Combustion behaviour of clean fuels in power generation ('BIO FLAM')

A Computer Controlled Scanning Electron Microscopy (CCSEM) procedure has been developed to determine the mineral speciation of coal-biomass mixtures, as a basis for ash formation and ash deposition predictions. For coal, such procedures are existent and well known. CCSEM for coals is used to determine the Mineral Size Distribution (MSD), the output of such an analysis being the normalised mass distribution of some 25 mineral types, divided over particle size bins of, e.g., 2-4, 4-8, 8-16, 16-32, 32-64 and 64-128µm. Many biomass materials also contain, specific, typically calcium or silicon-based bio-minerals. Agricultural (harvested) or waste biomass materials are often contaminated with sand and clay particles. Since the biominerals and the external contamination may both significantly contribute to the total inorganic matter, a new CCSEM procedure was developed to accommodate the analysis of biomass materials.

Work package 4 of the BioFlam project includes the establishment of a web based dissemination system for the BioFlam Results. The IFRF Research Station BV has commissioned the IFRF Communications Centre to set up a group of BioFlam web pages within its existing web site (www.ifrf.net). These pages are accessible through the main IFRF home page, or directly via the URL http://www.powerflam.ifrf.net/bioflam. The BioFlam home page is visible to the general public, including a general description of the project and the titles of reports and publications linked to it. Access to the content of the reports is currently limited to BioFlam partners only. A process of wider dissemination will be rolled out commencing with publication of the IFRF Research Station reports to the IFRF Members throughout the EU and the RoW who co-sponsored the IFRF RS work. At a later phase, additional reports will be opened up to IFRF Members, and in timescales to be agreed by the partners, to the EU and then the world Power Generation community. The IFRF Communication Centre will also look into the processing of the project's Excel based fuel-properties data-sheets into a searchable on-line database using the existing facility of the IFRF's on-line Combustion Handbook. Although the existence of this data will be visible to the public, access to the downloadable Excel files will initially by confined to the BioFlam Partners, with a broader roll-out of accessibility to IFRF Members and the EU/World Power Generation sector to follow that of the corresponding reports on the BioFlam web site.

Fuels which were ground separately and together with coal at different shares were ashed at 815°C, 550°C and analysed towards water, volatile, CHNS, Cl and ash components by XRF (550°C ash and 815°C ash). Analysis of the ash fusion behaviour (AFT) was done. The influence of the ashing temperatures, comparing the ash composition and ash fusion behaviour and also the connection of the fuel composition with the ash fusion characteristics is evaluated. To investigate the impact of the ashing conditions the pure fuels brown coal, paper sludge, mushroom substrate and cacao shells were ashed in a low temperature plasma asher (LTA). This ash and the ashes produced at 550°C and 815°C were analysed by XRF and XRD. Co-combustion tests at an electrical heated drop tube (BTS) and a 0.5MWth pulverised fuel facility were performed. Combustion and emission behaviour was monitored. Fly ashes of different size fractions and deposits on cooled and un-cooled probes were taken and analysed by TGA, CHNS, XRF, ash fusion and SEM-EDX.

Optical fibre flame monitoring system based on analysis of flame pulsation parameters has been developed. The system has been tested in laboratory stand (0.5 MWt) for coal combustion equipped with scaled down (1:10) low-NOx burner. Obtained results show that developed optical fibre system can be applied for characterisation of the ignition and combustion behaviour of different pulverised fuel mixtures Properties of pulverised fuel, like LHV, volatile content, reactivity, particle size distribution, and other factors influence behaviour of the flame produced in the combustion chamber downstream the model burner. In particular, addition of secondary fuel to the coal affects flame behaviour. Different pulverised fuels fired in similar turbulent flow conditions produce flames of different brightness, length, pulsation characteristics, and stability limits. The optical signals from the flame can be recorded by optical fibre system and processed, in order to obtain numerical values, which can be used for flame characterisation. Such approach can also answer the question whether certain amount of secondary fuel mixed with coal substantially changes flame parameters. Any new fuel ore mixture of the fuels can be investigated and compared with already investigated one.

The result is the parameters and procedure for the application of the CPD (Chemical Percolation Devolatilisation) model to describe the devolatilisation of biomass materials. The CPD model was developed in the USA and is one of the three more fundamental models to describe the pyrolisis process. The result of the Bioflam project is the adaptation of this model for the case of biomass materials. The model is used for the three main biomass components e.g. Lignin, cellulose and hemi-cellulose. Based on models for the structure of these components, parameters were derived for the CPD model and these are made available in the open literature for use by other researchers and engineers.

The IFRF Pilot Scale solid fuel milling facility was used to mill a number of raw coals and biofuel blends to a typical industrial particle size specification. The results indicate that there are no major problems with milling coal in this facility or similar industrial scale mills up to secondary fuel additions of 10% by weight. It was found that mill settings needed to be modified from those used for raw coal in order to deliver the same size distribution when milling blends. Data on a small number of coals and blends is available to BioFlam partners and IFRF Members. As a result of the BioFlam work, the IFRF solid fuel milling facility is now able to offer a grinding service for coals and coal secondary fuel blends to meet the particle size requirements of potential clients in laboratory and pilot scale quantities.

A numerical code calculating the burnout time of a single secondary fuel particle was developed. Combustion modelling was performed by numerical calculation of energy and mass balances on the particle. Kinetic data for pyrolysis and char combustion stages, as determined by thermogravimetric analysis, were used. The accrued results and conclusions hold for all secondary fuels studied, while the predictions of the simulation model were in agreement with the results of other researchers efforts. The effect of uncertain parameters on model predictions was investigated and it was proved that the particle diameter, moisture content, oxygen concentration and especially the gas temperature have the greater influence on the calculated results. Combined use of the developed combustion model with thermogravimetric studies provides a comprehensive view of biomass combustion behaviour and this methodology can be incorporated in advanced computational tools for the design and operation of large scale fluidised bed combustion installations.

Two water-cooled sampling probes - 3.5m an 4.5m long - were designed and manufactured. The first test of the probes were performed during full scale co-combustion experiments at Lagisza power plant. One sampling probe enables two functions, depending on applied exchangeable probes tip: - Deposit collection at given deposition surface temperatures - Measurement of flue gas composition and temperature. Temperature on the deposition surface of the first tip can be controlled in range of 500 degrees Celsius to 800 degrees Celsius by cooling air. The second tip enables simultaneous measurement of gas temperature with suction pyrometer and flue gas composition. Diameter and weight of the probe as well as water supplying hoses have been minimised in order to easier operation at power plants. An experience from the power plant tests has been used for improvement of the probes and preparing their final version.

A new method has been developed to determine the particle size and shape distribution (PSSD) of milled mixtures of coal and biomass. The outcome of the method (measured PSSD) can be utilised to evaluate grindability of heterogeneous fuel blends, or as input for Computational Fluid Dynamic (CFD) calculation of particle trajectories and combustion. The latter is specifically desired to describe the deviating behaviour of large, non-isotropic biomass particles, and the influence thereof on e.g. furnace heat distribution. A special sample preparation procedure was developed to be able to use the sample with an optical microscope; particle recognition software has been applied to determine the particle size distribution and particle morphology (shape factor). The method is useful to obtain data on the size and shape properties of especially the larger size fractions, which is where biomass fuel particles are expected to differ mostly from pulverised coal particles.

A small-scale ash deposition test has been developed as a low-cost tool for the screening of potential fuels for co-firing with coal. Deposition of ash can become a serious problem in coal fired boilers when co-firing secondary fuels at a high rate (more than a few percent). Due to nonlinear interactions between fuels, the composition and properties of the ash particles formed cannot be reliably predicted. As a consequence, the temperature range in which the ash particles soften and eventually melt, is not well known. Full-scale experience is scarce and the evaluation of ash deposition in short-term trials is not simple. A small-scale test can be performed under well known conditions to compare the ash deposition rate for a fuel blend containing a secondary fuel with the behaviour of a known base fuel. In addition, any samples collected can be analysed to evaluate the deposit in terms of attachment, sintering (relevant for strength) or composition. Samples can also be taken off-line for subsequent deposit-induced corrosion testing. Tests can be carried out to simulate typical slagging (uncooled refractory surface) as well as fouling (cooled metal alloy) conditions. In line with this (comparative) deposition test, the method was extended with an on-line measurement of heat flux through the ash deposit. The principles of this measurement formed the basis for the design of a probe with integrated heat flux sensors.

Three groups of secondary fuels and coals were used during the experiments. Proximate and ultimate analyses, determination of heating values and ash characterisation were carried out. The devolatilisation and char combustion experiments were realised in a Thermogravimetric Analyser by Perkin Elmer (precision of temperature measurement +/-2 degrees Celsius, microbalance sensitivity <5μg). Char samples were produced during the devolatilisation experiments in a fixed bed reactor. Based on the experimental results, the kinetic studies for both stages of the combustion process were accomplished. The fuel blends were tested under the same experimental conditions, i.e. particle size less than 250μm and low heating rate of 10degrees Celsius/min. The results showed that the main characteristics of the different secondary fuels were dependent on their chemical composition and especially the distribution of cellulose, hemicellulose and lignin. The additive properties of the separate materials were valid when investigating the pyrolytic behavior of the fuel blends. Some interaction occurred in the high temperature region when testing the German lignite / paper sludge blend and this was attributed to the mineral matter effect. The experiments were replicated at least twice to determine their reproducibility, which was found to be very good.

NSPCM (Non Spherical Particle Combustion Model) has been developed and tested. The NSPCM can be used for modelling combustion of non-spherical particles of materials like biomass. The model is based on well-known coal combustion model, describing spherical particle combustion. Two modifications are implemented in order to include non-spherical shape of the particles: - Reaction surface is multiplied by the particle shape coefficient PSC. - Real Nusselt number is used for oxygen iffusion to the particle instead of Nu=2 used in the coal model. The model assure accurate results in full range of particle burnout when the PSC is corrected during combustion. Hyperbolic function with exponent parameter p has been proposed for description of changes of the PSC. The computational cost of the model application is low due to a fact that particle is considered as a sphere and necessary model parameters PSC, Nu, and p are determined in advance before basic calculation.

The reference secondary fuels that were examined during the round robin test were wood pellets, paper sludge and waste wood. Their behaviour in the devolatilisation and char oxidation stages was investigated either alone or in combination with coal. The kinetics of biomass and biomass/coal blends for both stages were determined in a relatively simple and straightforward manner, through the thermogravimetric analysis (TGA) and based on multi-parallel reaction models. Knowledge of this information provides useful data for the combustion process control. Among the main conclusions was that biomass addition in the fuel blend increases the devolatilisation rate and the char reactivity. The behaviour during pyrolysis and char oxidation of waste materials with high ash and moisture content, such as paper sludge, was significantly different. Future research activities should focus on the combustion behaviour of other waste species and the mineral matter effect on the process evolution.

In line with the European energy related goals and accounting for environmental and socio-economic aspects the thermal utilisation of recovered fuels in pulverised fuel (pf) fired power plants is a shortly available way to achieve the White Paper target of 6Mtoe of biomass fuels being used in co-firing plants. Recovered fuels are inhomogeneous fractions with a large share of CO2-neutral or less carbon intensive amounts. Considering this the utilisation of recovered fuels offers a CO2 reduction potential of up to 90% in comparison to carbon intensive fuel such as coal. The available potential of recovered fuels is suitable to replace a share of coal in existing coal-fired power stations shortly. Currently most of the waste materials in Europe are deposited on landfills (efficiency 0%). Consumption of land, long-term reactions on the landfills and emission of green house gases (GHG) have a negative influence on the living conditions in Europe. This mirrored the need of a sustainable modern waste management system encouraging reduction, reuse, recycling and recovery measures. The state-of-the-art alternative is the thermal treatment of waste in a waste incineration plant at costs of 90Euros/t at best. The cost savings of the proposed co-combustion against waste incineration are above 70%. A further benefit of the co-combustion approach is that the energy content of the recovered fuels will be transformed into electricity in pulverised fuel fired power plant at high efficiencies (> 40%) compared to about 20% in a waste incineration and 0% in the landfill. This reduces CO2 emissions in electricity production. The status report based on information of the project partners on publications and information available by internet. The report include information about 16 European countries and focus on country specific criteria, laws and experience by pf co-combustion of recovered fuels. The status report could only give an extraction of the European co-combustion activities. This depends on the confidentially of experience and mirrored the competition on the energy market.

A small-scale test has been developed for the a priori, low-cost screening of potential fuels with respect to their impact on the flash produced when co-fired with coal. Economic considerations with respect to the potential of a secondary fuel should include, in addition to fuel price, the impact it may have on the economic value of any residues produced. Since the quality of a fly ash can only be evaluated once it has been produced, reliable economic considerations can be made only after a large-scale conformity test. A small-scale test would allow for a much wider screening of potential fuel candidates against low cost. And, in addition, various fuel blends can be tested easily in an attempt to identify the better fuel combinations. A small-scale combustion facility was modified to deliver relevant fly ash samples for subsequent off-line analysis and testing. The test addresses criteria as laid down in a Dutch guideline CUR 70 (later replaced by CUR 83 and 94) which follows EN-450 for the application of fly ash in concrete and has additional criteria when the fly ash is produced by co-firing of biomass.

Part A presents the evaluation of the sizing systems for different coal / secondary fuel mixes. The applicability of different particle size distribution systems and methods was tested for several secondary fuels and their mixtures with primary fuels. The investigated systems were different sieve methods, laser diffraction, optical microscopy and SEM. The screen sizing systems (shaker and air-jet) are equally applicable for non-fibrous materials. For strongly fibrous materials the air-jet screener should be used. The laser-optical methods used showed good results when particle size distribution was within the measuring range of the analyser. Optical methods like optical microscopy and scanning electron microscopy in combination with particle detection software was not successful. Reasons are the segregation of the particle and fuel components during sample preparation due to the size and density differences and the inaccurate detection and isolation of single particles in a mixed fuel with the software available. The particle size investigations of the grinding products show that all supplemental fuels shift the overall particle size distribution of the fuel mix towards larger particle sizes. Qualitative evaluation concerning the impact of co-grinding is possible with the tools applied. For a quantitative determination of the impact of co-grinding more detailed information about composition of the fuel and sieve fractions is necessary. Part B provides detailed information on the fuel composition and the distribution of the inorganic matter in the size fractions in form of a substantial database. Four coal / secondary fuel systems were investigated in terms of element distribution in pure materials and in fuel mixtures. Two coal / secondary fuel systems were investigated in terms of the element distribution in the sieve fractions. In order to determine the composition of the fuel and the inorganic matter in the fuel, a comprehensive database of the components in the sieve fractions had to be set up. This included Thermo Gravimetric and Proximate Analysis as well as Ultimate and XRF-Analysis of all fuels and sieve fractions produced, see also Part A. Proximate, ultimate, heating value and chlorine analysis were done on the fuels and selected sieve fractions. Ultimate and XRF analysis were performed on lab ashes of all fuels and selected sieve fractions. The ashes were produced in a lab oven at 550°C and 815°C. The bio-waste fuels show the typical characteristic in the proximate and elemental analysis with high volatile, lower fix-C and lower carbon content. There is almost no chlorine and little to no sulphur in the bio-waste. Nitrogen is not present in paper sludge, ranges in the same level as coal for wood pellets and is clearly increased for cacao-shells with a four times higher value compared to coal based on the energy input. Characteristic properties of the single fuels could be found, which indicate that a determination of the presence of one or the other fuel in a certain size fraction might be possible. However, the selective and inhomogeneous behaviour of some properties, for example the ash content, in the different size fractions has to be considered. TGA-characteristics provide information in how much the fuels behave homogeneously and therefore give helpful information on which assumptions are justifiable when carrying out the determination of the composition of the sieve fraction and on the evaluation of the results obtained. To compare the results of XRF Analysis, especially for 550°C lab ash samples of biomass, the data has to be normalised. This was done by a special approach in order to determine the loss of sample weight due to the release of CO2 from CaCO3.

A range of coals and secondary fuels and their blends have been characterised using the IFRF's Isothermal Plug Flow Reactor. Measurements include devolatilisation rates, char burnout rates and ignition distances. The data may be used directly to assess the comparative behaviour of the BioFlam fuel blends, or to derive kinetic data for use in CFD models.

Deliverables

Share this page

Download