This appendix summarizes techniques for reducing emissions of the principal pollutants that impair visibility.1The sources of emissions are discussed in approximate order of their contribution to haze, as set forth in Table 6-2.
Increasing environmental regulation in many industrialized countries has spurred the advancement of power-plant pollution control technologies over the past two decades. Research and development to render the technologies more efficient, more reliable, and less costly has accelerated in the United States in anticipation of new emission reduction requirements for existing power plants resulting from the 1990 Clean Air Act Amendments (Torrens, 1990). Power plants nationwide with 40,000–60,000 MW of electricity generation capacity probably will be retrofit with some form of control of SO2 or NOx before the year 2000 in response to new legislation (Dalton, 1991; EPRI, 1991a).
The Clean Air Act Amendments of 1990 were not aimed specifically at visibility. SO2 reductions required by the legislation were aimed
primarily at reducing emissions that cause acid deposition (i.e., acid rain). However, SO2 reductions also should improve visibility in Class I areas in the United States, particularly in the East.
Table D-1 shows the results of an analysis funded by the U.S. Department of Energy (DOE) of power plants in relation to Class I ''prevention of significant deterioration'' (PSD) areas. The percentage of total coal-fired power-plant capacity affected by an SO2 reduction requirement would be 31% within a 100-mile radius of a Class I area, increasing to 61% within 150 miles and 75% within 200 miles; 85% of total SO 2 emissions arise from boilers within 200 miles of Class I areas (Trexler, 1990).
TABLE D-1 Power Plants and Class I Areas
Capacity, 103 MW
SO2 Emissions, 106 tons
Total coal-fired boilers
Within 100 miles of Class. I area
Within 150 miles of Class I area
Within 200 miles of Class I area
Source: Trexler, 1990.
Sulfur Dioxide Control Technologies
Over 150 flue gas desulfurization (FGD) systems in power plants with approximately 72,000 MW of capacity are now operating in the United States to control SO2 emissions (Dalton, 1990). About one-fifth of the total coal-fired capacity is covered by the FGD systems, of which about
92% are wet scrubbers and 8% are spray dryers. By the end of the decade, new FGD-equipped plants with an estimated 27,000 MW of generating capacity are expected to begin operating.
Wet Flue Gas Desulfurization
The predominant scrubbing technology used by U.S. utilities is lime or limestone wet FGD (scrubbing) and landfill disposal of by-products. The reagent is prepared (limestone is ground and lime is slaked) and mixed with water in a reagent preparation area. It is then conveyed as a slurry (approximately 10% solids in water) to an absorber (typically a spray tower) and sprayed into the flue gas stream. SO2 present in the flue gas is absorbed in the slurry and collected in a reaction tank, where water is removed from the resulting calcium sulfite or sulfate crystals.
The U.S. utility industry's early experience with scrubbers was difficult. Inadequate understanding of the chemical reactions within the process led to frequent plugging and scaling of the scrubber components, corrosion and erosion of the construction materials, poor handling characteristics and large land requirements for sludge by-products, high capital, and high operating costs. Better understanding of system chemistry has led to increased reliability and improved performance. Additives permit SO2 removal to exceed 90% at an added cost that is usually not prohibitive (compared with the average cost of removing the first 90%).
Future improvements in conventional lime or limestone FGD are illustrated by three examples:
Process improvements such as the jet bubbling reactor used in the Chiyoda CT-121 process;
Application of spray drying to high-sulfur coal;
An advanced limestone FGD system producing gypsum with no reheat and no spare modules, for use at compact sites (i.e., space-constrained).
Each of the designs could achieve SO2 control of 90–95% or better. The advanced system would offer 20–50% capital cost savings and 20–40% operating cost savings over conventional FGD systems available
today. The jet bubbling reactor and the advanced limestone and gypsum designs could be used to make wallboard-grade gypsum, for sale or disposal.
Engineering and other improvements have reduced costs over the past decade, but wet FGD remains relatively expensive. Capital costs for conventional systems vary depending on fuel sulfur content, unit size, and other factors. Preliminary results from a recent Electric Power Research Institute (EPRI) cost estimation indicate that a state-of-the-art wet FGD system for medium-sulfur coal on a new plant could be built for less than $200/kW and annual operating costs would range from 5 to 10 mills/kW-hr (Torrens and Radcliffe, 1990).
Depending on existing plant conditions, especially available space and accessibility, retrofitting an FGD system can cost up to three times more than installing it in a new plant. For the different FGD systems now available and a moderately difficult installation (approximately 30% more costly than in a new plant), the range of capital requirements and total levelized costs over 30 years with no inflation are shown in Table D-2.
TABLE D-2 Capital Costs and Levelized Costs in 1990 of Retrofitting a FGD Systema
Capital Cost per kW
Cost per ton of SO2 (constant dollars)
Wet FGD (range of system types)
a The figures are in 1990 dollars and are subject to a ± 20% uncertainty. Source: EPRI, 1991b.
Spray Dry Flue Gas Desulfurization
Spray dry FGD is the other principal method of SO2 control used today. Calcium oxide (quicklime) mixed with water produces a calcium
hydroxide slurry, which is injected into a spray dryer, where it is dried by the hot flue gas and reacts with the gas to remove SO 2. The dry product is collected both at the bottom of the spray tower and in the downstream particulate removal device, where more SO2 may be removed. Capital costs for dry FGD can be substantially less than for wet systems, especially for low-sulfur coal applications. Seventeen spray dry FGD systems were operating as of mid-1987, all on relatively low-sulfur coal (less than 2%).
High-sulfur spray dryer applications have not been demonstrated on a long-term commercial scale. Pilot testing has indicated that SO 2 removal of 80–90% is possible, and over 90% removal is possible under certain conditions. However, a fabric filter may be added to maintain particulate emission standards if SO2 removal exceeds 90%, since the mass flow of solids to the electrostatic precipitator (ESP)—the principal particulate control technology in U.S. power plants—will at least double if a spray dryer is used.
Control of Plume Opacity
Stack-plume visibility, also known as plume opacity, has become a concern to the utility industry because some coal-fired stations with operating FGD systems have been cited for opacity in excess the New Source Performance Standards under the Clean Air Act, even though particulate mass emissions are within regulated limits. At other utilities, with scrubbed or unscrubbed stack emissions, the visible emissions have been higher than expected, and there is concern about visibility reduction in the local environment. Unacceptable opacity may be caused by scrubber-generated particulate matter, condensible particulate matter such as sulfuric acid in the flue gas, fine particles penetrating the particulate control device, or colored gases such as NO2 in the flue gas. EPRI field tests have shown some contribution from all the above causes. However, the primary contributor to plume opacity at most units firing medium-(1–3%) and high-(greater than 3%) sulfur coal appears to be condensed sulfuric acid mist.
Sulfuric acid is formed in the furnace when sulfur from coal is converted to SO2 and some of the SO2 is oxidized to SO3. The amount of SO2 thus oxidized is generally about 0.5%, depending on excess oxygen and the presence of conversion catalysts in the fly ash. The process is
not understood completely. As the flue gas cools, the SO3 combines with water vapor in the gas to form H2SO4 but does not condense unless the temperature drops below the acid dew point (usually between 250 and 350°F depending on the H2SO4 concentration and the moisture content of the gas).
When the gas temperature drops below the acid dew point, condensation begins as molecular clusters of H2SO4 agglomerate and attach colliding water molecules to form submicron droplets. That can occur at the stack exit, where the gas cools because of mixing and heat transfer with the ambient air, or at the inlet to an FGD system, where the flue gas is reduced in temperature to saturation (typically 125–130°F) by water evaporation. Because of their small size, the H2SO4-based droplets are not removed by the FGD system.
There are three options for controlling acid mist plume formation: switching to lower-sulfur fuel, injecting an alkaline additive to react with the H2SO4, or adding a wet ESP. Switching to lower-sulfur fuel might require major changes to the boiler system and other plant operating systems. Because, coal contracts usually are of long term and lower-sulfur coal might cost much more than coal obtained locally.
The utility industry has applied additives to enhance ESP performance (without regard to H2SO4 removal) for many years. The additive injection systems are normally simple, requiring only a storage facility, transport system, and injection ports. A variety of reagents have been used to control acid mist, including magnesium oxide (MgO), organic amines, and ammonia (NH3). MgO and NH3 were evaluated during an EPRI field testing program. MgO appeared limited to approximately 50% removal efficiency with injection into both high-and low-temperature regions of the boiler flue gas system. NH3, however, demonstrated consistently high removal efficiency. Care should be used when NH3 is injected. Some precipitators have been disabled by the use of NH3. It is not well understood why NH3 injection works well at some sites and not at others. Also, ammonium compounds in the fly ash or scrubber waste can make sale or reuse more difficult. EPRI is also conducting research on other additives to control H2SO4.
The wet ESP has potential for acid mist control but has a high capital cost and has no proven application to technology for U.S. utilities.
As a result of installing FGD systems in response to the Clean Air Act Amendments, plants burning high-sulfur coal are likely to form sulfuric acid mist under certain operating conditions. That possibility
should be anticipated in the design, so that appropriate control measures can be taken after start-up if necessary.
Combustion modification in the boiler can reduce NOx emissions by about 50%, depending on the type of boiler, at moderate cost ($10–$30/kW of capacity for capital cost, or $200–$400/ton of NOx removed for total annual cost). To achieve greater reductions requires some form of postcombustion NOx control (downstream of the boiler) at a cost of $100–$150/kW or $1,500–$4,000/ton of NOx removed (Kokkinos et al., 1991).
NOxReduction by Combustion Modification
Combustion modification (adjusting the fuel mixture so that there is enough oxygen to support combustion but not enough to combine with nitrogen) can reduce NOx emissions by 40–60% and has been applied widely to new coal-fired boilers. Twelve full-scale demonstrations for retrofit combustion modification also are in progress or planned, covering the range of boiler types in operation in the United States and sizes from 40 to 500 MW. Combustion modification is the most cost-effective method of meeting NOx emission reduction requirements for the majority of the U.S. boiler utilities.
Another combustion modification technique, and the only one suited to cyclone boilers, is reburning, which involves redirecting 10–20% of the total fuel to the upper furnace region of the boiler to create a zone for chemical reduction of NOx formed in the primary zone of combustion in the boiler. Full-scale demonstrations of reburning with either natural gas or pulverized coal are planned. Potential NOx reductions are in the range of 30–50%.
Capital costs for retrofit low NOx burners are estimated to be $10–$30/kW. Reburning costs depend on the choice of reburning fuel, and capital costs are $25–$45/kW.
The main technology for postcombustion NOx reduction, selective catalytic reduction (SCR), involves injection of NO3 in the presence of a catalyst between the boiler and the air heater. NO3 reacts with NOx, reducing it to nitrogen and water. Europe has over 30,000 MW of coal-fired generating capacity equipped with SCR, and Japan has 6,000 MW. Retrofit installations in Europe and Japan typically achieve 60–80% NOx removal, and residual NH3 emissions of less than 5 ppm (usually 1–2 ppm). Capital costs in Europe average $125/kW. The costs are consistent with EPRI's capital cost estimates for hypothetical retrofit installations, which range from $100 to $150/kW. Levelized cost projections for both U.S. and Japanese plants are estimated at 4–9 mills/kW-hr (Cichanowicz et al., 1990).
The United States and Europe are paying increased attention to selective noncatalytic reduction (SNCR) technologies. The principle is similar to that of SCR, but no catalyst is involved. Results with urea injection show the potential for NOx reductions of 30–50%, perhaps up to 75%, and NH3 emissions of 5–10 ppm or less. Capital costs are estimated to be $5–$15/kW, and operating costs of less than 3 mills/kW-hr.
Clean Coal Technologies
Initial operational problems with FGD, combined with the high cost of the systems, stimulated development of less costly methods of SO2, reduction at coal-fired power plants. That led to the introduction, during the mid-1980s, of clean coal technologies, which have evolved into a family of precombustion, combustion and conversion, and postcombustion technologies designed to improve technical pollution-control capabilities and flexibility at lower costs (Torrens, 1990).
Sorbent injection removes SO2 during combustion. The simplest technique is furnace sorbent injection (FSI) of lime or limestone. The pulverized coal and sorbent mixture is maintained suspended as a fluid
in the boiler by a stream of air flowing upward. SO2 removal of 35–55% has been measured in FSI demonstrations using lime. Injection of sorbent downstream of the furnace at approximately 1,000°F (540°C) shows similar removal efficiency but has not been demonstrated at full-scale. The U.S. Environmental Protection Agency (EPA) and the U.S. Department of Energy (DOE) have sponsored FSI programs at the 60–180 MW scale. A key issue may be the handling and disposal of highly alkaline waste material generated by FSI through the use of existing particulate control devices, usually an ESP.
The main advantage of sorbent injection is its low capital cost (estimated at $70–$120/kW). However, the overall cost per ton of sulfur removed is comparable to that of FGD at higher load factors: the economic benefit of sorbent injection would be greatest, therefore, for plants with lower capacity.
Fluidized Bed Combustion
Atmospheric fluidized bed combustion (AFBC) is now an established technology for industrial boilers (10–25 MW) and is being demonstrated for utility boilers (75–350 MW). The pulverized coal and sorbent mixture is maintained suspended as a fluid in the boiler by an upstream of air. AFBC boilers can meet the New Source Performance Standards (NSPS) for both SO2 and NOx,2 without additional control equipment. Much greater reductions in SO2 emissions from AFBC plants probably would require additional postcombustion technology. Like FSI, AFBC results in additional wastes that might prove difficult to handle in existing particulate control devices.
In pressurized fluidized bed combustion (PFBC), the boiler operates under a pressure of 10 atmospheres. The increased energy of the exit gases can drive both a gas turbine and a steam turbine (combined cycle),
potentially boosting generating efficiency to over 40%. Four commercial demonstration units, each of 70–80 MW capacity, are being built at utility sites in Sweden, Spain, and the United States; each is repowering an existing plant. Hot gas cleanup is still under development and is needed if the combined cycle is to fill its potential for improving generating efficiency.
Combined NOxand SOxReduction
Combined NOx and SO2 processes offer the potential to reduce SO2 and NOx emissions for less than the combined cost of SCR and conventional FGD. Most processes are under development and are not commercially available, although several are being demonstrated in the DOE Clean Coal Technology Program (DOE, 1991).
Integrated Gasification Combined Cycle
Several integrated gasification combined cycle (IGCC) processes are now in the developmental and demonstration stage. All result in a synthetic gas that is cooled by generating steam and desulfurized before combustion in a turbine. Heat is recovered as steam from the combustion turbine and is used to generate additional electricity. The advantages of IGCC include higher generating efficiency, the possibility of phased construction, and very low SO2 and NOx emissions (99% removal of SO2 and one-tenth of new source requirements for NOx emissions).
Integrated Gasification Fuel Cell
A further conversion technology under study is the integrated gasification fuel cell (IGFC) power plant. IGFC involves substituting a molten carbonate fuel cell for the combustion turbine, potentially increasing overall generating efficiency to above 45%. Optimal chemical and thermal integration of the fuel cell and a catalytic gasifier, with hot gas cleanup, could increase efficiencies even further, to 55–60%. To do so would require that both gasification and gas cleaning take place at
temperatures near the 1,300°F operating temperature of the molten carbonate fuel cell.
Thirty-eight IGFC demonstrations are under way or planned with cost-sharing by government in the DOE Clean Coal Technology Demonstration Program (DOE, 1991).
Alternative Emission Reduction Methods
Switching to Low-Sulfur Coal
The 1990 Clean Air Act Amendments have raised the question as to whether the supply of low-sulfur coals is adequate to meet substantially higher demands from electric utilities and others seeking to reduce emissions. In the near term, there are concerns about the utilities' ability to develop mines and the transportation industry's capacity to ship the coal. At least until the strategies of utilities in response to the legislation become clearer, predictions of the course of low-sulfur coal prices are highly uncertain (EPRI, 1991a; NERC, 1991).
Even without fuel price increases, switching to low-sulfur coal could be costly, depending on boiler compatibility with the available low-sulfur coals and on the new infrastructure required to handle the switch. Also, reliance on low-sulfur coal alone is not sufficient to meet present NSPS.
Switching to Natural Gas
The capital costs of installing natural gas generation technologies are relatively low, and the ease of adding capacity in small increments gives gas-fired electricity an added advantage. Emissions reductions can be achieved quickly by using natural gas, especially since many emitters already have gas available on site. A substantial amount of new gas-fired generating capacity, mainly gas turbines for power to meet peak demands, is being planned by utilities and independent power producers.
Like oil markets, natural gas markets are uncertain. The natural gas market is now going through a major structural change, especially with regard to the rules governing transmission pipeline access. It is difficult to predict how gas supply, demand, and price will interact.
Another potential natural gas alternative is co-firing coal with natural gas in a coal-fired boiler. As the percentage of gas increases above 15–30%, the capital improvement costs could increase because of greater engineering changes required for co-firing capability.
For some utility systems, environmental dispatching (i.e., sending out electricity from the lowest-emitting plants first and the highest-emitting plants last) could reduce emissions. By allowing gas-fired units to be dispatched ahead of higher-emitting stations (especially during the summer when gas prices are seasonably low), a utility system could lower emissions without capital investment or modification of its coal-fired units. This shifting of load, which is a kind of fuel switching, generally would increase gas usage and fuel costs.
Energy Efficiency and Demand Management
Improvements in the way electricity is produced and used reduce pollutant emissions by preventing future growth in demand and by reducing present emissions. In recent years, many U.S. utilities have shifted toward demand-side management (DSM), which includes both energy efficiency and load shifting. Virtually all U.S. utilities are pursuing DSM to some degree. In 1990, nearly 15 million residential customers participated in DSM programs. Existing and planned DSM programs are expected to reduce summer peak demand in the year 2000 by approximately 43,000 MW—the equivalent output of 43 large power plants. That represent a reduction of about 6.5% in the demand forecast for 2000 (EPRI, 1990).
Much of the technology discussed above for power plants can also be used to control emissions from industrial combustion sources. However, there are some important differences. The size, distribution, and fuel
type of industrial boilers are quite diverse. Apart from the steel industry, industrial boilers generally are small emission sources compared with power-plant boilers, but they are often located in clusters. Nearly 9,000 combustors are in operation, mostly in the eastern United States. Coal is the economic fuel choice for the larger units, and natural gas and fuel oil are used widely in the smaller units. Less common fuels, such as agricultural waste (bagasse), sewage sludge, wood waste, residual and waste oil, and refuse (garbage), also are burned.
As in the utility industry, wet scrubbing technologies are available in industrial boilers for controlling SO2 emissions from industrial combustion sources. Although most large utility boilers use lime or limestone scrubbing methods, about 90% of wet scrubbers installed on industrial boilers are sodium-based. These scrubbers use a solution of sodium hydroxide or carbonate to absorb the SO2 from the flue gas. The absorbers are simple, easy to control, and require little maintenance. Over 95% removal efficiency is possible. Dual-alkali scrubbing is the next most common method for small industrial units. That process also uses a sodium solution for absorption but incorporates a precipitation step where lime is added to remove sulfate and regenerate sodium for recycling. The waste stream is gypsum. About 90% removal efficiency is typical. Capital costs for the sodium solution method are $40–$60/kW of heat input; dual-alkali investment costs are about twice as high.3 Operating costs, including capital recovery, for sodium scrubbing are $400–$1,000/ton of SO2 removed by sodium scrubbing, depending on the sulfur content of the fuel. With dual-alkali scrubbing, operating costs are similar because lower raw-material costs offset the higher capital recovery.
Lime spray-drying systems also are practical for use on industrial boilers and municipal waste burners. Removal efficiency of 80–90% is possible. Removal costs are $400–$800/ton of SO2. Lime spray-drying is particularly useful in garbage combustion applications where the high-chloride content of the flue gas is corrosive to wet scrubbing systems; both SO2 and HCl are removed. As an alternative to spray drying, the sorbent can be injected dry either directly into the combustor or
into the ductwork upstream of the dust collector. Injection is inexpensive compared with spray drying, but removal efficiency is low (only 30–40%).
Processes are available to recover SO2 emissions in a useful form. MgO scrubbing appears to have good potential for industrial application because removal and recovery are separate operations. A centralized recovery plant can serve a number of small emission sources. Over 90% removal efficiency is possible. Costs are highly dependent on plant size, SO2 concentration, proximity of the operation that regenerates MgO, and credit for recovered product (sulfuric acid or elemental sulfur). In many situations, MgO scrubbing would be competitive with sodium scrubbing.
Atmospheric fluidized bed combustion (AFBC) is an alternative to conventional coal burning that removes SO2 during the combustion process as described earlier. AFBC is a well-developed technology for industrial use. In the United States, more than 80 units with capacities up to 65 MW are in operation on a wide variety of fuels. One method, reaching the commercial stage involves high-velocity flow to recirculate entrained solids to the combustor. Costs for SO2 control by AFBC are comparable to those for wet scrubbing.
NOx control methods for industrial sources are similar to those for large sources: combustion modification and flue gas treatment. Flue gas recirculation involves returning a side stream of furnace exhaust to the burner. Both temperature and oxygen concentration in the combustor are reduced, resulting in less favorable conditions for NOx formation; reductions up to 50% are possible, depending on fuel type and practical limitations on recirculation rate. Improvements in burner design can reduce NOx emissions by reducing flame temperature and adjusting air flow to the burner. Reductions of 30–70%, compared with conventional burners, have been achieved—the lower values with coal burners and the higher values with oil or gas.
Flue gas treatment by ammonia injection has been used in industrial boilers. The reaction rate is temperature dependent; reduction under optimal conditions (900–1100°C) is about 40%, but drops below 10% outside the favorable range. An effective temperature regime in the combustion systems is essential, and retrofit might not be feasible. When catalytic conversion is used in conjunction with ammonia injections, removal efficiencies up to 90% are possible. Industrial application has been mainly on oil and gas combustion sources.
New and developing technologies being demonstrated by the DOE
Clean Coal Technology Demonstration Program mentioned above will provide some additional options for industrial boilers (South et al., 1990).
Production of copper, lead, and zinc from natural ores historically has been a principal source of industrial SO2 emissions. Metal sulfide in the ore is converted to SO2 during the smelting process. If left uncontrolled, the concentration of SO2 in the off-gas can be as high as 12–14%. However, controls have been developed to reduce emissions, and the smelting industry is no longer a major contributor to the national SO2 problem. SO2 emissions from the copper industry, which accounts for over 80% of smelter emissions, have been reduced from nearly 2 million tons in 1975 to less than 300,000 tons in 1988.
Use of sulfuric acid plants to recover SO2 has become common. Smelter acid plants are distinguished from commercial plants that burn sulfur by the wide variability in inlet SO2 concentration and the need for extensive gas cleaning. The off-gas from smelter equipment is collected in a variety of hood and duct arrangements and typically is passed through a cyclone collector for coarse dust removal, then cooled in a spray chamber before removal of small particles in an electrostatic precipitator. The gas is then scrubbed, cooled, and passed through a mist separator before it is dried by contact with recirculated acid. The clean, dry gas containing 4–8% SO2 is then heated to reaction temperature in a series of heat exchangers ahead of the catalyst (vanadium pentoxide) beds, where the SO2 is oxidized to SO3. Heat liberated in the process is transferred to the inlet gas so that no additional energy is needed, provided that the SO2 concentration is in the proper range. Strong gases, such as those from a fluidized bed roaster, must be diluted to within the operational limits of 4–8% SO2.
The gas from the catalytic converter flows to an absorption tower where the SO3 is absorbed in strong recirculating sulfuric acid to produce commercial 98% acid. The drying tower that treats the inlet gas produces 93% acid, also a commercial product. A single contact (one-phase absorption) acid plant is 97% efficient; a double contact unit (two towers in series) is over 99% efficient. Effective collection of acid mist is essential.
Use of new smelter technology also will reduce emissions from the
smelting industry. For example, integration of steps into continuous processes reduces variation in SO2 concentration in fugitive streams, and use of flash smelting technology, such as the Outokumpu flash furnace instead of the reverberatory furnace, reduces emissions and improves efficiency (ACSPCT, 1977; South et al., 1990; E. Trexler, pers. comm., MSCET, DOE, Washington, D.C., 1990).
PETROLEUM AND CHEMICAL INDUSTRIES
Crude oil contains from 0.1% to 5% or more sulfur by weight (Carrales and Martin, 1975). As crude oil is broken down into products by the refining process, much of that sulfur is converted to either elemental sulfur or sulfuric acid, thereby reducing the potential for SO2 emissions to the atmosphere. However, sulfur recovery and sulfuric acid plants are not 100% efficient, and some emissions to the atmosphere do occur during sulfur recovery. In addition, certain refinery processes (such as fluid catalytic cracking), coking operations, flares, and heaters or other fuel-burning devices lead to SO2 emissions (Bond, 1972; Danielson, 1973).
The processing, storage, and marketing of petroleum products lead to emission of volatile organic compounds (VOCs). At petroleum refineries and petrochemical plants, losses occur from process streams, refinery pipe flanges and valves, cooling towers, wastewater treatment units, and storage tanks (Danielson, 1973; EPA, 1985c). During marketing of petroleum products, organic vapors are displaced into the atmosphere whenever tank trucks, bulk terminal storage tanks, gasoline-station tanks, and eventually automobile and truck fuel tanks are filled.
Sulfur Oxide Control at Refineries and Chemical Plants
Substantial emissions of sulfur oxides from refinery processes occur during fluid catalytic or thermal cracking processes. Available control techniques include (1) careful selection of cracking catalysts that will minimize the accumulation of sulfur-bearing coke on the catalyst, thereby reducing SO2 emissions during catalyst regeneration, (2)
desulfurization of the feedstock to the cracking units, and (3) installation of wet scrubbers on the exhaust of the cracking units (Hunter and Helgeson, 1976; SCAQMD, 1978). In the case of feedstock desulfurization or the use of scrubbers, control efficiencies of 80–90% or more are possible (see the earlier discussion of SOx scrubbers). Catalyst replacement is less expensive but might remove less sulfur. Both catalyst and feedstock should be considered carefully.
Sulfur recovery systems at refineries gather gases bearing hydrogen sulfide (H2S) and convert H2S to elemental sulfur, usually via the Claus process (EPA, 1985c). H2S is not completely converted to elemental sulfur in one-, two-, or three-stage Claus systems, which results in release of SO2 in the effluent. A variety of Claus tail-gas treatment technologies have been developed. Approximately 99% control of SO 2 can be obtained with their use relative to uncontrolled Claus plants. The cost of such technologies is attractive compared with that of many other SO2 reduction systems (Hunter and Helgeson, 1976; EPA, 1985c).
Elemental sulfur, H2S, or spent acid from refinery systems can be used as feedstock for sulfuric acid production (Hunter and Helgeson, 1976; EPA, 1985c). A conventional uncontrolled single-stage contact sulfuric acid plant can emit about 2–5% of its sulfur input in the form of unabsorbed SO2 or SO3. By adding of a second set of catalytic converters or absorbers to the process stream, overall control can be raised above 99.7%. SO2 scrubbers also can be used to meet similar control objectives.
Control of VOC Emissions from Petroleum Refining and Marketing
Noncondensable hydrocarbons generated by refinery processes generally are gathered at the refinery and burned to fuel process heaters. Fugitive emissions from leaking valves, flanges, compressor seals, wastewater treatment plants, and cooling towers, along with product spills lead to significant hydrocarbon release to the atmosphere. Through careful selection and maintenance of seals and packing materials, leaks from piping and compressors can be reduced. The degree of control varies greatly from site to site (Danielson, 1973). Approximately 90% reduction in uncontrolled VOC emissions from wastewater treat-
ment plants is possible through use of vapor stripping systems and through covering of tanks (EPA, 1985c). Cooling-tower emissions are controlled by preventing hydrocarbon leaks from heat exchangers and condensers into cooling water lines.
Storage tanks at petroleum refineries and petrochemical plants can release hydrocarbon vapors in two ways (EPA, 1985c). First, as fixed-roof tanks are filled, hydrocarbon vapors that occupy the empty volume of the tanks are displaced into the atmosphere. Some losses also occur when tanks are emptied. Second, diurnal temperature changes cause the vapors in tanks to expand and contract, causing tanks to ''breathe'' even though their liquid levels have not changed. Hydrocarbon losses can be reduced if floating-roof tanks or similar devices are used to eliminate the vapor space above the liquid product in the tanks.
Hydrocarbon losses from filling fuel tanks during petroleum transport and marketing can be reduced by submerged filling of the tanks, which creates fewer vapors than does allowing the product to splash in the tank while filling. Further control is possible by forcing vapors displaced by the rising fuel to be condensed or returned to the tank. Submerged filling of service station gasoline tanks can reduce vapor losses by about one-third. Submerged filling combined with vapor return to the tank from which the gasoline is disposed can achieve up to 97% control relative to emissions that result from splash filling (EPA, 1985c).
Vehicle refueling at gasoline stations also generates hydrocarbon vapors as the rising fuel displaces vapors from an empty tank. Vapor recovery systems that return those vapors to the gasoline-station storage tanks offer control efficiencies of about 90% if used properly (EPA, 1985c).
DIESEL-FUELED MOTOR VEHICLES
An uncontrolled diesel engine emits 30–100 times more particulate matter than a comparable-sized gasoline engine. Because particles impair visibility and threaten human health, EPA has established limits on particle emissions from light-and heavy-duty diesel engines. Many of the limits were codified in altered form by the 1990 Clean Air Act Amendments.
Manufacturers are expected to have little difficulty complying with the
0.25 gram per brake horsepower hour (g/bhph) standard, which went into effect for 1991 model-year trucks. Compliance includes use of the following:
Advanced, low-emitting, and fuel-efficient high-swirl direct injection engines.
Higher fuel injection pressures and more precise control over the fuel injection process.
Computerized electronic engine control systems, which will improve trade-offs between NOx and particles by continuously adjusting the fuel injection timing. Reductions in particle emissions of up to 40% are possible with this approach.
Turbocharging, which increases the amount of fuel that can be burned without excessive smoke, accompanied by intercooling, which reduces the adverse temperature effects of turbocharging and further increases maximal power potential. Turbocharging and intercooling can reduce NOx and particle emissions and increase fuel economy and power output. Most heavy-duty diesel engines are now equipped with turbochargers, and most have intercoolers.
Plans for reducing oil consumption, thereby reducing emissions of oil-derived particles.
Though sufficient to meet the 1991 standard, these improvements probably will not be adequate to achieve the higher standards scheduled to take effect later in the decade. Compliance with the 1993 urban bus standard of 0.1 g/bhph probably will require either alternative fuels or exhaust treatment systems, such as trap oxidizers. The 0.1 g/bhph standard will apply to all heavy-duty trucks in 1994; most manufacturers have indicated their intention to comply by using oxidation catalysts that require low-sulfur fuel. The measures are described below.
A trap oxidizer consists of a durable particulate filter in the engine exhaust stream and some means of cleaning the filter by burning off the collected particulate matter. The most challenging problem in developing trap oxidizers has been devising a system to burn accumulated par-
ticulate matter off the trap without damaging the exhaust system. There are two types of systems for doing so. So-called passive systems regenerate during normal vehicle operation. The most promising approaches require the use of a catalyst (either as a coating on the trap or as a fuel additive) to reduce the ignition temperature of the collected particulate matter. Regeneration temperatures as low as 420°C have been reported with catalytic coatings, and even lower temperatures are possible with fuel additives. Those temperatures, however, may not be low enough for regeneration to occur during normal operation of diesel trucks.
Active systems monitor the buildup of particulate matter in the trap and trigger actions to regenerate the trap when needed. Many approaches to triggering regeneration have been proposed—from diesel fuel burners and electric heaters to catalytic injection systems. Catalytic coatings also have several advantages in active systems and might make possible a simpler regeneration system.
Due to reductions in the solid soot fraction of particulate emissions from diesel engines, the soluble organic fraction now accounts for 30–70% of the emitted particulate matter. A catalytic converter can be used to treat that emission. Particulate control efficiency of even 25–35% would be enough to bring many engines into compliance with the 0.1 g/bhph 1994 standard. The oxidation catalyst also greatly reduces emissions of VOCs, CO, odor, and gaseous and particle-bound toxic air contaminants, such as aldehydes, PNA, and nitro-PNA. Unlike the trap oxidizer, the catalytic converter is a relatively mature technology; millions of catalytic converters are used in gasoline vehicles, and diesel catalytic converters have been used in underground mining applications for more than 20 years.
The catalytic converter requires low-sulfur fuel; otherwise, the increase in sulfate emissions from the catalyst's conversion of SO 2 would more than counterbalance the decrease in the soluble organic fraction. Regulations mandating low-sulfur fuel (0.05% by weight) have been promulgated by EPA and are scheduled to take effect by October 1993. In addition to reducing direct emissions of SO2 and sulfate particles, lowering the sulfur content of diesel fuel decreases formation of sulfate particles from SO2 in the atmosphere.
GASOLINE-FUELED MOTOR VEHICLES
Control technology for gasoline-fueled vehicles continues to advance rapidly. Controls generally target NOx, VOCs, and carbon monoxide (CO) emissions. VOCs and to a lesser extent NOx contribute to visibility impairment.
Current federal emission standards allow no more than 0.41 g of VOCs per mile, 3.4 g of CO per mile, and 1.0 g of NOx per mile. The standards usually are met through use of a three-way catalytic converter which oxidizes VOC and CO into water and carbon dioxide by using a platinum or palladium catalyst and reduces NOx into elemental nitrogen and oxygen by using a rhodium catalyst.
The Clean Air Act Amendments of 1990 follow California's lead by tightening standards for new vehicles. For instance, beginning in model year 1994, a more stringent VOC standard (0.25 g of non-methane hydrocarbon per mile) will be phased in for light-duty vehicles. According to the California Air Resources Board, which has adopted the same standards with an earlier phase-in, the standards will not require any fundamental change in existing technology. Rather, manufacturers are expected to comply through reduced oil consumption, more precise electronic control and diagnostics, engine improvements, and improved catalytic coatings. Additional emission reductions are expected as a result of lower trace lead levels in unleaded gasoline and more advanced emissions control components, particularly more durable catalysts, better air-fuel management systems, and improved electronics.
The Clean Air Act Amendments also phase in a standard for NOx of 0.4 g/mile already in force in California. In adopting this standard, the California Air Resources Board noted that it can be met by using three-way catalytic converters on engines if excess oxygen is controlled and the air to fuel ratio is kept at its ideal or stoichiometric ratio.
At present, manufacturers must show that their new vehicles will meet emissions standards for 5 years or 50,000 miles. The 1990 amendments follow California's lead by requiring manufacturers to demonstrate that their vehicles will meet slightly less stringent standards for 10 years and 100,000 miles. (Full warranty coverage and recall testing, however, will not initially apply for the whole of this period.) To meet the proposed 100,000-mile standards, manufacturers would have to ensure that deterioration of the emission control system in the second half of the vehicle's life is not greater than in the first half. California has conclud-
ed that low emission levels can be maintained for 100,000 miles by improving fuel quality and using advanced control systems, such as on-board diagnostics and increased precious metal loadings in catalysts.
The Clean Air Act Amendments of 1990 also establish a second tier of emissions standards for light-duty vehicles and trucks. Those so-called Tier II standards (non-methane hydrocarbons at 0.125 g/mile, NOx at 0.2 g/mile, and CO at 1.7 g/mile) will take effect in model-year 2003 unless EPA makes specified findings. To some extent, the Tier II standards follow the lead of California, which decided in 1989 to adopt a second phase of tightened standards. California, however, will go further by requiring that some zero-emissions vehicles be produced. The 0.125 g/mile standard is based on possible emissions reductions from use of fuels with lower ozone-forming potential than is used today. Those fuels, along with electrically heated catalysts and other improvements, are expected to allow compliance with the 0.2 g/mile standard for NOx.
In addition, the 1990 Clean Air Act Amendments establish new requirements for reformulated gasoline, which are expected to lower emissions of VOCs. Special requirements for fleet vehicles also might encourage development of alternative, less-polluting fuels, such as diesel-fuel substitutes discussed above.
The amendments supplement the requirements with other measures to decrease in-use emissions. Inspection and maintenance requirements have been strengthened; for instance, maximal expenditures for repair have been increased. EPA will now be able to require annual centralized inspections in areas with the worst ozone problems. EPA is also required to revise certification test procedures to determine whether 1994 and later model-year passenger cars and light-duty trucks are capable of passing state inspection emission tests. EPA also must review and revise certification procedures to ensure that motor vehicles are tested under conditions that reflect actual driving, including fuel condition, temperature, acceleration, and altitude. Recent evidence indicates that high acceleration testing might be critical. For example, the California Air Resources Board found one car that met the 0.41 g/mile standard for VOCs under the present test procedure but that emitted over 15 g/mile when accelerating.
The requirements of the Clean Air Act Amendments are expected to lower total mobile-source emissions at least for the rest of the century.
Much will depend, however, on how EPA and the states implement the amendments. For example, enhanced inspection and maintenance requirements could lower emissions more than tightening the tailpipe standards. States also could reduce emissions by adopting California standards for motor vehicles, which will continue to be stricter than the federal standards.
Substituting cleaner-burning alternative fuels for diesel fuel and gasoline has drawn increasing attention during the last decade. Alternative fuels now under consideration include natural gas, methanol made from natural gas, and, in limited applications, liquid petroleum gas. See NRC (1991b) for additional information.
Clean burning, cheap, and abundant in many parts of the world, natural gas already fuels many vehicles in several countries. The major disadvantage of natural gas as a motor fuel is its gaseous form at normal temperatures and poor self-ignition qualities. That makes it less promising as a fuel for diesel engines. Natural-gas engines are likely to use up to 10% more energy than the diesels they replace.
Liquified Petroleum Gas
Liquified petroleum gas is already widely used as a vehicle fuel in the United States, Canada, The Netherlands, and elsewhere. As a fuel for spark ignition engines, it has many of the same advantages as natural gas, and an additional advantage of being easier to carry aboard the vehicle. Its major disadvantage is the limited supply.
Like natural gas, liquid petroleum gas in spark ignition engines is expected to produce essentially no particulate emissions (except for a small amount of lubricating oil), very little CO, and moderate VOC emissions. NOx emissions are a function of the air-to-fuel ratio. Liquid
petroleum gas does not burn as well under lean conditions as does natural gas, so the NOx emission reductions achievable through lean-burn technology are expected to be somewhat lower.
Methanol has many desirable combustion and emissions characteristics, including good lean-combustion characteristics, low flame temperature, and low photochemical reactivity. Methanol cannot be used in a diesel engine without some supplemental ignition source. Investigations to date have focused on the use of ignition-improving additives, spark ignition, glow plug ignition, or dual injection with diesel fuel. Converted heavy-duty diesel engines using each of these methods have been developed and demonstrated.
Methanol combustion does not produce soot, so particulate emissions from methanol engines are limited to a small amount of lubricating oil. Although flame temperature is lower for methanol than that for hydrocarbon fuels, it is not clear whether methanol use would lead to lower NOx emissions (NRC, 1991b).
The potential for large increases in formaldehyde emissions with the widespread use of methanol fuel has raised considerable concern. Efforts to resolve that problem are focusing on developing special low-formaldehyde catalysts and on minimizing unburned methanol emissions. Those efforts, although promising, have not yet provided a solution to the problem in methanol diesel engines under all conditions.
PRESCRIBED FORESTRY AND AGRICULTURAL BURNING
Three factors contribute to the importance of prescribed burning as a source of visibility impairment in Class I areas: (1) the emissions are often from forests near Class I areas; (2) the emissions are often comparable in magnitude to those from stationary industrial sources; and (3) the emissions are effective in scattering light because of the fine size of smoke particles.
The use of fire as a land management tool has evolved only over the
past 50 years (Walstad et al., 1990). Land managers use fire in reforestation programs to clear debris after timber harvest, improve rangeland characteristics, reduce fuel levels that might create a fire hazard, and improve wildlife habitats. Over 7 million acres of forest and range lands are burned annually in the United States, generating about 1.8 million metric tons (megagrams) of fine particles. Much of the burning occurs in the southeastern United States, far from Class I areas. However, prescribed fire emissions (about 150,000 metric tons of fine particles annually) are an important contributor to Class I area visibility impairment in Oregon and Washington (State of Washington Department of Ecology, 1983; State of Oregon Department of Environmental Quality, 1986). As a result, restrictions on forestry burning play a key role in both states' Visibility Protection State Implementation Programs (Core, 1989b; Pace, 1990).
Prescribed Forestry Burning
Emission Control Measures
Control measures to reduce prescribed burning emissions include (1) reduction of the number of acres burned; (2) reduction in fuel consumption; (3) burning under conditions of increased fuel moisture; (4) use of helitorch ignition methods; and (5) application of rapid mop-up techniques to expedite fire suppression during the smoldering phase.
Reduction in acres burned can be achieved through alternative treatment methods described below. Several land managers in Washington State have stopped using prescribed fire entirely because of concern about public sensitivity to smoke (State of Washington Department of Natural Resources, 1989).
Reduction in fuel consumption can be achieved through increased residue use. As forest conservation programs reduce timber harvest levels, the demand for chipped fuelwood for use in industrial boilers will increase. That demand could be met by increased use of residues, which have become an increasingly source of energy over the past 15 years. Nationally, the use of wood waste has increased from less than 1 quad (1015 British thermal units) in 1972 to 2.1 quads in 1982. Much of the residue is taken as firewood. In Oregon alone, it is estimated that
as much as 300,000 tons per year of slash could be used as firewood rather than burned as slash. Hogged fuel boilers also are likely to use more wood residues as other energy sources become more costly.
Burning under conditions of increased fuel moisture has been shown to reduce emissions by about 30% (Sandberg, 1983). In most cases, the land manager's intent is to eliminate small residue to create an adequate number of seedling planting spots. There is little value in reducing the amount of large residue on the site. Successful burns eliminate small residue but do not burn large logs or the duff layer of twigs, needles, and leaves on the forest floor. Burning under increased fuel moisture reduces the amount of fuel consumed and therefore emissions.
Increased use of helitorch (or aerial) ignition can reduce emissions by up to 20% by achieving mass burn fire behavior (in which small-sized combustible material is burned quickly without sufficient radiative heat to combust large logs) (D.V. Sandberg, pets. comm., USDA Forest Service, Pacific Northwest, 1985).
Rapid mop-up of residual smoke following the active phase of a fire can also reduce emissions. Mop-up typically is conducted to minimize the risk of new ignitions of smoldering fuels that could result in an escaped fire; it also eliminates smoldering emissions that might be carried downslope into valley floors. Because emissions from the smoldering phase of a fire account for about 40% of the total emissions from a prescribed burn, mop-up can reduce emissions without compromising the land manager's objectives. Mop-up within 8 hours typically reduces overall emissions by about 10% (Freeburn, 1986).
The costs of alternative measures to reduce prescribed burning emissions were evaluated during the development of the Oregon Visibility Protection Program (Freeburn, 1986). That study showed that the cost of prescribed burning varies greatly as a function of harvest unit conditions, fuel characteristics, and land ownership but averages about $102 per acre on private land and $150 per acre on federal land (Marcus, 1981). The standard deviation of cost per unit is as high as 98% for western Oregon and Washington (Mills et al., 1985), but about one-half of the actual unit costs appear to lie within ±70% of the nominal burning cost (Freeburn, 1986).
The cost of mopping up a typical 20-acre unit with a crew of 15 persons working for 4 hours is in the range of $100–$200 per acre, depending on fuel characteristics. In comparison, the cost of alternative treatment, such as using herbicides, manual methods, or mechanical methods, is $250 per acre for private land and as much as $400 per acre for federal lands (Freeburn, 1986).
Alternative Treatment Methods
Several available alternatives to prescribed burning include the following:
Manual methods (e.g., chain saws) to remove competing vegetation or create conditions favorable to a desired plant;
Mechanical methods (e.g., tractors, cable systems, or crawlers equipped with circular blade devices);
Biological methods (e.g., animals or insects) to control vegetation
Explosives and herbicides.
The U.S. Forest Service has recently completed a thorough evaluation of each method and of the relative health risks to workers and the public (USDA, 1988).
The burning of straw stubble following the harvesting of cereal grain, grass-seed fields, and other crops is different from forestry burning in that agricultural lands are more accessible, are typically level enough to be worked by farm machinery, and have much lighter fuel loading than forested lands.
Large reductions in particulate emissions are possible through burning grass-seed fields on alternate years (rather than annually), and through growing crops that do not require burning. The removal of stubble from the field before treatment by using tractor-mounted, propane-fueled torches also reduces emissions.
The use of straw to produce cattle feed, hardboard products, and paper has not proved to be economically feasible, nor has straw inciner-
ation for energy production. Mobile field sanitizers, designed to burn the stubble at high temperature as they pass over the field, have not proved to be either economically feasible or practical (Oregon State Department of Environmental Quality, 1988). As a result, the primary emphasis has been on improving smoke management programs to minimize effects of burning on the public and (in Oregon) on wilderness air quality and visibility.
Other Biomass Burning
Open burning of biomass is a common way to dispose of brush, stumps, and other residues following land clearing and highway right-of-way projects. In most of the United States, open burning outside urban areas is largely unregulated, requiring only a permit issued by a local fire district. Alternatives, such as open-pit incineration using air curtain destructors and grinding or chipping residues for use as mulch or boiler fuels, are seldom used.
There is a growing awareness that biomass burning can impair visibility in national parks and wilderness areas. Both the Oregon and Washington visibility protection programs include strategies to control emissions from biomass burning. In recognition of the importance of establishing ''best available control measures'' (BACM) for biomass burning, EPA's Office of Air Quality Planning and Standards has undertaken a program to describe BACM for biomass burning that affects serious PM10 nonattainment areas. Development of BACM documents is required by the 1990 Clean Air Act Amendments.
RESIDENTIAL WOOD COMBUSTION
Particulate emissions from residential space heating with wood, especially in urban areas, create a noticeable pall of smoke over many western communities during the winter months. Many ski resorts in the West (e.g., Colorado) and small towns located in forested regions near
Class I wilderness areas have a wood-smoke air-quality problem. Wood smoke is a significant contributor to PM10 nonattainment in many western communities. Regional emissions from wood smoke can greatly impair visibility in Class I areas during the fall, winter, and early spring.
A large, steady decline in emissions of residential wood combustion since the 1940s was reversed in 1973–74 when prices for oil, natural gas, and electricity increased sharply as a result of the Arab oil embargo. The strong resurgence in residential use of cordwood as a space-heating fuel has continued over the past decade, and about 1 million new wood stoves were sold each year during 1975–85. National fireplace and wood-stove PM10 emissions have been estimated to be at about 1 million tons per year (EPA, 1986b).
Pollutant emission rates from wood stoves are influenced by the following factors:
Wood-Stove Design. The design of the appliance is very important (EPA, 1988b). Older, conventional wood-stove emissions are typically 21 g/kg (i.e., 21 g of pollutant per kilogram of wood burned) for PM10, as compared with 4 g PM10/kg for stoves with the newest "best existing stove technology" (BEST) (Crane, 1989).
Wood Moisture Content. For wood stoves, cordwood that produces the lowest particulate emissions contains about 20–26% moisture. Wood moisture above or below this range results in higher emissions (EPA, 1988b).
Burn Rates. Generally, the higher the burn rate, the lower the particulate and carbon monoxide emissions.
Heat Output Requirements. The higher the home owner's heat output requirement, the greater the amount of fuel burned. Energy conservation through home weatherizing programs is the key to reducing fuel consumption, although the use of wood stoves with smaller fireboxes also reduces total emissions.
Wood-Burning Control Technologies
There are two approaches to reducing wood smoke from stoves and fireplaces: (1) improving the performance of wood heating systems through programs such as certification testing; and (2) burning less wood through wood-stove curtailment, home weatherization, and fuel-switching programs. Some of those strategies have multiple advantages. Woodstoves that have been certified, for example, reduce the amount of wood smoke per cord of wood burned while improving energy efficiency. Other examples are public information programs to teach proper wood-burning techniques and firewood-seasoning programs that result in better combustion (lower emissions) and increased energy efficiency.
To assist the states, EPA has issued a guidance document describing emission control measures for residential wood combustion (EPA, 1989b). That document describes four basic strategies: public information and awareness, improvements in wood-burning appliance performance, reduced dependence on wood, and wood-burning curtailment programs.
Public Education Programs
Local programs to educate the public about the wood smoke problem, to promote good burning practices, to urge reduced reliance on wood and mandatory wood-burning curtailment programs are considered essential as a space-heating fuel, and to promote compliance with voluntary to any control of residential wood combustion. EPA guidance allows a 50% or more emissions reduction credit for mandatory curtailment and up to a 50% credit for voluntary programs.
Wood-Stove Certification Programs
In 1983, Oregon became the first state to adopt a wood-stove certification program that required all new wood stoves sold in the state to be laboratory tested for emissions and efficiency to assure compliance with newly adopted emission standards (Kowalczyk and Tombleson, 1985). As a result, stoves sold after July 1986 were required to emit 50% less
wood smoke than conventional stoves. After July 1988, new stoves were required to emit 70% less smoke.
After the Oregon program was adopted, EPA adopted a slightly more restrictive national certification program, which became effective in July 1990 (CFR Title 40). The national certification program will result in substantial emission reductions as old stoves are replaced with newer certified models.
Further emission reductions are possible by increasing the durability of the stoves to reduce sheet metal warpage that allows flue gases to bypass catalytic converters and by increasing the durability of the converters themselves (Crane, 1989).
Reduced Dependence on Wood
In some mountain communities, reduced dependence on wood through weatherization and fuel switching might be the only long-term strategy that will assure compliance with the PM10 NAAQS. Programs have been adopted that limit installation of new wood stoves, require phaseout of stoves, and prohibit stove use (EPA, 1989b).
The most immediate short-term strategy to achieve the PM10 NAAQS is often adoption of voluntary or mandatory wood-burning curtailment programs. Mandatory curtailment programs are in operation in Boise, Idaho; Denver, Colorado; Juneau, Alaska; Missoula, Montana; Seattle and Yakima, Washington; Reno, Nevada; and Medford, Oregon; and voluntary programs operate in many other communities. Emission reductions of as much as 80% have been documented on winter days as a result of curtailment programs.
As the federal wood-stove certification program and the state PM10 control strategies are implemented, emissions from wood stoves might
decline if the energy market continues to offer home owners low prices on natural gas, fuel oil, and electricity. However, wood use depends on fuel prices, and reductions achieved through the above strategies may be offset by population growth and changing economic conditions.
Fugitive particulate emissions originating from a variety of sources can severely impair visibility in Class I areas in arid parts of the western United States or near major agricultural areas. Global emissions of wind-blown dust are estimated to be of the order of 2 million tons per day, or about one-tenth of total global tropospheric particle emissions.
Fugitive emissions may be separated into process sources (those associated with industrial operations) and open dust sources. Process sources include emissions from storage and transfer of raw, intermediate, and waste aggregate materials; open dust sources include agricultural tilling, paved and unpaved road dust, wind erosion of soils in areas without ground cover, and construction activities.
Agricultural tilling is often the largest anthropogenic source of fugitive dust. Emissions from tilling depend on the silt content of the soil (nominally 18%), wind speed, soil erodibility, soil moisture, and the portion of total particulate emissions that fall within the PM10 or fine particle (PM2.5) fraction (EPA, 1988c).
Over the years, the U.S. Department of Agriculture Soil Conservation Service has taken a leading role in reducing topsoil loss by wind erosion through improved land management techniques, such as reducing the need for tilling and using wind barriers and strip-cropping farming methods. The Food Security Act contains two provisions to reduce dust emissions from agricultural tilling. The first required development of conservation plans by 1990 for all lands designated as "highly erodible" by wind. The second provision, the Conservation Reserve Program, has taken highly erodible cropland out of production and covered it with vegetation.
Paved and Unpaved Road Dust
In Class I areas downwind of major urban areas, fugitive emissions from paved and unpaved roads might impair visibility. Haul roads at open pit coal mines near Class I areas also might be of concern.
Fugitive dust is emitted whenever a vehicle travels over a paved or unpaved road surface. Soil dust loading, silt content, vehicle traffic volumes, and soil moisture are critical in determining emission rates of road dust.
Control strategies for reducing paved road dust include road sweeping and flushing, reducing soil tracked out onto the road network from construction sites or unpaved roads, reducing spills from haul trucks, and reducing wind erosion from lands adjacent to the roadway. Unpaved road dust strategies include roadway surface improvements such as watering, chemical stabilization, and paving.
Storage piles, raw material handling, and transfer operations also are sources of visibility-impairing dust emissions. Dust is emitted at several points in the storage cycle, including material-loading onto piles, disturbances by strong wind currents, and movement of trucks and loading equipment. Emission rates vary with the volume of the aggregate passing through the storage cycle, the age of the pile, moisture content, silt content, and friability of the material. Control strategies include improved material handling to reduce transfer needs, wind sheltering, moisture retention, chemical stabilization or the use of water sprays, and enclosure of the materials.
Although measures to reduce fugitive dust emissions are available and are being applied nationwide, it is impractical to control natural wind-entrained soils in much of the West. Where agricultural operations are an important (and controllable) source of dust, emissions are being reduced. The reductions can be calculated and verified. They are also cost effective in reducing wind erosion of topsoils.
The technology of fugitive dust emission control is relatively well known, especially as it applies to agricultural operations. It is difficult, however, to differentiate visibility impairment associated with anthropogenic soil dust emissions from that associated with natural causes.
FEEDLOTS AND OTHER SOURCES OF AMMONIA
Of the 830,000 metric tons of ammonia (NH3) emitted in the United States in 1980 by anthropogenic sources, emissions from livestock-waste management dominate at 540,000 metric tons/yr (Placet and Streets, 1987); animal excrement is estimated to be the major terrestrial source of ammonia worldwide. Other sources (in metric tons per year) are fertilizer production (110,000), agricultural application of anhydrous NH3 (50,000), NH3 synthesis (40,000), petroleum refineries (40,000), motor vehicles (30,000), stationary fossil fuel combustion (20,000), and coke manufacture (10,000). In polluted urban areas such as Los Angeles and Denver, where nitric acid concentrations are high (greater than 1 ppb), NH3 emissions from feedlots can influence urban airborne particle composition and concentration. Studies made in 1978 in the Denver "brown cloud" showed that about 20% of the fine particle mass (i.e., particles less than 2.5 µm in diameter) was composed of ammonium nitrate (NH4NO3) (Countess et al., 1980); these particles accounted for about 17% of the visibility reduction (Wolff et al., 1981).
Söderlund and Svensson (1976) estimated that wild and domestic animals and humans produce globally about 27–50 million metric tons/yr of NH3, but only a small faction of those emissions (2–7 million metric tons/yr) is attributable to wild animal wastes. Therefore, NH3 emissions in national parks and wilderness areas should be very low.
It is possible, but probably unlikely, that visibility in the vicinity of national parks and wilderness areas could be impaired by NH3 emissions from feedlots and other nearby anthropogenic sources. In such cases, control of the emissions might be desired but would be difficult because the openness of most feedlots prevents capture of NH3 by the usual air treatment procedures employed in industrial operations (Bond, 1972). Chemical additives to the animal waste can lower NH 3 emissions; for example, added natural zeolites can lower NH3 emissions by 50% (Miner, 1984).