Sulphur trioxide emissions have been the subject of increasing attention from a spectrum of observers, but a new injection technology is working to help wipe plumes from the sky.
James H. Wilhelm, Codan Development LLC, USA
Control of sulphur trioxide (SO3) emissions from coal fired power plants has recently been the subject of attention from utility management, the public living near the plants, and environmental regulators. What once was an aggravating problem of sulphuric acid corrosion of ducts and electrostatic precipitators (ESP) has become a problem of visible plumes, plume touchdowns, increased operating costs, and increased scrutiny from regulators and the public. Ironically the main cause of the recent attention has been the advent of selective catalytic reduction (SCR) for control of nitrogen oxides (NOx). These SCR units actually reduce the amount of air pollution coming from the plants. However, they also increase SO3 concentrations in flue gas enough to make significant changes in the sulphuric acid dew point and in the visible plume.
When burned, about one per cent of the sulphur in coal is oxidized in the boiler to form SO3. The installation of SCR units for NOx control can more than double the amount of SO3 in the flue gas by oxidation of sulphur dioxide (SO2) across SCR catalysts. When the flue gas is subsequently cooled, the SO3 in the gas is converted to sulphuric acid. This sulphuric acid can condense in air heaters, in ducts and ESPs, and it can become a very fine mist or aerosol in the plume from the stack. Depending on atmospheric conditions, the visible plume appears as a blue-white haze, or a brown cloud carried for miles downwind. At elevated SO3 concentrations, plume buoyancy can be influenced to the extent that touchdowns of the plume in the vicinity of the plant have occurred at several plants upon startup of the SCR units, leading to public outcry and some regulatory concerns.
SBS injection technology
Control of SO3 emissions can be achieved by contacting the flue gas with a clear solution of sodium bisulphite and/or sodium sulphite (SBS solution). The technology is patented by Codan Development LLC and is offered, in conjunction with URS Corporation, Codan’s engineering partner.
Ideally the clear SBS solution is injected into the duct before the flue gas is cooled below the acid dew point for sulphuric acid. The sprayed solution dries upon contact with the hot flue gas and forms millions of small alkaline particles that react with the SO3 in the flue gas according to the following chemical reactions:
(Sodium bisulphite) à‚ à‚ à‚ à‚ (Sodium bisulphate)
(Sodium sulphite) à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚
When the sodium to SO3 molar addition ratio exceeds 1.0, the reactions become:
à‚ à‚ à‚ à‚ à‚ à‚ à‚ à‚ (Sodium sulphate)
Alternately, Codan has developed and patented a process known as In-Situ SBS Injection. In this process, a clear solution of a variety of sodium-based reagents (carbonate, bicarbonate or hydroxide) is injected into the duct where it is converted to sodium sulphite or bisulphite by reaction with SO2 in the flue gas. Then the same chemical reactions listed above occur to replace SO2 in the SBS particles with SO3 absorbed from the flue gas. The end result is the removal of SO3 and the formation of particles of sodium sulphate and sodium bisulphate, which are then removed with the fly ash.
In all cases, the alkali used for SO3 control is not consumed by reacting with SO2. Since SO2 is not absorbed, very low molar ratios of active sodium to SO3, typically 1.3 to 2.5, can produce the desired SO3 removal results. Other chemicals such as lime, limestone and magnesium hydroxide must be used at very high molar ratios, since SO2 in the flue gas consumes most of the alkali.
Figure 1. The optimum SBS injection point is dependent on many parameters, which differ from plant to plant
Figure one indicates the various potential locations for injecting the SBS solution. Since a solution is being injected, it is necessary to have a three to eight metre section of duct downstream of the injection point and free of obstructions, to allow adequate drying time before the solids contact any surfaces to prevent solids deposition. The duct work and objectives at each plant must be evaluated to select the most desired injection location.
Figure 2. SBS reagent solution is injected into the flue gas through a series of lances
Figure two is a simplified illustration of the equipment required for the SBS Injection technology. The heart of the system is the injection grid itself. SBS reagent solution is injected into the flue gas through a series of lances with a number of dual-fluid nozzles on each lance. Each nozzle is supplied with both SBS solution and compressed air. The lance itself is cooled by an internal flow of ambient air that is forced into the lance shroud by the differential pressure between the duct and the shield air supply.
After initial pilot testing of the SBS injection process on simulated flue gas in 2000 and 2001, a 265 MW demonstration test was conducted at the AB Brown station Unit 2 of Vectren Corporation, Indiana, USA. The Electric Power Research Institute (EPRI) studied the process in two phases between August and December 2002. URS Corporation directed all aspects of process application, including nozzle and lance development.
Three additional full-scale demonstration programmes were conducted in 2002, at 400 MW, 430 MW, and 750 MW scales, leading to the first commercial systems starting up in 2003.
The first permanent SBS Injection system was installed at the Bruce Mansfield plant of FirstEnergy, which began operation at Unit 1 in March 2003. A total of eight full-scale systems (totaling 5300 MW) are now operational, and several additional full-scale systems (totaling 4000 MW) are being planned for installation in 2005. Some of the SBS Injection systems are installed ahead of the air heater, while others are installed between the air heater and the ESP. The injection point is selected based on evaluation of many parameters, including duct configuration, residence time ahead of major obstacles and major equipment, balance of plant impacts, cost of reagent and objectives for additional process benefits at the plant.
Costs and performance
Capital costs for the technology are inversely related to unit size, and have ranged from $4 to $10 per kW. Operating costs are highly dependent on the level of SO3 in the flue gas, the SO3 removal required, and the reagent contract. Although site specific, operating costs in the US are estimated to fall between $250 to $600 per ton of SO3 removed.
Data from full-scale demonstration tests and many of the permanent installations show that over 90 per cent removal efficiency can be achieved at molar ratios of 1.0 and above, and removal efficiencies above 95 per cent are possible at higher reagent ratios (Figure three). Full-scale installations have generally achieved reliable, consistent operation at molar ratios of sodium to SO3 of about 1.3 to 2.5, depending on site-specific considerations.
Figure 3. Removal efficiency of 90 per cent can be achieved at molar rates of 1.0 and above
The reagent cost is by far the largest component of the annual operating cost for the SBS Injection technology. Commercial products are available for reagent (sodium bisulphite solution, sodium sulphite solution, and sodium sulphite solids), but are relatively expensive compared to the use of a flue gas desulphurisation (FGD) byproduct solution. Since the product of the scrubbing of SO2 with caustic or sodium carbonate is a solution of sodium sulphite and bisulphite, a low cost reagent for the SBS Injection process can be made at any plant site willing to install and operate a scrubber for the purpose of producing their own reagent for SBS Injection. Alternatively, the In-Situ SBS process using soda ash or caustic can be considered. If a utility elects to produce its own SBS reagent or use the In-Situ SBS process, the reagent cost could be reduced by 30 to 70 per cent.
The effective removal of SO3 from the flue gas can provide valuable operation and maintenance (O&M) benefits, in addition to the highly desirable environmental benefit of eliminating the plume opacity aspects of sulphuric acid aerosol emissions. Removal of 90 to 95 per cent of the SO3 will reduce the acid dew point of the flue gas by somewhere between 4à‚°C and 16à‚°C (depending on the pre-injection SO3 level), thereby providing significant protection against acid corrosion for all equipment and ductwork downstream of the injection point. This lower acid dew point allows for flexibility in the temperature of operation of the air heater, and in some cases can result in substantial savings by improving heat rate.
SBS Injection technology is promising for effectively and selectively removing SO3 from flue gas. The technology is currently being applied at eight power plants totaling approximately 5300 MW, and is being considered for a number of additional applications in the US utility sector. Demonstrations are planned for SO3 mitigation applications for a utility oil fired boiler and a petrochemical industry fluidized catalytic cracker unit (FCCU). The capability to effectively remove SO3 has been consistently demonstrated. As the commercial applications accumulate operating experience, it will be possible to better document long-tem reliability, maintenance and operating costs as well as the potential for other plant benefits like reduced corrosion and improved heat recovery.