Energy Fuels 2010, 24, 5464–5473 Published on Web 09/08/2010
: DOI:10.1021/ef1002843
Dry Lignite Cofiring in a Greek Utility Boiler: Experimental Activities and Numerical Simulations Michalis Agraniotis,*,† Dimitris Stamatis,† Panagiotis Grammelis,†,‡ and Emmanuel Kakaras†,‡ † National Technical University of Athens, Laboratory of Steam Boilers and Thermal Plants, Heroon Polytechniou 9, 15780 Athens, Greece, and ‡Institute for Solid Fuels Technology and Applications/Centre for Research & Technology Hellas (CERTH/ISFTA), Fourth km. N.R.Ptolemais-Kozani, 50200 Ptolemais, Greece
Received March 10, 2010. Revised Manuscript Received August 20, 2010
Lignite predrying and dry lignite firing is considered as an important development step for the next generation of lignite power plants. The integration of a predrying system in future lignite power plants, which utilizes low temperature steam for the drying process, combined with the further thermal utilization of lignite’s evaporated moisture, may bring an efficiency increase of four to six percentage points compared with today’s state of the art. Firing predried lignite is however expected to cause certain changes in the combustion behavior of large scale boilers. The investigation of these changes in an existing Greek utility boiler through experimental activities and numerical simulations is the scope of the present work. The investigations take place in a 75 MWth lignite fired Greek boiler. The specific unit is equipped with dedicated dry lignite burners. The measurements are performed with a dry lignite thermal share of 6%. Higher cofiring thermal shares of up to 20% are further simulated and the effect of cofiring on the combustion behavior is evaluated by specific parameters including temperature fields, wall heat flux, fuel’s burnout, and NOx emissions. No clear effects of cofiring on boiler operation are observed during the experimental campaign indicating that the operational behavior in low cofiring shares remains constant. Some trends are observed in the simulations of the dry coal cofiring cases. Increased temperature peaks in the near burner region and higher furnace outlet temperatures are predicted. A clear increase of the total fuel burnout when firing dry coal is also foreseen by the simulations, which is a strong argument for the coutilization of dry lignite as a supportive fuel. The overall examinations imply that dry lignite cofiring for low thermal shares up to 20% is feasible in the case of Greek boilers without major technological difficulties. For the realization of a lignite predrying concept in an existing Greek boiler however, a detailed study on each particular boiler is necessary.
waste heat utilization (“Wirbelschichttrocknung mit interner Abwaermenutzung”, WTA). The concept has been developed by RWE and demonstrated for more than 10 years in an industrial scale pilot plant in Frechen, Germany.3,4 The large scale realization of the particular concept is currently taking place in a 1000 MWe brown coal power plant in Niederaussem, Germany, with the integration of a prototype fluidized bed dryer into the current steam cycle. The prototype dryer will produce dry coal to be cofired with raw coal to a thermal share of a maximum of 30% of the boiler’s thermal input. The vaporized coal moisture produced from the drying process will be utilized as a heating medium for feedwater in the boiler’s water preheaters.5 The application of a predrying concept toward increasing efficiency is of prime importance for Greek power plants firing low-quality lignite with a high water and ash content. Thermodynamic studies on the integration of a WTA dryer in a Greek power plant have been carried out in the past6 as well as for existing power plants for the next generation of power plants
Introduction It is generally expected that coal will continue to play a key role in the future energy mix as the most abundant and cheapest fossil fuel source. Brown coal especially, as a domestic fuel for many European countries, guarantees the efficient and cost-effective power generation in the base load range. Since a reduction of CO2 emissions produced from fossil fuel power plants is imposed by the Kyoto-protocol, the increase of their efficiency and the minimization of their emissions are some of the main challenges for the existing and future generations of coal power plants.1,2 In the case of brown coal particularly, the optimization of the drying process in future brown coal power plants through the utilization of steam predrying technologies is expected to lead to an efficiency increase of 4-6% age points. Besides, the development of an efficient brown coal drying process is a necessary step toward the implementation of oxy-fuel firing in future generations of brown coal power plants. One of the brown coal predrying technologies, currently under development, is the atmospheric fluidized bed drying concept with
(3) Ewers, J.; Klutz, H. J.; Renzenbrick, W.; Scheffknecht, G. VGB Conference “Power Plants in Competition, Cologne, Germany, 2003. (4) Klutz, H. J.; Moser, C.; Block, D. VGB PowerTech 2006, 11. (5) Schwendig, F.; Klutz, H. J.; Ewers, J. VGB PowerTech 2006, 12. (6) Kakaras, E.; Ahladas, P.; Syrmopoulos, S. Fuel 2002, 81, 583– 5930.
*To whom correspondence should be addressed. Telephone: þ30 210 7722865. Fax: þ30 210 7723663. E-mail:
[email protected]. (1) Tanaka, N. Clean Coal Technologies Conference, Dresden, Germany, May 18-21, 2009. (2) Kjær, S.; Bugge, J.; Blum, R. Energy 2006, 31, 1437–1445. r 2010 American Chemical Society
5464
pubs.acs.org/EF
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
Figure 1. (a) Drawing of the Liptol boiler and (b) boiler cross section at level 26.5 m.
with elevated steam parameters7 including also future oxy-fuel power plants.8,9 The present work focuses on the investigation of dry coal cofiring in an existing Greek boiler through experimental activities and numerical simulations. The large scale tests took place in the 75 MWth “Liptol” power plant operated by the Public Power Corporation of Greece (PPC). In order to evaluate the operational behavior of the plant in reference and dry coal cofiring conditions, specific parameters are monitored and measured during the tests including furnace outlet temperature, unburned carbon in ash, produced electric power, and emissions. A dry lignite cofiring thermal share of a maximum 6% is demonstrated during the campaign. The influence of dry coal cofiring on the combustion behavior in increased cofiring shares is investigated by numerical simulations. The parameters taken into account for the evaluation of the numerical simulations are the average furnace exit temperature and the average flue gas temperature along the furnace height, the fuel burnout, the wall heat flux, and the NOx emissions. The trends of the considered parameters when increasing the cofiring share are finally evaluated as well in the numerical simulations as in the experimental measurements.
Table 1. Main Operational Parameters of the Liptol Boiler parameter
unit
value
raw coal mass flow raw coal heating value thermal input (Qth-input)
t/h MJ/kg MW
48 5.6 75
steam mass flow superheated steam pressure superheated steam temperature enthalpy superheated steam temperature feeding water enthalpy feeding water useful heat (Quseful) boiler thermal efficiency
t/h bar °C kJ/kg °C kJ/kg MW %
80 64 495 3406 154 645 61 82
maximum dry coal mass flow dry coal heating value thermal share of substitution corresponding to maximum dry coal mass flow
t/h MJ/kg %
2 14.5 10
Each boiler has two hammer mills firing at a full load of 48 t/h of raw lignite with a heating value of 5.6 MJ/kg and moisture of 58% from the front wall side. Both boilers are capable of cofiring dry lignite dust to a mass flow of 2 t/h per boiler. Dedicated dry lignite burners are used for this scope, which are located at the left and the right boiler walls at two different levels (Figure 1a). Dry lignite dust with a heating value of 14.5 MJ/kg and moisture of about 12% is collected in the ESPs of the nearby lignite drying and briquetting plant and is fed into the boilers through separated, opposed fired dry lignite burners located at two different levels. The maximum thermal share of dry lignite is 10%. The maximum thermal share of dry lignite is calculated as the fraction of the maximum thermal input of dry lignite to the total thermal input into the boiler in reference conditions. Because of the increased heating value of dry lignite, a dry lignite mass fraction to the total fuel input of about 4% corresponds to a thermal fraction of about 10%. The detailed boiler operational data can be found in Table 1. It is also pointed out that in the last years a steady deterioration of the lignite quality extracted from the open cast mine nearby “Liptol” power station is observed. In order to ensure the continuous and stable boiler operation and prevent unexpected flame instabilities, due to the aged firing system, dry lignite firing to a thermal share of about 1% is continuously
Methodology Experimental Investigations at the “Liptol” Boiler. “Liptol” power station consists of two lignite fired boilers working in parallel and producing 80 t/h of steam each with parameters 485 °C and 64 bar and feeding two steam turbines. The nominal electrical power of the first turbine is 10 MWe with a back pressure of 4.5 bar. A lignite drying and briquetting plant together with the district heating system are fed from the produced steam of the first turbine. The second turbine unit is a 33 MWe condensing unit. No reheat steam cycle is applied. (7) Kakaras E.; Koumanakos A.; Doukelis A.; Giannakopoulos D. Ultra-supercritical power plant fired with low quality Greek Lignite. ECOS Conference, Padova, Italy, June 25-28, 2007. (8) Kakaras, E.; Koumanakos, A.; Doukelis, A.; Giannakopoulos, D.; Vorrias, I. Fuel 2007, 86, 2144–2150. (9) Kakaras, E.; Doukelis, A.; Giannakopoulos, D.; Koumanakos, A. Fuel 2007, 86, 2151–2158.
5465
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
applied during the standard boiler operation. The baseline firing mode therefore includes dry lignite cofiring in this low thermal share. The dry coal cofiring tests are performed in boiler no. 2 of the Liptol power station. The scope of the measurement campaign is the investigation of the effects of dry coal cofiring on following aspects: (1) the operational behavior of the boiler including (a) the steam production, (b) the steam parameters (pressure, temperature), and (c) the operation of the feeding system for raw and dry lignite; (2) the temperature profiles at the superheater (SH) section of the boiler; (3) the flue gas emissions in the stack; and (4) the residues’ quality. The cofiring thermal shares tested during the tests are 6% and 3%, while raw cofiring is also applied for a short period of time, due to the low quality of raw lignite and the possible flame lift off. A measurement time of about 30 min is set for the cofiring tests, while the intermediate time to get to steady combustion conditions before each measurement point is 10-15 min. Because of the small size of the boiler, the changes in the firing system directly affect boiler operation and a new steady state is achieved within the regarded period of time. The performed measurements during each test point include (1) profile measurements of the flue gas temperature at the super heater section; (2) records of flue gas temperature between superheaters 2 and 1 and after superheater 1; (3) records of steam temperature at superheaters 2 and 1; and (4) measurement of flue gas emissions at the stack. Ceramic shielded thermoelements are used for flue gas temperature monitoring. They are placed inside the furnace, at a depth of about 0.5 m from the furnace wall at the according levels. The operation of the feeding systems of raw and dry lignite is also monitored during the tests. No closer mechanical inspection is performed after the tests, since both systems stayed in the normal operation range. The temperature profile measurements are performed at the 26.5 m level between superheater 2 and superheater 1 (Figure 1a,b). The furnace exit level is considered to be the most important measurement level; however, no suitable measurement location is found at that point, and for this reason the next level is chosen. A 6 m long water-cooled suction pyrometer is used equipped with a Ni-CrNi thermoelement and shielded from radiative heat transfer by a proper ceramic part. An ejector is additionally utilized, in order to generate suction pressure and extract flue gas out of the furnace. The pyromener is inserted at 0.5 m steps inside the furnace up to the depth of 3.5 m. The measuring time for each point is 4-5 min. The measurement error of the specific measurement method is (10 K. A portable gas analyzer is used for the flue gas measurements at the stack. Oxygen concentration in the flue gas and CO, NOx emissions are continuously monitored for more than 10 min in the test cases. CO, NOx are postprocessed to a reference oxygen concentration of 6% volume. Numerical Simulations of the Liptol Boiler. The second part of the investigation includes the combustion simulations of the furnace of the Liptol boiler under reference and dry coal cofiring conditions. A commercial CFD code is used for that purpose. Because of the aged control system of the plant and the false air insertion into the boiler, the boiler’s boundary conditions for the simulations could not be determined in high accuracy. For this reason, a point to point comparison between simulation results and the experimental data obtained from the profile measurements at the superheater section is not tried. Average values are used instead for the evaluation of the numerical results. Average profiles of temperature, O2 and NOx concentration along the furnace height, as well as overall coal burnout and total heat flux to the evaporator are calculated for this scope. Cofiring thermal shares of 5%, 10%, and 20% are simulated, although the maximum thermal share realized during the experiments was 6%, and the nominal value in the particular boiler is 10%, due to limitations of the dry lignite feeding system. Despite the
Figure 2. Overall view of the boiler’s numerical mesh.
increased age of the particular boiler and its planned shutdown in the near future, the obtained experience by the experimental and numerical investigations will be used in future investigations of dry coal cofiring in other Greek boilers, for which the issue of dry coal coutilization is of high importance. The boiler is front-wall-fired and is equipped with two jet burners, each consisting of a lower and an upper part. The combination of the front wall firing system and the jet burners leads to the generation of large recirculation zones in the furnace region. These recirculation zones are not comparable with the central vortex, which is typical in modern tangential fired boilers or with the small recirculation zones, which can be found in the burner outlets in wall fired boilers equipped with swirl burner firing systems. Because of these recirculation zones, the flow field acquires unsteady characteristics. The numerical convergence is hindered in this way, and an optimized numerical mesh in combination with the determination of proper inlet boundary conditions for the distribution of the secondary air was necessary in order to achieve satisfying numerical convergence. The final version of the mesh composed of 357 000 hexahedral cells is presented in Figure 2. A detail of the placement of the burners and the secondary air nozzles in the particular firing system is given in Figure 3. The proximate and ultimate analysis of the dried lignite, used in the simulations, is given in Table 2. A mass balance in the recirculation ducts and the lignite mills is also carried out in order to calculate the composition of the carrier gas. The primary and secondary air velocities are given in Table 3. The commercial computational fluid dynamics code Fluent is used for the analysis. An overview of the available submodels applied in a comprehensive CFD code that used pulverized fuel combustion simulations is given in the literature.10 The Realizable k-ε turbulence model is used to account for the large recirculation regions which are dominant in the specific flow. The P1 radiation model is used for the calculation of the radiative heat flux. Particle trajectories are calculated by the (10) Eaton, A. M.; Smoot, L. D.; Hill, S. C.; Eatough, C. N. Prog. Energy Combust. Sci. 1999, 25, 387–436.
5466
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
Figure 3. Detail of burners and air nozzles.
Furthermore, regarding NOx emissions, dry lignite cofiring is expected to lead to an increased NOx emission level, due to the thermal NOx formation mechanism. This is an important technical aspect to be addressed in future large-scale boilers that will cofire predried lignite. The accurate prediction of NOx emissions and the investigation of possible reduction measures is therefore one of the key research objectives for the future dry lignite-fired boilers. Numerical simulations may play a key role in these investigations, since experience of more than a decade in this field14-16 has proven that CFD is a reliable design tool for retrofitting existing power plants toward the optimization of combustion related aspects such as temperatures, fuel burn out and NOx emissions. In the NOx postprocessing calculations performed, the “thermal NOx” and “fuel NOx” formation mechanisms are applied. The “thermal NOx” formation model is based on the extended Zeldovich mechanism,17 and the principal reactions governing the formation of thermal NOx from molecular nitrogen are given in the work of Diez et. al.,16 while the appropriate reaction rates are obtained by previous experimental investigations.18,19 For the “Fuel NOx” formation mechanism, the nitrogen in the fuel is considered. Nitrogen is assumed to be distributed between coal volatiles and char. Volatile nitrogen is converted first to HCN and NH3 and then to NO or N2 according to the reaction path given in ref 20. A HCN/NH3 ratio of 1:9 is used for the calculations. According to available literature,21,22 solid fuels of younger age like low rank brown coals, lignites, and biomass fuels tend to produce up to 10 times more NH3 than the level of HCN in contrast with fuels of older age like bituminous coals. Char nitrogen is assumed to be converted to HCN before
Table 2. Proximate and Ultimate Analysis of the Greek Lignite proximate analysis (% a.r.)
ultimate analysis (% daf)
before mill water ash volatiles fixed C Hu (kJ/kg K)
mean
56.25 13.35 18.33 12.07 5656
C H N O S
63.81 4.87 2.07 27.60 1.68
Table 3. Velocity Values of Primary and Secondary Air combustion air velocities (m/s) baseline primary air burner inlets (carrier gas) secondary air front side left-right side rear side hopper
dry coal dry coal dry coal 5% 10% 20%
6.1
5.7
5.4
4.6
42.4 24.9 13.5 2.2
42.4 24.9 13.5 2.2
42.4 24.9 13.4 2.2
35.5 20.9 11.3 1.9
integration of the particle force balance in a lagrangian reference frame. One discrete phase calculation and update of the discrete phase source terms is performed every 50 continuous phase calculations, while about 50 000 particles are injected and tracked in the boiler. The single rate kinetic model is used to account for coal devolatilization and the kinetics/diffusion limited rate model is used to account for char combustion. The kinetic constants applied in the present simulations are obtained from previous simulation work,11 where a 1 MWth combustion facility was simulated under raw lignite and dry lignite firing conditions and the simulation results were compared with respective experimental data.12 The Eddy-Break Up (EBU) model by Magnussen and Hjertager13 is applied for volatiles’ combustion, and the global two step reaction mechanism is taken into account. The volatiles are expressed as a hydrocarbon C xH yO z , where the corresponding x, y, z coefficients are derived from the ultimate analysis of the fuel.
(14) Epple, B.; Schneider, R.; Schnell, U.; Hein, K. Combust. Sci. Technol. 1995, 108 (4), 383–401. (15) Le Bris, T.; Cadavid, F. Fuel 2007, 86, 2213–2220. (16) Diez, L. I.; Cortes, C.; Pallares, J. Fuel 2008, 87, 1259–1269. (17) Zeldovich, Y. B.; Sadovnikov, P. Y.; Kamentskii, D. A. Oxidation of Nitrogen in Combustion; Academy of Sciences of USSR: Moscow, 1947. (18) Monat, J. P.; Hanson, R. K.; Kruger, C. H. 17th International Symposium on Combustion, 543, 1979. (19) Hanson, R. K.; Salimian, S. In Combustion Chemistry; Gardiner, W. C., Jr., Ed.; Springer-Verlag: New York, 1984; p 361 (20) Boardman, R. D.; Smoot, L. D. Pollutant Formation and Control. In Fundamentals of Coal Combustion: For Clean and Efficient Use; Smoot, L. D., Ed.; Elsevier: Amsterdam, The Netherlands, 1993; pp 433-438. (21) Nelson, P.F.; Buckley, A.N.; Kelly, M.D. Functional Forms of Nitrogen in Coals and the Release of Coal Nitrogen as NOx Precursors (HCN and NH3). 24th Symposium (Int’l) on Combustion, The Combustion Institute, 1992; p 1259. (22) Liu, H.; Gibbs, B. M. Fuel 2002, 81, 271–280.
(11) Agraniotis, M.; Stamatis, D.; Grammelis, P.; Kakaras, E. Fuel 2009, 88, 2385–2391. (12) Agraniotis, M.; Grammelis, P.; Papapavlou, Ch.; Kakaras, E. Fuel Process. Technol. 2009, 90, 1071–1079. (13) Magnussen, B. F.; Hjertager, B. H. Sixteenth Symposium (International) on Combustion, Pittsburgh, PA, 1976; pp 719-729.
5467
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
the standard operation in the whole day shows that dry coal cofiring does not have any direct impact on plant operation, since the variation of the monitored values stays within this high variation range in the entire day and is not affected by the cofiring tests. As a general result, it is stated that dry coal cofiring does not have a significant effect on the operational behavior of the specific boiler. The aged firing and control system seems to have a more visible impact on the observed high variation range of the operational data. Profile Measurements. The temperature profiles measured are presented in Figure 6. Although a slight increase of the flue gas temperature could be expected in the dry coal cofiring cases, this estimation is not confirmed by the measurement results. No clear tendency is noticed by the measured temperature profiles, while the standard deviation in the measured temperature values is generally high. The design of the particular firing system and the unstable character of the generated flow and temperature field, due to this firing system, may be the reason for this behavior. It is further argued that the level of the temperature values recorded in the profile measurements is representative for the given location and therefore that the particular experimental data should be taken into consideration and not disregarded. Emission Measurements at the Stack. The average values of the flue gas emissions during the tests are presented in
reacting to N2 or NO or oxidized directly to NO according to the mechanism described in refs 23 and 24.
Results and Discussion Experimental Activities. Operational Behavior. Average values of steam production and power production data of the two turbines for the tests performed are given in Table 4. They remain almost constant during the tests, indicating that steam production and heat and power production is not significantly affected by the firing mode. The flue gas and steam temperatures monitored are presented in Figures 4 and 5. The variation range of the steam temperatures presented is relatively high, more than 30 K at some periods of the day. This high range is not typical for modern, state of the art, large-scale boilers and is mainly justified by the small boiler size and the aged firing and control system. The comparison of monitored data during the testing period and Table 4. Stream Production and Produced Power during Tests
steam production (t/h) steam pressure (bar) MWe (back pressure turbine) MWe (main turbine) MWth (back pressure turbine)
raw lignite firing
3% cofiring
6% cofiring
79 60 4.3 8.1 21.8
76 61 4.2 8.2 21.7
78 62 4.3 8,1 21.8
Figure 4. Records of flue gas temperatures.
Figure 5. Records of steam temperatures.
5468
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
Figure 6. Temperature profiles at the main cofiring tests. Table 5. Mean Values and Standard Deviation of the Measurements Performed
O2 (% vol dry) NOx (mg/N m3 dry) 3
CO (mg/N m dry) SO2 (mg/N m3 dry)
mean std deviation mean std deviation mean std deviation mean std deviation
raw lignitefiring
3% cofiring
6% cofiring
6.1 0.6 290 11 60 5 984 82
7.1 0.7 319 18 67 5 1040 134
6.4 0.4 340 15 58 3 991 65
Figure 7. Average and maximum flue gas temperature along the furnace height.
directly extrapolated to typical Greek boilers of the 300 MWe class. Numerical Simulations. Temperature and O2 Profiles. The mean temperatures along the furnace height for the reference and the dry coal cases are presented in Figure 7. With an increase in the dry coal cofiring share, the average temperature at the near burner region at the level between 5 and 10 m increases. In particular, according to the simulation results, a potential increase of the dry coal cofiring share up to the high percentage of 20%, which however cannot be reached in the current operation mode of the particular boiler, would lead to a rise of the average temperature at the level of the lower raw and dry lignite burners (þ4 to þ5 m) up to 80 K compared to the reference case. At the level of the upper raw and dry lignite burners (þ7 to þ8 m), the predicted temperature rise for 20% cofiring is about 30-40 K, while at the furnace exit level is 20-30 K. These values should be considered as indicative for the expected influence of dry coal on the temperature fields in the furnace, when increasing the thermal share up to a very high percentage. Even in the 20% cofiring share, which is not realistic for the current boiler operation, no drastic increase is expected in the furnace exit temperature that could lead to overheating problems of the superheaters or even to melted deposits and fouling phenomena. This is an indication that moderate changes in the temperature fields are expected when increasing the dry coal thermal share. This moderate behavior was also noticed in the experimental investigations, where lower thermal shares were applied, and implies that dry coal cofiring in the investigated thermal shares is not expected to add any additional risks to the boiler’s operational behavior in regards to the expected temperature levels. In the same figure, the calculated maximum temperatures along the furnace height are also presented. A similar behavior as with the average temperature is also observed in the case of the maximum temperatures calculated. With an increase in the dry coal cofiring share, the maximum temperatures increases in the main burner region and slightly
Table 5. The increased percentage of O2 in the flue gas is an indication of the increased amount of false air entering into the boiler and the electrostatic precipitators. NOx values tend to increase by increasing the cofiring share. Although an increase tendency of flue gas temperature at the furnace exit section was not observed during the experimental investigation, local temperature picks, due to the increase of the dry coal cofiring share, could be expected primarily in the regions near the dry lignite burners. These temperature picks may provide an explanation for the increase tendency of NOx levels after taking into account the potential effect of the thermal NOx formation mechanism. Besides, an increase tendency of NOx is also predicted by the numerical simulations when increasing the cofiring share and will be presented below. CO emissions remain at a low level during the entire tests and do not seem to be influenced by the different cofiring modes. SO2 emissions are measured in all tests, and no clear effect of cofiring on SO2 emissions is noticed. No effect of the dry coal on SO2 formation is also expected since it is the sulfur content of the fuel that plays the main role in SO2 formation and not the combustion conditions in the furnace. The lack of a desulphurization unit in the particular power plant is the reason for the increased SO2 emissions. To sum up, the relative increase of NOx emission values up to 16% compared to the reference case is the main influence of cofiring on the plant’s emission behavior. This trend is characteristic for the particular boiler only and cannot be (23) Arenillas, A.; Backreedy, R. I.; Jones, J. M.; Pis, J. J.; Pourkashanian, M.; Rubiera, F.; Williams, A. Fuel 2002, 81, 627–636. (24) Jones, J. M.; Patterson, P. M.; Pourkashanian, M.; Rowlands, L.; Williams, A. 14th Annual International Pittsburgh Coal Conference (Clean Coal Technology and Coal Utilisation), Taiyuan, Shanxi, PRC, September 23, 1997.
5469
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
Figure 8. Temperature contours (°C) at the left raw coal burners’ plane: (a) reference case, (b) cofiring thermal share 5%, (c) cofiring thermal share 10%, and (d) cofiring thermal share 20%.
decreases in the region above the burner belt. In particular, an increase of the maximum temperature values up to 100 K is calculated for the case of 20% dry lignite cofiring compared to the reference case. These elevated temperature peaks are expected to influence thermal NOx formation. The former observations are confirmed by the presented temperature contours. The vertical plane of the left raw coal burner is used as a reference plane in Figure 8, while the horizontal dry lignite burners’ planes are used as reference planes in Figure 9. Simulation results are shown for the baseline and the three dry coal combustion cases (5%, 10%, and 20%). Hot zones in the near burner region increase by raising the dry coal cofiring thermal share indicating faster fuel ignition. What is more, additional hot spots in the area of the dry lignite burners are visible in the horizontal contours (Figure 10) when increasing the dry lignite cofiring share. The average values of O2 concentration in the flue gas at different levels along the furnace height are presented in Figure 10 for the baseline and the dry lignite cofiring cases. Dry lignite cofiring leads to a decrease of mean O2 concentrations at the main burner region and also to a subsequent decrease in the region above the burner belt. This is a hint for
a quicker fuel ignition during dry coal firing, which also leads to an improvement of fuel burnout. Burnout. The percentage of unburned carbon calculated in each simulated case is presented in Figure 11. The burnout is clearly improved by applying dry lignite cofiring. In the maximum dry coal firing case of 20%, the unburned carbon calculated is reduced to half of the reference value. The improvement of the fuel burnout is also one of the main arguments speaking in favor of the utilization of dry lignite as a supporting fuel in existing Greek lignite boilers. Boiler efficiency is expected to increase in the dry coal cofiring cases due to the reduced burnout losses. Furthermore, in cases where Greek boilers have to fire extremely low lignite qualities, cofiring of predried lignite as a supporting fuel would assist keeping the nominal load and avoiding potential flame instability problems. The increase of the fuel burnout predicted could not be measured in the experimental campaign, since the duration of the tested cofiring cases was too short and no representative bottom ash and fly ash samples could be collected for each cofiring mode tested. Nevertheless, the predicted trend of improvement for coal burnout is considered as realistic, although the calculated absolute values of coal burnout could not be verified by respective measurements in the large scale. 5470
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
Figure 9. Temperature contours (°C) at the dry lignite burners’ planes: (a) reference case, (b) cofiring thermal share 5%, (c) cofiring thermal share 10%, and (d) cofiring thermal share 20%.
Figure 11. View of unburned carbon percentage.
Heat Flux. In order to evaluate the impact of dry coal cofiring on the boiler’s heat balance, the wall heat flux to the evaporator walls is calculated for each of the simulated cases. The heat flux to the furnace walls is composed of two different parts, the radiative one, which is the major part and is generated by the radiation of the high flame temperature to the walls, and the convective one, which represents the convective heat flux from the hot flue gas to the walls. Both parts are taken into account. The heat flux integral value (in kilowatts) in the three main furnace zones: (a) the hopper,
Figure 10. Average O2 concentration along the furnace height.
5471
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
Figure 12. Radiative heat flux integral (kilowatts) on furnace zones.
Figure 15. Average NOx concentration along furnace height (parts per million).
furnace zones; however, this rise is in relative terms less than 5% in most of the cases considered. Moreover, the values calculated for the average heat flux density in the burner belt, of about 50 kW/m2, are considered as low and are only representative of the specific furnace. Typical values for modern lignite fired furnaces are about 100 kW/m2. The change of the heat flux distribution due to dry coal cofiring is, therefore, expected to have a low impact on the performance of the evaporator of the specific boiler. More generally speaking, however, there is the possibility of potential damage on the evaporator surfaces due to overheating when cofiring dry lignite has to be investigated in detail for any future potential application. In summary, the impact of dry coal cofiring on the evaporator heat flux and, in general, on the overall boiler heat balance is expected to be low in the case of the particular furnace. NOx Emissions. The calculated profiles of NOx concentration are presented in Figure 15. With an increase in the dry coal cofiring share, a clear increase of NOx concentration along the whole furnace height is observed. Previous experimental investigations in the lab scale25,26 and the experimental investigations in the particular boiler confirm this tendency. Further studies on possible NOx reduction methods, like air and fuel staging practices, cannot be performed in the specific boiler, due to its aged design and limited options for retrofitting works. Nevertheless, if dry coal cofiring is considered for a modern Greek boiler of the typical 300 MWe class, the investigation of the possible NOx reduction measures is one of the key aspects which have to be addressed in detail.
Figure 13. Total heat flux integral (kilowatts) on furnace zones.
Figure 14. Distribution of total heat flux (kilowatts/meters squared) on furnace zones.
(b) the main burner zone, and (c) the zone above the burner belt as well as the overall heat flux to the whole furnace are presented in Figures 12 and 13 for the radiative and the total heat flux, respectively. The total heat flux value to the furnace walls increases by raising the dry coal cofiring share; nevertheless, the relative changes are small. In particular, in the 20% cofiring case the calculated total heat flux is about 900 kW higher compared to the reference case, which corresponds to a relative increase of less than 4%. Moreover, the ratio of the radiative heat flux to the total heat flux almost remains constant indicating that dry coal cofiring does not extensively affect the heat transfer in the particular boiler. In order to get detailed information on the possible effect of dry coal cofiring on the distribution of the wall heat flux along the furnace height, average values of total heat flux density (in kilowatts/meters squared) are calculated for six different zones of the evaporator. Each zone corresponds to a different furnace level. According to the numerical results presented in Figure 14, dry coal cofiring has a moderate effect on the heat flux distribution and no major changes are expected. The specific heat flux density increases when increasing the dry coal firing share in most of the considered
Conclusions Lignite predrying and dry lignite firing is considered as a significant technological development, which will play an (25) Maier, J.; Kluger, F.; Hocquel, M.; Spliethoff, H.; Hein, K. R. G. 23rd International Conference on Coal Utilization and Fuel Systems, Clearwater, FL, March 9-13, 1998. (26) Maier, J.; Heinzel, T.; Spliethoff, H.; Hein, K. R. G. Proceedings of the International Symposium on Clean Coal Technology, Xiamen/ China, 1997.
5472
Energy Fuels 2010, 24, 5464–5473
: DOI:10.1021/ef1002843
Agraniotis et al.
important role in future lignite power plants. The influence of dry lignite cofiring on several technical aspects of an industrial scale boiler is evaluated by on site measurements and numerical simulations. No clear effect of dry coal coutilization on the boiler’s operation and performance is experimentally observed, due to the low cofiring shares applied. Furthermore, the high variation of some operational parameters during the tests is assumed to be more related with the aged firing and control system than with the effect of dry coal cofiring. A definite increased tendency of NOx emissions is measured by increasing the cofiring share, which is expected and has to be further examined in the future, in the case that lignite predrying and dry lignite cofiring will be applied in one of the existing, modern Greek lignite boilers. Regarding numerical simulations, clear trends are observed in most of the combustion related parameters. The average temperatures along the furnace height, as well as the fuel’s burn out, increase by increasing the cofiring share. This is also an argument for the further utilization of dry lignite as a supporting fuel, in a time
when the raw lignite quality fired is extremely poor. The overall results indicate that dry coal cofiring is feasible in the particular plant, while the expected changes of the combustion behavior are moderate and within the ordinary operation range. The current limitation on the cofiring share is based on the existing feeding system for dry lignite and the boiler’s aged firing system. In order to evaluate the possible application of dry lignite cofiring in modern Greek boilers, further investigations are necessary. CFD modeling is a useful tool in this field, since it may assist in the evaluation of several firing concepts for dry lignite, which are under consideration, before the installation of the firing system for dry lignite on site. In this way, valuable time and resources can be saved. Acknowledgment. The support of the European CommissionResearch Fund for Coal and Steel on the work of the Project Drycoal (Grant RFCP-CT-2004-00002) is gratefully acknowledged. The authors would like also to thank Mr. Papapavlou, Mr. Nikolaidis, and Mr. Logothetis from PPC for their support during the industrial tests at the Liptol power plant.
5473
