Energy from Waste (EfW) facilities are almost exclusively based on grate fire technology. Grate combustion technology is often considered a robust technology which is insensitive to disturbances and quality changes in the feeding of combustible matter to the furnace. This ability to handle temporary disturbances comes partly from the fact that the feeding of combustible matter and the combustion process are significantly detached in time. As there is a significant amount of burning maintained in the bed and this bed inventory may correspond to half an hour of full load operation the impact of short term upset in fuel feed does not necessarily impact the combustion performance. The flip side to this resilience to feed disturbances is that if a change or other disturbances in the feed does occur this is not necessarily registered until after a significant time has lapsed, after which the original disturbance may or may not have already passed.

A key aspect of accurately controlling a grate EfW boiler and achieving optimized performance is to ensure that the bed is maintained in an optimal state. If too little combustible material is fed to the bed, the bed inventory will decrease, and it will not be possible to maintain maximum load. If too much material is fed the bed inventory will increase, and the bed may become excessively long and lead to improper burnout and unburned material in the bottom ash.

One special disturbance situation may come from a sudden increase in feed density coupled with increased moisture level; as this material may not ignite as quickly it may lead to a temporary reduction in boiler load, which in turn, depending on how the boiler load controller is set-up, may lead to the loaf controller reacting my feeding even more fuel. If too much dense material is fed onto the grate it may lead to a situation where the grate is partly choked, and the primary air does not properly penetrate the bed and boiler maximum load cannot be maintained. By correctly and quickly identifying situation where there’s a risk of overfilling the grate it is possible to take corrective actions by adjusting primary airs, changing grate speed, and possibly temporarily limit the feed rate to correspond to what the grate can handle.

As a visual aid to plant operators there is typically a camera installed in the back end of the grate, providing a visual input for the operators to evaluate the current status of the bed and determine e.g. the burnout position. These cameras do not however provide numerical outputs or any quantifiable information of the process state, making a systematic follow-up of the connection between visually available information and the measurable process performance almost impossible. By applying image processing it is possible to extract numerical values for the key items that a human operator may use to determine the process status, and additionally feed this data to the control system where this numerical data can be incorporated into an more advanced control structure. These key items may include burnout position and flame intensity, both as average values as well as separate values for each grate section.

In order to achieve stable optimised performance in grate fired unit it is important to continuously ensure that the correct amount of combustible material is being fed in order to maintain long term load and bed inventory, while grate speed and combustion airs are controlled to make shorter term load adjustments and optimize e.g. emission performance. To handle the time delay between certain control action and process response it is useful to incorporate a predictive control scheme where the near future development is estimated based on process values and control actions in the near past. The feed of combustible material should in most cases not be directly connected to maintaining boiler load but focussed on maintaining the bed inventory in such a state that other load control features such as primary air and grate speed are constantly in a position to make quick and accurate process adjustments whenever necessary.

An advanced process control scheme may be set up solely utilizing conventionally measured process data. On top of this a furnace camera and data available from image processing can however significantly improve the possibilities to achieve a significantly improved operation, as detailed online insight in the bed status may not be available from other sources. The typical expectation of a successfully implemented performance optimization project is an load stability improvement with a 50% reduced variability, which in turn may be translated to improved emission performance and combustion efficiency, and a throughput increase of 5-10%.

Valmet are sponsoring Operational Optimisation 2022 on 27-28 April at the Conference Aston in Birmingham. Matts Almark, Production Manager, Valmet Automation will be taking part in a session to share his vast expertise about 'Innovation developments - Performance optimisation using Advanced Process Controls & Furnace Imaging'.

For more information and to book your place at Operational Optimisation visit opopworkshop.com