Optimizing a batchcontinuous process: the Suiker Unie case

1
Optimizing a batch-continuous process the Suiker
Unie case

• Joost Smeets

2
Overview

• Project goals
• The optimisation/control problem
• Approach
• Literature
• Modelling approach
• Status

3
Project goals

• Opimisation/control of heat supply of the sugar
  house at Suiker Unie Groningen
• Application and evaluation of Model Predictive
  Control to a process consisting of continuous and
  batch-driven units
• Testing of a demonstrator operator support system
  for managing heat

4
The optimisation and control problem

• Process suffers of fluctuations because of
  varying heat demands
• Variations arise because of
• starting/stopping batch units
• Heat interactions between batch and continuous
  units
• Heat interactions between cristallisation process
  and raw materials ( juice evaporator)
• Recycles with different cycle times

5
The optimisation and control problem

• From D. Vermeulen, SU

6
Approach

• Investigation of current Energy management status
  at SU Groningen
• Which units are critical to energy management/
  project scope
• Occurrence of recycles
• Available measurements
• Available measurement data / PFD
• Development of energy model
• Use of mass-balance model developed by SU (Ad
  Muller)
• Include energy and/or dynamics in model
  (TNO-TPD/TU Delft)
• Validation of model
• Use campaign measurement data for validation
• Simulate process with measurement data and model,
  compare with measured output data

7
Approach optimisation

• Off-line optimisation
• Adjust heat supply batch-units to match
  continuous units
• Go/No-go
• Design of demonstrator for operator support at
  TNO-TPD
• Off-line testing
• Off-line simulation of system using campaign data
  
• Presentation of diagnosis/recommendations
• Implementation on-line (if possible)
• Interface to Process control system at SU
  Groningen
• Testing in upcoming campaign

8
Demonstrator Example
Product Quality Mean res. time Min res.time
(1) Mean fining index Min fining index Mean
top temp Min top temp
26.2 h 4.1 h 96 0.31 1500C 1377C
Batchcoverage Batchvolume Batch pos east
Batch pos west
47 7.1 6.5 m 7.7 m
Energie/ton
2052MJ
1404C 1216C
Temp forward flow Temp back flow
Pull Crown temp Boosting Cullet Throat east
302 t/d 1500C 1.85 MW 84 31
T bottom east T bottom west
1170C 1176C
Mean forward flow Mean back flow
5.4 kg/s 1.9 kg/s
Temp forward flow Temp back flow
1403C 1212C
Temp feeder 51 Temp feeder 52 Temp feeder 53
1218C 1231C 1189C
9
Literature

• From Urbaniec, 1989

10
Literature