Combined 1D-Fuel-Bed- and 3D-CFD-Model as Design Tool for Biomass Combustion
Authors: G. Barroso, S. Roth, T. Nussbaumer (HSLU) and R. Schmid (Schmid AG energy solutions)
Bioenergy has a significant potential to substitute fossil fuels and to reduce greenhouse gas emissions for heating, high temperature processes and electricity production. The use of biomass according to the cascade principle, first for food and goods and then the use of resulting residues and waste materials for energy generation is uncritical and therefore favoured. Today 36 PJ/a of woody biomass are energetically used in Switzerland corresponding to 3.3% of the primary energy consumption of 1108 PJ/a[i].
There is an additional sustainable potential of woody biomass in Switzerland of 14 PJ/a enabling an increase of 38%. Beside wood, agricultural residues and other non-woody biomass stocks can potentially be used for energy purposes. The remaining potential is, however, mostly related to low-grade fuels with increased moisture and ash content, which are difficult to handle due to their tendency to slag formation on the grate and due to condensation and deposition of fly ash in the furnace.
In addition, biomass combustion is related to the following air pollutant emissions: On the one hand, health relevant organic compounds (OC) as volatiles (VOC) and condensables (COC), and soot particles as carrier of organics are emitted due to incomplete combustion. These pollutants can be reduced by improved burn-out conditions. On the other hand, nitric oxide emissions (NOX) are formed from fuel-bound nitrogen at high temperature and sufficient oxygen availability. Hence there is a target conflict between OC and NOX.
Mastering the clean combustion of these difficult fuels is essential to use this potential for the energy turn around and leads to the following questions: What type of technology is today used to handle this energy source? Is there a possibility to introduce a disruptive new boiler design to deal better with these low-grade fuels? Which tools would assist industry to design furnaces capable to handle these fuels? Which physical and chemical processes are not well understood and implemented in design tools? These questions are addressed by the HSLU bioenergy research group and the team of the industrial partner Schmid AG energy solutions with financial support of the SCCER BIOSWEET, the Swiss National Science Foundation NRP 66, and the Swiss Federal Office of Energy[ii].
As a consequence, a generic model which describes the solid fuel conversion on a moving grate and the subsequent gas-phase chemistry, both considering the nitrogen chemistry, was further developed and validated. The model is applied for the optimisation of the boiler design in particular with respect to reduce both types of pollutants, i.e. OC and NOX. For the solid phase conversion a one-dimensional transient model based on a two-phase porous medium is used whereas the gas phase is simulated using CFD with reduced chemical mechanisms[iii].
Figure 1 Methodology of Combined 1D- 3D -Model
In parallel to the modelling work, a prototype moving grate boiler was further improved (e.g. new grate-elements and flue gas recirculation) and numerous operating points were measured.
Figure 2 Pyrolysis gas measurements were performed by introducing an oil-cooled lance through the door viewer of the Schmid AG energy solutions prototype furnace
To validate the simulations, a gas sampling equipment with an oil cooled lance was built to measure gas profiles above the fuel bed in a moving grate boiler. The gas analysis covers the concentration of VOC, H2O, CH4, CO, CO2, H2 and O2 as main indicators of the combustion. Figure 3 shows a comparison of calculated and measured data for the bed height and the water vapour release. The results show reasonable agreement of the simulation with the measurements.
Figure 3 Comparison of the simulated bed height (left) and the water vapour above the fuel bed (right) with measured data at part load
The validated model was used to analyse the influence of fuel properties on the fuel bed behaviour. Figure 4 shows an analysis of the influence of the moisture content of the wood chips on the fuel bed at 30% load. The moisture content was varied from 25% to 50%. The results show that an increased moisture content extends the fuel bed and leads to a delay in the pyrolysis and gasification reactions.
Figure 4 Mass fraction of dry wood for the case of wood chips with moisture content 25% (left) and 50% (right)
The combined 1D-3D-simulation was validated, with the measured flue gas emission of the 150 kW prototype moving grate boiler. The simulation shows that the largest share of the NOX emissions is released from the fuel bed and that the NOX production mainly occurs in the first half of the bed (Figure 5).
Table 1 Comparison of simulated and measured emissions at part load. CO and NOX in [mg/mn3]@13% O2
Figure 5 NOx concentration in [mg/mn3 @ 13% O2] in various planes — part load
The presented combined 1D-3D-simulation method using reduced chemistry shows that the developed method can be used as powerful design tool to analyse and improve existing furnaces and “virtually” develop new combustion processes. Consequently, the model is currently applied to develop a new screw-burner within the framework of the SCCER BIOSWEET and a Swiss Federal Office of Energy project (within the frame of an Eranet project) by HSLU and Schmid AG energy solutions. The specific target of the screw-burner is to enable the use of low-grade biomass fuels with high ash content in a size range of 100 to 300 kW thus enabling a broad application. To use experiences from a 35 kW prototype, a methodology for the scale-up of biomass combustion is being developed and implemented. Beside this, investigation and model development is continued to further improve the model predictability for a broad number of operating conditions.
[i] Thees, O., et al., Biomassenpotenziale der Schweiz für die energetische Nutzung, WSL Bericht, Heft 57, 2017
[ii] Nussbaumer, T., Final Report, Clean Technologies for Wood Combustion from 500 kW to 50 MW, https://www.aramis.admin.ch/Texte/?ProjectID=32982
[iii] Barroso, G., et al. Proceedings 25th EUBCE 17, 358–362