Publications

Publications / SAND Report

Closed-loop biomass co-firing in a laboratory reactor and in a full-scale boiler

Blevins, Linda G.

Co-firing tests were conducted in a pilot-scale reactor at Sandia National Laboratories and in a boiler at the Hawaiian Commercial & Sugar factory at Puunene, Hawaii. Combustion tests were performed in the Sandia Multi-Fuel Combustor using Australian coal, whole fiber cane including tops and leaves processed at three different levels (milled only, milled and leached, and milled followed by leaching and subsequent milling), and fiber cane stripped of its tops and leaves and heavily processed through subsequent milling, leaching, and milling cycles. Testing was performed for pure fuels and for biomass co-firing with the coal at levels of 30% and 70% by mass. The laboratory tests revealed the following information: (1) The biomass fuels convert their native nitrogen into NO more efficiently than coal because of higher volatile content and more reactive nitrogen complexes. (2) Adding coal to whole fiber cane to reduce its tendency to form deposits should not adversely affect NO emissions. ( 3 ) Stripped cane does not offer a NO advantage over whole cane when co-fired with coal. During the field test, Sandia measured 0 2 , C02, CO, SO2, and NO concentrations in the stack and gas velocities near the superheater. Gas concentrations and velocities fluctuated more during biomass co-firing than during coal combustion. The mean 0 2 concentration was lower and the mean C02 concentration was higher during biomass co-firing than during coal combustion. When normalized to a constant exhaust 0 2 concentration, mean CO concentration was higher and mean NO concentration was lower for biomass co-firing than for coal. The SO2 concentration tracked the use of Bunker C fuel oil. When normalized by the amount of boiler energy input, the amounts of NO and SO2 formed were lower during biomass co-firing than during coal combustion. The difference between NOx trends in the lab and in the field are most likely a result of less effective heat and mass transfer in the boiler. Particles were sampled near the superheater tube using an impaction probe and were analyzed using scanning electron microscopy. Particle loading appeared higher for biomass co-firing than for coal combustion, especially for the smaller particle diameters. Laser-induced breakdown spectroscopy (LIBS) was used to detect silicon, aluminum, titanium, iron, calcium, magnesium, sodium, and potassium concentrations near the superheater. LIBS provided an abundant amount of real-time information. The major constituents of the fuel ash (silicon and aluminum) were also the major measured inorganic constituents of the combustion products. The combustion products were enriched in sodium relative to the fuel ash during all tests, and they were enriched in potassium for the biomass co-firing tests. Alkali metals are enriched because compounds containing these elements are more readily releasable into the combustion products than refractory components that remain in large particles such as silicon, aluminum, and titanium. Relative to the measured deposit chemistry, the combustion flows were enriched in iron, sodium, and potassium, constituents that are known to form fumes laden with fine particles and/or vapors. The LIBS results yield insight into the deposition mechanism: Impaction of larger particles dominates over fume deposition. The present application of LIBS reveals its potential to provide real-time field information on the deposition propensity of different fuels and the effects of different fuels and boiler operating conditions.