Year: 2007
Type: dissertation
Abstract: Biomass combustion can be an economic way to contribute to the reduction of CO2 emissions, which are a main suspect of the so-called greenhouse effect. In order to promote a widespread utilization of biomass combustion, operational problems like fuel treatment, slagging, fouling and corrosion have to be solved. There are few research initiatives focused on the improvement of design and operation, for example combustion modeling, material research or equipment optimization. In this work, two aspects are considered: performance improvement due to the optimization of cleaning cycles to reduce slagging and the use of predictive dynamic models as an aid for control and equipment design. These objectives can be accomplished by using process models that share a number of characteristics. The first part includes the development and application of an online process-monitoring tool to recognize deposit phenomena during biomass combustion. Slagging and fouling refers to deposits of solids on heat exchanger surfaces. Since slagging and fouling is a local process highly dependent on time and temperature, the analysis of the heat transfer diagrams is used to study the deposit tendency in an early stage of the process. In this work, heat transfer transients of the evaporator, the economizer and the superheaters are analyzed on the basis of a physical model. Measured data from a biomass-fired power plant, e.g. from an acoustic pyrometry and the process control system, were used as input for the monitoring. A thermodynamic steady state model was developed and validated with data from several boiler types, i.e. pulverized fuel, fluidized bed and grate. The monitoring campaign was combined with fuel and ash measurements. The fuel from the power plants showed clear slagging tendencies. Measured material properties were analyzed and used to improve the modelâs accuracy. Deposits on the heat exchangers could be accurately detected and the overall soot blowing strategy could be optimized, something not possible without a process-monitoring model. The model has been verified and validation of the results proved the correct diagnosis of the deposit status inside the furnace. A predictive dynamic model was developed, based on the steady state results and is presented in the second part of this thesis. The steam boiler is one of the most complex components of a thermal power plant as far as the process control is concerned. A common boiler configuration is the once-through arrangement and this is the type of boiler considered in this work. A problem in two-phase systems modeling is the correct calculation of the phase boundary, because the position of the phase transition changes rapidly, depending on load conditions and temperature distribution along the walls. The prediction of the correct spatial distribution of the phases is crucial with respect to the accuracy of the model. A lumped parameters model with a moving boundary approach is developed instead
Author: F. Ahnert
Language: English
Subfield: Fuel Technology
Field: Energy
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1959965858