This page has been archived on the Web
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.
Technical Assessment of Environmental Performance and Emission Reduction Options for the Base Metals Smelters Sector - Final Report
Techniques Available to Achieve Gazette I Targets
3.2 SO2 Reduction Options
3.2.1 Sulphuric Acid Plants
3.2.2 Noranda Reactor with an Acid Plant
3.2.3 Acid Plant Tail Gas - Conversion to Double Absorption
3.2.4 Liquid Sulphur Dioxide Plants
3.2.5 Diverting Feeds to Smelters With Controls (i.e., Acid Plants) Already- In Place
3.2.6 Increased Pyrrhotite Rejection
3.2.7 Alkali Scrubbing
3.2.8 Eliminating Feeds With High Sulphur or High Mercury Contents
3.2.9 Hydrometallurgical Production Process
3.3 PM Emission Reduction Options
3.3.1Electrostatic Precipitators (ESPs)
3.3.2 Fabric Filters
3.3.3 Options to Reduce PM Emissions from Area Sources
3.3.4 Improving PM Emissions Data
3.4 Mercury Emission Reduction Options
A number of technically feasible options for further reducing SO2 emissions are being applied among Canadian smelters. These as well as other technically feasible options are available to further reduce emissions and achieve the Gazette I targets for 2008 and 2015. Technologies and operating practices available for addressing sulphur dioxide emissions include:
- capture of various process gases (e.g., roasters, furnaces, converters) and sulphuric acid production plants;
- increasing converter gases concentrations such as a single blowing converter;
- continuous converters (technical feasibility needs to be further proven for some metals);
- liquid sulphur dioxide plants (in combination with sulphuric acid plants);
- alkali scrubbing of converter gases (e.g., lime, limestone, caustic, dual alkali);
- acid plant conversion from single absorption to double absorption process (to address acid plant tail gases);
- scrubbing of acid plant tail gases;
- increased pyrrhotite rejection;
- diverting some feeds to smelters with controls (i.e., acid plants) already in place;
- restricting metal production (e.g., low level reduction to achieve 2008 target);
- reducing feeds with high sulphur or mercury contents; and
- hydrometallurgical production process (technical feasibility needs to be proven for some metals).
The set of emission reduction options that can be applied to achieve the Gazette I target at each facility is unique in most cases. The table below shows the options identified in this study and their applicability to each of the smelters. These technologies are briefly described in the next sub-section.
|Main Option(s) |
|Main Options |
Hudson Bay Mining & Smelting
Flin Flon, MB
Technologies and operating practices to achieve the 2008 Gazette I targets for PM emissions are the same as the techniques currently employed by some of the smelters. These technologies and practices could be upgraded (e.g., upgrading electrostatic precipitators and baghouses) or extended in terms of their application to the various sources. This includes technologies and operating practices to address process as well as area PM sources. However, the Gazette I targets present challenges, especially the 2015 targets for some facilities.
|EC 2015 Target|
* Includes only stack process emissions. Does not include area sources and fugitive process sources.
Environment Canada established the PM Gazette I targets using historical PM emissions estimated and reported by the base metal facilities, as reference points. PMT historical emissions may not include estimates from process fugitives or area sources for all smelters. They only included stack emissions from processes in most cases.
Technologies and operating practices were identified in this study to achieve the forecast estimates for PM emissions shown above. In general, the "business-as-usual" forecast shows that facilities can meet the 2008 Gazette I target (for process emissions only), but the 2015 target presents technical issues for 4 of the 6 smelters. Technologies and operating practices were not identified that could be applied to achieve and demonstrate compliance with total facility emission targets as contained in Gazette I for those 4 facilities. Some industry representatives expressed concern regarding the technical ability to achieve and demonstrate compliance with the Gazette I 2015 PM targets.
Environment Canada may wish to re-examine the Gazette I targets. It may also be possible to re-state the targets distinguishing between process stack emissions and process fugitives. It may not be feasible to include estimated area source emissions as parts of specific numerical targets, since it may be technically difficult to demonstrate compliance. Estimates of fugitives and area source emissions should first be collected from each of the facilities to use as reference points.
HBMS is the only facility identified that faces an expected requirement to meet the mercury emission target level of 2.0 grams-emissions/tonne metal produced. Forecast 2015 mercury emissions from the facility are expected at 5.4 grams/tonne of product, which implies a requirement to reduce by about 62%. Activated carbon injection into the furnace off-gases, with subsequent collection of resultant particulate matter, is a technically feasible option that has been applied to address mercury emissions. The furnace gases comprise approximately 80% of the mercury sources, and activated carbon injection can achieve 80% reduction in mercury, therefore presenting an option for achieving 64% reduction for the facility as a whole.
Dioxins and furans emissions have only been identified as an issue at the Falconbridge Sudbury smelter. The electric furnace has been identified as the main emission source. Dioxins sampling was carried out on September 20, 2005 during which the electric furnace freeboard temperatures were maintained above 450 °C. Dioxin tests were carried out simultaneously in the furnace off gas ducts and the main stack to test the hypothesis that operating with furnace freeboards temperatures above 450°C will avoid de novo synthesis. The results are expected early in 2006. Options to reduce emissions include furnace modifications and process control changes that would reduce dioxin formation. The adoption of these or additional options may become evident after the trails are assessed.
Technically feasible options to reduce SO2 emissions at smelters include pollution prevention as well as end-of-pipe controls. Examples of pollution prevention options include pyrrhotite rejection, rejection of feeds with high sulphur (or mercury) content, or diversion of concentrates with high sulphur content to facilities equipped with appropriate pollution controls. These controls include sulphuric acid plants and liquid SO2 plants, (which yield a marketable product), and alkali scrubbing. Pollution controls are more economically attractive if the emission streams they are addressing have high concentration of the pollutant and low flow rates. Some of the options described below deal with first increasing the concentration of the SO2 and reducing the flow rate of the stream to be addressed. Some SO2 emission reduction technologies are described below. Additional options unique to facilities may be available.
Sulphuric acid plants are technically feasible and have been applied in various base metal smelters in Canada and around the world. Various sizes of acid plant could be installed to address emissions from one or more sources. The SO2 conversion efficiency of single absorption acid plants is typically 97-98%. High conversion efficiencies (99%+) can be achieved using the double absorption process.
Dilute SO2 gas concentrations present challenges for conventional acid plant technologies and could lead to higher operating costs. Typically, facility designers would take steps to minimize the gas flow rate and increase the concentration of streams being routed to an acid plant. While process off-gases from initial smelting units (e.g., roasters and furnaces) generally present high concentration streams, concentrations from converters or finishing vessels (and some furnaces) may be too low, intermittent and variable, such that improved hooding may be required on these sources to increase concentration and reduce gas flow.
Sulphuric acid plants generate acid that needs to be sold on the North American market. The assumption applied to provide cost estimates in this study is that there is a negative netback (selling price of acid less transportation, less selling/marketing costs) of C$65 to $85 per tonne of acid produced.8 Netbacks for sulphuric acid will depend on the market conditions (which affect the price of acid), and the actual transportation costs between the smelter and the acid customer.
Installation of a Noranda Reactor is an option that can result in the replacement of roasters and reverberatory furnaces at some smelters. The Noranda reactor takes on the functions of these two units. The relatively high SO2 concentration stream from the Noranda Reactor would be captured and routed to an acid plant. This option also allows a portion of the SO2 that is being emitted from converters to be "transferred" to the Noranda Reactor. This reduces the amount of SO2 being emitted in dilute form from the converters. The remaining SO2 in the converters would need to be addressed separately.
SO2 in acid plant tail gas streams can be reduced by increasing the conversion efficiency within the acid plant. Sulphuric acid plants using the single contact absorption process typically achieve 97.5% conversion of SO2 to sulphuric acid, while those using the double contact process typically achieve 99.0% conversion. Using these two conversion averages, an acid plant conversion can achieve a theoretical 60% reduction in tail gas SO2,
(1.5/2.5 x100%) but in practice, the reduction may be somewhat less.
In the single contact process, the entire SO2 conversion (oxidation to SO3) step occurs in sequential catalytic converter beds before the absorption stage, where the SO3 is reacted with water ("contacted" or "absorbed") to form sulphuric acid. In the double contact process, there are two absorption stages. An intermediate absorption stage is inserted into the process, typically in between the third and fourth converter beds. This intermediate stage removes most of the SO3 produced as liquid acid, so that when the mixture returns to the fourth converter bed, the reaction equilibrium is shifted so that most of the remaining SO2 (the last 1.5%) can be converted to SO3. A second and final absorption stage removes this last fraction as sulphuric acid.
The conversion of a sulphuric acid plant from the single contact to the double contact process requires an additional absorption tower and an additional gas/gas heat exchanger to remove the heat from the absorption reaction. In some cases, it may be necessary to replace or modify the large converter vessel, as well.
Stack gases may be cleaned and concentrated to make pure sulphur dioxide, which can then be compressed to liquid form and sold. Alternatively, the liquid SO2 can be routed to an acid plant to make sulphuric acid. Liquid SO2 plants can act as an acid plant raw material buffer, smoothing out variations in process SO2 concentrations and flow rates.
Liquid sulphur dioxide is used in the pulp and paper sector, and other application markets. However, the basic assumption is that the SO2 made at smelters in substantial quantities would need to be routed to an acid plant, since the North American liquid SO2 market is limited.
One option for smelters is to sell feeds that contain high sulphur levels (as well as suitable metal contents) to smelters that have acid plants. In that way over 97% of the SO2 can be captured and converted into acid rather than being emitted from the facility that does not have the acid plant. This option is suitable for some low volume feeds.
Pyrrhotite (Fe1-xS) is the sulphur-bearing iron mineral that is combined with pentlandite (a complex iron-nickel sulphide, (Fe-Ni)9S8) in most nickel/copper sulphide ores. When concentrating ores for further processing in smelters, the separation and rejection of pyrrhotite from pentlandite is a key process operation at the mill. Using fine regrinding techniques, magnetic separation, selective flotation and the use of flotation reagents, pyrrhotite crystals can be separated from ultra-fine pentlandite crystals. Pyrrhotite rejection is utilized by smelters to improve their overall smelting performance. Application of this technique has a multi-purpose impact on the operation of a smelter. The percent of nickel content of the concentrate is increased; the volume of the material to be smelted is reduced, resulting in reduced energy use and in the amount of silica flux required for the slag formation. The practice has also the desirable effect of removing a significant amount of sulphur prior to smelting. However, since the oxidation of sulphur is usually the most significant energy source in smelting processes, energy balance has to be carefully watched so that high pyrrhotite rejection levels do not result in energy penalties (additional electricity or coke use) within the smelter. There can also be penalties associated with loss of metals in the concentrates with related throughput considerations at the smelters.
There are various technically feasible alkali scrubbing technologies available to address concentrated as well as dilute SO2 streams. These include, but are not limited to: wet limestone forced oxidation (LSFO) scrubbing (with calcium carbonate), dry or semi-dry lime scrubbing (calcium oxide), caustic (sodium hydroxide), ammonia, and dual alkali scrubbing.
Alkali scrubbing could be applied to remove sulphur dioxide from roasters, furnaces, converter gases, and acid plant tail gases. A large portion of the annual costs for these scrubbing systems involves the purchase of alkali such as limestone rock (that must be processed - crushed, etc.), lime, or caustic. Actual costs for reagent materials will vary and depend on the proximity of the facility to sources of reagent. If local limestone deposits are available and can be accessed, the cost of this material will be lower. Another important cost element is the disposal of by-product (e.g., gypsum - CaSO4, when lime or limestone are used), which holds the sequestered SO2. Although there are markets for by-products (e.g., gypsum used in wallboard), the general assumption is that the by-products would need to be disposed of locally as they may be contaminated with metals, and possibly have other technical or quality issues, which may restrict their sale.
There are opportunities among the 6 base metal smelters to achieve cost-effective reductions by eliminating or rejecting feeds with high sulphur or high mercury contents. The ability of smelters to practice this option will depend on the specific operational design features of the facility, its market requirements, and economic viability. Facilities with control options already in place, such as sulphuric acid plant, may be able to better accommodate different concentrates, including feeds with high sulphur.
Hydrometallurgical processes for recovering metals would virtually eliminate the sulphur dioxide, mercury and PM process emissions. Hydrometallurgical processes involve a complete redesign of the manufacturing process, eliminating the roasters, furnaces, and converters. There are several hydrometallurgical processing technologies being developed. However, there is technical uncertainty as to whether the hydrometallurgical processes would be feasible for some specific concentrates being processed in the smelters. Extensive laboratory and pilot plant tests would need to be conducted to assess the technical feasibility of hydrometallurgical processing for each of the facilities.
There are a variety of technologies and operating practices available to address potential PM process emissions. Electrostatic precipitators (ESP) and baghouses are already in place at most facilities. However, their performance may not be at maximum levels. Rebuilding and adding new ESPs and baghouses to enhance performance are options to further reduce PMT emissions. Other options such as wet scrubbers, cyclones offer lower PM removal efficiencies, and are not described here. There are also many options available to reduce PMT emissions from area sources.
Electrostatic Precipitators (ESP) use electrical fields to remove particulate matter in gaseous streams. In an ESP, an intense electric field is maintained between high-voltage discharge electrodes, typically wires or rigid frames, and grounded collecting electrodes, typically plates. A corona discharge from the discharge electrodes ionizes the gas passing through the precipitator, and gas ions subsequently ionize particles. The electric field drives the negatively charged particles to the collecting electrodes. Periodically, the collecting electrodes are rapped mechanically to dislodge collected particulate, which falls into hoppers for removal. Wet electrostatic precipitators use a water spray to remove the particulate. A typical wet configuration has (vertical) cylindrical collecting electrodes, with discharge electrodes located in the centres of the cylinders.
Electrostatic precipitator overall (mass) collection efficiencies can exceed 99.9%, and efficiencies in excess of 99.5% are common. Older, poorly maintained, undersized, or poorly-operated ESPs will show lower efficiencies. Options for improving the performance of existing precipitators begin with simple rebuilds. These normally include the replacement of electrodes, rappers, and other internal elements, and modernization of the precipitator power supply and control system. An upgraded control system allows for improved voltage control, so that the voltage in each field may be maintained at the highest level possible without sparking. Precipitator rebuilds also include improvements to the ductwork, casing, and flow devices to improve the flow distribution, seal leaks, etc.
Fabric filter collectors (baghouses) are conceptually simple: by passing flue gas through a tightly woven fabric, particulate in the flue gas will be collected on the fabric by sieving and other mechanisms. The dust cake, which forms on the filter from the collected particulate, can contribute significantly to collection efficiency. Practical application of fabric filters requires the use of a large fabric area in order to avoid an unacceptable pressure drop across the fabric. To provide a large fabric area in a small space, the fabric is formed into cylindrical bags (hence the term baghouse). Baghouses often are capable of 99.9% removal efficiencies. Baghouse removal efficiency is relatively level across the particle size range, so that good or excellent control of PM10 and PM2.5 can be obtained. Determinants of baghouse performance include the fabric chosen, the cleaning frequency and methods, and the particulate characteristics.
There are many technologies and operating practices currently being employed by the base metal smelter facilities to reduce area source PM emissions. These have been listed above in this report. The options that are currently being applied would be applicable to achieving greater reduction levels of PM emissions from area sources in the future.
Most of the base metal facilities have yet to complete annual emission estimates for area and process fugitive PM sources. In addition, it would be useful to develop estimates of potential emissions (pre-controls), so that the reduction efficiencies or performance of the control equipment already in place can be determined. These datasets should be developed for each source and related control devices. Such data along with further analysis could be used to develop baseline performance levels for total PM emissions for each smelter.
There are several techniques that can be utilized for removing mercury from base-metals smelter off-gases. The Boliden-Norzinc process is one of the most widely used, including adoption at two Canadian Base Metals smelter facilities (Teck Cominco and Noranda CEZ).
The mercury removal process involves scrubbing the gas in a packed bed tower (Mercury Tower) with a liquid containing an acidified solution of mercuric chloride. The elemental mercury in the gas reacts with the mercuric chloride and precipitates as insoluble mercurous chloride (calomel).
Falconbridge, at its Horne facility, utilizes a multi stage mercury removal process, which includes a proprietary technology and a lime injection technique in addition to the Boliden-Norzinc process to remove mercury from its smelter off-gases.
In a base metal smelter equipped with an acid plant, the mercury removal step is normally placed after the washing and cooling steps in the acid plant, so the gas temperature is below 40°C, and is free of dust and sulphur trioxide
(SO3). This means, for smelters without an acid plant and whose dusty and hot furnace off-gas may contain mercury, a stand-alone Boliden-Norzinc mercury scrubbing system must include a dedicated gas cleaning system.
8 Pentasul Inc., Byproduct Sulphuric Acid at Smelters in Manitoba from 2007, Marketing and Disposal Options with Associated Costs, (prepared for Chemtrade Logisitics Inc.), Feb. 2002.
- Date modified: