. - Critical load is not exceeded.
1 Maximum distance at which deposition exceeds the 25th percentile critical load (CL25) is based on comparison to the following CL25s (mg/m2/a): Cu=6.2, Zn=77, Ni=61, Pb=47, Cd=1.6 and As=27. These effects thresholds are discussed in Section 2.4.1.1.3.
Figure 2 Fiftieth percentile of total soluble deposition rates (mg/m2/a) estimated by dispersion modelling for copper emitted from a generic copper smelter
Figure 3 Ninety-fifth percentile of total soluble deposition rates (mg/m2/a) estimated by dispersion modelling for copper emitted from a generic copper smelter
Table 24 Annual average concentration of As, Cd, Cr, Ni and Pb in ambient air near copper smelters and refineries and zinc plants in Canada
Note: In calculating the summary statistics for each site, a value of one-half of the detection limit was assumed, wherever possible, for those samples that did not contain detectable levels. However, it was sometimes necessary to use the data as provided by the companies, and there was some inconsistency in the values that they assumed for samples that did not contain detectable levels of metals. Analyses in which the assumed values for such samples were varied systematically (as zero, half the detection limit, and the detection limit) indicated that the effect of this assumption was minimal except when the concentrations of the metals were relatively low.
2.3.1.3 Particulate matter
2.3.1.3.1 Fate of particulate matter in air
The diameter of PM released to the atmosphere from copper smelters and refineries and zinc plants can range from <1.0 mm up to about 20 mm.
The following paragraphs, describing the fate of PM in the atmosphere, are largely summarized from EC/HC (2000a).
Particulate matter is generally subdivided into a fine fraction of particles 2.5 mm or less (PM2.5) and a coarse fraction of particles larger than 2.5 mm. PM may be "primary" (emitted directly into the atmosphere) or "secondary" (formed in the atmosphere through chemical and physical transformations). The principal gases involved in secondary particulate formation are generally SO2, nitrogen oxides, VOCs and ammonia. Primary particles are present in both the fine and coarse fractions, whereas secondary particles, such as sulphates and nitrates, are present predominantly in the fine fraction. Particulate matter may include a broad range of chemical species, including elemental carbon and organic carbon compounds; oxides of silicon, aluminum and iron; trace metals; sulphates; nitrates; and ammonia.
Particle size is considered to be one of the most relevant parameters in characterizing the physical behaviour of PM in the atmosphere. Extremely small ("ultrafine") particles less than 0.1 mm in diameter (the nuclei mode) are formed primarily from the condensation of hot vapours during high-temperature combustion processes and from the nucleation of atmospheric species to form new particles. While the greatest concentration of airborne particles is found in the nuclei mode, these particles contribute little to overall particle mass loading due to their tiny size. They are subject to random motion and to coagulation processes in which particles collide to quickly yield larger particles. Consequently, these tiny particles have short atmospheric residence times.
Particles in the size range of 0.1-2.0 mm (the accumulation mode) result from the coagulation of particles in the nuclei mode and from the condensation of vapours onto existing particles, which then grow into this size range. These particles typically account for most of the particle surface area and much of the particle mass in the atmosphere. The accumulation mode is so named since atmospheric removal processes are least efficient in this size range. These fine particles can remain in the atmosphere for days to weeks. Dry deposition and precipitation scavenging are the primary processes by which these fine particles are eventually removed from the atmosphere. It is calculated that precipitation scavenging accounts for about 80-90% of the mass of particles in the accumulation mode removed from the atmosphere (Wallace and Hobbs, 1977).
Particles larger than 2.0 mm (the sedimentation or coarse mode) are typically associated with mechanical processes, such as wind erosion and grinding operations. Grinding operations result in the physical breakdown of larger particles into smaller ones to yield particles such as wind-blown soil and dust from quarrying operations. These particles are efficiently removed by gravitational settling and therefore remain in the atmosphere for shorter periods of a few hours to a few days. They contribute little to particle number concentrations but significantly to total particle mass. While particles resulting from metal smelting are usually relatively small, recent studies have identified spherules larger than 2.0 mm that are believed to be of pyrometallurgical origin (Kliza et al., 2000).
2.3.1.3.2 Concentrations of particulate matter in ambient air
Data on the levels of PM in ambient air were available for a small number of monitoring sites near each of the Canadian copper smelters and refineries and zinc plants. A summary of the data, which were obtained from the companies and, in some instances, from the provinces, is presented in Table 25. For each site, the table includes summary statistics for concentrations of PM, including the arithmetic mean, standard deviation, minimum and maximum, and various percentiles, as well as the identity, location and type of site (e.g., residential).
In most cases, the data obtained were for TSP collected using high-volume samplers and measured gravimetrically (although respirable fractions [PM10 and PM2.5] were determined near a few of the facilities). Because the health impacts of PM have been most extensively quantified based on the respirable fraction, the TSP concentrations were converted to estimated PM10 concentrations using the following regression: (0.826 x log TSP). This equation PM10 = 10 was derived based on monitoring of TSP and PM10 at 14 urban sites across Canada in the NAPS network between 1986 and 1994 (WGAQOG, 1999). The use of this approach to estimate PM10 concentrations near the facilities is supported by the results of parallel surveys of TSP and PM10 near the HBM&S copper smelter and zinc plant in Flin Flon and the Cominco lead-zinc smelter in Trail, supplied by the companies in response to requests for data on levels of PM in ambient air. In monitoring conducted at three locations over 10 months in 1998 in Flin Flon, the mean ratio of the measured PM10 to that estimated from TSP using the above regression was 1.04, 1.13 and 1.15, respectively, and was 1.11 overall. The corresponding ratios of the annual mean for 1998 at four sites in Trail were 1.38, 1.33, 0.81 and 1.18, with an overall mean ratio of 1.18.
Table 25 Summary of estimated or measured concentrations of PM10 (µg/m3) near copper smelters and refineries and zinc plants in Canada
Based on the data summarized in Table 25, in those cases where there was more than one monitoring site, the ambient concentrations of PM were generally increased the most at those sites nearest the facility. Further, the mean concentrations near most of the facilities were elevated above the background levels measured at remote sites in North America, which averaged between approximately 4 and 11 mg/m3 (WGAQOG, 1999). However, the impact of the copper smelters and refineries and zinc plants on ambient levels of PM was not as marked as for the metals (Section 2.3.1.2.4) or for SO2 (Section 2.3.1.1.2), likely because of the wider variety of sources from which PM originates.
