Prediction of Sorbent Performance in a Circulating Fluidized Bed Boiler Based on Petrographic Properties

A method for prediction of sorbent consumption is presented here and has been developed based on plant operating data for a boiler in which several limestone and dolostone products were tested under similar firing conditions. The method considers the characteristic partitioning of calcium and sulfur between the flyash and bottom ash stream for the boiler, the feed particle size distribution of the sorbent, and petrographic properties of the sorbents. The predictions of sorbent usage were compared to plant operating data for five sorbents, of two distinct petrographic types. The plant operating data used featured full load operation. The five sorbents tested were all from Pennsylvania, and each contained greater than $40wt.%$ CaO. In four of the five cases, the predicted sorbent usage was within $10wt.%$ of the average full load sorbent usage by the boiler.