A transient 3D model for two main zones, namely the fuel bed and the freeboard, of biomass packed bed combustion systems was developed. It integrates the models for the biomass conversion sub-processes and solves the governing equations for the gas and solid phase and their interactions. The intra-particle gradients are included by considering the biomass particles as thermally thick particles. The shrinkage of the packed bed and the variations of the bed porosity due to the uneven consumption of the fuel are taken into account. Detailed kinetic mechanisms are used for the simulation of homogeneous gas phase reactions. To verify the model and to increase the understanding of packed bed combustion, laboratory-scale fixed-bed batch experiments have been performed in a reactor with 9.5 cm diameter and 10 cm length. The model performance was extensively validated with gas phase measurements (CO, CO₂, CH₄, H₂, H₂O and O₂) above the fuel bed, temperatures at different heights in the bed and in the freeboard, and the propagation rate of reaction front. The simulation results are in a good agreement with the measured values. © 2014 Elsevier Ltd. All rights reserved.

The release of potassium during biomass combustion leads to several problems as the emissions of particle matter or formation of deposits. K release is mainly described in literature in a qualitative way and this work aims to develop a simplified model to quantitatively describe it at different stages. The proposed model has 4 reactions and 5 solid species, describing K release in 3 steps; during pyrolysis, KCl evaporation and carbonate dissociation. This release model is coupled into a single particle model and successfully validated with experiments conducted in a single particle reactor with spruce, straw and Miscanthus pellets at different temperatures. The model employs same kinetic parameters for the reactions in all cases, while different product compositions of the reactions are employed for each fuel, which is attributed to differences in composition. The proposed model correctly predicts the online release at different stages during conversion as well as the final release for each case.

Holzpellets werden in zunehmendem Maße aus Hackschnitzel hergestellt. Es ist davon auszugehen, dass diese Veränderung der Rohstoffbasis zu erhöhten Aschegehalten im Brennstoff führt und zu Schwierigkeiten bei der Nutzung führen kann. Die vorliegende Arbeit kommt zum Schluss, dass moderate Anteile (< 5%) sauberer Rinde zu keinen wesentlichen Verschlackungen führen. Bei Verunreinigung oder / und sehr hohen Rindenanteilen ist mit
Verschlackungsproblemen jedenfalls zu rechnen.

Pellet boilers and pellet stoves are widely used for heat production. But in most cases, only specific wood pellets with a low ash content are approved due to the increased risk of slagging and limited deashing capacity. The ash fusion test (AFT), according to prCEN/TS 15370-1, is currently the only standard method for the prediction of slagging. This method is not feasible for all biomass fuel types, since sometimes the characteristic temperatures cannot be determined or the characteristic shapes do not occur for temperature determination. Furthermore, the method is costly and requires complex instrumental infrastructure. Hence, a demand for more expressive or more rapid methods to characterize slag formation potential of fuels is often claimed. Based on a literature study, four such laboratory test methods were chosen, partly adapted, and then experimentally investigated. These methods included thermal treatment of the fuel itself or the ashes of the fuel and were the rapid slag test, CIEMAT, the slag analyzer, and the newly developed pellet ash and slag sieving assessment (PASSA) method. Method performance was practically assessed using 14 different biomass fuel pellets, which were mainly from different assortments of wood, but also herbaceous or other nonwoody fuels. The results from the tests with these four alternative methods were evaluated by comparing to both results from standard AFT and results from full-scale combustion tests performed over a maximum of 24 h. Seven different pellet boilers were assessed, of which one boiler was used to apply all 14 test fuels. According to the granulometric ash analysis (i.e., the ratio of >1 mm-fraction toward total ash formed), the sensitivity of the new test methods to depict slagging phenomena at a suitable level of differentiation was assessed. Satisfactory conformity of the boiler ash assessment (reference) was found for both, the slag analyzer and the PASSA method. The latter may, in particular, be seen as a promising and relatively simple low-input procedure, which can provide more real-life oriented test results for fixed-bed combustion. The standardized AFT could, however, not sufficiently predict the degree of slag actually formed in the reference boiler, particularly when only wood fuels are regarded.

A method for deriving kinetic models of gas–solid reactions for reactor and process design is presented. It is based on the nonparametric kinetics (NPK) method and resolves many of its shortcomings by applying tensor rank‐1 approximation methods. With this method, it is possible to derive kinetic models based on the general kinetic equation from any combination of experiments without additional a priori assumptions. The most notable improvements over the original method are that it is computationally much simpler and that it is not limited to two variables. Two algorithms for computing the rank‐1 approximation as well as a tailored initialization method are presented, and their performance is assessed. Formulae for the variance estimation of the solution values are derived to improve the accuracy of the model identification and to provide a tool for diagnosing the quality of the kinetic model. The methods effectiveness and performance are assessed by applying it to a simulated data set. A Matlab implementation is available as Supporting Information.

The increased biomass utilization leads to the need of an efficient and flexible usage of available sources. Therefore, it is necessary to combust low-cost biogenic residues, which inherently have higher nitrogen contents that lead to increased NOx emissions. In order to tackle this issue a new combustion technology with double air staging and flue gas recirculation is under development. The technology also features an increased fuel bed height and very low oxygen concentrations in the fuel bed to reduce fuel bed temperatures. This work focuses on the CFD simulation of the formation and reduction of NOx emissions of in a small scale boiler (35 kWth). Compared to previously applied models, major modification concerning the heat and mass transfer in the fuel bed as well as the subsequent conversion in the freeboard were made. The fuel bed is modelled via representative fuel particles with a Lagrangian approach and a thermally thick particle model considering intra-particle
gradients. Due to the increased fuel bed height and the relatively low oxygen concentration the formation and cracking of tars has to be considered in the simulation. This heavily influences the formation and reduction of NOx and its precursors. The fuel bound nitrogen is released via the particle model in the form of NO during char burnout and via a lumped tar species during pyrolysis. The cracking of the lumped tar species is modelled via two global gas phase reactions that releases the NOx precursors NH3 and HCN. The cracking reactions are added to a skeletal reaction mechanism with 28 species and 102 reactions that includes the fate of the N species. The simulation results are compared to experimental data from test runs with spruce wood chips and Miscanthus pellets as fuels. The comparison showed good agreement for the test runs with wood chips, where the temperature distribution inside the fuel bed and the released species above the fuel bed were predicted well. The test runs with Miscanthus showed a greater deviation between the measured and simulated values. For both fuels the NOx reduction that was experimentally observed in the secondary combustion zone could not be predicted with reasonable agreement. Therefore, it is necessary to further investigate the cracking of the tars and the subsequent formation of the NOx precursors. The presented work forms the basis for further improvements of the numerical models and subsequently the optimization of the new technology.

In this study, we investigated the emissions, including odor, from log wood stoves, burning wood types indigenous to mid-European countries such as Austria, Czech Republic, Hungary, Slovak Republic, Slovenia, Switzerland, as well as Baden-Württemberg and Bavaria (Germany) and South Tyrol (Italy). The investigations were performed with a modern, certified, 8 kW, manually fired log wood stove, and the results were compared to emissions from a modern 9 kW pellet stove. The examined wood types were deciduous species: black locust, black poplar, European hornbeam, European beech, pedunculate oak (also known as “common oak”), sessile oak, turkey oak and conifers: Austrian black pine, European larch, Norway spruce, Scots pine, silver fir, as well as hardwood briquettes. In addition, “garden biomass” such as pine cones, pine needles and dry leaves were burnt in the log wood stove. The pellet stove was fired with softwood pellets.

The composite average emission rates for log wood and briquettes were 2030 mg MJ⁻¹ for CO; 89 mg MJ⁻¹ for NO_x, 311 mg MJ⁻¹ for C_xH_y, 67 mg MJ⁻¹ for particulate matter PM₁₀ and average odor concentration was at 2430 OU m⁻³. CO, C_xH_y and PM₁₀ emissions from pellets combustion were lower by factors of 10, 13 and 3, while considering NO_x – comparable to the log wood emissions. Odor from pellets combustion was not detectable. CxHy and PM10 emissions from garden biomass (needles and leaves) burning were 10 times higher than for log wood, while CO and NO_x rise only slightly. Odor levels ranged from not detectable (pellets) to around 19,000 OU m⁻³ (dry leaves). The odor concentration correlated with CO, C_xH_y and PM₁₀. For log wood combustion average odor ranged from 536 OU m⁻³ for hornbeam to 5217 OU m⁻³ for fir, indicating a considerable influence of the wood type on odor concentration.