Economics of an Integrated Approach to Control SO 2, NO x, HCl, and Particulate Emissions from Power Plants

Transcription

1 Journal of the Air & Waste Management Association ISSN: (Print) (Online) Journal homepage: Economics of an Integrated Approach to Control,, HCl, and Particulate Emissions from Power Plants Brooke E Shemwell, Ali Ergut & Yiannis A Levendis To cite this article: Brooke E Shemwell, Ali Ergut & Yiannis A Levendis (00) Economics of an Integrated Approach to Control,, HCl, and Particulate Emissions from Power Plants, Journal of the Air & Waste Management Association, 5:5, , DOI: / To link to this article: Published online: 7 Dec 011 Submit your article to this journal Article views: 64 View related articles Citing articles: 4 View citing articles Full Terms & Conditions of access and use can be found at

2 TECHNICAL PAPER Shemwell, Ergut, and Levendis ISSN J Air & Waste Manage Assoc 5: Copyright 00 Air & Waste Management Association Economics of an Integrated Approach to Control,, HCl, and Particulate Emissions from Power Plants Brooke E Shemwell, Ali Ergut, and Yiannis A Levendis Northeastern University, Boston, Massachusetts ABSTRACT An integrated approach for the simultaneous reduction of major combustion-generated pollutants from power plants is presented along with a simplified economic analysis With this technology, the synergistic effects of high-temperature sorbent/coal or sorbent/natural gas injection and high-temperature flue gas filtration are exploited Calcium-based (or Na-based, etc) sorbents are sprayed in the post-flame zone of a furnace, where they react with S- and Cl-containing gases to form stable salts of Ca (or Na, etc) The partially reacted sorbent is then collected in a high-temperature ceramic filter, which is placed downstream of the sorbent injection point, where it further reacts for a prolonged period of time With this technique, both the likelihood of contact and the length of time of contact between the solid sorbent particles and the gaseous pollutants increase, because reaction takes place both in the furnace upstream of the filter and inside the filter itself Hence, the sorbent utilization increases significantly Several pollutants, such as, H S, HCl, and particulate (soot, ash, and tar), may be partially removed from IMPLICATIONS This manuscript presents laboratory results from an integrated approach for the simultaneous reduction of major combustion-generated pollutants from power plants (such as, H S, HCl, and particulates), along with a simplified economic analysis Calcium/sodium-based sorbents are sprayed in the post-flame zone of a furnace, where they react with S- and Cl-containing gases The partially reacted sorbent is then collected in a high-temperature ceramic filter, where it further reacts for a prolonged period of time, enhancing its utilization If a limestone/coal mixture is used, capital costs for the filter/sorbent combination were estimated to be in the range of $61 $105/kW for a new plant At projected removal efficiencies for HCl/ / of about 40% and 97 99% for PM 5 particulate removal, based on laboratory results, the levelized costs were projected to be $03 $61/ton of combined pollutant /HCl/ and particulates removed from coal-fired power plants the effluent The organic content of the sorbents (or blends) also pyrolyzes and reduces Unburned carbon in the ash may be completely oxidized in the filter The filter is cleaned periodically with aerodynamic regeneration (back pulsing) without interrupting furnace operation The effectiveness of this technique has been shown in laboratory-scale experiments using either rather costly carboxylic salts of Ca or low- to moderate-cost blends of limestone, lime, or sodium bicarbonate with coal fines Injection occurred in the furnace at 1150 ºC, while the filter was maintained at 600 ºC Results showed that 65 or 40% removal was obtained with calcium formate or a limestone/coal blend, respectively, at an entering calciumto-sulfur molar ratio of A sodium bicarbonate/coal blend resulted in 78% removal at a sodium-to-sulfur molar ratio of HCl removal efficiencies have been shown to be higher than those for reductions of 40% have been observed with a fuel (coal)-to-air equivalence ratio, φ, around The filter has been shown to be 97 99% efficient in removing PM 5 particulates Calculations herein show that this integrated sorbent/filter method is costeffective, in comparison with current technologies, on both capital cost ($/kw) and levelized cost ($/ton pollutant removed) bases, if a limestone/coal mixture is used as the sorbent for fossil fuel plants Capital costs for the filter/sorbent combination are estimated to be in the range of $61 $105/kW for a new plant Because current technologies are designed for removing one pollutant at a time, both their cost and space requirements are higher than those of this integrated technique At the minimum projected removal efficiencies for HCl/ / of about 40%, the levelized costs are projected to be $03 $61/ton of combined pollutant /HCl/ and particulates removed from coalfired power plants INTRODUCTION An integrated technique for capturing multiple pollutants in the effluent of coal-fired power plants and waste-toenergy plants is discussed herein (see also ref 1) Currently, most power plants are fitted with devices that Volume 5 May 00 Journal of the Air & Waste Management Association 51

3 capture particulates [eg, cyclones, electrostatic precipitators (ESPs), and baghouses] However, emissions of, HCl, and are often not controlled because of the associated costs Title IV of the US Clean Air Act Amendments (CAAA) of 1990 calls for an approximately 40% reduction of from the 1980 levels (then practically uncontrolled) by the year 010 Currently, most power plant operators opt for purchasing emission allowances for, depending on the market price, or they minimize emissions by burning low-sulfur coal or burning the cleaner natural gas, depending on its price Burning lowsulfur coal, however, may be detrimental to the particulate capture efficiency of common ESPs because existing ESPs are undersized for fine particulate collection for these coals 3 Upon implementation of the year 000 Phase II of Title IV of the CAAA, power plants may require more drastic controls of emissions Power plant operators that install devices for flue gas desulfurization (FGD) typically choose from currently available technologies such as wet lime/limestone scrubbers, spray dry scrubbers, in-duct or in-furnace sorbent injection, regenerable sorbent techniques, and so on Wet scrubbers have been the most popular method [see the Department of Energy (DOE) report on advanced technologies for the control of emissions from coal-fired boilers 4 ] The 1995 reduction requirements of Phase I of Title IV of the CAAA were met by low- burners and other combustion modifications It is estimated that, in 000, more than one-third of the coal-fired generating capacity in the United States has low- burners installed 5 However, to meet the Phase II standards beyond 000, additional post-combustion treatment may be necessary Techniques for reducing include reburning of fuels in the boiler and selective noncatalytic or catalytic reduction (SNCR or SCR), where reacts with other nitrogen-containing compounds such as NH 3, urea, isocyanic acid, and so on to form stable N Because the air pollution regulations in the United States and their implementation historically changed from year to year and varied from state to state, major coalcombustion pollutants (particulates,, and ) are either not captured at all or are controlled with separate processes compounded in series Both the cost and the space requirement for such compounded processes are high For instance, an effective arrangement of proven technologies would be to place a wet limestone/lime scrubber (FGD) in series with a selective catalytic reactor Efficiencies for these devices can be high (upper 90% for FGD and mid to upper 80% for SCR, see Flagan and Seinfeld 6 ), but their capital costs have been typically in the range of $15 $00/kW for FGD (Soud, personal communication, 000) and $60 $95/kW for SCR, 7,8 depending on boiler size, coal type, and so on For reduction by reburning with either natural gas or coal, capital costs have been reported in the range of $6 $66/kW, while efficiencies are 8 63% 5 A DOE report on reburning technologies gives capital costs in the range of $43 $66/kW for cyclonefired boilers 9 Hence, combined capital cost is estimated to be in the range of $150 $95/kW for FGD/ reburning or FGD/SCR For instance, Felsvang et al 10 quote $06/kW capital costs for an FGD/SCR system fitted to a 90-MWe coal-fired unit A comparative representation of the capital and levelized costs is represented in Table 1 Levelized costs for a wet limestone FGD system, fitted to a 300-MWe new coal-fired unit, are given by Hirano et al 11 to be $37/ton removed To this, costs of $674 $1768/ton removed by SCR should be added 7 If, instead of SCR, fuel reburning is used, these costs are in the range of $1075 $1187/ton removed 5 Hence, the addition of the levelized costs results in a range of $1000 $000/1 ton of plus 1 ton of removed To all these figures, the capital and levelized costs of the ESP/baghouse should be added, to reflect removal of particulates Typical capital costs for pulse-jet baghouses are $60/kW and for reverse air baghouses are $85/kW 1 Hence, compounded capital costs for wet limestone FGD/SCR/baghouse systems are currently expected to be in the range of $45 $380/kW Actually, Holmes et al 13 give this figure to be in the even higher range of $360 $400/kW Compounded levelized costs may exceed the range of $1000 $000/1 ton of plus 1 ton of plus 1 ton of particulates removed, because such processes are additive Integrated and removal systems have been in development or are at the demonstration level Such methods, however, have not yet been completely adapted at the commercial scale Demonstration of some of these technologies has taken place throughout the 1990s, funded by the DOE Clean Coal Technology Program 5 Such systems include the SNOX flue gas cleaning project (by ABB), the LIMB project (by B&W), the SO x - -Rox Box or SNRB project (by B&W), the Gas Reburning Sorbent Injection project (by EERC), the Milliken project (by NY Electric & Gas Corp), and the Integrated Dry - Emissions Control System (by the Public Service Company of Colorado) These systems employ disparate methods to achieve capture of,, and particulates from the flue gases Out of these combined systems, only the SNRB incorporates all three processes in one unit Details of this system, as well as a pertinent economic analysis, are given by Redinger and Corbett 14 and Martinelli et al 15 Briefly, hydrated lime duct injection (at a Ca/S ratio of 19) was combined with anhydrous NH 3 injection (at a NH 3 / ratio of 09) upstream of a catalyzed baghouse, to obtain removal of 85% and reduction of 90% in pilot plant tests of the SNRB system The capital cost 5 Journal of the Air & Waste Management Association Volume 5 May 00

4 Table 1 A list of current pollution control technologies for power plants and associated costs (a) for separate systems and (b) for combined systems (a) Separate Systems Toxic Gas Method Used Notes (Removal Capital Cost Levelized Cost Reference Number Removed Efficiency) SCR Dry-bottom boilers (group 1) $8 $116/kW $1600 $1768/ton 7 (80 90%) Wet-bottom boilers $73 $15/kW $674/ton (50%) Cyclone and cell-fired boilers (80%) 15-MWe $61/kW $811/ton 8 (60%) 50-MWe $54/kW $500/ton (New installation) (60%) 700-MWe $45/kW $165/ton (60%) 50-MWe $5/kW $350/ton (40%) 50-MWe $57/kW $036/ton (80%) 100-MWe (60%) $87/kW MWe (60%) $59/kW (Retrofit installation) 50-MWe (40%) $75/kW Approximately $4050/ton 50-MWe (80%) $86/kW Approximately $500/ton Reburning (cyclone- 110-MWe bituminous coal $66/kW $1075/ton 9 fired boiler) 115% S, 14% N (5%) 605-MWe $43/kW $408/ton Gas reburning MWe coal 3% S; $6/kW ($1/kW $908/ton low- burners 15% heat input by natural for GR, $14/kW (wall-fired boiler) gas (65%) for LNB) Low- cell burner 605-MWe bituminous coal $9/kW $9848/ton Low- cell burner 1% S, 13% N retrofit by B&W, 1999 (>50%) (>90%) $15 $00/kW Soud, 000 (personal communication) Advanced flue gas 500-MWe coal 3% S $94/kW $30/ton 4 desulfurization (>90%) (15-year project life) CT-11 FGD 100-MWe (>90%) $80 $95/kW Saarberg-Holter- 300-MWe coal 3% S $300/kW $534/ton Umwelt (S-H-U FGD) (>90%) (15-year project life) FGD 90-MWe $06/kW 10 Dry scrubber $154/kW Wet limestone FGD 300-MWe $37/ton 11 Particulates Pulse jet baghouses $60/kW 1 Reverse air baghouses $85/kW Volume 5 May 00 Journal of the Air & Waste Management Association 53

5 Table 1 (cont) (b) Combined Systems Toxic Gas Method Used Notes (Removal Capital Cost Levelized Cost Reference Number Removed Efficiency) Dry sorbent injection (>80%) 15 SCR (90%) Approximately $390 $550/ton NO + SO x (New installation) $/kw Particulates Fabric filtration (<003 lb/mbtu) Dry sorbent injection Ca/S = (85%) SCR NH /NO = 09 3 x $450 $600/ton NO + SO x (Retrofit installation) (90%) Particulates Fabric filtration (<003 lb/mbtu) Dry sorbent injection 50-MWe 14 (85 90%) SCR Coal 35% S Approximately (90%) $60/kW Particulates Fabric filtration 1 lb NO /MBtu x (<003 lb/mbtu) Dry sorbent injection 100-MWe (85 90%) SCR Coal 15% S Approximately (90%) $30/kW Particulates Baghouse (<003 lb/mbtu) Wet scrubber (85 90%) 13 SCR (90%) $360 $400/kW Particulates Fabric filtration (<003 lb/mbtu) FGD 90-MWe $06/kW 10 SCR Dry scrubber $154/kW SCR Wet limestone FGD 300-MWe $37/ton 11 SCR projection for a 150-MWe SNRB plant was $60/kW, and the levelized costs were estimated to be $553/ton of and removed, or $70/ton of all three: plus plus particulate emissions reduction HCl emissions from coal-fired power plants are typically an order of magnitude lower than those of, 16 but because large quantities of coal are burned daily, the overall amounts of HCl emitted are of concern In the case of municipal or hospital waste-to-energy plants, the HCl emissions are actually higher than those of (see Case 3 in the following sections) Moreover, HCl is a precursor, along with ash and unburned carbon, to the formation of polychlorinated dibenzo-dioxins/ furans Concerns over such emissions have caused the closure of many older hospital and municipal incinerators Furthermore, such environmental concerns, combined with the reasonably low cost of fossil fuels, have discouraged the construction of new waste-toenergy plants The integrated technique discussed herein combines the injection of Ca-based sorbents (or blends of sorbents) with the installation of high-temperature ceramic filters to effectively reduce,, and HCl gaseous emissions simultaneously, while trapping fine particulate emissions 54 Journal of the Air & Waste Management Association Volume 5 May 00

6 All these processes are integrated in one unit, which is installed at the exit of the furnace This technique is noncatalytic; thus, it avoids problems typically associated with the disposal of spent catalysts that are toxic If needed, however, the filters may be catalyzed The ceramic filters may be placed at the highest-temperature location that is technically possible in the flue gas stream of the boiler This location should be chosen to minimize obstruction of direct radiation of the flame to heat-exchanger surfaces in the furnace The sorbent (blend) is injected in the post-flame region (such as the reburn zone) of the boiler upstream of the filters Partially reacted sorbent particles are captured in the filters, where they keep reacting with the flow of gaseous pollutants for a prolonged period of time The filters are periodically cleaned with bursts of reverse-flow compressed air, releasing the trapped particles to a hopper, possibly fitted with a fiber bag, without disturbing the operation of the furnace This method has been shown to be effective in experiments conducted in our laboratory (see refs 16 18) These references describe in detail the sorbents used and their properties, the experimental apparatus and procedure, the experimental conditions, and all the data points collected They also describe a theoretical effort to interpret the results using and further developing Simon s pore-tree model for transport and reaction in porous media That experimental effort is succinctly discussed in the following section Having completed the experimental/theoretical work, this manuscript presents engineering/economic considerations on scaling up the laboratory setup depicted in Figure 1 to estimate the dimensions of the filter for any furnace size and flue gas flow rate, given an allowable pressure drop (back pressure) A simplified economic analysis is conducted to estimate the capital costs for the filters and casings, as well as their installation These costs are then levelized, and the maintenance, labor, sorbent, and disposal costs are added To approximate the required sorbent tonnage, experimentally determined values of pollutant removal efficiencies were also used Capital costs are presented as $/kw electric capacity, and levelized costs are presented as $/ton pollutant removed/year RESULTS FROM LABORATORY EXPERIMENTS Early experiments in this laboratory showed that sorbents based on carboxylic salts of Ca are effective in the simultaneous reduction of,, and HCl 19 Such sorbents included calcium formate (CF) and calcium magnesium acetate (CMA), among others The organic fraction of the carboxylic salts decomposes into hydrocarbon fragments, at ~400 ºC, that eventually react with oxygen at higher temperatures and, thereby, generate an oxygen-lean atmosphere The remaining hydrocarbon radicals then reduce NO to mainly N 3,4 These sorbents calcine to CaO, at or above ~700 ºC, forming highly porous cenospheres that react with or with HCl according to the overall reactions While such sorbents are effective triple -HCl- sorbents (capable of achieving removal efficiencies on the order of 90% as documented in the aforementioned references), they are costly, because they are produced in small quantities as high-purity chemicals for limited markets Thus, processes have been developed for producing acetic acid (or formic or propionic acid, etc) from renewable organic or biomass substrates, instead of using natural gas Such substrates include woody biomass and wastewater treatment sludge 5 7 If such production becomes commercially viable, the cost of CMA could drop drastically An economic study by Palasantzas and Wise 8 estimated that the price of CMA may change from a current optimized cost of approximately $350/ton to an actual revenue, if appropriate credits for accepting and using waste biomass are taken into account However, because low-cost carboxylic salts of Ca have not yet been massproduced, they are not taken into consideration by the economic analysis herein Instead, this economic analysis concentrates on the use of conventional low- to moderate-cost sorbents such as limestone (CaCO 3 ) or sodium bicarbonate (NaHCO 3 ) blended with bituminous coal or lignite coals to provide the necessary organic pyrolizates for reduction Calcium carbonate reacts with directly according to the reaction CaCO3 + SO + 1 O CaSO4 + CO (3) In the case of NaHCO 3, the decomposition and sulfation reactions occur simultaneously and 1 CaO(s) + SO( g) + O( g) CaSO4( s) CaO() s + HCl(g) CaCl() s + HO(g) NaHCO + SO Na SO + CO + H O 3 3 Na SO3 + 1 O Na SO4 NaHCO3 Na CO3 + CO + H O (6) 1 Na CO3 + SO + O NaSO4 + CO (7) To enhance the effectiveness of these less costly, less porous sorbents (CaCO 3, NaHCO 3 ), a high-temperature (1) () (4) (5) Volume 5 May 00 Journal of the Air & Waste Management Association 55

7 ceramic filter was used to prolong their residence times in high-temperature regions and, thus, somewhat compensate for their unfavorable physical structure The sorbents that were used in the laboratory experiments were reagent-grade from Aldrich 16,18 Two types of coal fines (<38 µm) were used in the laboratory: a bituminous coal (PSOC-1451 HVA or Pittsburgh #8) and a lignite coal (PSOC-1507), both from the Penn State University coal sample bank A brief description of the laboratory tests that demonstrated the combined, HCl,, and particulate removal is given in the following paragraphs Detailed descriptions of the apparatus, procedure, test conditions, and results are available in the literature Experiments were performed in a 5-cm-long, electrically heated isothermal, laminar flow, drop-tube furnace at a gas temperature of 1150 ºC and a flow rate of 4 L/min It was in this furnace that the sorbent coal blends were injected; residence times therein were on the order of 1 sec A small ( cm) honeycomb silicon carbide ceramic filter was placed in a second heated zone below the droptube furnace (see Figure 1) The filter temperature was set at 600 ºC A gas mixture of, NO, and N was also fed into the furnace at the point of sorbent injection Upon mixing with the gas carrying the sorbent blends, the resulting partial pressures of and in the furnace were on the order of 1100 and 500 ppm, respectively In all tests, the reaction (ie, NO + NO ) was monitored A stream of air was introduced at the entrance of the filter to create an overfire-type zone inside the filter element and promote oxidation of any unburned coal accumulated therein The composition of exhaust gases exiting the filter was monitored by,, CO, CO, and O on-line analyzers Results indicate that 40% reductions in both and can be achieved with injection of CaCO 3 at a Ca/S molar ratio of and of pulverized coal at sufficient quantities to result in an injection zone stoichiometric ratio, λ, of 05 (an equivalence ratio, φ = 1/λ, of ) (see Figure ) Please note that the equivalence ratio φ is defined by eq A6 in the Appendix The filter also proved to be more than 99% efficient in particulate matter removal for particles of all sizes, even submicron (see Figure 3) This technique may be applied to most new and retrofit boilers Title IV of the CAAA has established different emission limits for various types of coal-fired boilers 5 For instance, while most uncontrolled cyclone/ cell-fired boilers produced at rates of 1 15 lb/mbtu, 7 the Phase II (year 000) limit is set at 086 lb/mbtu This would constitute a reduction of of about 40%, which can be readily achieved with these sorbent/coal blends, as indicated by the laboratory tests, as long as the overall conditions in the injection zone are oxygen-lean Indeed, reduction efficiencies as high as 95% have Figure 1 The electrically heated, laminar flow, drop-tube laboratory furnace During these tests, a small wall-flow honeycomb filter was inserted below the furnace and was separately heated Provisions for aerodynamic regeneration (back pulsing) of the filter were made The operation of the ceramic honeycomb filter is also illustrated 56 Journal of the Air & Waste Management Association Volume 5 May 00

8 Figure Results of laboratory tests for and reduction Sorbents were blended with pulverized bituminous coal or lignite coal The sorbent/coal blends were injected at 1150 ºC while the filter was maintained at 600 ºC 18 been recorded in this laboratory with carboxylic sorbent injection at 100 ºC 19,0 ECONOMIC ANALYSIS Case Studies An economic analysis is given based on three case studies involving two coal-fired power plants and a waste-toenergy incinerator Table summarizes the specifications of the three plants used in this analysis Case Study #1 A coal-fired power plant in Davies County, KY, was selected as the first case study The facility is the Elmer Smith Power Plant operated by Owensboro Municipal Utilities It is a cyclone-fired boiler and burns mediumsulfur coal with an output of 151 MWe The airflow rate for this plant was given as 15,700 scfm The plant emits at a rate of 10,347 tons/year and at a rate of 3396 tons/ year 9 These emission rates were verified with calculations outlined in the Appendix Other pertinent calculations, including the conversion of the pollutant yields to partial pressures in the effluent, are also included in the Appendix A simplified economic analysis is shown in the following paragraphs, aiming at comparing this technique Figure 3 Photographs of paper filter samples taken (a) at the entrance and (b) at the exit of a silicon carbide monolith filter, fitted in the tailpipe of a diesel engine with competitive technologies and at showing that substantial capital and levelized cost savings may be achieved over the state-of-the-art control technology, assumed to be the compounded wet limestone FGD/SCR/fabric filter with capital costs in the range of $50 $400/kW These calculations are based on the specifications of the aforementioned 151-MWe Unit #1 at the Elmer Smith Power Plant Having completed the background calculations, attention was then directed to the calculation of the proposed high-temperature ceramic filter volume requirement for this plant Extrapolations were based on our lengthy experience of using ceramic monoliths to filter the exhaust of diesel engines as well as the experiences gathered at Penn State University on a pilot-scale, coal-fired furnace where ceramic filters were installed in the effluent duct at a gas temperature of 478 K Over the last ten years, studies have been conducted in this laboratoty on the filtration of particulates from diesel engines using ceramic honeycomb monoliths in conjunction with aerodynamic regeneration It was extrapolated that at the same temperature of 478 K, the airflow rate of 15,700 scfm of the Elmer Smith boiler may be handled with a number of 3400 cylindrical filters (±100) of a size of 16 cm diameter 30 cm height (65 1 in) and maintain an acceptable pressure drop (back pressure) not exceeding 5 mbars, as suggested in the Penn State tests 33 This pressure drop is in line with the 30 mbar reported by Martinelli et al 15 for the B&W SNRB pilot plant It should be noted that, in diesel engine operation, back pressures as high as 00 mbars can be afforded 36 In the proposed application, the filters are installed to accept hot flue gas at temperatures around 600 ºC (873 K); thus, on average, 600 such filters will be needed for this boiler These figures were estimated by multiplying the number of filters (80) used in the Penn State pilot plant by the ratio of flue gas flow rates (15,700 ft 3 /min/5088 ft 3 /min) at the two furnaces and by the ratio of average proposed filter temperature herein to that actually used in the Penn State tests (873 K/478 K) At a price of approximately $400/filter (SiC filters from Ibiden), the total filter cost would be $,480,000 If the filter is mass-produced, then the price would be drastically lowered If the price of the matting and steel casings is $1 million and the installation costs another $,480,000 (equal to the price of the filters), then a capital cost of $5,960,000 is estimated Two approaches have been applied for the estimation of the cost for the sorbent injection system Martinelli et al 15 reported that they invested $34/kW for reagent storage and injection system Noting that they used NH 3 for reduction along with sorbent injection and knowing that storage and injection of NH 3 needs a lot of additional expenses because of the corrosiveness and flammability of NH 3, $1/kW has been estimated for the Volume 5 May 00 Journal of the Air & Waste Management Association 57

9 Table Plant specifications Power Plant Capacity Fuel Flow Airflow Rate SO Emissions NO Emissions x HCl Emissions (MWe) Rate (ft 3 /min) (tons/year) (tons/year) (tons/year) (tons/year) Elmer Smith ,30 15,700 10, N/A Paradise 1150,700,000,333,333 1,860 53,000 N/A Lawrence 53 37,15 14, cost of the dry-sorbent injection system alone (an estimation that is between one-half and one-third of the cost of the conbined ammonia and sorbent injection system) The second approach is adapted from Redinger and Corbett, 14 who estimated the cost of the injection system to be roughly 15% of the total emission reduction system Although the cost of the injection system should not be dependent on the number of filters used, this provides an opportunity to examine the effect of the cost of the injection system on the total cost Using the first approach and adding the sorbent injection system s cost to the other capital costs, a total capital cost of $7,77,000 was obtained If a 10% technology contingency factor and a 5% project contingency factor, utility indirect charges, and allowance for funds during construction are accounted for, 7,8,15 then the capital cost will climb to $10,49,00 (ie, $70/kW for a new 151-MWe plant) If a retrofit factor of 5% is also included, as recommended by Martinelli et al, 15 then the capital cost will be $8/kW for a retrofit Using the second approach, the cost for injection system is lower by a factor of and the capital cost becomes $9,5,900 (ie, $61/kW for a new plant and $73/kW for a retrofit) Thus, lowering the cost of the injection system by 50% lowers the total cost by 1% Calculations from this point on will use the 15% approximation of Redinger and Corbett 14 for the cost of the injection system Assuming a 15-year loan at 7% interest, the levelized capital costs are calculated to be $998,01/year based on a commercial loan-calculator program Maintenance and labor costs are calculated according to the data obtained from Ellison 9 Operating labor is $75/hr and maintenance (labor and materials) is $100/hr for a 5-MWe plant in Ellison s report 9 These values are normalized for the Elmer Smith Power Plant by multiplying this cost by 151 MWe/ 5 MWe, and so the annual cost is calculated to be $666,198 for labor and maintenance To this, the costs of the sorbent and disposal must be added Two cases are examined: (1) a blend of CaCO 3 and coal, and () a blend of NaHCO 3 and coal The flow rate of these sorbents was calculated based on the amount of emitted, as given in the plant specifications, to provide a molar Ca/S ratio of in the post-combustion gases The amount of coal that must be injected as reburn fuel was calculated according to the airflow rate into the boiler, assuming a typical oxygen concentration in the post-flame zone effluent of 3% Therein, there should be sufficient organic pyrolyzates from the injected reburn coal to create an overall fuel-rich environment A conservatively high equivalence ratio, φ, of was targeted for the reburn zone (λ = 1/φ = 05) Upon consumption of the background oxygen, the remaining pyrolyzates will reduce the From laboratory experiments, such conditions are expected to be optimum It is calculated that 3,335 tons of CaCO 3 per year are necessary (ie, equivalent to 9% of the coal flow rate to the boiler) Most of the coal requirement to the sorbent blend may be subtracted from the total coal input, as is normally done in reburning boilers; hence, the cost of the coal is not accounted for in this analysis At an assumed price of $5/ ton for limestone, 37 the sorbent cost per year is $808,375 Assuming a 0% Ca utilization for CaCO 3, based on the experimental data, 18 it is calculated that 8796 tons of CaSO 4 will be produced every year The remaining unreacted sorbent, which is mostly converted to CaO at high temperatures, is estimated to be 14,486 tons/year The combined solid product of 3,8 tons will need to be disposed of every year For a disposal cost of $93/ton, 15 the total disposal cost is calculated to be $16,515/year Note that the disposal of the coal ash is not included in this calculation, because this process uses no extra coal A fraction of the coal that would be used by the power plant to produce electricity is diverted to this reburn zone If calcium formate (CF) is injected, then 4,035 tons/ year will be required, which at a current price of $600/ ton amounts to $5,8,30/year Of course, this price is prohibitively high but, as shown by Palasantzas and Wise, 8 the price of carboxylic salts of Ca may be reduced drastically if they are mass-produced from waste biomass Assuming an optimized cost of $350/ton, the levelized cost for CF is $14,716,50 Thus, such sorbents may still be of technological interest For CF, the measured Ca utilization efficiency was 30% Hence, 13,193 tons of CaSO 4 may be produced every year The rest of the sorbent is disposed of as CaO, which is 1,675 tons/year Again, with $93/ton, the total disposal cost is $40,57/year for CF If NaHCO 3 is used as the sorbent, then 7,160 tons of bicarbonate will be needed per year With an estimated cost of $60/ton for NaHCO 3, 37 the cost for the sorbent will be $7,063,940/year With a 50% Na utilization for NaHCO 3, it is assumed that,958 tons of Na SO 4 are produced every year The rest of the sorbent is assumed to be 58 Journal of the Air & Waste Management Association Volume 5 May 00

10 Table 3 Levelized cost calculations for the Elmer Smith Power Plant (151 MWe) Ca-Based Levelized Maintanance + Sorbent Cost Disposal Cost Total Cost $/ton $/ton $/ton Sorbent Capital ($) Labor ($) ($) ($) ($) Emissions Emissions NO + SO + x (tons/year) (tons/year) Particulates Carbonate 998,01 666, ,360 16,515,689,085 10, Formate 998,01 666,198 5,8,30 40,57 7,133,10 10, , Bicarbonate 998,01 666,198 7,063,940 37,887 9,101,037 10, disposed of as Na CO 3, at a rate of 17,138 tons/year The total annual disposal cost for NaHCO 3 is $37,887/year The total levelized costs (including capital plus operation and maintenance costs plus sorbent and disposal costs) are given in Table 3 As shown in Table 3, the estimated levelized cost to remove a ton of each pollutant ( and ) per year would be this total levelized cost divided by the amount of the pollutant removed, which is total emission of the pollutant multiplied by removal efficiency If the efficiency is 40% for removal and 40% for removal in the case of CaCO 3 + coal blend, the levelized cost is $489/ton of and A 100% power plant load is assumed Given that this technique also removes particulates, the real cost is much lower, even by one-half; if the approximation of Martinelli et al 15 is adapted, then the levelized cost will be $45/ton each of,, and particulates removed, with efficiencies of 40, 40, and 100%, respectively But it must be noted that the values shown in Figure are the averages over the tests done with both of the coals It was seen that all of the sorbents reached higher efficiencies when mixed with this particular lignite coal For example, CaCO 3 reached removal efficiencies as high as 50% when blended with lignite at a Ca/S molar ratio of The same calculations are repeated for mixtures assuming a 78% removal efficiency for sodium bicarbonate (SCH), a 65% removal efficiency for CF, and a 40% removal efficiency for both of the coals (see Table 3) After accounting for the efficiency, the levelized cost of the integrated technique presented herein is quite lower than the costs of competitive technologies Case Study # The second facility selected for our study was a larger coal-fired plant in Muhlenberg County, KY It is the Paradise Steam Plant Unit Number Three and is operated by the Tennessee Valley Authority Each of the steam plant s three main units are cyclone-fired and burn high-sulfur bituminous coal for power production Units One and Two produce 704 MW of electrical energy each, while Unit Number Three provides 1150 MWe Unit Number Three burns 7 million tons of coal/year and emits 1,860 tons of and 53,000 tons of /year The airflow rate for this plant was given as 140 million ft 3 /hr, or,333,333 ft 3 /min Following the same procedure as the Elmer Smith power plant, a capital cost of $91/kWe is estimated for Paradise Steam Plant Unit Number Three This is higher than the specific capital cost ($/kwe) of the much smaller Elmer Smith plant, because the major part of the capital cost is the cost of the filters, and the number of filters is proportional to the airflow rate through the plant Because the ratio of airflow rates of the two facilities is not proportional to the ratio of the electric power outputs, the Paradise Steam Plant is more expensive on the basis of capital costs However, if economies of scale were applied to this much larger plant, then the value of the unit cost of the filters should be smaller because the number ordered will be larger This would reduce all capital costs and levelized costs Levelized costs are estimated to be $03/ton each of,, and particulates using CaCO 3 as the sorbent and assuming a 40% removal efficiency for and This is a better result compared with the Elmer Smith plant, but one must note that this value is obtained by dividing the total levelized costs by the amount of pollutants emitted Using NaHCO 3 as the sorbent, the levelized cost is estimated at $447/ton each of,, and particulates assuming 78% removal efficiency for NaHCO 3 and 40% removal efficiency for the coals Case Study #3 The third plant was a waste-to-energy facility in Lawrence, MA, that produces 53 MWe The plant reports a standard airflow rate of 14,00 ft 3 /min (at 300 K) Pollutant emissions for this facility were reported as follows: 194 tons /year, 584 tons /year, and 59 tons HCl/year Using the same method as the previous cases, a capital cost of $105/kWe is calculated for this unit The levelized cost for this plant is $933/ton of,, HCl, and particulates using CF as a sorbent and natural gas as reburn fuel Because the reburn process cannot be applied by injecting a portion of the fuel into the reburn zone as in previous cases, reburn fuel has to be supplied separately to this facility, which will be an extra cost To illustrate the effect of introducing highly porous Ca salts, CF is chosen as a sorbent Assuming a Ca/S molar ratio of and a Ca/Cl ratio of 1, 96 tons of CF are needed per year, which, at a price of $600/ton, amount to $1,755,54/year Assuming that the organic part of CF reacts with the effluent gas to convert a portion of the into nitrogen Volume 5 May 00 Journal of the Air & Waste Management Association 59

11 Table 4 Calculated levelized and capital costs of the system using different sorbents assuming 40% SO and 8% HCl removal efficiency for CC, 65% SO and HCl removal efficiency for CF, 73% SO removal efficiency for SHC, and 40% NO removal efficiency for all of the coals or the natural gas x Case 1 Elmer Smith CaCO 3 CaCO 3 CF CF NaHCO 3 NaHCO 3 Capital Cost (151 MWe) (New Plant) (Retrofit) (New Plant) (Retrofit) (New Plant) (Retrofit) ($/kwe) Total levelized costs,689,085,873,90 7,133,103 7,317,90 9,101,038 9,85,854 New plant Retrofit ($/year) Levelized costs $/ton Levelized costs ,974 0, $/ton Levelized costs $/ton NO + SO x Levelized costs $/ton SO + particulates Levelized costs , $/ton NO + particulates x Levelized costs $/ton NO + SO + x particulates Case Paradise CaCO 3 CaCO 3 CF CF NaHCO 3 NaHCO 3 Capital Cost (1150 MWe) (New Plant) (Retrofit) (New Plant) (Retrofit) (New Plant) (Retrofit) ($/kwe) Total levelized costs 8,491,61 30,571,65 318,651,973 30,731, ,60, ,68,355 New plant Retrofit ($/year) Levelized costs $/ton Levelized costs $/ton ,031 15, Levelized costs $/ton NO + SO x Levelized costs $/ton SO + particulates Levelized costs $/ton NO + particulates x Levelized costs $/ton NO + SO + x particulates CF + CF + Case 3 Lawrence Wte CaCO 3 CaCO 3 Natural Gas Natural Gas NaHCO 3 NaHCO 3 Capital Cost (53 MWe) (New Plant) (Retrofit) (New Plant) (Retrofit) (New Plant) (Retrofit) ($/kwe) Total levelized costs ($/year),709,155,819,859 $4,367,096 4,477,800 New plant Retrofit Levelized costs $/ton 1,484,36 34,63 35, Levelized costs $/ton 11,597 1,071 18,695 19,169 Levelized costs $/ton HCl 16,344 17,01 11,349 11,637 Levelized costs $/ton + + HCl Levelized costs 10,74 11,181 17,316 17,755 $/ton SO + particulates Levelized costs , $/ton NO + particulates x Levelized costs $/ton HCl + particulates Levelized costs $/ton NO + x SO + particulates + HCl 530 Journal of the Air & Waste Management Association Volume 5 May 00

12 gas, 7443 tons of additional natural gas will be necessary every year to obtain an equivalence ratio, φ, of Assuming the price of natural gas to be $5/1 million Btu (ie, $37/ ton) the gas will cost $1,763,73/year, which will bring the total sorbent plus natural gas cost up to $3,519,74/year Assuming a 65% and HCl removal efficiency for CF and a 70% removal efficiency for natural gas, a levelized cost of $933/ton each of,, HCl, and particulates is obtained One must note that using CF will achieve an removal efficiency of more than 65% at a Ca/S molar ratio of and an even higher HCl removal efficiency at a Ca/Cl molar ratio of 1 38 A similar calculation using CC with removal efficiencies of 40% for (at Ca/S = ) and 8% for HCl (at Ca/Cl = 1) 38 results in a levelized cost of $578/ ton each of,, HCl, and particulates removed (see Table 4) for a new plant Included in this cost is the price of the natural gas for reduction of ; 765 tons of natural gas/year are needed for this case Because natural gas does not contain nitrogen, it is expected that its removal efficiency is higher than coal s removal efficiency In fact, Bowman 39 reports that at the conditions where coal achieves a 40% removal by reburning, natural gas removes almost 70% of the emitted The values for capital and levelized costs for all three plants are listed in Table 4 One should also realize that by burning this additional natural gas fuel, the thermal input to this power plant will increase by ~1 MWth, and its electrical output will increase accordingly General Comments It is evident from the previous cases that the use of CF or other carboxylic salts of Ca at their current high prices are noncompetitive for coal-fired power plants They may be considered, however, for niche markets such as those of municipal waste or hospital incinerators, where the small volume of flue gases and high Cl/low S content result in a small sorbent tonnage requirement, 40 especially if such sorbents are mass-produced at a lower price In this work, CF is used merely to provide a benchmark for the performance of the less costly sorbent blends The capital cost of this process may be less than 0% of the total capital costs for compounded wet limestone FGD/SCR/baghouse systems, which are reported to be in the range of $50 $400/kW (see Table 1) The calculated levelized costs are also less than 0% of those of current processes, estimated to be in the range of $1000 $000/ ton of each of the three major pollutants While the calculations presented herein are simplified and are subject to omissions, they are illustrative of the potential of the investigated integrated process If a different cost method called the Levelized Revenue Requirement (LRR) method is used, as outlined by the Electric Power Research Institute (EPRI), 41 levelized costs are obtained ~40% higher than the values calculated by the levelized cost calculation presented previously for the three plants The LRR method uses factors that can be found in the EPRI Technical Assessment Guide (EPRI TAG) For fossil-fueled power generation, EPRI specifies a tax recovery period of 0 years and a book life of 30 years Instead of this method, the levelized cost method was used to compare with reports on other competitive technologies that used the same method to conduct their economic analyses CONCLUSION The reduction of gaseous pollution created by coal-fired and waste-to-energy power plants is important for minimizing human and ecological health risks and is mandated by current CAAA guidelines Currently, removal of requires different equipment than does removal of or particulates High costs are involved in the installation, operation, and maintenance of pollution control equipment; therefore, an economical alternative is desirable A method for simultaneously removing,, HCl, and particulates is described herein This method combines the high-temperature injection of a dry (or wet) Ca-based sorbent and the installation of a high-temperature ceramic filter in the effluent stream Laboratory experiments have shown that this method is capable of removing up to 95% of the,, and HCl in combustion gases, depending on the sorbent and the conditions involved (with 40% readily achieved), and 99% of PM 5 particulates in the effluent Because this sorbent/filter combination is a promising alternative to current pollution control technologies, a cost analysis was performed to explore the economic feasibility of installing such a device The capital costs (filter plus installation), operation and maintenance, and sorbent and disposal costs were calculated and amortized over a 15-year period and broken down into yearly expenses Because the sizes and emissions of power plants vary widely, the costs were normalized to the price per kilowatt of energy produced (capital cost) and to price per ton of pollutant removed (levelized cost) Results showed that if a CaCO 3 /coal mixture is the chosen sorbent blend, this method of simultaneous removal appears to be much less expensive than implementing conventional methods For example, the capital cost of the filter/sorbent combination is calculated to be $61/kWe for a 151-kWe unit for a 40% reduction in and emissions and 99% reduction in PM 5 reductions, as compared with $0 $400 for wet limestone FGD/SCR/fabric filter combination, which, however, removes these pollutants with efficiencies on the order of 80 and 90% Thus, the comparison of the capital costs is not direct A more complete comparison may be obtained by contrasting the levelized costs Typical levelized costs for Volume 5 May 00 Journal of the Air & Waste Management Association 531

13 current technologies range from $1000 $000 per ton removed of plus plus particulate pollutants Levelized costs in this analysis include capital costs, operation and maintenance, project/process contingency, utility indirect charge(uic), allowance for funds during construction (AFDC), and sorbent and disposal costs A coal/caco 3 blend exhibited good,, and HCl reduction in laboratory tests If this blend is used as the sorbent for the coal-fired power plants together with a ceramic filter,,, HCl, and particulates can be simultaneously removed from effluent gases for approximately $03 $61/ton of each,, and particulates actually removed (ie, after accounting for expected removal efficiencies) This appears to be more cost-effective than existing technologies The main advantage of the proposed technology is that it can simultaneously capture several pollutants with an efficiency that varies from moderate to high values, depending on the sorbent type and price It requires hightemperature regenerable filters, such as ceramic monoliths, and a sorbent injection system The filters may be commercially available silicon carbide or cordierite/mulite multichannel wall-flow monoliths The cordierite or mulite filters have a safe maximum temperature of around 1000 ºC, while SiC filters can be heated safely to temperatures of 000 ºC; the only issue is their higher thermal expansion coefficient Tests with both kinds of filters were conducted SiC Ibiden filters capture the PM 5 particulates with an efficiency of 99% 35 Micromembrane-coated Corning/Ceramem filters also have efficiencies of 99% Moreover, there is an advantage in capturing particulates at a high temperature and retaining them therein for a certain residence time: high carbon and hydrocarbon burnout can be achieved, eventually producing ash with greater potential for subsequent utilization value Of course, engineering challenges associated with the installation of the filters in the effluent sections of boilers are to be expected, particularly in retrofit cases Perhaps this process is suited best to small or moderate-size units (like the 151-MWe unit exemplified herein), where the number of filters would be manageable Moreover, one should keep in mind that, according to Cox et al, 8 lower levelized costs ($/ton) may be achieved when pollutant removal systems are applied to larger, highly utilized units because of economies of scale and the fact that larger amounts of pollutants are removed therefrom Also, greater inlet and levels significantly decrease the levelized costs Finally, better-processed limestone or lime sorbents, even if more expensive, may increase the pollutant removal efficiency and lower the levelized costs ACKNOWLEDGMENTS This research was supported by National Science Foundation grant BES , Dr Ed Bryan, Technical Director APPENDIX Combustion calculations were conducted to verify the published operating parameters for the power plants used in this study Elmer Smith 151-MWe Plant Assuming the 151-MWe coal-fired plant is 45% efficient, the thermal power input, Qin was calculated by W elec Qin = (A1) ηth which gives a thermal power input, Qin, for this plant of 336 MWth Using this thermal work input value and a typical heating value, Q HV, of 30,000 kj/kg for mediumsulfur bituminous coal and making the necessary time conversions shows that the plant burns 353,30 metric tons of coal/year Qin m coal = (A) QHV Assuming that the coal contains 15% S, then 5300 tons of S are produced per year Then, tons coal tons S MWSO tons SO m SO = 353, , 600 year tons coal MW = S year (A3) This compares well with the value of 10,347 tons / year, as given in the plant specifications Assuming N content in coal is 13% by mass, which is a typical value for coals, then 353,30 tons of coal contain 4593 tons N Typically, for φ 1, ~30% of the N contained in the fuel is converted to NO m NO Fuel tons coal tons N MWNO tons NO =, year tons coal MW N = 953 year (A4) Additionally, an equivalent of 5% of the amount of NO generated by the fuel N is typically produced by atmospheric N m NO Total tons NO tons NO = year = 3691 year (A5) The value given in the plant specifications was 3396 tons NO/year; hence, the agreement is satisfactory Assuming a mildly fuel-lean bulk equivalence ratio, φ, of 095, and using the mass flow rate of the coal found previously, the airflow rate was calculated in actual cubic feet per minute (acfm) The bulk equivalence ratio, φ, is defined as mfuel 1 mair φ = = actual λ mfuel mair stoichiometric (A6) The actual mass flow rate of coal was converted from 53 Journal of the Air & Waste Management Association Volume 5 May 00