By Brad Buecker, Contributing Editor
Stricter environmental regulations are forcing many utilities to install flue gas desulfurization (FGD) systems to control sulfur dioxide (SO2) emissions below levels that can be attained by burning Powder River Basin (PRB) coal alone. The choice for many applications is wet-limestone scrubbing, a proven technology. Startup of the new scrubbers, combined with the many workforce retirements that are coming or have already occurred, will force many new personnel to learn FGD details. Properly controlling chemistry in these systems is vital for issues such as scale control, good reagent utilization and corrosion prevention. This article examines important concepts of wet-limestone scrubbing.
A generic wet-limestone flow diagram is outlined in Figure 1. (The diagram also applies for systems using hydrated lime-Ca(OH2)-as the reagent, where equipment and vessel sizes are smaller.) Wet-limestone scrubbing is a classic example of an acid-base chemistry reaction applied on a large industrial scale. Simply stated, an alkaline limestone slurry reacts with acidic sulfur dioxide. As flue gas passes through the scrubber and is contacted by the limestone slurry sprays, sulfur dioxide absorbs into the liquid. Theoretical chemists argue that sulfur dioxide forms only a hydrated compound, where individual SO2 molecules are surrounded by water. However, when SO2 is introduced to water, a pH depression occurs, where the following equilibrium reactions are representative:
Limestone, whose primary components are calcium carbonate (CaCO3) with lesser amounts of magnesium carbonate (MgCO3), when introduced to water will raise the pH according to the following mechanism.
However, CaCO3 is only very slightly soluble in water, so this reaction is minor in and of itself. In the presence of acid, calcium carbonate reacts much more vigorously. It is the acid generated by absorption of SO2 into the liquid that drives the limestone dissolution process.
Equations 1, 2 and 3 when combined illustrate the primary scrubbing mechanism.
In the absence of any other factors, calcium and sulfite ions will precipitate as a hemihydrate, where water is actually included in the crystal lattice of the scrubber byproduct.
However, oxygen in the flue gas has a major impact on chemistry and in particular on byproduct formation. Aqueous bisulfite and sulfite ions react with oxygen to produce sulfate ions (SO4-2).
Approximately the first 15 mole percent of the sulfate ions co-precipitate with sulfite to form calcium sulfite-sulfate hemihydrate [(CaSO3 • CaSO4) • ½H2O]. Any sulfate above the 15 percent mole ratio precipitates with calcium as gypsum.
Calcium sulfite-sulfate hemihydrate is a soft, difficult-to-dewater material that previously has had little practical value as a chemical commodity (although interest is beginning to develop in agricultural benefits of the material). For this reason, many scrubbers are equipped with forced-air oxidation systems to introduce additional oxygen to the scrubber slurry. A properly designed oxidation system will convert all of the liquid sulfite and bisulfite ions to sulfate ions. Sulfate, of course, precipitates with calcium as gypsum, which typically forms a cake-like material when subjected to vacuum filtration. In many cases 85 to 90 percent of the free moisture can be removed by this relatively simple mechanical process. Gypsum is the primary ingredient of wallboard. A number of FGD systems throughout the world produce wallboard-grade byproduct. (To read more about combustion by product markets, see the July 2006 issue of Power Engineering.)
Problems that plagued first- and second-generation wet-limestone scrubbers included poor SO2 removal, scale formation in the scrubber vessels and poor utilization of the limestone reagent. Spray nozzle efficiency, scrubber vessel configuration, limestone reactivity and particle size are all factors that influence these processes.
Adequate mixing of the flue gas and slurry is critical. Early scrubber towers usually were equipped with internal packing or trays to enhance gas-liquid contact. While the theoretical concept behind these mixing devices was valid, the material would often become plugged with scale, necessitating periodic cleaning, replacement or laborious control methods.1 In some early designs, the packing consisted of plastic balls, which often would “cement” together and cause a degradation in scrubber performance. Spraying technology has greatly improved in the last few years and open spray towers are now becoming popular.2 Spray nozzle design is critical in these systems, as droplet size must be optimized to provide the best contact. The slurry spray pattern also must be such that channeling of the flue gas does not occur. (An excellent article on spray nozzle types may be found in reference 3.) A still-common technique is to introduce the flue gas in a tangential pattern to the scrubber tower. This imparts a centripetal motion to the gas and forces it to swirl around the tower as it passes upwards. The swirling action improves slurry-gas mixing and increases gas residence time in the vessel.
Limestone reactivity is another key factor. In general, limestones with 94 percent or greater calcium carbonate content provide suitable alkalinity for reaction. Impurities in the stone may cause significant operating difficulties. Magnesium, a common substitute for calcium, can be either helpful or harmful depending upon its chemical makeup within the stone. If the magnesium exists as homogenous magnesium carbonate (MgCO3), it can enhance SO2 removal by providing extra alkalinity to the scrubbing solution. However, magnesium often co-exists with calcium in a crystal matrix known as dolomite (CaCO3 • MgCO3). Dolomite is rather un-reactive and stones containing a significant dolomite content may require excess feed to achieve the required SO2 removal. Limestones typically also contain inert materials, including siliceous compounds such as quartz. These have different densities than the scrubber byproducts and may negatively affect slurry separation device performance. Iron in limestone can form oxides that plug vacuum filter cloths. Iron can also influence gypsum scale formation on scrubber vessel internals, although this is usually not a problem in forced-oxidation systems.
Limestone reactivity is greatly influenced by particle size. A typical method of preparing limestone slurry is to grind the raw limestone with water in a ball mill. This produces a suspended solution of fine limestone particles (slurry), which is then pumped to the reaction vessel. Smaller particle size increases the total surface area of the limestone reactant. Grind size is determined by passing a slurry sample through progressively smaller sieves. A typical specification for grind size in first-generation scrubbers was 70 percent passage through a 200-mesh screen. However, scrubber designers, operators and chemists came to realize that this size was too coarse to promote good utilization. Nowadays, 90 percent or greater passage through a 325-mesh screen is more desirable.
Even with a well-ground, high-purity limestone, utilization may fall short of expected levels. A recent approach is the use of additives to enhance performance. One of the most popular of these is adipic acid (HOOCCH2CH2CH2CH2COOH), which goes by the common name of dibasic acid (DBA). DBA functions by assisting limestone dissolution, which in turn increases sulfur dioxide removal kinetics. Supplemental DBA feed represents a practical approach for enhancing the SO2-removal performance of existing scrubbers.
Improvements have also been made in scrubber vessel construction material. Chlorine in coal converts to hydrogen chloride (HCl) during combustion. HCl is an acid that reacts with limestone to produce calcium and magnesium chloride (CaCl2 and MgCl2), both of which are soluble salts. Chloride concentrations may reach several thousand milligrams per liter. Many first- and second-generation designs incorporated stainless steels in system components. These materials proved unstable when exposed to high chloride concentrations, as chlorides penetrate the protective oxide layer on stainless steels and initiate pitting.
Various inorganic and organic linings have been tested over the years. These often failed due to poor application or simply the stressful nature of the scrubber environment. More exotic materials are not always the answer. Even titanium will fail in the presence of porous slurry deposits, which allow chloride to concentrate at the metal substrate. These conditions are prevalent at the wet-dry interface where flue gas first contacts the slurry sprays. A retrofit technique for some scrubber components, such as scrubber vessel outlet ducts, is overlay (commonly termed wallpapering) of the base metal with a corrosion-resistant material. The most common choices have been the nickel-based alloys C-276 and C-22.
An issue of continuing importance is byproduct disposal. At plants equipped with forced-air oxidation systems and filter drying systems to produce high grade gypsum, land requirements and costs for byproduct disposal are greatly reduced when plant manufacturers can sell the product to wallboard manufacturers. Other options include gypsum production with landfill of the byproduct, or no forced oxidation with disposal of the byproduct slurry in retention ponds. Some utilities own enough land so that the retention ponds can serve as evaporation ponds, eliminating liquid discharge as an issue. Disposal requirements will undoubtedly become more important due to water conservation issues. Plant personnel are facing regulations that require minimized or zero liquid discharges. No longer can a scrubber be planned without giving thought to liquid discharge issues.
One potential drawback of wet-limestone systems is that they can emit very fine particulates and aerosols. Health and regulatory officials are becoming increasingly concerned about the effects of fine particulates on human well-being. Regulations are becoming increasingly strict with regard to particulate discharge.
Other drawbacks of wet-limestone scrubbing are large up-front capital costs, large equipment size and substantial predictive and corrective maintenance requirements. Substituting hydrated lime as the reagent reduces equipment size and costs, but increases reagent costs and material handling issues. Thus, limestone is more popular as a reagent for wet system handling equipment. This eliminates the need for expensive and maintenance-intensive dewatering and sludge disposal equipment. Also, the drying process does not expose the scrubber materials to chlorides as in wet systems. This relaxes requirements for materials of construction, which in turn lowers capital and building costs.
- Buecker, B., “Gypsum Seed Recycle in Wet-Limestone Scrubbers”; unpublished paper presented at the Utility Representatives FGD Conference, June 10-12, 1986, Farmington, New Mexico, USA.
- Weilert, C., Basel, B., and P. Dyer, “Startup and Initial Operation of the Wet Limestone FGD Retrofit at V.Y. Dallman Station Units 31 & 32”, from the 2001 Conference Proceedings Power-Gen International, December 11-13, 2001, Las Vegas, Nevada, USA.
- Pagcatipunan, C., and R. Schick, “Maximize the Performance of Your Spray Nozzle System”; Chemical Engineering Progress, Vol. 101, No. 12, Pgs. 38-44, December 2005.
Brad Buecker is an air quality control specialist at a large Midwestern power plant. He has previous experience as a chemical cleaning services engineer, a water and wastewater system supervisor and a consulting chemist for an engineering firm. He also served as a results engineer, flue gas desulfurization (FGD) engineer, and analytical chemist for City Water, Light & Power, Springfield, Illinois. Buecker has written more than 70 articles on steam generation, water treatment and FGD chemistry, and is the author of three books on steam generation topics published by PennWell Publishing. He is a member of the ACS, AIChE, ASME, and NACE.