Dry sorbent injection and activated carbon injection are evolving and improving as methods of SOx and acid gas removal at a low capital cost, write Conner Cox and Jon Norman


Dry sorbent injection modular test unit

Credit: UCC

The Council of the European Union has published the Industrial Emissions Directive (IED, 2010/75/EU) to set limitations for emissions.

This has been accompanied by a working draft of the best available techniques (BAT) from the European IPPC Bureau to define the available technologies to achieve those limitations.

This regulation reduces the amount of allowable emissions from industrial boilers, including SOx dust and mercury (Hg). In order to achieve the new regulations, plants are evaluating available technologies, such as dry sorbent injection (DSI) and activated carbon injection (ACI) systems for the reduction of acid gases and mercury.

DSI (BREF Section 10.8) is listed as an applicable BAT for hard coal or lignite units less than 300 MW, when used in combination with an electrostatic precipitator (ESP) or fabric filter. This proven technology is used extensively in the US for the reduction of SO2, SO3 and HCl emissions.

DSI systems have also been implemented throughout Europe for years, primarily for municipal waste combustors with sodium bicarbonate (SBC) and hydrated lime as the primary sorbents used to reduce acid gases.

SBC is typically used for high levels of SO2 removal with ESPs and fabric filters. When used in conjunction with in-line milling, removal rates up to 98 per cent have been recorded with fabric filters and over 90 per cent with ESPs.

In-line milling allows the SBC to be stored in a larger particle size, which promotes feeding from the silo. Particle size is then reduced through the mill to increase the surface area available to react in the ductwork.

Similarly, hydrated lime is regularly used for the reduction of acid gases from hard coal and lignite boilers. The equipment is very similar to an SBC injection system. However, the sorbent is delivered in a very fine and porous state. Advanced hydrated lime products have achieved moderate to high levels of SO2 removal when used in conjunction with a fabric filter.

Typically, hydrated lime injection rates required for high SO2 removals are higher than SBC injection rates for similar performance. The use of hydrated lime can allow fly ash to remain salable, whereas SBC cannot.

The traditional role for hydrated lime has been the removal of HCl and SO3. When compared to SO2, these acid gases appear in smaller quantities and are much more reactive with the hydrated lime product, and therefore injection rates are drastically reduced. Hydrated lime injection has been shown to remove >95 per cent HCl and SO3. This method optimizes activated carbon injection (ACI) through the reduction of SO3 interference.

Mercury reduction

The US has recently implemented the Mercury and Air Toxics Standard (MATS), as well as the Maximum Achievable Control Technology (MACT) regulations for the reduction of hazardous air pollutants, including mercury (Hg), from electric generating units and industrial boilers. As a result, hundreds of ACI systems have been installed and are successfully achieving compliance by injecting powdered activated carbon (PAC).

The BREF document states that the primary methods for removal of Hg from flue gas include Hg oxidation through an SCR and capture by a wet scrubber, or sorbent injection into a fabric filter, ESP or other device. Very similar to a DSI system, ACI technology is relatively simple, low capital cost, easy to install, and reliable. Problems with plugging, abrasion, and corrosion can occur; however, this takes place when the system is not designed properly for the product.

There is a large number of ever-improving sorbents available for Hg removal from flue gas. These include PACs with and without oxidizing agents, as well as non-carbon products. The type of sorbent required is dependent on the unit where ACI is applied. Mercury is much easier to capture on sorbents or in other pollution control devices when it is in the oxidized (Hg2+) state.

Oxidants can be fed into the system at a number of locations. For some, application of a calcium bromide (CaBr2) solution on the coal feeders provides the oxidation required and allows the use of a lower-cost, non-oxidizing PAC. Other plants find that the use of an oxidizing PAC is required in order to achieve the removals needed to meet the limits.

Figure 1. Dry sorbent injection modular test unit

Credit: UCC

Specialty sorbents, like Carbonxt’s CXT-2100, have been designed to provide the oxidation with little to no halogen content due to concerns around corrosion associated with bromine and other halogens. Other sorbents have moved away from carbon altogether and strive to remove Hg using bentonite or iron-based products. These products can vary dramatically in performance, depending on the application. The graph on page 4 shows different performance for a variety of PAC sorbents used in a recent test in Europe.

Injection locations

There is no one-size-fits-all solution for the multi-pollutant control that is required to meet the IED limits. Industry testing has shown that performance of the acid gas and Hg sorbents depends greatly on residence time, particulate removal device, other air pollution control equipment and mixing.

By injecting PAC upstream of the air pre-heater (GE Power’s patented Mercure Process), more residence time and mixing is provided before the PAC is removed from the flue gas. This is highly beneficial for overall mercury removal performance and sorbent usage optimization. Recent testing has shown up to a 50 per cent reduction in PAC usage is possible. Mixing for all sorbents is further improved through the use of proper distribution splitters and mixing devices. The use of a fabric filter for particulate removal also provides increased performance due to added residence time while the sorbent is on the filter cake.

For units with high SO3 levels (>5 ppm), typical on units with SCRs and/or using higher sulphur coals, the SO3 interferes with the PAC performance by blocking the pores that would otherwise adsorb Hg. To combat this, hydrated lime is often injected upstream of ACI to remove the SO3 prior to injection of PAC. Ideally, the injection of hydrated lime will occur a minimum of one second prior to the injection of PAC. When enough ductwork exists, both injections should occur prior to the air pre-heater.

For units also needing high levels of SO2 removal, the use of DSI to target SO2 will drastically reduce the SO3 and HCl levels. This can help the performance of PAC by removing the SO3, but it can also negatively affect Hg reduction by removing halogens like Cl that would assist in Hg oxidation.

When using SBC for SO2 removal on low sulphur coals, a beneficial strategy is to inject PAC upstream of the air heater and SBC downstream. This allows the PAC to adsorb Hg before the HCL is removed, and provides the ideal temperature window for SBC. If SBC is injected at temperatures above 350oC, the product softens and loses porosity. Conversely, on high sulphur coal units (with high SO3 levels), it is usually best to inject the DSI sorbent at the air heater inlet and PAC downstream as described in the previous paragraph. Firms experienced with DSI and ACI should be consulted to determine the best strategy for a given unit/fuel.

Figure 2. Activated carbon injection modular unit

Credit: UCC

It is also recommended to test at the specific unit to demonstrate the best application for the unique situation.

The equipment required for a DSI system is relatively simple, low capital cost and reliable, but when engineered incorrectly the systems will regularly plug, wear and underperform. Therefore, only experienced, reputable DSI firms should provide these systems. The basic system consists of a storage silo equipped with one or more feed trains. The feed trains typically include measurement for injection rate, as well as a variable speed feeder to meter the sorbent into the conveying line. A positive displacement blower provides the motive air to convey the sorbent in dilute phase from the feed point, through the splitter to lances, and into the ductwork.

Each of these components must be properly designed and sized to minimize pressure drop, plugging, and wear on the conveying line, while maintaining proper velocities and distribution. A system that accurately distributes sorbent throughout the ductwork will provide the optimum usage of sorbent, significantly reducing the operating cost of the system. The use of an in-line mill, like UCC’s VIPER Mill, further improves sorbent usage, while minimizing storage and feed issues. These systems can be custom engineered or supplied in modular fashion based on the application.

ACI equipment is very similar to the DSI equipment. Due to the relatively small injection rates required and the increased cost of the sorbent, the equipment is typically smaller and designed for finer control of injection rate.

This equipment is usually designed to maintain injection rates to within 2 per cent of desired, and the splitters and lance designs are even more important for proper distribution. PAC is also less affected by temperature and humidity in the conveying line, so heat exchangers and dehumidifiers are not commonly used. This allows the equipment to be installed inside the silo skirt, or in a modular unit with or without a silo.

A frequent method for entraining PAC into the conveying line is the use of an air eductor. This device provides a suction at the feed point and is powered by the positive displacement blower that then provides the motive air for conveying the product to the duct. This alleviates difficult venting required with other entraining methods.

With the increased cost per kg of typical PACs, distribution through well-designed splitters and lances is even more important. It is recommended to design the injection grid based on computational fluidized dynamics (CFD) modelling to achieve the best sorbent dispersion. This is true for both DSI and ACI systems.

Project cost

Capital cost can vary depending on project scope, injection rate and degree of modularization and it can typically be completed in less than 12 months.

Operating costs for these systems are driven mainly by the type of sorbent selected, injection rates required, sorbent price and freight. Injection rates and operating costs can be closely estimated by experienced DSI/ACI suppliers for a specific application, given the details of the unit.

Although DSI and ACI are well understood technologies, testing on specific units is often recommended to understand the optimum injection rates and balance of plant effects prior to implementation of the permanent system. UCC currently maintains mobile testing equipment, similar to the modular systems available for permanent installation, in both Europe and the US. Testing equipment is usually installed the week prior to testing and demobilized the week following. Testing durations vary, but typically average between one and three weeks.

Dry sorbent injection is a proven technology that provides SOx and acid gas removal at a low capital cost. For most smaller units, DSI will also provide the lowest annualized cost solution. With a range of additional benefits including reduction of PAC usage, air preheater plugging and backend corrosion, DSI is a flexible technology with many applications. Activated Carbon Injection has continued to evolve and improve as well, making systems and sorbents more effective and better performing. ACI has become a standard, proven technology for Hg reduction.

Conner Cox is Process Engineer and Jon Norman is DSI Technology Manager at United Conveyor Corporation.