United States Environmental Protection Agency Industrial Environmental Research Laboratory Research Triangle Park NC 2771 Research and Development EPA-600/S7-84-017b Apr. 1984 Project Summary Development of Criteria for Extension of Applicability of Low Emission, High Efficiency Coal Burners: Second Annual Report A.R. Brienza, S.L Chen, M.P. Heap, J.W. Lee, W.H. Nurick, D.W. Pershing, and D.P. Rees This report describes the second year's effort under EPA Contract 68- 02-2667, which concerns the develop- ment of criteria for the evaluation and applicability of low-emission, high- efficiency coal burners. The report describes progress in three major areas: (1) bench scale studies, (2) distributed mixing burner development. and (3) comparison with commercial practice. The bench scale studies concern the impact of fuel characteris- tics on fuel NO formation during the combustion of pulverized coal. Although fuel NO emissions generally increase with increasing fuel nitrogen content, coals with the same rank and similar nitrogen contents may produce markedly different fuel nitrogen levels. Prelimi- nary results indicate that a simple procedure, evaluating reactive volatile nitrogen, can be used to assess the impact of coal type on fuel NO forma- tion. Data relating to the development of the distributed mixing burner have been reviewed, and data have been obtained with both single and multiple burners. No operability problems were encountered with different fuel types, but NO emissions were fuel dependent. Several commercial burners have been tested satisfactorily in the test facility. One gave severe flame impingement. Carbon monoxide emissions were generally higher and nitrogen oxide (NO>) emissions were slightly lower than commercial practice. This Project Summary was developed by EPA's Industrial Environmental Research Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report ordering information at back). Introduction NOX formation in pulverized coal flames can be reduced by using the burner to control the rate of fuel/air mixing to minimize fuel nitrogen conversion. This report describes the second year's progress on EPA Contract 68-02-2667. The program is aimed to: • Expand the fuel capability of low NOX burners to include the major types of solid fossil fuels projected for use by the utility industry. • Explore additional burner concepts and configurations that show poten- tial for improving the emission and thermal performance of pulverized coal burners. • Determine the effects of multiple burner configurations that are encountered in utility boilers. • Provide for direct comparison be- tween the experimental burners being developed here and the current state-of-the-art for commer- cially available coal burners. • Arrange for coordination of the technology development program with boiler manufacturers and users. • Provide testing in support of planned application of the burner technology. Figure 1 shows the relationship of the ------- Planning/Approach Taskl Program Definition Experimental Burner Development Task 2 Screening Experiments / \ Task 3 Single Burner Experiments Task 4 Multiburner Experiments TaskS Comparison with Commercial Practice Relate to | Technology Practical Units Transfer Task? Data Analysis and Criteria Development Figure 1. Overall program logic. various tasks that are planned to meet these objectives. In this, the second year of the program, efforts were expended on the following tasks: • Bench scale fuel screening studies to assess the impact of fuel properties on fuel NO formation (Task 2). • Tests with the distributed mixing burner and different fuel types (Tasks 3 and 4). • Tests with commercial burners to allow a comparison with commercial practice. Bench Scale Studies In the first year's effort, a bench scale reactor was constructed with a design input firing rate ranging from 50 to 150x 103 Btu/hr.* This facility, to be used for fuel screening studies and concept development, had several versatile features including: • The ability to burn the fuel with nitrogen-free oxidant mixtures com- posed of 02, COz, and Ar, to define fuel NO production. • Complete control over fuel input rate and fuel size distribution. • A modular combustion chamber to permit probe access and variable heat extraction rates. *To convert nonmetric units to their metric equiva- lents, please use the conversion factors at the back of this Summary: Table 1 lists the 25 coals tested to date in this furnace and their properties. Initially, the impact of fuel properties on NO formation under excess air is discussed; then data obtained to date on the evaluation of advanced concepts for NO* control will be presented. The Influence of Coal Properties on Fuel NO Formation Pulverized coal was burned under premixed and diffusion flame conditions in nitrogen-free atmospheres to evaluate fuel nitrogen conversion. Figure 2 summarizes the data obtained with the coals tested to date under premixed and axial diffusion conditions. Fuel NO emissions are shown in pounds of NOa per 10s Btu as a function of percentage nitrogen in the coal on a dry ash-free basis for overall 5 percent excess oxygen in the combustion products. The data presented illustrate two major points: 1. Fuel NO emissions tend to increase with increasing fuel nitrogen content. This is most pronounced for premixed conditions. However, for any given nitrogen content there is a consider- able spread in fuel NO emissions, indicating that some property other than total fuel nitrogen content influences fuel NO formation. 2. Fuel nitrogen conversion decreases with decreasing fuel/air mixing rates. Since oxygen availability controls fuel NO formation, this is expected. Some anomalous behavior was observed associated with fuel/air mixing effects and fuel nitrogen formation. Some coals show a very strong dependence of fuel nitrogen formation on fuel/air mixing characteristics. However, this is not always the case; even for coals of the same rank, the influence of fuel/air mixing appears to be also coal-dependent. Two fuels tested to date show very different characteristics. The anthracite shows almost no impact of fuel/air mixing rate. Since anthracite is very low in volatile content, it can be assumed that most of the fuel NO formed is produced from solid nitrogen, and that the impact of fuel/air mixing characteristics is mini- mized. Alternatively, the Australian lignite showed a decrease in fuel NO from 950 to 675 to 175 ppm as the combustion conditions changed from premixed to radial diffusion to axial diffusion. It can be argued that the fraction of the coal fuel nitrogen that is volatile (i.e., emitted with the volatile fractions during thermal decomposition) has a major impact on the formation of fuel NO. Therefore, two simple experiments were carried out to determine if volatile fuel nitrogen (determined by either the ASTM volatile matter procedure or an inert pyrolysis experiment developed under EPA sponsorship) can be used to predict fuel NO formation in pulverized coal flames. The char residue from an ASTM volatile test was analyzed for nitrogen content to determine the fraction of the nitrogen that had been liberated with the coal volatiles. The results suggest that this method of assessing volatile nitrogen evolution is not even capable of ranking fuels, and is certainly not suitable as a predictive tool. Under EPA sponsorship, a pyrolysis reactor was developed to investigate the conversion of fuel nitrogen to HCN under both inert and oxidative pyrolysis conditions. Gas-phase kinetics experiments suggest that HCN is a reasonable model for volatile fuel nitrogen compounds; therefore, the amount of fuel nitrogen that is converted to HCN could indicate the potential for a particular coal to produce fuel NO. The pyrolysis experi- ment was operated as a two-stage system. The fuel was placed in an initial reactor whose temperature could be varied, and the pyrolysis compounds were swept into a secondary reactor (maintained at 1373 K) with an inert Ar flow. Experiments have demonstrated ------- Fuel Symbol Fuel Source Proximate Analysis 1% as received) Moisture Ash Volatile Matter Fixed Carton Ultimate Analysis l%Dry) C H N S Ash 0 Heating Value IBtu/lb. Wet) CLASSIFICATION (ASTMD388) O Haze/ton PA 5.13 5.74 4.39 84.74 88.45 2.14 0.79 0.47 6.05 2.10 13. 124 Anthracite O Upper Cliff AL 3.0O 9.49 20.44 67.O7 79.32 4.47 1.47 1.3O 9.78 3.66 13.254 Medium Volatile Bituminous O Rosa AL 8.02 6.79 21.81 63.38 81.23 4.73 1.54 1.04 7.38 4.08 13.394 Medium Volatile Bituminous 0 Black Creek AL 2.25 4.45 28.28 65.02 8212 5.21 1.79 0.76 4.56 5.56 14.284 Medium Volatile Bituminous 0 W.KY 5.43 7.53 37.79 49.25 71.58 5.43 1.55 3.21 7.96 10.27 12.082 High Volatile A Bitu- minous a wv 7.57 12.05 29.12 51.26 74.68 4.96 1.38 0.86 13.04 5.08 12.228 High Volatile A Bitu- minous O Elkay WV 0.60 11.61 33.35 54.44 73.96 4.93 1.39 2.47 11.69 5.56 13.115 High Volatile A Bitu- minous 0 Gauley Eagle WV 3.72 26.89 26.65 42.74 59.43 4.11 1.35 0.84 27.94 6.33 10.110 High Volatile A Bitu- minous O Price UTfl) 6.39 7.40 38.89 47.32 73.52 5.52 1.44 0.61 7.89 11.29 12.340 High Volatile B Bitu- minous k. Price UT(II) 7.41 8.83 38.84 44.92 72.24 5.75 1.55 0.76 9.54 10.16 11.877 High Volatile B Bitu- minous 0 Utah (Coal) 4.20 9.62 42.97 43.21 69.36 5.32 1.50 1.04 10.05 12.73 11.718 High Volatile B Bitu- minous e Utah (Char) 2.70 16.83 10.29 70.18 75.39 1.28 1.22 1.12 17.30 3.69 11.185 — a Cadiz OH 4.29 18.05 34.87 42.79 62.08 4.44 1.07 7.40 18.86 6.15 11.038 High Volatile B Bitu- minous Fuel Symbol Fuel Source Proximate Analysis /% as received) Moisture Ash Volatile Matter Fixed Carbon Ultimate Analysis (% Dry) C H N S Ash 0 Heating Value (Btu/lb. Wet) CLASSIFICATION (ASTMD380) Farmington NM 7.76 17.96 34.95 39.33 63.06 4.65 1.40 O.81 19.47 10.61 10.391 High Volatile C Bituminous Four Corners NM 5.24 24.06 35.94 34.76 55.99 4.71 1.23 1.03 25.39 11.65 9.425 High Volatile C Fruit/and NM 11.02 20.72 31.66 36.60 58.71 4.21 1.30 0.93 23.29 11.56 9.344 High Volatile C Bituminous Bituminous Co/strip MT 21.27 9.58 30.82 38.33 67.52 4.36 1.38 0.63 12.17 13.94 9.169 Sub- Hardin MT 22.70 11.26 31.26 34.88 65.54 4.15 0.95 0.79 14.57 14.00 8.603 Sub- Hardin MT 20.49 8.62 33.24 37.65 67.88 4.65 0.99 1.07 10.84 14.57 9.229 Sub- bituminous bituminous bituminous B B B Shell TX 25.23 10.28 35.31 29.18 60.99 4.49 1.13 1.O2 13.74 18.63 8.131 Sub- bituminous C Scranton NO 34.96 7.50 28.85 28.69 64.61 4.17 0.83 1.52 11.53 17.34 6.446 Lignite A Beulah ND 33.10 7.12 28.65 31.13 65.29 3.96 0.99 1.14 10.64 17.98 7.245 Lignite A Beulah ND 34.63 4.97 27.02 3338 66.15 4.20 0.96 0.37 7.60 20.72 7.245 Lignite A Savage MT 36.36 4.61 28.48 30.55 64.99 4.04 1.00 0,42 7.25 22.30 6.995 Lignite A Morwell Australia 9.07 3.38 48.79 38.76 66.25 5.01 0.65 0.28 3.72 24.09 10.051 (wet) — Saar Germany 216 8.04 36.23 53.57 76.13 5.25 1.37 0.83 8.21 8.21 13.657 (dry) — almost quantitative conversion of nitrogen in pyridine to HCN at 1373 K. Thus, in this experiment, the fraction of fuel nitrogen capable of conversion to HCN could be assessed as a function of first-stage pyrolysis temperature. The results shown in Figure 3 compare the fraction of fuel nitrogen converted to NO for the three combustion conditions as a function of the fraction of the fuel nitrogen converted to HCN under pyrolysis conditions at 1373 K. It can be seen that the pyrolysis data agree qualitatively with the experi- mental results obtained in the premixed and radial diffusion flames. The Savage lignite (A) gives the highest percentage conversion of fuel nitrogen to NO under combustion conditions, and also has the highest fraction of volatile nitrogen yield in the inert pyrolysis study. Optimization of Staged Combustion Conditions It is generally recognized that staged combustion offers the most cost-effective control technique for minimizing fuel NO formation. The basis for staged combus- tion as an NOX control technique involves the competition between two reaction paths: one forms N2, and the other NO from the fuel-nitrogen species. The path producing N2 is favored under fuel- rich conditions. Thermodynamic limita- tions indicate that the total fixed nitrogen species (NO, NH3, HCN) are minimum at about 65 percent of the theoretical air requirements for complete combustion for most hydrocarbon fuels, and that this minimum concentration is of the order of 10 ppm under combustion conditions. Thus, it could be concluded that if all the nitrogen species were in the gas phase then this thermodynamic limitation would represent the minimum NO achievable under staged combustion conditions, since it is reasonable to assume total conversion of these small TFN concentrations during second-stage burnout. Gas-phase kinetics would also indicate that the approach to thermody- namic equilibrium is accelerated at high temperatures, even though the minimum TFN concentration would tend to increase. Staged combustion also has an impact on the formation of thermal NO because the heat release process is extended in time allowing enthalphy loss, thus reducing peak temperatures. The primary control options in a staged system are: ------- / .u 1.4 1.2 1.0 0.8 s 5 0.6 ° O 4 ^ 0 0.2 « n g/.. Q) - i 3 1.2 U. /.O 0.8 0.6 0.4 0.2 O „ o i2 - Premixed ^ a ^ 0 ° - r * ° 4° ^ 70x/03fltu/"r " V' 5% O2 (Stack) Open Symbols— Bituminous - - - ^ - _ Half Shaded Symbols — Subbituminous Solid Symbols— Lignite , * — Anthracite \\i\\\\l . Axial Diffusion . - * * A* Q*o t • fc vO O Q (f> - r \ \ \ i i i i i - . - - - - - 0.6 0.8 1.0 1.2 1.4 Fuel Nitrogen, lb/10e Btu Figure 2. Fuel NO emissions as a function of fuel nitrogen content and mixing conditions. • How long and at what temperature must the fuel products remain under fuel-rich conditions? • What is the optimum stoichiometry or distribution of stoichiometries in fuel-rich zone? Two series of experiments have been carried out: one involved three-stage combustion; and the other, heat extrac- tion from the primary and/or secondary zone to assess the various options. Experiments were carried out in which the fuel-rich stage was divided into two sections: the first fuel-rich stage stoichi- ometry was maintained constant (as was the overall excess air level), and the stoichiometry of the second fuel-rich stage was varied. Exhaust NO emissions were reduced by about 15 percent when operating under three-stage conditions compared to two-stage operation. In another experiment the bench scale reactor was modified to allow removable space cooling coils to be added to the fuel- rich first stage, or the second stage of a two-stage combustion system. Maximum emissions were obtained without cooling, and minimum emissions were obtained when heat was extracted from the primary section. This is surprising since a reduction in temperature of the first stage would be expected to increase the concentration of total fixed nitrogen at 15 0.4 .C w Q as (£> O 0.3 s 1 0.2 0.1 - Premixed Radial Diffusion\ -tt- Axial Diffusion 0.1 -if- -If—• •— 0.2 0.1 0.2 0.1 Fraction of Fuel N Converted to HCN at 1373 K 0.2 Figure 3. Fuel nitrogen conversion can be correlated by HCN produced during pyrolysis for well mixed combustion conditions. the exit of the first stage, thus increasing (not decreasing) final emissions. Commercial Burner Testing Since a primary goal of the program is to demonstrate the distributed mixing burner (DMB) technology in field operating boilers, some assurance is required that the DMB, which has been demonstrated only in the large watertube simulator (LWS) research furnace, will also perform in the field as predicted. Also, the basis for the prototype design is the LWS furnace, not a steam raising system. Although indirect, the approach taken was to evaluate commercial burner operation in the LWS and compare it against field experience. This approach provides some assurance that, if the burner operates satisfactorily in the LWS, it will also operate satisfactorily in the field. The burners selected for testing under this program are: • Babcock and Wilcox dual-register burner. • Peabody Engineering Corporation standard pulverized-coal burner with low-NOx modification. • Steinmuller low-NOx burner (similar to burners currently used at the Weiher plant in West Germany). All the burners were designed for a nominal firing rate of 50 x 106 Btu/hr, and the installation in the LWS was similar to that in the field. To date, the B&W and Peabody burners have been tested with the baseline Utah bituminous coal. The Steinmuller burner will soon be tested with a bituminous coal from Germany (similar to coal used in the field) and baseline Utah coal. In support of the industrial demonstration program (EPA Contract 68-02-3127), an 80 x 106 Btu/hr, Foster Wheeler intervane burner (pre-NSPS) was also evaluated; results are included in this report for complete- ness. All of the burners gave stable flames and performed satisfactorily. However, the Babcock and Wilcox dual- register burner produced a flame that was too long for the LWS firing depth and gave severe flame impingement on the back wall. Thus, it is very difficult to compare field and LWS experience because impingment prevented complete heat release of the input coal. Several general conclusions can be drawn from the studies with commercial burners: • The flames observed in the LWS were judged, by experienced engi- neers, to be very similar to those observed in the field. • CO levels were consistently higher than measured in the field. • NO levels were slightly lower than measured in field operating units. Conclusions • Bench Scale Studies. These studies concerned the development of a data base on the effect of coal type on NO formation and the definition of optimum conditions for minimum NO formation under idealized fuel/ 4 ------- air contacting. Although fuel NO emissions increased with increas- ing coal nitrogen content, coals with the same rank and similar fuel nitro- gen content produce markedly dif- ferent fuel NO levels. Preliminary re- sults suggest that a simple proce- dure (that evaluated reactive volatile nitrogen) could be used to assess the impact of coal type on NO pro- duction. Studies on the optimization of the conditions in the fuel-rich stage of a staged system indicate that minimum gas-phase nitrogen species (HCN + NH3 + NO) does not necessarily give minimum emis- sions after burnout. • Distributed Mixing Burner Develop- ment. Data relating to the develop- ment of the distributed mixing burner have been reviewed and summarized. Single and multiple burner data have been obtained for two different fuels. The distributed mixing concept can be operated with different fuels, and it appears that emission levels are sensitive to fuel type. However, no attempt was made to reoptimize the burner for different fuels. • Comparison with Commercial Prac- tice. Several commercial burners have been tested in the research facility. All the burners performed satisfactorily and produced stable flames. However, one burner gave a flame length that caused severe flame impingement on the rear wall of the test furnace. Since severe impingement cannot be tolerated in the field, comparison with commercial practice is invalid. Carbon monoxide emissions produced with the com- mercial burners were slightly higher than those usually found in field operating units, and NOX emissions were slightly lower. Conversion Factors Readers more familiar with metric units may use the following factors to convert the nonmetric units used in this Summary: Nonmetric Btu/hr Btu/lb lb/106 Btu Times Equals Metric A. ft. Brienza. S. L Chen. M. P. Heap, J. W. Lee. W. H. Nurick, D. W. Pershing, and D. P. Rees. are with Energy and Environmental Research Corp., Irvine, CA 92714. G. Blair Martin is the EPA Project Officer (see below). The complete report, entitled "Development of Criteria for Extension of Applicability of Low Emission, High Efficiency Coal Burners: Second Annual Report," (Order No. PB 84-163 898; Cost: $19.00, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, V'A 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Industrial Environmental Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 2.93 2.33 430 W, J/0 ng/J ------- United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Official Business Penalty for Private Use $300 U o Kfc'ol d 5 U C H 1 C u t A K D u K «^ 5 [ ^< 1 1 I -. U i L o U D 0 4 U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/922 ------- |