EUBCE 2026 - Arkya SANYAL - Early Prediction of Bed Agglomeration in Stationary Fluidized Beds Using High-Frequency Pressure Analysis and Image Processing

Early Prediction of Bed Agglomeration in Stationary Fluidized Beds Using High-Frequency Pressure Analysis and Image Processing

Short Introductive summary

This presentation introduces an industrially-validated system for the early detection of bed agglomeration in the fluidized bed combustion (FBC) of sewage sludge, a critical challenge driven by German AbfKlärV regulations. Our approach, developed in the ProKläR-mission project, uses high-frequency pressure sensors to extract a "characteristic frequency" as a robust predictive indicator. At the commercial TVM Mainz sewage sludge plant, this method provided a clear 8-day advance warning before an agglomeration event. This sensor data was scientifically validated in the lab using a novel camera-based image tool, which visually correlated the frequency drop with the physical onset of defluidization. This proven, low-cost technology is now being scaled in the Klär-vision follow-up project with over 12 major industrial partners (including BASF, RWE, and Martin GmbH) to build an AI-driven, Big Data optimization platform.

A. Sanyal, Chair of Energy Process Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg, GERMANY
S. Leimbach, Chair of Energy Process Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg, GERMANY
J. Karl, Chair of Energy Process Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg, GERMANY
S. Salzmann, KMT, Munich, GERMANY