Boiler fouling management

Pipeline name

Boiler fouling management

Pipeline overview

Pipeline components

The following components are used in the pipeline:

Digital Twin Data Acquisition/Collection:

  • Sensors: Acoustic Sensing
  • SFW Digital Service Platform
  • OPC DA

Digital Twin Representation:

  • Prosys OPC UA Simulation Server

Hybrid/Cognitive Digital Twin:

  • MATLAB-Simulink, incl. MATLAB OPC Toolbox
  • COGNITWIN MATLAB tools: Physical model tuning (PMFIR), Unscented Kalman filtering (UKF), Ensemble Kalman filtering (EnKF), subspace identification (SID), dynamic modelling (Prony)
  • COGNITWIN physical model for a heat exchanger (HE model, for MATLAB)
  • SFW model for CFB furnace (hotloop, for MATLAB)
  • COGNITWIN MIP tool for sootblowing optimization

Digital Twin Visualisation and Control:

  • Apache StreamPipes
  • Sumitomo plant UI/UX

The tools are packaged into five components following the WP3 pilot-driven approach:

  • FUSE, aiming at solving the fuel characterization problem, including application independent PMFIR and UKF tools.
    • The FUSE OPC-UA component supports the FUSE by demonstrating WP3 data communication.
    • The SubFUSE component provides an alternative application independent data-driven identification tool, applied in solving the fuel characterization problem.
  • FouMon collects tools for solving the fouling monitoring problem, including the HE-model, EnKF and ODtool application independent tools.
  • FouCon provides the optimization tool for solving the WP3 sootblowing problem: MIP/sb. The application independent Prony-tool is provided for data-driven modelling.

Pipeline example description

Data sources

  • Process data is collected on-line from plant
  • Acousting sensor is operating on-line

Data processing

  • Data is transferred from plant to SFW services and to analytics, on-line or in off-line batches
  • Analytics for interpretation of acoustic sensoring data.
  • Analytics contains Matlab-tools for data quality improvement, physical models, model tuning, process identification (data-driven modelling) and optimization/predictive maintenance.
  • OPC-UA can communicate data and computation results between SFW services, Matlab and StreamPipes

Data sinks

  • Sumitomo plant UI/UX is used for on-line plant operation and guidance

Defined in COGNITWIN WPs

  • WP3 SFW pilot problem and solution
  • WP5 hybrid and cognitive digital twin analytics & optimization
  • WP4 OPC and data quality