COAL COMBUSTION IN O2/CO2 MEDIUM FOR POWER GENERATION: NUMERICAL MODELING OF CHAR BURNOUT, NOX AND CO EMISSIONS

Open Access
Author:
Naredi, Prabhat
Graduate Program:
Energy and Geo-Environmental Engineering
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
April 16, 2009
Committee Members:
  • Sarma V Pisupati, Dissertation Advisor
  • Sarma V Pisupati, Committee Chair
  • Alan W Scaroni, Committee Member
  • Ljubisa R Radovic, Committee Member
  • Semih Eser, Committee Member
  • Daniel Connell Haworth, Committee Member
Keywords:
  • Char reactivity
  • thermogravimetric analysis
  • drop tube reactor
  • modeling
  • CFD
  • oxy-coal combustion
  • NOx
  • CO emission
  • burnout
Abstract:
Oxy-fuel combustion is a promising technique to achieve significant reduction of carbon dioxide (CO2) emissions into the atmosphere from coal-fired power plants. In this approach, coal particles are burned in a medium of oxygen (O2) and recycled flue gas (RFG) consisting primarily of CO2 and/or H2O so that the product is a near-pure stream of CO2. The CO2 rich flue gas obtained using this approach, as compared to flue gas from existing conventional power plants, can be more readily sequestered or utilized for oil recovery because of lesser amount of impurities. One of the significant barriers for implementing this approach is the need for an expensive oxygen generation unit. Currently, the RFG approach is being evaluated for its feasibility in existing and new coal-fired power plants. In order to realize the full potential of oxy-fuel combustion, researchers are applying commercially available computational fluid dynamics (CFD) tools to predict the carbon burnout, CO and NOx emissions, flame stability and heat transfer to boiler walls. However, in numerical studies, the effect of higher concentrations of CO2 on heat transfer is typically accounted for, but its effect on coal pyrolysis behavior and char burnout is usually neglected. In experimental studies, conflicting results have been published and the discrepancies remain unresolved. The overall objective of this thesis is to modify the existing char-oxidation kinetics sub-model in a commercially available CFD tool FLUENTTM by including user defined functions (UDFs) and more appropriate intrinsic activation energy values to capture the effect of elevated levels of CO2 on char burnout and CO emissions. This study’s experimental investigation attempts to identify coals that are more suitable for oxy-coal combustion for further reducing un-burnt carbon and NOx emissions compared to combustion in air. In order to achieve these objectives, pyrolysis and combustion tests were conducted in a lab-scale drop tube reactor (DTR) at furnace wall temperatures of 1,173-1673 K using high-volatile (hvCb) and low-volatile (lvb) bituminous coal samples. The results show that pyrolyzing the coal particles in CO2 changes the physical properties of resultant chars and the split of coal-N between the volatile and char fractions. However, these changes appear to be caused by the contribution from char-CO2 reaction and not due to a change in gas phase chemistry. The modified char-oxidation model uses a more appropriate intrinsic approach to account for the transition in rate controlling regime based on differences in char reactivity towards O2 and CO2. In the model, activation energy for the char-O2 and char-CO2 reactions is taken as an average of activation energies estimated from the onset of maximum to 80% conversion. Model calculations at lower heating rates and temperature condition in TGA using intrinsic rate parameters and physical properties of char accurately capture the transition from a kinetically controlled regime (Zone I) to an intra-particle diffusion controlled regime (Zone II). However, at higher heating rates and temperature condition in DTR, char burnouts are predicted to be ~5 times lower than the measured values. This under-prediction of the char burnout is attributed to the uncertainty in measured activation energy value (by about 13 kJ/mol) which was not reflected in TGA condition due to lower operating temperature. During combustion tests, a higher CO, lower NOx, and lower char burnout are observed for combustion in a 21% O2/79% CO2 mixture, as compared to combustion in air. The computational prediction showed that the majority of the differences in NOx emissions between the two combustion media are the result of a decrease in gas temperature due to higher specific heat of CO2 compared to N2 and not due to coal-N retention in char in a CO2-rich environment. As expected, burnout measurements in the DTR showed a rapid increase in char burnout with temperature for the lvb coal compared to that of the hvCb coal during combustion in a 21% O2/79% CO2 mixture in contrast to combustion in air. Reactivity and activation energy measurement of the two coal chars towards O2 and CO2 in the TGA showed that the observed increase in char burnout in DTR during combustion in a 21% O2/79% CO2 mixture arises from higher activation energy and reactivity of lvb coal char toward CO2, as compared to hvCb coal. Model prediction from a pilot-scale facility shows that a higher char burnout during oxy-coal combustion occurs because of higher O2 partial pressure (30%) compared to combustion in air. A significant increase in char burnout due to contribution from char-CO2 reaction is predicted only for lvb coal. A higher reduction in NOx emissions is predicted for hvCb coal, as opposed to lvb coal because of differences in flame temperatures during oxy-coal combustion, as compared to combustion in air. Overall, the results obtained in this study indicate that for oxy-coal combustion approach, a high-volatile coal is more suitable to achieve lower NOx emissions and a low-volatile coal is beneficial to achieve lower un-burnt carbon.