Experimental and numerical investigations of ... - IEAGHG
2nd Oxyfuel Combustion Conference
Experimental and numerical investigations of devolatilisation and char oxidation in oxycombustion conditions Aleksandra Milewskaa, Jarosław Hercoga* , Bartosz Świątkowskia a
Institute of Power Engineering, Mory 8, Warsaw 00-784, Poland
Keywords: devolatilisation, char oxidation, CFD modeling, drop-tube furnace
1. Introduction The oxycombustion has already been developed for several years, however the understanding and mathematical description of that process is not completed and available experimental data are scarce. The knowledge of coal devolatilisation and char combustion i n O2 /CO2 must be still developed [1, 2]. Preliminary experiments and simulations made at Institute of Power Engineering during project “Supercritical Pulverised Coal Boilers” [3, 4] proved that oxycombustion process requires further fundamental investigations. The aim of the present study is to validate the existed devolatilisation and char combustion models in oxycombustion conditions and proposed possible modifications. The numerical simulations of coal combustion in laboratory drop tube facilities were done using the commercial simulation software (Ansys/Fluent 12). The mathematical model is the Euler-Lagrangian approach with heat and mass transfer between continuum and dispersed phases. The pressure-velocity coupling is solved using the SIMPLEC method and turbulent model is calculated by k -epsilon algorithm. The finite-rate / eddy dissipation model was applied to calculate the behaviour of volatiles combustion. Radiative heat transfer was modelled by discrete ordinates (DO) model. The non-standard devolatilisation and char combustion models were implemented via UDF (User Defined Functions). The kinetic parameters were calculated based on carried out experiments in laboratory scale experiments and compared with the mathematical modeling. 2. Facility Two drop -tube fu rnaces, for evaluation of devolatilisation and char combustion kinetic parameters were used. They are made of heat -resistant, electrically heated, steel tubes of 89 mm inner diameter. Reactors are equipped with electrical gas pre-heaters, gas mass-flow controllers, thermocouples mounted along their height and acquisitioncontrol unit. Solid fuel is transported into the furnace by use of screw feeder and feeding probes. Fuel flows through the reaction zone and is subsequently sampled by sampling probes and separated in cyclone and filt er. First, 1.5 m. long, drop-tube, is designed to perform devolatilisation tests and is equipped with movable axial, water-cooled probe, which enables instantaneous quenching and collection of samples at different residence times. The second, 4
*
Corresponding author. Tel.: +48223451411; fax: +48226428378 E-mail address: [email protected]
2
m long, tube is designed for char combustion and emission experiments and is equipped with horizontal, watercooled probes, which can be inserted into the furnace through the sampling ports. Estimated heating rates from CFD 4 simulations are in the range of 0.8 to 1·10 K/s at adopted conditions. Maximum operational temperature of both o reactors is 1150 C [5]. The schematic and picture of the drop-tubes are presented in Figure 1 and 2. 3. Experimental Devolatilisation and char oxidation tests were conducted for two hard coals, which were dried, grinded and sieved to particle size between 63-90 m. The proximate and ultimate analysis of the fuels is presented in Table 1. Table 1. Proximate and ultimate analysis of the fuels (all values in %wt.). Fuel
M
A
VM
C
H
O
N
S
LHV, kJ/kg
RC
3,1
14,9
30,0
66,04
4,18
9,50
1,99
0,33
26054
SAC
4,0
13,0
23,5
69,83
3,86
7,04
1,64
0,61
26582
In order to compare pulverized fu el behavior between air and oxycombustion conditions, devolatilisation experiments were conducted in gaseous atmospheres containing of 100 %vol. N2 or CO2, whereas char oxidation tests were performed in mixtures of either N2 or CO2 with 10, 20, and 35 %vol. O2 . Devolatilisation experiments under 100 %vol. N2 or 100 %vol. CO2 were conducted fo r both coals at two temperatures (1023, 1123 K) and three residence times ranging from 0.15 to 0.35 s. Thermal histories of the particles were obtained from the CFD simulations in order to determine accurate particles residence times and temperatures for kinetic parameters calculations.
Figure 1. Schematic of the experimental stand Figure 2. Experimental stand overview with two drop -tubes Experimental results were compared with CFD results, based on the typical devolatilisation models, i.e. Single First Reaction Order - SFOR [6], 2-step Kobayashi [7] and Distributed Activation Energy Model - DAEM [8]. Results from the experiments together with SFOR model predictions, based on the kinetic parameters obtained in the tests, are presented in Figure 3. SFOR model well predicts the devolatilisation process either in oxycombustion or aircombustion conditions. It can also be observed that the potential maximum volatile matter release – HVM is higher under N2 atmosphere, and the overall devolatilisation rate is higher at CO2 atmosphere (but not significantly). Effect
Author name / Energy Procedia 00 (2011) 000–000
3
o f lower HVM in CO2 atmosphere can be explained by crosslinking reactions at the surface of the particle, which can prevent a more extend devolatilisation.
Figure 3. Weight loss versus residence time at 1023 and 1123 K reactor temperature for Russian Coal and South African Coal (points - experiments results, lines- predictions). Derived parameters of the SFOR model (ie. Q parameter, frequency factor – k0 and activation energy – E) are presented in Table 2. Table 2. Kinetic parameters for examined fuels and conditions Atm., Q, HVM, k0 , 1/s E, J/mol Fuel %vol. %wt. daf
k(1000K), -
RC RC
100 CO2 100 N2
1.53 1.70
58.61 61.46
44 18529
21436 73818
3.31 2.58
SAC SAC
100 CO2 100 N2
1.45 1.58
47.18 51.55
39 1156
18452 52735
4.19 2.03
4. Summary and conclusions Devolatilisation and char combustion tests for 2 coals were studied under air (O2 /N2 ) and oxycombustion (O2 /CO2 ) conditions. Results were applied for validation of mathematical models. Good ag reement between experimental and numerical simulations results was obtained for most models. Single First Reaction Order and Pseudo-Chemical Reaction [9 ] models for devolatilisation and char oxidation respectively well predict fu el behaviour also in oxycombustion conditions and can be used fo r quick simulations of the combustion processes and facility design operating in modified atmospheres.
4
5. References 1. Murphy, J. J.; Shaddix, C. R., Combustion kinetics of coal chars in oxygen-enriched environments, Combustion and Flame 2006, 144, (4), 710 -729. 2. Toporov D., et. al., Detailed investigation of a pulverized fuel swirl flame in CO2 /O2 atmosphere, Combustion and Flame 155 (2008), 605-618. 3. Milewska A., Swiatkowski B., Jovanovic R., Simultaneous devolatilisation and char combustion in oxyfuel coal combustion. Flame stand-off distance, 1st Oxyfuel Combustion Conference, Cottbus, Germany 2009. 4. Bocian P., Świątkowski B., “Ignition and coal combustion in oxycombustion conditions” in: “New Technologies of combustion and flue gas cleaning” (in Polish), ed. Nowak W., Pronobis M., Wydawnictwo Politechniki Śląskiej, Gliwice 2010, p. 30-36. 5. Biomass Oxy fuel and Flameless Combustion (BOFCom), Final Report, 2010, Project carried out with a fi nancial grant of the Research Programme of the Research Fund for Coal and Steel. 6. Badzioch S., Hawksley G.W., Ind. Eng. Chem. Process Design Develop., vol. 9, no. 4, p. 521, 1970. 7. Kobayashi H., Howard J. B. and Sarofim A. F., Coal devolatilization at high temperatures, 16th Symposium (Int.) on Combustion, p. 411 - 425, The Combustion Institute, Pittsburgh, Pennsylvania, 1976. 8. Pitt, G. J., The kinetics of the evolution of volatile products from coal, Fuel, 41, p. 267 (1962). 9. Baum M. M., Street P. J., Combustion Science and Technology, vol. 3, p. 231-243, 1971.
Experimental and numerical investigations of devolatilisation and char oxidation in oxycombustion conditions Aleksandra Milewskaa, Jarosław Hercoga* , Bartosz Świątkowskia a
Institute of Power Engineering, Mory 8, Warsaw 00-784, Poland
Keywords: devolatilisation, char oxidation, CFD modeling, drop-tube furnace
1. Introduction The oxycombustion has already been developed for several years, however the understanding and mathematical description of that process is not completed and available experimental data are scarce. The knowledge of coal devolatilisation and char combustion i n O2 /CO2 must be still developed [1, 2]. Preliminary experiments and simulations made at Institute of Power Engineering during project “Supercritical Pulverised Coal Boilers” [3, 4] proved that oxycombustion process requires further fundamental investigations. The aim of the present study is to validate the existed devolatilisation and char combustion models in oxycombustion conditions and proposed possible modifications. The numerical simulations of coal combustion in laboratory drop tube facilities were done using the commercial simulation software (Ansys/Fluent 12). The mathematical model is the Euler-Lagrangian approach with heat and mass transfer between continuum and dispersed phases. The pressure-velocity coupling is solved using the SIMPLEC method and turbulent model is calculated by k -epsilon algorithm. The finite-rate / eddy dissipation model was applied to calculate the behaviour of volatiles combustion. Radiative heat transfer was modelled by discrete ordinates (DO) model. The non-standard devolatilisation and char combustion models were implemented via UDF (User Defined Functions). The kinetic parameters were calculated based on carried out experiments in laboratory scale experiments and compared with the mathematical modeling. 2. Facility Two drop -tube fu rnaces, for evaluation of devolatilisation and char combustion kinetic parameters were used. They are made of heat -resistant, electrically heated, steel tubes of 89 mm inner diameter. Reactors are equipped with electrical gas pre-heaters, gas mass-flow controllers, thermocouples mounted along their height and acquisitioncontrol unit. Solid fuel is transported into the furnace by use of screw feeder and feeding probes. Fuel flows through the reaction zone and is subsequently sampled by sampling probes and separated in cyclone and filt er. First, 1.5 m. long, drop-tube, is designed to perform devolatilisation tests and is equipped with movable axial, water-cooled probe, which enables instantaneous quenching and collection of samples at different residence times. The second, 4
*
Corresponding author. Tel.: +48223451411; fax: +48226428378 E-mail address: [email protected]
2
m long, tube is designed for char combustion and emission experiments and is equipped with horizontal, watercooled probes, which can be inserted into the furnace through the sampling ports. Estimated heating rates from CFD 4 simulations are in the range of 0.8 to 1·10 K/s at adopted conditions. Maximum operational temperature of both o reactors is 1150 C [5]. The schematic and picture of the drop-tubes are presented in Figure 1 and 2. 3. Experimental Devolatilisation and char oxidation tests were conducted for two hard coals, which were dried, grinded and sieved to particle size between 63-90 m. The proximate and ultimate analysis of the fuels is presented in Table 1. Table 1. Proximate and ultimate analysis of the fuels (all values in %wt.). Fuel
M
A
VM
C
H
O
N
S
LHV, kJ/kg
RC
3,1
14,9
30,0
66,04
4,18
9,50
1,99
0,33
26054
SAC
4,0
13,0
23,5
69,83
3,86
7,04
1,64
0,61
26582
In order to compare pulverized fu el behavior between air and oxycombustion conditions, devolatilisation experiments were conducted in gaseous atmospheres containing of 100 %vol. N2 or CO2, whereas char oxidation tests were performed in mixtures of either N2 or CO2 with 10, 20, and 35 %vol. O2 . Devolatilisation experiments under 100 %vol. N2 or 100 %vol. CO2 were conducted fo r both coals at two temperatures (1023, 1123 K) and three residence times ranging from 0.15 to 0.35 s. Thermal histories of the particles were obtained from the CFD simulations in order to determine accurate particles residence times and temperatures for kinetic parameters calculations.
Figure 1. Schematic of the experimental stand Figure 2. Experimental stand overview with two drop -tubes Experimental results were compared with CFD results, based on the typical devolatilisation models, i.e. Single First Reaction Order - SFOR [6], 2-step Kobayashi [7] and Distributed Activation Energy Model - DAEM [8]. Results from the experiments together with SFOR model predictions, based on the kinetic parameters obtained in the tests, are presented in Figure 3. SFOR model well predicts the devolatilisation process either in oxycombustion or aircombustion conditions. It can also be observed that the potential maximum volatile matter release – HVM is higher under N2 atmosphere, and the overall devolatilisation rate is higher at CO2 atmosphere (but not significantly). Effect
Author name / Energy Procedia 00 (2011) 000–000
3
o f lower HVM in CO2 atmosphere can be explained by crosslinking reactions at the surface of the particle, which can prevent a more extend devolatilisation.
Figure 3. Weight loss versus residence time at 1023 and 1123 K reactor temperature for Russian Coal and South African Coal (points - experiments results, lines- predictions). Derived parameters of the SFOR model (ie. Q parameter, frequency factor – k0 and activation energy – E) are presented in Table 2. Table 2. Kinetic parameters for examined fuels and conditions Atm., Q, HVM, k0 , 1/s E, J/mol Fuel %vol. %wt. daf
k(1000K), -
RC RC
100 CO2 100 N2
1.53 1.70
58.61 61.46
44 18529
21436 73818
3.31 2.58
SAC SAC
100 CO2 100 N2
1.45 1.58
47.18 51.55
39 1156
18452 52735
4.19 2.03
4. Summary and conclusions Devolatilisation and char combustion tests for 2 coals were studied under air (O2 /N2 ) and oxycombustion (O2 /CO2 ) conditions. Results were applied for validation of mathematical models. Good ag reement between experimental and numerical simulations results was obtained for most models. Single First Reaction Order and Pseudo-Chemical Reaction [9 ] models for devolatilisation and char oxidation respectively well predict fu el behaviour also in oxycombustion conditions and can be used fo r quick simulations of the combustion processes and facility design operating in modified atmospheres.
4
5. References 1. Murphy, J. J.; Shaddix, C. R., Combustion kinetics of coal chars in oxygen-enriched environments, Combustion and Flame 2006, 144, (4), 710 -729. 2. Toporov D., et. al., Detailed investigation of a pulverized fuel swirl flame in CO2 /O2 atmosphere, Combustion and Flame 155 (2008), 605-618. 3. Milewska A., Swiatkowski B., Jovanovic R., Simultaneous devolatilisation and char combustion in oxyfuel coal combustion. Flame stand-off distance, 1st Oxyfuel Combustion Conference, Cottbus, Germany 2009. 4. Bocian P., Świątkowski B., “Ignition and coal combustion in oxycombustion conditions” in: “New Technologies of combustion and flue gas cleaning” (in Polish), ed. Nowak W., Pronobis M., Wydawnictwo Politechniki Śląskiej, Gliwice 2010, p. 30-36. 5. Biomass Oxy fuel and Flameless Combustion (BOFCom), Final Report, 2010, Project carried out with a fi nancial grant of the Research Programme of the Research Fund for Coal and Steel. 6. Badzioch S., Hawksley G.W., Ind. Eng. Chem. Process Design Develop., vol. 9, no. 4, p. 521, 1970. 7. Kobayashi H., Howard J. B. and Sarofim A. F., Coal devolatilization at high temperatures, 16th Symposium (Int.) on Combustion, p. 411 - 425, The Combustion Institute, Pittsburgh, Pennsylvania, 1976. 8. Pitt, G. J., The kinetics of the evolution of volatile products from coal, Fuel, 41, p. 267 (1962). 9. Baum M. M., Street P. J., Combustion Science and Technology, vol. 3, p. 231-243, 1971.
