BIOS - Bioenergy
  YOUR PARTNER FOR ENERGY FROM BIOMASS AND ENERGY EFFICIENCY  
 

CFD Simulations

Fundamentals and overview

Computational Fluid Dynamics (CFD) is a tool increasingly used for the solution of flow related engineering problems. The applications have a broad variety: Simulations in automobile and aerospace industry, biomedical industry, electronics cooling, chemical engineering, turbo machinery, combustion, heat and power generation as well as heat and cold pipes are possible applications.

In the field of combustion CFD is applied for the optimisation of gas and oil burners as well as pulverised coal furnaces. CFD modelling of biomass grate furnaces is especially difficult due to the complexity of the combustion process in the solid biomass fuel bed on the grate, as well as due to the turbulent, reactive flow in the combustion chamber. BIOS in co-operation with researchers at the Graz University of Technology, Institute for Process and Particle Engineering, has successfully developed a CFD model which is especially designed for the development and optimisation of biomass grate furnaces. The CFD model consists of an in-house developed empirical grate combustion model, as well as of modified and lab-scale tested CFD sub-models of the CFD code FLUENT for the turbulent reactive flue gas flue flow in the combustion chamber of biomass furnaces. The applicability of the CFD models for biomass grate furnaces, as well as the reliability of simulation results were tested at pilot-scale and industrial-scale furnaces.

The CFD routines developed for biomass grate furnaces and boilers were successfully applied in various cases regarding furnace and boiler development as well as regarding the optimisation of existing biomass combustion plants from small-scale to medium and large scale furnaces and boilers.

In recent years, BIOS increasingly applied CFD simulations in the field of small-scale furnaces to support the development of prototypes. Besides pellet and woodchips fired furnaces, also wood log fired furnaces and stoves were simulated. In order to consider the instationary combustion of wood logs the empirical grate combustion model was modified and tested successfully by comparison with test runs. Here, the basic grate combustion model was adapted in order to derive time-dependent profiles of the wood log combustion process from the profiles along the grate. Using these time-dependent profiles, the composition of a virtual fuel, which consists of the fuel components C, H, N, O and water vapour converted during solid fuel combustion, can be determined for any particular time.

The development and optimisation of biomass combustion plants via CFD analysis leads to a considerable reduction of investment and operating costs by a compact furnace design, by an increased availability of the plant, by reduced emissions, as well as by reduced air and flue gas fluxes in the furnace. This can be achieved by an appropriate design of the air nozzles and flue gas injection nozzles, as well as by adjusting the shape of the combustion chamber (e.g. barriers) in order to improve the efficiency of mixing of unburnt flue gas and air, as well as the utilisation of the furnace volume.
In order to keep investment and operating costs at a low level, it is especially important for small-scale furnaces, to keep the pressure loss over the nozzles for the supply of re-circulated flue gas and secondary air at an acceptably low level. By employing CFD simulations, the pressure losses over the supply pipes and nozzles for secondary air and re-circulated flue gas, as well as over the furnace and the boiler can be calculated in detail and the influence of the mixing of secondary air on CO burnout can be cross-checked.
Furthermore, CFD analysis helps to avoid velocity and temperature peaks in certain furnace sections which are of special relevance regarding material stress and deposit formation. For dry fuels (e.g. waste wood) additional measures, like furnace cooling, might be necessary and are investigated therefore by means of CFD simulation. In addition, the influence of particle laden flue gas flow on material stress caused by erosion can be evaluated using "particle tracking calculations" and considered by material selection or prevented by appropriate modifications of the furnace geometry. Furthermore, the precipitation of coarse fly ash particles in various zones of the plant can be investigated by the calculation of the particle trajectories. Actual residence times of flue gas and particle flows are interesting concerning minimum required reaction times in various furnace and boiler sections and can be determined by means of a spatially resolved simulation of residence time distributions (particle trajectories or scalar transport equations of residence times).
Finally, CFD simulations lead to an improved understanding of the fundamental physical and chemical processes in the furnace and, therefore to a considerably improved plant design. In this context a simulation-aided plant monitoring of already erected plants can be very beneficial, since it contributes to an efficient analysis and optimisation of the plant operation (see also Plant Monitoring.

To top

CFD Broschüre – PDF#0105

pdf CFD-IZ-BIOS-Keyinformation-NOV-2015-EN.pdfCFD Key Information (1.65 MB)

 

The following fields of activities are covered by BIOS using CFD:

Design and optimisation of the geometry of biomass combustion plants

Geometry of the nozzles for the injection of secondary air and re-circulated flue gas

The design of the secondary air and flue gas nozzles is a key factor in meeting the following requirements:

  • High turbulent mixing and homogenisation of the flow across the flue gas channel.
  • Minimisation of furnace volume (investment costs).
  • Reduction of excess air and flue gas recirculation ratio (efficiency, operation costs).
  • Reduction of CO and NOx emissions.
  • Reduction of temperature peaks (fouling and slagging) and flue gas velocity peaks (material stress and erosion).

An example of a furnace geometry optimised by CFD analysis is shown in the figure below. A significant reduction of CO emissions and temperature peaks was achieved by the appropriate arrangement of the secondary air nozzles, resulting in optimised mixing conditions.

To top

Geometry of the combustion chamber

The design of the geometry of the combustion chamber is of great importance in order to fulfil the requirements already stated for the design of the secondary air and flue gas injection nozzles.
Below, exemplary figures of a Low-NOx biomass grate furnace appropriate for a broad fuel assortment (waste wood, wood chips, bark) are shown. This type of furnace was realised as a pilot-scale plant and subsequently also as a large-scale plant. The combination of vertical barriers and a staged secondary air injection leads to a highly turbulent mixing, a homogeneous flue gas distribution and a good utilisation of the secondary combustion zone. Besides a significantly reduced and simplified furnace volume, the following advantages could be achieved:

Strong reduction of CO emissions

CO profiles [v-ppm] in the symmetry plane of the furnace

Lowering of temperature peaks (fouling & slagging)

Temperature profiles [°C] in a horizontal cross-section at the level of the vertical barriers and the secondary air nozzles

Lowering of flue gas velocity peaks (erosion & material stress)

Profiles of flue gas velocity [m/s] in a horizontal cross-section at the level of the vertical barriers and the secondary air nozzles

To top

Optimised design of wood log fired stoves

The basic grate combustion model, which was developed to describe the release of the flue gas components during biomass combustion on the grate in order to achieve boundary conditions for the CFD simulation of turbulent reactive flow in the combustion chamber, was modified for wood log fired stoves, wood log furnaces and wood log fired boilers operated in discontinuous batch mode. This modified model is used to calculate time-dependent profiles of wood log combustion by a transformation of the release profiles along the grate determined with the basic grate combustion model. Using these time-dependent profiles, the composition of a virtual fuel consisting of the fuel components C, H, N, O and water vapour converted during solid biomass combustion, can be determined at any point in time.  Furthermore, a mass balance and the amount and composition of flue gas released can be calculated at any point in time during batch operation.
In order to prevent the distortion of CFD simulations by the heat storage of the stove when defining virtual steady-state operating cases, it is necessary to determine the time-dependent profile of the heat fluxes over the stove surface based on test runs. By balancing energy, two virtual steady-state operating cases can be identified, which are characterised by the heat storage of the stove, which is zero.
The following processes can be analysed with the in-house developed CFD model for wood log fired stoves:

  • Flow of combustion air and flue gas in the stove and flow of convection air in the double jacket of the stove
  • Gas phase combustion in the stove
  • Heat transfer (conduction, convection, and radiation) between gas phase, stove materials (stove lining, metal sheets and glass sheets) and surroundings

This enables analysis of:

  • Velocities and temperatures of combustion air, convection air and flue gas
  • Path lines of air and flue gas
  • Concentrations of O2 and CO in the flue gas
  • Material and surface temperatures of stove lining, metal sheets and glass sheets (see figure below)
  • Heat transfer and efficiency
  • Pressure losses over different plant zones
Temperature profile at the outer surface of a wood log stove [°C]; 3D view of the stove from frontside (left) and backside (right)

In the figure below the development of a new stove is demonstrated by means of the CO concentrations. In the basic variant the emissions are rather high due to a bypass flow in the redirection baffle of the post combustion chamber. Furthermore, the post-combustion chamber was not insulated. In the pre-optimised variant (before the realisation as testing plant) first improvements could be achieved by a closure of the bypass flow and an insulation of the post combustion chamber. By these measures, the temperature in the post-combustion chamber was elevated and the CO burnout considerably improved. A further improvement could be achieved by the optimised variant which was realised as testing plant. Here, additional tertiary air nozzles have been installed in the rear part of the combustion chamber, which lead to an improved flue gas burnout already in the combustion chamber. Moreover, the CO emissions are a leading parameter for the burnout quality of the flue gas and can be used as an important indicator concerning organic fine particle emissions from incomplete combustion. Besides the considerably reduced CO emissions also the organic fine particle emissions could be reduced. Finally, the excess air could be reduced, leading to a higher plant efficiency.

Iso-surfaces of CO concentrations [ppmv w.b.] in the flue gas in the vertical symmetry plane of a stove
Modifications: closure of opening in the redirection baffle; additional tertiary air nozzles; larger transition to the chimney and insulation of the post-combustion chamber

To top

Optimised design of wood log furnaces

The basic grate combustion model, which was developed to describe the release of the flue gas components during biomass combustion on the grate in order to achieve boundary conditions for the CFD simulation of turbulent reactive flow in the combustion chamber, was modified for wood log fired stoves, wood log furnaces and wood log fired boilers operated in discontinuous batch mode (for further information see section “Optimised design of wood log furnaces”). Here, virtual steady-state operating cases are balanced with the empirical model for which CFD simulations of flow, gas phase combustion and heat transfer can be performed for any point in time during batch operation.
The CFD simulations enable an analysis of:

  • Quality of penetration of the primary combustion zone (space filled with log wood) with primary air
  • Utilisation of the secondary combustion zone, mixing of flue gas with secondary air and CO burnout
  • Velocity and temperature peaks for the best possible reduction of material erosion and ash deposit formation
  • Heat transfer in the primary and secondary combustion zones of the furnace as well as in the boiler tubes as a basis for the optimisation of thermal efficiency
  • Pressure losses over different plant zones

Results of a CFD analysis of a wood log furnace are shown in the following figure. The diagram on the left shows path lines of the primary air coloured by gas temperature. The path lines allow the penetration of the space filled with log wood to be analysed and optimised in order to achieve good and even combustion of the wood logs and to avoid bridging. The diagrams on the right show calculated CO concentrations in different cross-sections of the secondary combustion zone, which serve as a basis to analyse and optimise the mixing of flue gas with secondary air and the utilisation of the secondary combustion zone.

Path lines of primary air in the primary combustion zone coloured by gas temperature [°C] (left); CO concentrations [ppmv] in a horizontal cross-section (top right) and in a vertical cross-section (bottom right) of the secondary combustion zone

To top

Simulation of pulverised biomass furnaces / entrained flow reactors

Models have been developed for the combustion of thermally thin and thick biomass particles in order to utilise CFD for the design of pulverised combustion and gasification units. The models for thermally thick biomass particles take the intra-particle mass and heat transfer into account and thus enable a more accurate prediction (compared to available commercial CFD particle models) of the combustion process of the solid biomass particle and particle temperature. The model for entrained flow conversion was successfully tested by a comparison with measurements for a pulverised biomass flame and successfully applied to the simulation of a pulverised wood furnace.  This model allows the qualitative description of particle combustion along particle trajectories and thus provides qualitative information about the flow and combustion processes in the furnace, making it ideal for the development and optimisation of pulverised biomass furnaces.

The simulated CO concentrations in the vertical symmetry plane of the basic design (left) and the optimised design (right) are shown in the following figure. The results show that a considerable improvement of CO burnout in the furnace can be achieved by modifying the design of the nozzles for the supply of secondary air and re-circulated flue gas as well as the operating conditions.

CO concentrations [ppmv] in the vertical symmetry plane of a pulverised wood furnace

To top

Investigation of furnace cooling systems

As mentioned before, moderate and well controlled flue gas temperatures in the furnace are important in order to prevent slagging and deposit formation. Additional measures like cooled walls or tubes are recommended especially for dry fuels (waste wood) and biomass fuels with a high content of alkali metals (straw).

The figure below shows the temperature distribution in a biomass grate furnace for waste wood combustion with path lines of re-circulated flue gas injected by the lower nozzle row, indicating high turbulent mixing and increased flue gas temperatures. The highest flue gas temperatures near the wall are expected in the primary combustion zone and the regions around the flue gas and secondary air injection nozzles. Cooled furnace walls are recommended for this region in order to lower temperature peaks and to prevent slagging.

Temperature profiles [°C] in different horizontal cross-sections at the level of re-circulated flue gas and secondary air injection; the furnace wall section between and around the nozzles is cooled - Explanations: SA…secondary air nozzles; FGR…flue gas re-circulation nozzles

To top

CFD simulation of boilers including the convective section

CFD simulation also proved to be a powerful tool for the design of biomass boilers. Optimisation of flow results in an improved utilisation of the boiler volume, enhanced heat transfer and a more even temperature distribution, thereby also reducing deposit formation.

In view of this potential, BIOS BIOENERGIESYSTEME GmbH has carried out an R&D project in order to develop a CFD model for heat exchangers. This model allows the flue gas flow within the tube bundles of the convective heat exchanger section to be included in the CFD optimisation process. A detailed simulation of convective heat exchangers would be impossible in most cases, as the high spatial resolution needed to resolve the geometry of interest could not be covered by computer capacity.

The simulation models developed are able to predict the flue gas flow field including the pressure losses, as well as the heat transfer within the convective part of the boiler. Furthermore, the maximum temperatures at the flue gas side of the heat exchanger tubes are calculated, which are the crucial factor in the process of formation of solid deposit layers. Special attention is paid to the influence of radiative heat transfer on tube rows exposed to increased thermal radiation.
Consequently, the CFD simulation gives better and more valuable information for tube bundles in highly inhomogeneous flow fields, e.g. tube bundles positioned in regions of strong flow deflections, than conventional heat exchanger design methods based on one-dimensional assumptions. The following advantages for the optimisation of biomass boilers can be summarised:

  • Optimisation of flow through the heat exchanger tube bundles
  • Decision basis for arranging the evaporator tube bundles upstream the superheater sections
  • Identification of surfaces prone to formation of sticky ash deposit layers
  • Optimised positioning of soot blowers to improve the functionality of the boiler cleaning system
  • Optimisation of steam parameters to increase the efficiency of electricity production

The CFD model developed expands the capabilities of the present state-of-the-art commercial CFD software and is currently available for the most relevant boiler types (water tube steam boilers, thermal oil and fire tube boilers). Additionally, the model is able to simulate the flow of both primary and secondary heat carriers. It can thus be used to optimise heat transfer and flow field for both flue gas and water side in a fire tube boiler. Exemplary simulation results for different biomass boilers are shown in the following.

 

a) Flue gas temperature distribution [K] and (b) maximum temperature [K] at the flue gas oriented surfaces of fouled heat exchanger tube bundles in the convective section of a biomass fired steam boiler.

The figures above show CFD results of a real plant, which was selected for validating the CFD models developed by BIOS BIOENERGIESYSTEME GmbH. The inhomogeneous inlet flow into the convective part of the boiler leads to a strong fluctuation of the maximum tube surface temperatures at the first tube rows of the evaporator section. Deflections of the flue gas flow in front of the tube bundle heat exchangers can be optimised by means of CFD simulations in order to achieve a more even distribution of the incoming flow and a reduction of temperature peaks.

Temperature profiles [°C] in different horizontal cross sections of a thermal oil-boiler (radiative section)
Profiles of the flue gas side wall temperature [°C] of the radiative heating surfaces of a thermal oil boiler at different stages of fouling from clean walls (left) to walls covered with a considerable layer of fly ash deposits (right)

The figures above show simulation results of flue gas temperature profiles in the radiative section of a thermal oil boiler as well as simulation results based on an investigation of the effective heat transfer in the radiative section of the boiler under real operating conditions. The increased wall temperatures due to deposit layer formation lead to a dramatic reduction of the heat transfer in the radiative boiler section. This should be considered in boiler dimensioning and should be prevented as far as possible by an appropriate design and by appropriate automatic cleaning facilities.

Computational grid (above) and flue gas temperature profiles [°C] in the symmetry plane of a fire tube boiler; a) simulation of the convective section using the CFD heat exchanger model; b) detailed simulation of the convective section with spatially resolved tubes.

The figure above shows the results of a validation study of the developed CFD heat exchanger model for fire tube boilers. A comparison of a detailed boiler simulation with a simulation using the CFD heat exchanger model was performed for this study. The detailed simulation, used as a reference, was carried out with spatially resolved tubes (convective section). However, this is only possible for very small boilers. In the simulation using the CFD heat exchanger model the tubes were considered by the model (compare figures above). The comparison shows good agreement of the simulation results for both methods (see flue gas temperature shown). An additional comparison with experimental data of pressure losses and heat transfer in the boiler also showed good agreement of calculated and experimental results. The CFD heat exchanger model can thus be applied as a design tool for fire tube boilers.

For further details see:

To top

Simulation of ash deposit and fine particulate formation in biomass furnaces and boilers

It is of the utmost importance to avoid the formation of deposits in biomass furnaces and boilers. Furthermore, the reduction of fine particulate emissions is gaining increasing importance due to continuously stricter emission limits and an increasing market demand concerning „new“ biomass fuels with enhanced ash contents like short rotation coppice and agricultural residues.   A model accounting for deposit and fine particulate formation is currently under development in a R&D project. At the present stage of CFD modelling, the time-dependent formation of deposits at furnace walls and boiler walls can be predicted. The model currently considers the impaction of fly ash particles (silicate and salt particles) depending on their stickiness, as well as the condensation of ash forming vapours at furnace and boiler walls. Furthermore, the formation of fine particulates (basic model) and their deposition on furnace and boiler walls can be investigated. In addition, the erosion of the deposition layer by coarse fly ash particles can be studied. This model is characterised by high flexibility regarding the biomass fuel used, a detailed consideration of the ash chemistry and reasonable computing time even if applied for engineering applications.

Calculated thickness (mm) of deposit layer in furnace and radiative boiler section (fire tube boiler) of a biomass grate furnace (fuel: waste wood) after 1 hour of operation
Calculated deposit mass flux (kg/m2s) of coarse fly ash particles to the walls of furnace and radiative boiler section (fire tube boiler) of a biomass grate furnace (fuel: waste wood) after 1 hour of operation
Calculated condensation mass flux (kg/m2s) of K2SO4 to the walls of furnace and radiative boiler section (fire tube boiler) of a biomass grate furnace (fuel: waste wood) after 1 hour of operation

The figures above exemplify the deposit build-up calculated for a waste wood fired grate furnace including the radiative section of the fire tube boiler. The separate visualisation of the different deposit formation mechanisms (e. g. deposit build-up by impaction of coarse fly ash particles or by condensation of different ash forming vapours (e. g. K2SO4)) can be mentioned as a major advantage of this deposit formation model. The simulation results agree qualitatively with observations made in the biomass grate furnace investigated as well as in a number of other biomass fired boilers. The deposit formation model is currently being validated by comparing the simulation results with experimental data gained from deposit probe measurements performed at a biomass grate furnace. An enhanced model for the prediction of ash deposition formation on convective heat transfer surfaces was developed in cooperation with ANSYS Germany GmbH. The model was based on the CFD heat exchanger model and the deposit formation model described above. Additionally, models for a more accurate description of the release of ash-forming vapours during biomass combustion on the grate as well as the formation of fine particulates (regarding the distribution of particle sizes) and their influence on ash deposit formation will be developed. The extension of deposition formation modelling to convective heat exchanger sections is of special importance, since ash deposit formation causes problems in these regions especially if they are exposed to high temperatures (e. g. superheaters).

For further information see:

Total fine particle concentrations [mg/Nm³ dry flue gas, 13% O2); (left) and chemical composition of the fine particles (right)
Explanations: 70 kW pellet boiler; 1…first particle formation; 2…particle formation starts to dominate at the entrance into the heat exchanger; PCZ…primary combustion zone; SCZ…secondary combustion zone

In the figure below the simulation results concerning fine particulate formation in a 70 kW pellet boiler are shown. In the PCZ, the flue gas temperature and the wall temperatures are too high for a direct wall condensation or fine particle formation. The first formation was predicted at the exit of the PCZ. Simultaneously to the fine particle formation, condensation occurs at the cooled walls. The highest deposition mass fluxes have been calculated on the opposite side of the SCZ. In this region, mass transfer coefficients as well as concentration gradients at the wall are high in comparison to the other regions. In the heat exchanger, the condensation flux strongly decreases and the formation of fine particles dominates. For the purpose of model check, the simulation results have been compared with measurements during test runs. The predicted fine particle emissions are in good agreement with the measurement values (simulated: 9.92 mg/Nm³; measured: 7.65 mg/Nm³). Moreover, the predicted chemical composition of the fine particulate emissions is in good agreement with results from chemical analyses. Concluding, the selected results of a number of validation simulations showed that already at the present state of development the model is able to predict fine particulate formation and emissions even at a quantitatively acceptable level. Hence, the model under development can be already applied as efficient tool for the development of new low-dust combustion technologies since it predicts local fine particle formation in dependence of relevant influencing parameters and thus leads to a better and deepened understanding of the underlying processes.

Furthermore a model for the CFD simulation of ash deposition processes on convective heat-exchange surfaces is available. This enables the consideration of the release of ash-forming vapors during biomass combustion on the grate, as well as the formation of fine particulates (regarding the distribution of particle sizes) and their influence on ash deposit formation. Deposition processes are of special importance since ash deposit formation causes problems on the surface of convective heat exchanger tube bundles (e.g. superheater) especially if they are exposed to high temperatures.

For further information see:

To top

CFD-simulation of NOx formation in biomass furnaces

The reduction of NOx emissions is a very important issue due to ever stricter emission limits. In order to meet these requirements an extensive amount of R&D work has already been performed. This work focuses on the implementation of N-release functionalities and a NOx formation model in the CFD routines of BIOS in order to develop an efficient design and prediction tool. This model consists of:

  • Extension of the empirical grate combustion model to include the most relevant NOx precursor species NO, NH3 and HCN in the empirical combustion model
  • Eddy Dissipation Concept for turbulence / chemistry interactions
  • Detailed chemical kinetics (Kilpinen 92) and reduced kinetics (Kilpinen 97-skeletal)
  • ISAT (in-situ adaptive tabulation) algorithm for run-time tabulation of the reaction kinetics (reduction of CPU time)
Simulated mole fraction profiles of NH3 (right) and NO (left) in the symmetry plane of a pilot-scale biomass grate furnace and comparison of measured and simulated NOx emissions at boiler outlet

The figure above shows the comparison of measurements and simulations for the investigated biomass grate furnace with air staging (fuel investigated: fibreboard) concerning NO (main NOx component in a biomass furnace) and NH3 (usually the most important NOx precursor in a biomass furnace). Very good qualitative and quantitative agreement was obtained between the measured and simulation results regarding NOx as well as the precursors NH3 and HCN for two different operating conditions of the furnace (oxygen-lean and oxygen-rich conditions in the primary combustion zone). In order to save CPU time, a reduced mechanism was also applied and validated for lab-scale flames and for fixed bed biomass grate furnaces of different scale. Good qualitative agreement between simulated and measured NOx emissions was achieved for all applications with lower CPU time. Hence, the NOx formation model can be applied to simulations of the performance of sensitivity analyses concerning the influence of furnace geometry as well as plant operation and air staging on NOx formation.

For further information see:

Iso-surfaces of NOx concentrations [ppmv w.b.] (left) and of the local TFN/TFNin ratio [-] (right) in a vertical cross-section through the axis of a 20 kW underfeed multi-fuel boiler at nominal load

Explanations: fuel: straw; fuel-N = 0.54 wt.% d.b.; λtotal= 1.71; λprim = 0.69; TFN (total fixed nitrogen): sum of all moles of nitrogen contained in NO, NH3, NO2, HCN and N2O; TFN/TFNin: TFN in the flue gas related to TFN released in the fuel bed and introduced via the recycled flue gas (TFNin)

 

The figure above exemplifies the simulation results for a 100 kW multifuel furnace. It is shown that the formation and reduction of nitrogen oxides mainly takes place in the PCZ and in the region of the secondary air nozzles. Looking at the TFN/TFNin ratios it is possible to identify the regions of NOx reduction (the smaller the ratio, the more NOx precursors (HCN, NH3 and NOx) are reduced to N2). In the example shown below it can be seen, that NOx is mainly formed in the outer, oxygen-rich region of the PCZ and that it is reduced in the inner, oxygen-lean region of the PCZ.

To top

Simulation of cyclones, particle separators and filters

The range of simulation services of BIOS also includes the CFD-based design and optimisation of cyclones, particle separators and filters. In this case the simulations cover flow and temperature distributions, the calculation of particle separation rates as well as of particle impaction and erosion rates on the walls.
The following figure shows simulation results for a cyclone of a circulating fluidised bed combustion plant. The comparison of erosion simulation results with observations (photo) showed good qualitative agreement of the calculated zones of high erosion rates with locations at the plant where strong erosion was observed.

Erosion rates observed at the wall of a circulating fluidised bed furnace (left) qualitatively compared with calculated erosion rates (right)

Present enhancements and future objectives of CFD simulations

Gas phase combustion

Future developments should include models for the consideration of gas streaks caused by channelling in the fuel bed. Furthermore, the models for gas phase combustion should be adapted to low-turbulence zones in biomass furnaces (especially relevant for small-scale furnaces), in order to improve the prediction accuracy.

Corrosion in biomass fired boilers

Material corrosion of steel surfaces in biomass combustion and boiler plants is of major importance especially when firing biomass fuels with high contents of chlorine, sulphur and alkali metals (waste wood as well as agricultural fuels) but also for conventional wood fuels (wood chips, bark) with respect to increasing the steam parameters and thus the efficiency of future biomass CHP plants. Therefore, within the framework of the large FFG project (BioCorrSim), basic models for the prediction of the local corrosion potential in biomass fired combustion and boiler plants in dependence of relevant influencing parameters were developed. The simpler approach was an empirical model, which described the corrosion potential in dependence of relevant influencing parameters like the molar 2S/Cl ratio, flue gas temperature and surface temperature. The second and more sophisticated CFD-based model considered transport processes and chemical reactions between the steel surface, surrounding deposit layer and gas phase for the most relevant high temperature corrosion processes in biomass combustion plants. Both corrosion potential models were linked with an existing and comprehensive CFD based deposit formation model, which provided the local values of e.g. flue gas temperatures and species concentrations as input values for the corrosion potential models and further allowed for a 3D simulation of the local corrosion potential in dependence of the influencing parameters. These new models enabled the 3D simulation of the local corrosion potential in the plant in dependence of influencing values like fuel and furnace temperature.

Solid biomass conversion in fixed beds

As a result of the great complexity of the thermo-chemical processes involved in solid biomass conversion in fixed beds, no CFD models are commercially available at present that allow a spatially resolved process analysis. Thus, a grate design taking into consideration the supply of primary air and re-circulated flue gas below the grate is not possible. The single particle model for entrained flow reactors (furnaces and gasifiers), which is currently under development, will therefore be integrated in a 3D CFD model for fixed bed furnaces and gasifiers.
While CFD simulations are successfully being applied for the simulation of flow and gas phase combustion in biomass grate furnaces no engineering model for the 3D simulation of solid fuel combustion on the grate is available so far. Therefore, within the framework of several R&D projects, models for solid biomass conversion on the grate are being developed and linked with the gas phase CFD models. The basic model already developed is based a 2-step approach. In a first step the movement of the particles in the packed bed is described with a non-reacting multi-flow simulation (Euler-Granular Model). In a second step, the conversion of the particles along their trajectories is calculated with a layer model for thermally thick particles with temperature gradients inside which is embedded in a Lagrange model. This enables the 3D simulation of the processes during solid biomass combustion in dependence of relevant influencing parameters for the first time.

Iso-surfaces of flue gas temperature in the furnace axis [K] (left); particle trajectories coloured by particle temperature [°C] – top view (middle); photo of the fuel bed taken from the top of the furnace (right)
Explanations 20kW underfeed stoker pellet furnace

In the picture above the simulation results (flue gas temperatures and particle tracks coloured by particle temperatures) of a 20 kW pellet furnace as well as a photo of the fuel bed taken from above are depicted for the purpose of a qualitative comparison. The calculated peak temperatures of the fuel particles are in good qualitative agreement with measurements at a lab-scale packed-bed reactor. Moreover, the calculated carbon content of the ash is in the same range as empirical values from different test runs.

In a next step, a model based on the Discrete Element Method will be implemented in the simulation routines for the purpose of a more accurate description of movement and heat transfer of the particles in the packed bed and the linked processes of solid fuel conversion. Furthermore, release models for nitrogen species and ash forming elements will be implemented in order to simulate the influence of solid fuel combustion on NOx and fine particulate formation.

To top

Selected references - plant design

Small and medium-scale plants:

Moreover, BIOS achieved for its contribution "CFD ANALYSIS OF AIR STAGING AND FLUE GAS RECIRCULATION IN BIOMASS GRATE FURNACES" the poster award for the best scientific work in the section "Biomass Production and Utilisation R&D - Combustion" at the 1st World Conference and Exhibition on Biomass for Energy and Industry, June 2000, Sevilla, Spain.

For further references concerning CFD-based plant development see:
http://www.bios-bioenergy.at/en/references.html