EmiL - Emission Limited Biomass Combustion
Austria is one of the pioneers in the development and commissioning of biomass boiler systems and makes an environmentally friendly contribution to district heating. However, with the size of the biomass boiler plants, the requirements for combustion control and flue gas cleaning also increase. The EmiL project creates the prerequisites for efficient and low-emission biomass boilers on the basis of fundamental studies.
The goal of low-emission biomass combustion is to be achieved by combining primary measures in the area of combustion sensors and combustion control as well as secondary measures in the area of cost-efficient particulate matter separation technology. The desired project results are developed in close cooperation between the research institute, university and boiler manufacturers.
The research area Control Engineering and Process Automation investigates the modeling of a model-predictive control for combustion optimization, which is tested in experimental tests.
The methodology for researching the project results is based on experimental investigations on test benches and in real plant operation. An innovative model predictive control is developed, which is experimentally investigated by implementation in a LabVIEW® environment via an interface to the boiler control. In the field of fine dust separation, the integration of fine dust separators in the boiler body is being tested on the basis of CFD simulations and experimental fundamental investigations on a test vehicle.
In the first step, a MIMO model of the boiler was created on the basis of combustion tests on the boiler test bench. The modeling is based on the underlying physical principles, while the thermochemical reactions in the boiler were approximated using so-called gray box models. Unknown parameters and non-measurable quantities could be determined from existing measurement data with the help of non-linear optimization methods.
The non-linear boiler model was then subdivided into favorable operating points using a so-called “gap-metric”, a measure of the difference between two transfer functions, around which a linearization was carried out in a further step. The combination of the above methods makes it possible to implement a linear, model-based state controller on the test bench that covers the entire working range of the boiler. In particular, the previously mentioned model-predictive controller was selected, which, in addition to its predictive property, also elegantly takes into account manipulated variable limitations without the need for classic measures such as anti-windup. An important feedback variable for the control is the residual oxygen content in the exhaust gas, which is determined using lambda sensors.
Interested in more information?
Please visit the project video on this exciting topic!
More similiar projects of our Institute
April 2016 - September 2020