On the diagnosis of a fouling condition in a kraft recovery boiler: combining process knowledge and data-based insights

Abstract

Fouling is still a major challenge for the operation of kraft recovery boilers. This problem is caused by accumulation of ash deposits on the surfaces of heat exchangers in the upper part of the boiler over time. The first consequence is the reduction of steam production due to loss of heat transfer and, finally, the shutdown of the boiler due to clogging. The present work investigated the operational condition of a modern kraft boiler under a critical fouling condition. This boiler had even faced a manual cleaning due to a clogging event. This analysis combined process knowledge, plant team experience, and a data-driven approach, given the complexity of the process. In this sense, historical data covering this critical period of operation were collected. After a cleaning procedure, they were used to obtain a predictive neural network model for the flue gas pressure drop in the boiler bank, which is an indirect measure of ash deposit accumulation. Once validated, it was used for sensitivity analysis, with the aim of quantifying the effects of the model inputs. Five variables out of eighteen accounted for nearly 60% of the total effect on pressure drop. Namely, primary air temperature (21.6% of the total effect) and flow rate (11.1%), black liquor flow rate (9.9%) and temperature (8.4%), and white liquor sulfidity (8.6%). The analysis of these results mainly suggested an excess of carryover, which composes the ash deposits. Recommended actions to mitigate the fouling condition involved adjustments to the primary air system before the more drastic solution of reducing the boiler load.