Boilers, Water Treatment

SO2 Control Technology

Issue 6 and Volume 115.

 

Click to Enlarge

By Lindsay Morris, Associate Editor

Sulfur dioxide (SO2) emissions have long been an area of concern for the U.S. Environmental Protection Agency (EPA) and power producers. A number of regulations are currently targeting or soon are set to target SO2 emissions, including the Clean Air Transport Rule (CATR) and the SO2 National Ambient Air Quality Standards (NAAQS). In addition, SO2 is a surrogate for acid gas hazardous air pollutants (HAPs), poised for regulation under the Air Toxics Rule for utilities via the proposed Maximum Achievable Control Technologies (MACT) standard.

It seems as though SO2 is an all-encompassing concern in the scope of emissions control. As a result, if a utility can get a solid control on SO2, other air toxins will be controlled, said Andrew Byers, associate vice president of Black & Veatch Environmental Services.

“When you look at all the different things that can be addressed with SO2 controls, it casts a wide net over a lot of EPA air quality issues.”

But do any of the control technologies on the market meet all of the SO2 regulations—current or proposed—in EPA’s mix? And how can a coal-fired generator prepare for compliance while some of the regulations are still in the proposal stage?

According to the EPA, power plants are responsible for 66 percent of worldwide SO2 emissions, with the majority—more than 98 percent—coming from coal-fired power plants. Wet FGD, semi-dry FGD, dry sorbent injection (DSI) and other technologies have been responsible for cutting SO2 emissions by 57 percent between 1980 and 2008, according to the Edison Electric Institute. But there is still work to be done and utilities need to be aware of the many regulations likely to affect their plant’s financial and operational decisions.

What Are the Rules?

The Clean Air Transport Rule (CATR)—proposed by the EPA in July 2010—was created to reduce the interstate transport of emissions from power plants in the eastern U.S. as required by the Clean Air Act. The Transport Rule replaces the Clean Air Interstate Rule (CAIR) that EPA issued in March 2005. “Good neighbor” provisions of the Act require states to prohibit emissions that contribute significantly to a downwind state’s air quality problems. One example is West Virginia, where the EPA found that power plants significantly affect air quality statuses of counties in Ohio, Indiana, Kentucky, Pennsylvania and Michigan, preventing these states from achieving federal air quality standards.

The Transport Rule will go into effect under two phases: the Phase 1 compliance date of 2012 and the Phase 2 compliance date of 2014. The Transport Rule also establishes two independent trading programs for SO2: Group 1 states and Group 2 states (See figures 1 and 2). SO2 emissions from Group 1 states would be capped at 3.1 million tons per year beginning in 2012 and 1.7 million tons per year beginning in 2014. The 2012 cap represents a 13 percent reduction below 2009 emissions levels. SO2 emissions from Group 2 states would be capped at 0.8 million tons beginning in 2012. The 2012 cap for Group 2 states represents a 29 percent reduction below 2009 emissions levels. The rule will affect fossil fuel-fired power plants 25 MW and larger in 31 states and Washington D.C., encompassing both SO2 and NOx reductions.

The SO2 NAAQS are another concern for utilities. The Clean Air Act requires EPA to set national air quality standards for SO2 and five other emissions types. On June 2, 2010, the primary SO2 NAAQS was revised to set the SO2 standard at 75 parts per billion (ppb), which is attained when the three-year average of the 99th percentile of one-hour daily maximum concentrations does not exceed 75 parts per billion (ppb). SO2 emissions must now be recorded at a one-hour measurement, which raises a new set of challenges for utilities. The Clean Air Act directs states to submit their SO2 designation recommendation by June 3, 2011.

On March 16, 2011, EPA proposed new source performance standards and Maximum Achievable Control Technologies (MACT) standards for fossil fuel-fired units under the Air Toxics Rule. This ushers in the first-ever national standards for mercury, arsenic and other hazardous air pollutants (HAPs). Since SO2 is a surrogate for HAPs, the Utility MACT standard could also introduce or reinforce a push toward SO2 control installation.

Best Available Retrofit Technology (BART) guidelines under the Clean Air Visibility Rule will also take a toll on SO2 emissions (as well as NOx, ammonia and certain volatile organic compounds), affecting facilities built or reconstructed between Aug. 7, 1962, and Aug. 7, 1977, that have the potential to emit more than 250 tons a year of emissions and fall into one of 26 different categories. These include utility and industrial boilers and large industrial plants such as pulp mills, refineries and smelters. The EPA’s presumptive BART emission limit for SO2 is 0.15 lb/mmBtu, or 95 percent removal.

Dissecting the regulations

“The way I read into it, EPA’s ultimate goal is to have controls on every single coal-fired unit in the U.S.,” said Robert Nicolo, director of air quality for Hitachi Power Systems America.

This seems to be a common assessment among many utilities. Especially when it comes to SO2 emissions, it is questionable whether any coal-fired plant will survive the regulations sweep without control technology installations or switching to low sulfur coal or natural gas. Instead of undergoing extensive retrofits, a percentage of coal-fired generation—an estimated 50 to 60 GW by 2020—will likely be retired, according to a 2010 Credit Suisse report.

While a control technology installation is often the simplest, most cost-effective method of compliance for younger plants, retrofits on older plants are met with a number of challenges. For this reason, some industry leaders think the host of regulations from EPA is intended to be a push towards renewable energy or natural gas and away from coal.

“EPA is not necessarily driving legislation only from the emissions monitoring standpoint, but from the incentive side; the cost to burn coal vs. natural gas,” said Barney Racine, software development manger for the environmental solutions group of Honeywell Process Solutions.

An example of an EPA push toward natural gas is in Oklahoma, where three coal-fired plants have not met compliance of the Clean Air Visibility Rule. The plants are Oklahoma Gas and Electric (OG&E) Co.’s Muskogee and Sooner stations and AEP-Public Service Co. of Oklahoma’s Red Rock Station. The Oklahoma Department of Environmental Quality submitted a State Implementation Plan recommending Best Available Retrofit Technology switches for the plants. However, EPA identified the SIP as one that “does not meet one or more of the required elements.” EPA states that natural gas conversions, however, would be an acceptable control method.

While renewable energy or natural gas conversions may be acceptable compliance methods, they are not necessarily the simplest undertaking. So how does a utility decide which compliance steps to take in order to meet all of the SO2 compliance requirements in the works?

“They’re all getting ready to do something; they’re holding back to pull the trigger while evaluating further regulations that will be implemented in the months and years ahead,” Nicolo said of the position utilities find themselves in regarding SO2 control.

Transport Rule

EPA is expected to finalize the CATR by the end of 2011. Emissions reduction limits will then begin to take effect in 2012. EPA’s preferred approach would allow intrastate trading among covered power plants with some limited interstate trading.

Byers of Black & Veatch said it is understood that trading under the CATR will provide “a bit of cushion where operators can add SO2 control on the back end.” Given the levels that may be imposed under the rule, it may actually be more beneficial to add SO2 control on the back end and bank or sell credits back into the market, Byers said.

Ultimately, Byers said, the Transport Rule will drive SO2 controls on existing facilities in Group 1 states.

In an effort to compare statewide SO2 emission reductions required for CATR, Burns & McDonnell has compared state-by-state 2010 annual SO2 emissions to EPA’s proposed CATR SO2 allowance allocations for existing units. The data reflects the wide range of SO2 reduction requirements, which will have an obvious effect on the technology required in each area. Some states, like Nebraska, have no SO2 reductions required to meet 2012 or 2014 Transport Rule requirements. Other states fall closer to the opposite end of the spectrum. Ohio, for example, is projected to require a 21 percent reduction of SO2 by 2012 and a 70 percent reduction by 2014. Pennsylvania is expected to require a 4 percent reduction of SO2 by 2012 and a 65 percent reduction by 2014.

Brandy Johnson, manager of flue gas desulfurization project development for Babcock & Wilcox Power Generation Group Inc., said her company is seeing an increase in demand for scrubbers in comparison to previous years. “We expect the demand to continue to increase as a result of the CATR as well as NAAQS, consent decrees and state requirements.”

The effects of the CATR are viewed conversely by others, however; interpreted as a path for the EPA to encourage conversions away from coal on a state level. Since state incentives for renewables are increasing at the same time state-level SO2 control is being required, this could trigger conversions to renewables, said Racine of Honeywell. That scenario, in turn, would cause a reduction in the need for SO2 control technology.

“EPA is trying to get companies from a state incentive level to drive the conversion away from coal,” Racine said.

SO2 NAAQS: The One-hour Challenge

Perhaps the greatest concern for utilities facing compliance for SO2 NAAQS is the one-hour daily maximum concentration measurements, not to exceed 75 ppb. (The previous standard allowed no more than 140 ppb over 24 hours.) In contrast, many other EPA standards have concentration measurements taken over a span of eight hours, one day or even one year. For SO2 NAAQS, EPA argues that long-term measurements of SO2 NAAQS produce too much fluctuation. “While national SO2 air quality levels have improved, EPA remains concerned about short-term peak SO2 concentrations,” EPA said on its web site.

EPA has said no evidence exists that long-term exposure affects health, said Racine. The concern over short-term exposure has also prompted EPA to change data reporting requirements to include reporting the highest five-minute average from each hour. While there is no limit on the five-minute average, the reporting is being done to collect additional data so EPA can further study the health impacts of short-term exposure to high concentrations of SO2 based on official data from their monitoring network.

Click to Enlarge
McGill Air Clean’s spray dry scrubber and fabric filter operating on a Bubbling Fluidized Bed boiler. Photo courtesy McGill Air Clean.

Consequently, short-term SO2 concentrations have become the focus for utilities. Adding to the new challenge is the fact that a measurement can be conducted during a plant’s start-up or shut-down at the agency’s discretion, Byers said.

“Usually SO2 control isn’t operating at its fullest efficiency during start-up or shut-down,” Byers said. “Therefore, there might be more SO2 released during start up or shut down.”

For example, Byers said a typical SO2 control system might reduce SO2 by 98 percent. However, during start-up or shut-down, the technology might operate at only half the efficiency.

As a result of the new measurements, EPA is “expecting an increase in the number of non-attainment areas,” said Racine. SO2 NAAQS is a five-year program, so it will take time to “trickle down to the individual facilities.”

Racine said that SO2 NAAQS will likely have a positive impact on scrubber technology demand. After one-hour ambient measurements are collected, EPA can focus on identifying which facilities nearby may be contributing to an exceedance of the standard, which ultimately may result in requirements for still more emissions control installations.

“If a plant emits a lot of SO2 during start-up and it affects the non-attainment status, that means someone who doesn’t have a scrubber on now but has high SO2 emissions during start-up and shut-down may need to install scrubber technology.”

Charlie Moore, sales engineer for emissions control supplier McGill AirClean LLC, said the company is anticipating a strong demand for acid gas control technology for both HCl and SO2 based on SO2 NAAQS and Utility and Industrial Boiler MACT in the near future.

MACT: SO2 as Surrogate

The Air Toxics Rule is EPA’s replacement for the Clean Air Mercury Rule (CAMR), an interstate cap-and-trade program issued in 2005. On March 8, 2008, a federal court ruled that EPA violated the Clean Air Act when it sought to regulate mercury-emitting power plants through CAMR. EPA estimates the proposed Air Toxics Rule will reduce mercury emissions from covered power plants by 91 percent, acid gas emissions by 91 percent and SO2 by 55 percent. This is the first time federal limits have been established for hazardous air pollutants (HAPs) and mercury control. A consent degree with public health and environmental groups requires EPA to finalize the standards by Nov. 16, 2011. Compliance to the MACT ruling is expected by around 2015.

While the MACT ruling does not limit SO2 emissions directly, it provides an alternative wherein plants can meet an SO2 standard in place of the HCl limit. Plants can opt to meet the 0.2 lbs. of SO2 per million Btu instead of installing HCl monitoring technology.

In other words, there will be some allowances in the MACT standard for plants that have already installed or are installing SO2 control, said Byers. “These plants can use an SO2 limit as a surrogate for what they call acid gases or in place of monitoring the compliance limit for hydrogen chloride.”

Carl Weilert, principal air pollution control engineer in the energy division of Burns & McDonnell, said that if a source burns a low-chlorine fuel like Powder River Basin (PRB) coal, it might already be in compliance with the proposed MACT standard for HCl. A source burning Eastern bituminous coal, which typically has a higher chlorine content, however, will require as much as 99 percent HCl removal.

“Plants won’t want to comply with the HCl standard if they already meet the alternate SO2 standard,” Weilert said. “If they already have a dry scrubber on a PRB coal plant, they are probably already meeting the proposed SO2 surrogate MACT standard.”

While the MACT standard is still in the proposed stage, some in the industry believe it is already providing more regulatory certainty than in the past. “Over the past few years, utilities have been asking for regulatory certainty to allow them to make informed decisions around the investment of emissions reduction technologies,” said Tom Thompson, CEO of Eco Power Solutions.

Thompson said the increased regulatory certainty brought about by the MACT standard should allow power generators to embrace a compliance plan.

The technology produced by Eco Power Solutions, Comply 2000, is a multi-pollutant removal system included by EPA on a list of technologies that “offer the potential of reduced compliance costs and improved overall environmental performance” of the New Source Performance Standards and MACT.

Other Concerns

While greenhouse gas legislation remains a question mark for the time being, Hitachi’s Nicolo said that if such regulations do take effect, they could impact SO2 control technology implementations as well. Amine-based CO2 capture technology, for example, becomes “poisoned” if any SO2 is emitted in the process. Therefore, if utilities begin to install greenhouse gas control technologies, they may also need to install more stringent SO2 controls – such as a wet scrubber to drive SO2 emissions to single digit PPM levels.

The Clean Water Act’s Section 316(b) Cooling Water Intake regulation could also affect SO2 technology. Because 316(b) limits the amount of groundwater allowed in a power plant, some utilities may elect to use a dry scrubber as opposed to a wet scrubber. Many scrubber technology providers, such as Hitachi, are developing technologies to reduce the amount of water used by a wet scrubber, said Nicolo.

Control Technologies

Two basic methods exist for controlling SO2 emissions from coal-fired power plants: switching to a lower sulfur fuel or implementing SO2 control technologies. Switching to a lower sulfur coal can be a fairly simple undertaking. Certain coal types inherently low in sulfur, such as sub-bituminous coal from the PRB of Montana and Wyoming, can sometimes bring SO2 emissions within acceptable limits. However, some plants cannot burn 100 percent PRB coal without substantial modifications to the boiler or fuel-handling systems. In these cases, PRB coal can be blended with a bituminous coal to reduce emissions.

If a facility is unable to burn a lower sulfur coal or needs to achieve greater SO2 emissions reductions than is possible through a fuel switch, emissions reductions technology should be implemented.

Click to Enlarge

Eco Power Solutions’ Comply 2000 is a multi-pollutant removal system. Photo courtesy Eco Power Solutions.

Many technologies exist for controlling SO2, and the decision of which to implement must be conducted on a case-by-case basis. Utilities must take into consideration cost, regulatory impact, proximity and more. Control technologies capable of capturing SO2 include dry scrubbers, wet scrubbers and semi-dry scrubbers.

In the case of a typical wet scrubber, flue gas coming from the boiler is saturated with a slurry containing limestone reagent. This type of SO2 control is characterized by high capital cost, low operating cost and high performance.

Dry scrubber processes inject particles of alkaline sorbent into the flue gas, producing a dry solid by-product. The flue gas leaving the absorber is not saturated in this process. Dry scrubber costs have gone down in recent years due to technical innovations. Dry scrubber systems can be grouped into three categories: spray dryers, circulating spray dryers and dry injection systems. The circulating dry scrubber (CDS) technology operates at similar temperatures, but is based on separately feeding dry hydrated lime and water into a fluidized bed reactor. Here the SO2 removal takes place in a bed of moistened powder.

Another scrubber option known as dry sorbent injection (DSI) involves injecting a reagent in dry powdered form (hydrated lime, sodium bicarbonate or Trona) into the flue gas upstream of existing particulate control equipment. DSI is praised for its simplicity and low capital costs, but is limited in performance and requires reagent injection rates above stoichiometry that are often necessary for removal efficiencies.

Jim Dickerman, director of flue gas treatment applications for Lhoist North America, said a DSI system can cost about 10 percent of a standard wet scrubber. Dickerman said that while DSI is not as effective for SO2 removal, it is better suited for sulfur trioxide (SO3) and HCl removal. DSI can also aid in mercury removal and reduce PM emissions by removing the SO3 that condenses and forms PM, Dickerman said.

One semi-dry FGD process uses a circulating fluid bed (CFB) dry scrubber or a lime spray dryer. Lime spray dryer technology operates by spraying a slurry of slaked lime reagent into the flue gas. The flue gas is cooled to 30 to 40 degrees above its saturation temperature as the slurry droplets are dried. As a result, when flue gases come out of the spray dryer, they are present as a dry powder product that is collected in a bag house. Semi-dry FGD technologies are characterized by capital costs that are about half that of wet FGD. They have higher operating costs than wet FGD, but lower operating costs than DSI.

McGill AirClean is a provider of DSI technology as well as spray dry scrubbers. SO2 removal efficiencies are process-specific depending on temperature, moisture, and other factors for typical DSI, Moore said. For example, 50 to 70 percent emissions reductions are achievable with lime injection, 70 to 80 percent reductions with Trona injection and 90 percent reductions or more are possible with sodium bicarbonate injection, assuming a baghouse is used for the PM collection device.

Dickerman said it’s not likely a large demand will arise for DSI as a result of SO2 regulations, at least not for hydrated lime systems. “However, in certain cases where relative low SO2 removal rates are needed for compliance, DSI may be a good solution as it is low capital and can be turned on and off with ease.”

With spray dry scrubbers, however, removal efficiencies are not as process-dependent as DSI, Moore said, and greater than 90 percent SO2 removal is achievable. Spray dry scrubbers can achieve higher acid gas removal efficiencies at lower stoichiometric ratios, but there is maintenance associated with the lime system and dual fluid nozzles or atomizers to ensure consistent lime slurry drop size.

One of the main drawbacks of wet scrubbers is the wet effluent or blow down that is necessary, Moore said. High operating costs are also associated with the large pumps required to recycle water. According to a report by the Utility Air Regulatory Group, capital cost for FGDs escalated over the past six years. For example, a retrofit of wet FGD to a 500 MW plant between 2004 to 2007 cost an average of $342/kW. A similarly sized unit retrofit with FGD between 2008 to 2010 cost $407/kW, a 19 percent cost escalation.

Robert Nicolo of Hitachi said the company has installed and commissioned six wet scrubbers in the U.S. between 2009 to 2011. Hitachi’s wet scrubbers are an SO2 control technology with an open spray tower design capable of achieving single digit PPM levels of SO2 (equating to over 99 percent SO2 control). Nicolo said the units have also delivered aggressive emissions limit guarantees for mercury, SO3 and PM control. Nicolo said that Hitachi’s wet limestone scrubber is able to treat 100 percent of the flue gas from a 1,000 MW-plus facility.

Nicolo said wet scrubbers are advantageous to dry scrubbers in certain cases because they are able to remove a higher percentage of SO2, and are capable of treating flue gas for a variety of higher sulfur fuels delivering fuel flexibility for utilities, whether it’s PRB, bituminous or lignite.

Comply 2000, a multi-pollutant system produced by Eco Power Solutions, reduces SO2 emissions by injecting a fogging spray mixed with a hydrogen peroxide solution that is condensed concurrently with other pollutants over coils to remove all combustion emissions from the exhaust gas stream. This process converts NOx and SO2 to nitric and sulfuric acid in the wastewater stream, resulting in 99 percent removal of SO2, mercury, halogens including fluoride, chlorine and bromide, heavy metals include arsenic and cadmium, 2.5 and 10-micron PM, as well as 20 percent removal of CO2. Thompson of Eco Power Solutions said that almost 100 percent of the utility generators his company works with have SOx concerns of some sort: SO2, SO3, or both. “We’ve seen an uptake inbound in our activity.”

Thompson said the Comply 2000 has advantages over traditional scrubber technology due to its capital cost (30 percent less than traditional technologies) and retrofit characteristics (about half the size than a typical FGD). Also, Comply 2000 generates supplemental energy by recovering waste heat that normally escapes into the atmosphere, reducing a plant’s operating costs.

Technology Implementation

Each regulation will affect power generators differently, depending on where their plants are, what type of coal is burned and other factors.

When it comes to compliance with SO2 NAAQS, since the states are responsible for developing their own SIPs, technology will likely be different for each state, depending on which states have SO2 non-attainment areas, said Weilert of Burns & McDonnell.

For CATR, Weilert said the kind of technology needed to comply with the rule will depend on what state a utility is physically located. For states that need “a very small reduction,” which Weilert indicated to be 20 percent or less, DSI with Trona could be an acceptable control technology. Another option could be if one source in the state installed a highly efficient scrubber and if other plants bought allowances from that source. This scenario assumes that some type of trading will be allowed in the final version of CATR. For states with higher reduction requirements, a greater number of scrubbers will likely need to be installed.

For the MACT ruling, the EPA has predicted the need for 166 GW of baghouse installations, 25 GW of dry scrubber installations and 1 GW of wet scrubber installations. Weilert said the low number of wet scrubbers in EPA’s prediction is “hard to believe” because it’s unlikely that most bituminous coal-fired units, which produce relatively high sulfur emissions, will install dry scrubbers. Weilert said that units burning Eastern coal, in general, will need to install wet scrubbers designed for at least 95 percent removal efficiency.

Lignite-fired plants, however, have their own subcategory in the MACT ruling, at least with regard to mercury emissions. Lignite fuel contains high mercury levels, but features a combination of low chlorine in the fuel and high alkalinity in the ash. Therefore, lignite plants may “get a break on the HCl emission limit,” Weilert said. Those plants may need to install only a baghouse for PM control and activated carbon for mercury control.

According to the Institute of Clean Air Companies, about 60 GW of scrubbers were brought online from 2008 to 2010. However, CATR, SO2 NAAQS and the proposed MACT standard will aim to cut SO2 emissions regionally and statewide at an even greater clip than previous requirements, encouraging the installation of even more SO2 control technologies. Since all of these regulations will likely be enforced within the next five years, many power producers are already exploring technology implementation options.

More Power Engineering Issue Articles
Power Engineerng Issue Archives
View Power Generation Articles on PennEnergy.com