Low NOx Combustion with Natural Gas

Page 8

ADDENDUM LOW NOx COMBUSTION WITH NATURAL GAS Glenn C. Neff, Sr., M. L. Joshi, Marvin E. Tester and Sherry K. Panahi Current Test Results (August 29-31. 1990) Using The Original Test Equipment With Some Modifications: The existing cracker (Table-l) was modified to fire in tangential direction using a series of oxygen and natural gas ports in the combustion chamber. At two tangential locations (180° apart), four (4) natural gas ports and eight (8) oxygen ports, gave essentially a swirling flame with initial gas velocities of about 30 ft/sec. This flame enabled a uniform temperature profile in the combustion chamber, which was not attained using an axial firing in the previous tests. In the new firing configuration, the cracking gas was introduced axially through the center of the combustion chamber front face. It was necessary to inj ect the cracking gas in the chamber with lo~ enough velocity (and sufficient residence time) needed for thermal cracking (-45 ft/sec). Preliminary tests indicated a successful campaign wi th significant progress on many fronts. First, the new firing configuration offered a very high (2500°F) and uniform temperature profile, which enabled accelerated cracking with a reduced level of reforming. Secondly, the soot yields wi th the new configuration were as high as 25 - 30% on the weight basis. An average soot rate was about 0.003 lb/cf of flue gas. The enhanced cracking was partly accomplished by the operation at a much higher cracker operating temperature (-2500°F). Third, the operation of burner at a nearly stoichiometric operation was feasible wi thout any undesirable refractory damage. This was done using an appropriate amount of cracking gas, which attenuated the chamber temperature to approximately 2500°F. The Orsat data indicated, C02 8.4%, CO = 26.2%, H2 = 58.5%, CH4 = 6.6% and N2 = 0.3%, which does indicate some reforming. The total oxygen to total natural gas ratio was approximately 0.67, the cracking gas to total natural gas ratio was approximately 0.70, and the overall stoichiometric ratio was 0.33. Future tests are planned to fine tune the various process variables and establish a record of continuous running for a reasonable period.