ERA-Net Evaluation of technology status for small-scale combustion of pellets from new ash rich biomasses - combustion tests in residential burners

Full text

(1)ERA-Net Evaluation of technology status for small-scale combustion of pellets from new ash rich biomasses - combustion tests in residential burners. SP Technical Research Institute of Sweden. Marie Rönnbäck, Mathias Johansson, Frida Claesson. Energy Technology SP Report 2008:31.

(2) ERA-Net Evaluation of technology status for small-scale combustion of pellets from new ash rich biomasses - combustion tests in residential burners Marie Rönnbäck, Mathias Johansson, Frida Claesson. SP Sveriges Tekniska Forskningsinstitut SP Technical Research Institute of Sweden SP Report 2008:31 ISBN 978-91-85829-48-4 ISSN 0284-5172 Borås 2008. 2.

(3) 3.

(4) SUMMARY In this project, pellets with higher ash content compared to the wood pellets used today on the Swedish market were tested in three domestic-scale burners in the combustion laboratory at SP. The tests were carried out based on EN 303-5. In the flue gas, combustion parameters as carbon monoxide, carbon dioxide, oxygen and hydro carbons were measured, and also more fuel specific parameters such as nitrogen oxides, sulphur dioxide, hydrogen chloride, total dust and particle mass- and number concentration. The dust (fly ash) and bottom ash were characterized chemically. The implications of high ash content on combustion performance are discussed in the report. Altogether five pellets with 8 mm diameter were tested: oilseed straw pellet, reed canary grass pellet (RCG), barley straw pellet, bark pellet and wood pellet. All fuels were dry ranging from 6.5 – 12 % moisture. The ash content varied from 0.3 weight-% dm in wood to 7.9 % in RCG. Barley straw has a noticeable low ash melting temperature, IT is < 980 ˚C, and could not be combusted in any of the burners. The nitrogen content varied nine times (between wood and RCG) and sulphur more than 10 times (between wood and oilseed and barley straw). The chlorine content was very low in wood and bark and more than 20 times higher in oilseed and barley. The composition of inorganic species in the fuel ash was dominated by calcium, potassium and silica in wood, bark and oilseed pellet, while RCG and barley straw were dominated by silica. The three burners used were commercial and known to fulfil high quality requirements. They represented different burner techniques. These kind of burners that can be applied to an existing boiler is a common technique in Sweden. Burner A is a pellet burner where fuel is supplied on top of the grate with no mechanical mean for moving bottom ash on the grate during combustion. Bottom ash is blown away, and any slag remaining on the grate is removed with a scrape before ignition. Burner B is an upward burning pellet burner where fuel and ash is pushed upwards and the glow bed is exposed to the surrounding combustion department. Burner C is a forward burning grain burner that pushes fuel and ash forwards, inside a cylinder. In burner A it was possible to combust bark pellets with an ash content of 3.4 %, though the load had to be reduced and the grate had to be cleaned more frequently than with wood pellet. The combustion was not optimal and indicated insufficient combustion of the char with low temperature in the glow bed, leading to low temperature also in the gas phase which resulted in incomplete combustion of CO. The ash piling up on the glow bed acts as a hindrance for oxygen to reach the char surface. With fuels with even higher ash content this effects was even more accentuated, and combustion was not possible to sustain. Normally, when burner A is used in a house, the combustion periods are shorter (when during these tests) and cleaning of the grate with the scrape with shorter intervals (compared to wood pellets) should not be a problem for using a fuel with several percentage of ash (as for bark), provided that the load and the cleaning periods are adjusted to the fuel. Still, to achieve really good combustion and thorough burnout of char, actions have to be taken to enhance oxygen diffusion and/or temperature in the flue bed. Burner B continuously pushes the char and ash upwards and, finally, above the rim of the burner. Depending on the volume and structure of the ash it piles up to a certain height and covers the char before it falls off. The piled up ash acts at a “hat” that effectively covers the char. The oxygen is hindered from reaching the char and the temperature is lowered. Neither Reed Canary Grass pellets nor oilseed straw pellets were possible to combust in this burner. In this design, the glow bed is exposed to the surrounding.

(5) combustion compartment, and therefore the temperature in the glow bed is lower compared to the glow bed in burner C. Nonetheless, the barley straw pellets reached a temperature where they sintered into hard pieces that suffocated the combustion. Burner C continuously pushes the char and ash forward. In this process, fresh fuel and glowing char is not covered by ash and can more easily be reached by oxygen compared to the glow bed in burner B. Also, the closed design of the burner keeps the temperature in the glow bed higher compared to a design where the glow bed is exposed to the surrounding. All fuel, except barley straw, was possible to combust in burner C. From the combustion tests in this project it was concluded that an increase of ash content may lead to a poorer combustion and even extinction of the glow bed. Poorer combustion is manifested by high CO emissions and (in one case) low flue gas temperature. Ash can act as a hindrance if it is not transported away from the grate but piled up above the fresh fuel, or if it forms a shell around each fuel particle and decrease the oxygen diffusion velocity. CO emissions from the ash rich fuels were higher that should be expected from good combustion of wood pellets. The OGC emissions, as well as the combustible parts of the fly ash particles, measured as weight-% of dust after filter sampling of total dust, were low, showing a generally good combustion performance with low amount of unburned hydrocarbons and soot in the flue gas. Measured emissions of SO2 and HCl in this project are low. Only combustion of RCG resulted in emission values exceeding the order of combustion of wood pellets. Because of the low emissions of SO2 and HCl measured from the fuels in this project, no urgent measures are to be taken to avoid corrosion. Nevertheless, it is important to continuously survey these corrosive emissions, because variations in fuel content and ash composition may influence the formation of corrosive gases. Let out of the chimney, these gases are also acidifying and harmful for the environment. Therefore, if there will be a massive expansion of combustion of fuels containing sulphur and chlorine in the future, it might be necessary to reduce these emissions. Emissions of nitrogen oxide follow the fuel content with: wood 137, bark 410, oilseed straw 478 and RCG 940 mg/Nm3 at 10 % O2 (full load). At minimum load, CO emissions are higher and NO2 emission lower than at full load, though the difference in all cases but bark is quite small. The conversion of fuel nitrogen to NOx decrease with increasing fuel content. Emissions of total dust are for RCG pellets in the same order as for wood pellets (29 – 75 mg/Nm3 at 10 % O2), and for bark pellets slightly above 100 mg/Nm3 at 10 % O2. For oilseed straw pellet the dust emissions are considerably higher, at full load 639 and at minimum load 359 mg/Nm3 at 10 % O2. Emission of total dust do not directly correspond to fuel ash content. Measured particle mass concentrations show that the fly ash consists mainly of submicron particles. Chemical analyses show that submicron particles from wood combustion are dominated by potassium as positive ion, and sulphur and chlorine as negative ions, which is typical for wood fly ash. Particles from bark combustion are quite similar to particles from wood combustion, but potassium is now accompanied by sodium, and the share of chlorine is higher. Particles from Reed Canary Grass pellets exhibit phosphor as positive ion together with potassium. Also here, sulphur and chlorine are main negative ions. Unfortunately, the chemical analysis of fly ash from oilseed straw was corrupt. The comparably high fuel chlorine content of oilseed straw, the low content of chlorine in 5.

(6) bottom ash and the low emission of HCl indicates that most of the chlorine leaves as particles and are found in the fly ash. From the experiments it was concluded that appliances optimized for wood pellets will have to be further adapted to and optimized for ash rich pellets. To succeed with combustion of ash rich pellets the following has to be ensured: –. The ash needs to be hindered from piling up and cover the glow bed.. –. The ash has to be removed from the grate.. –. High ash content acts as a hindrance for oxygen diffusion and thus – char residence time has to be long enough for complete combustion of the char, – the temperature in the glow bed has to be high enough for complete char combustion.. –. Combustion of the char bed is crucial for complete combustion of CO in the gas phase. Therefore temperature and mixing in the gas phase have to be secured.. 6.

(7) CONTENTS SUMMARY. 4. CONTENTS. 7. INTRODUCTION. 8. 1.1 1.2 1.2.1. Background Objectives Scientific and technical objectives of SP. 8 8 8. 2. EXPERIMENTAL PROCEDURES. 9. 2.1 2.2 2.2.1 2.2.2. Experimental procedures at SP Experimental set up at SP Combustion Laboratory Fuels Measurement equipment and analyses. 9 10 12 15. 3. RESULTS. 16. 3.1 3.2 3.3 3.4. 16 21 23. 3.5. Used burners and combustion results Sintering and fouling Results of SO2, HCl, NOx and total dust Particle mass- and number concentration and composition of inorganic contents Contents in bottom ash. 27 30. 4. DISCUSSION. 31. 4.1 4.2 4.3. Implications on combustion by high ash content The burners Emissions from the different fuels. 31 32 33. 5. CONCLUSIONS AND RECOMMENDATIONS. 35. 5.1 5.2. Combustion performance Fuel specific emissions. 35 35. REFERENCES. 37. ACKNOWLEDGEMENT. 37. APPENDIX A. 38. 7.

(8) INTRODUCTION 1.1. Background. The use of small-scale biomass burners is rapidly increasing, and at the same time the competition for materials from the forests increases. Today’s pellets are made from residues from the saw mills and are homogenous and have low nitrogen and ash content. With increasing demand, the fuel of tomorrow will consist of a variety of material from energy crops, residues from agriculture, sorted waste etc. as well as virgin wood and wood residuals. Many of the fuels of tomorrow have higher ash content and other properties that the combustion technology must handle. A motive for use of bio fuels is that they are CO2-neutral. But high ash content may lead to emissions of dust, and high nitrogen content may lead to emissions of nitrogen oxides. Also, sulphur and chlorine in the fuel may lead to emissions of sulphur oxides and hydrogen chloride that are corrosive and acidifying. Therefore it is important to obtain information regarding the performance of small-scale technology burning tomorrow’s fuels. At the same time, it is a challenge to develop the small-scale technology to meet the demand for easy handling and high availability as well as emission standards of tomorrow using varying pellet qualities.. 1.2. Objectives. 1.2.1. Scientific and technical objectives of SP. In this project, pellets with higher ash content compared to the wood pellets used today on the Swedish market were tested in three domestic-scale burners (< 50 kW). The tests were be carried out based on EN 303-5 and complemented to evaluate the demand on the appliances by high ash content and ash properties. Besides the flue gas measurements required by EN 303-5 (CO, CO2, O2, OGC and total dust) also NOx, SO2, HCl and particle mass- and number concentration were measured. The fly ash and bottom ash were characterized chemically. The implications of high ash content on combustion performance are discussed in the report.. 8.

(9) 2. EXPERIMENTAL PROCEDURES. 2.1. Experimental procedures at SP. Five fuels were tested in three domestic-scale burners in the combustion laboratory at SP. The tests were carried out based on EN 303-5. The test rig is set up as shown in annex A.6 in EN-304 (stated in EN-303-5). In the standard, the test periods are 6 h at nominal load and 6 h at minimum load (minimum as stated by the manufacturer, not above 30 % of nominal). Flue gas measurements required in the standard are CO, CO2, O2, OGC and total dust. In this project, also NOx, SO2, HCl and particle mass- and number distributions were measured, and the measurements were complemented by observation of the combustion such as slagging, combustion performance etc. Fly ash and bottom ash were characterized chemically. The measurements followed the test plan in Table 1. As seen in the table, SO2, HCl and particle mass- and number distributions were measured only at nominal load. It was not always possible to carry through 6 hours of continuous combustion, and deviations from the test plan are described in relation to each measured case. Table 1. Test plan at SP Energy Technology.. Fuel Flue gas measurements Instruments on-line Wet chemical analysis Filter sampling Particle number concentration and size distribution Particle mass concentration and size distribution Other measurements Other observations Influence of high ash content. Measured parameter Moisture content. Comments All fuels. CO/CO2, O2, OGC, NOx SO2, HCl Total dust ELPI. Nominal and minimum load Average of 6 hours Nominal load Nominal and minimum load Nominal load. DLPI. Nominal load. Heat output (kW). Nominal and minimum load. Sintering in fuel bed and in Visual observation after each bottom ash, foaling of heat test1 exchange surfaces. Analyses In total dust Unburnt material Nominal load Dust from DLPI, six particle Main elements (Na, K, Ca, Nominal load sizes Mg, Zn, Al, Si, Fe, Mn, Ba, P, S, Cl) Nominal load Bottom ash Unburnt material Main elements (Na, K, Ca, Mg, Zn, Al, Si, Fe, Mn, Ba, P, S, Cl). 1. Bottom ash was categorized according to the following: Category 1: Only slightly sintered ash that falls apart when touched. Category 2: Somewhat sintered ash that keeps together when touched but can be broken apart. Granules are easily distinguished in the material. Category 3: Sintered ash still possible to brake into pieces. Granules are still possible to distinguish, but melted material/parts can be seen by eye. Category 4: Totally sintered ash, not possible to break apart by hand. The ash has melted and formed larger blocks. No individual granules are possible to distinguish by eye. 9.

(10) 2.2. Experimental set up at SP Combustion Laboratory. Two burners for wood pellet and one burner for cereal grain were used for the experiments. The burners were connected to a Combifire boiler from Ved & Solteknik, Långshyttan. The boiler was connected to a test stand with pump, flow meter, valves and heat exchanger. The same type of boiler is used for certification and P-marking of burners for wood pellets and energy grain. These kind of burners that can be applied to an existing boiler is a common technique in Sweden. The three used burners represent different burner techniques. Burner A is a Janfire NH burner for wood pellets with a nominal power of 20 kW (for wood pellets). The burner is P-marked and fulfils high quality requirements. The pellets fall down onto a grate. The grate is equipped with a scrape that cleans the grate from any residual ash or sintered parts before every start of the burner. The pellets are ignited by the heated air from an electrical coil. The burner cup is quite big to secure a thorough burnout of the fuel before it leaves the cup. The glow bed is exposed to the walls of the burner cup and the flame director, keeping the temperature of the glow bed high. Figure 1 shows the burner and Figure 2 shows a close-up of the grate with the ash scrape.. Figure 2. Close-up of the grate with the ash Figure 1. Burner A: Janfire NH wood pellet scrape. burner. The flame is directed upwardsforwards. A principle figure of the technique represented by burner A is shown in Figure 3. The air is supplied partly through the glow bed and partly to the combustible gas through holes at the upper part of the burner. When wood pellets are combusted, inorganic material mainly follows the gas flow out of the cup and is found at the bottom of the boiler.. Figure 3. A principle figure showing the burner technique represented by burner A. 10.

(11) Burner B is an Ecotec burner for wood pellets with a nominal power of 15 kW (for wood pellets), see Figure 4. The burner is P-marked and fulfils high quality requirements. In this burner, the fuel is pushed all the way into the burner by a screw, and ash and any sintered material is pushed over the edge of the burner cup. The pellets are ignited by the heated air from an electrical coil. The glow bed is exposed to the inner roof of the burner compartment of the boiler, and depending of the boiler, the glow bed is more or less cooled by heat exchange with the surfaces. A principle figure of the burner technique represented by burner B is shown in Figure 5.. Figure 4. Burner B: Ecotec Bioline 20 Figure 5. A principle figure showing the wood pellet burner. The flame is directed burner technique represented by burner B. upwards. Burner C is an AgroTec burner for energy grain, see Figure 6. The burner is not Pmarked but it was tested thoroughly in an earlier project [1] and fulfilled high quality requirements. With grain the power was 14 kW. In this burner, the fuel is pushed all the way into the forward oriented burner and ash and any sintered material is pushed out through the front part of the burner cylinder. The pellets are ignited by a sustained glow bed. During combustion, the glow bed is exposed to the inner sides of the cylinder and therefore the temperature in the glow bed is quite high. A principle figure of the burner technique represented by burner C is shown in Figure 7.. Figure 6. Burner C: AgroTec energy Figure 7. A principle figure showing the burner grain burner. The flame is directed technique represented by burner C. forwards.. 11.

(12) 2.2.1. Fuels. Three fuels were pelletised at BTK (Unit of Biomass Technology and Chemistry): the oilseed straw pellet, the reed canary grass pellet (RCG) and the barley straw pellet. The bark pellet were bought from Södra Cell AB. The pellets had a diameter of 8 mm. The fuels were analysed by BTK, see Table 2. Composition of inorganic contents in fuel ash is also shown in Figure 8 and contents in fuel ash as weight-% of dry matter in Figure 9 to facilitate comparison of the fuels. •. All fuels were dry ranging from 6.5 % of moisture in wood to 12 % in RCG.. •. The ash content as weight-% of dry matter was: wood 0.3 %, bark 3.4 %, oilseed straw 4.7 %, barley straw 6.6 % and RCG 7.9 %.. •. Volatile content in dry, ash free matter was 75.8 % for bark and 80.5 – 85.4 % for the other fuels.. •. Wood and oilseed straw had ash melting temperatures > 1500 ˚C. RCG had an initial ash melting temperatures IT of 1350 ˚C and bark had an IT of 1250 ˚C. Only barley straw had a noticeable low ash melting temperature; IT was < 980 ˚C.. •. The composition of inorganic species in the ash was similar in wood, bark and oilseed pellet, dominated by calcium, potassium and silica, see Figure 8. In wood and bark also aluminium and iron are found. RCG and barley straw were dominated by silica. The share of calcium and potassium were lower in RCG than in wood, bark and oilseed straw, see Figure 8, while the absolute value of potassium was comparable between bark and RCG, see Figure 9. In Barley straw the absolute value of calcium was higher than in oilseed straw, and the absolute value and the share of potassium were considerably higher. The low content of potassium and chlorine in RCG was typical for a crop that is harvested in spring.. •. The difference in nitrogen content was 9 times. Nitrogen as weight-% of dry matter was: wood 0.1 %, bark 0.4 %, oilseed straw 0.5 %, barley straw 0.7 %, RCG 0.9 %.. •. The sulphur content was very low in wood and bark and about 10 times higher in RCG, oilseed and barley straw. Sulphur as weight-% of dry matter was: wood < 0.01 %, bark 0.03 %, RCG 0.11 %, oilseed straw and barley straw 0.13 %.. •. The chlorine content was very low in wood and bark and only slightly higher in RCG. In oilseed straw the chlorine content was 20 times higher than in wood and bark. In barley straw the chlorine content was noticeable high. Cl as weight-% of dry matter was: wood < 0.01 %, bark 0.01 %, RCG 0.04 %, oilseed straw 0.18 % and barley straw 0.71 %.. 12.

(13) Table 2. Fuel analyses. All fuels were pellets with 8 mm diameter. Unit Wood Bark Oilseed straw Dry matter % 93.5 90.9 90.4 Moisture % 6.5 9.1 9.6 Ash % of d.m. 0.3 3.4 4.7 Net calorific value MJ/kg d.m. 19.13 17.77 17.57 Bulk density kg/m3 5741 679 590 Sulphur % of d.m. <0.01 0.03 0.13 Nitrogen % of d.m. 0.1 0.4 0.5 Hydrogen % of d.m. 6 5.9 6 Carbon % of d.m. 50.6 52.5 47.2 Oxygen % of d.m. 43 37.8 41.3 Chlorine % of d.m. <0.01 0.01 0.18 Volatiles % of d.m. 85.1 73.2 78.9 Volatiles % of a.f. d.m. 85.4 75.8 82.8 Ash melting IT °C 1550 1250 1590 Ash melting ST °C 1550 1340 1590 Ash melting HT °C 1550 1400 1590 Ash melting FT °C 1550 1420 1590 Inorganic ash components Si 0.01 0.37 0.30 % of d.m. Ca 0.07 0.85 1.18 % of d.m. Al 0.02 0.08 0.01 % of d.m. Fe 0.004 0.04 0.01 % of d.m. K 0.03 0.20 0.51 % of d.m. Mg 0.01 0.08 0.08 % of d.m. Mn % of d.m. 0.01 0.05 0.003 Na 0.001 0.03 0.06 % of d.m. P 0.005 0.05 0.07 % of d.m. Ti 0.0001 0.03 0.0005 % of d.m. 1 Measured by VTT. RCG 88 12 7.9 17.31 568 0.11 0.9 5.9 45.9 39.4 0.04 76 82.5 1350 1530 1580 1590. Barley Straw 90.7 9.3 6.6 17.41 535 0.13 0.7 5.7 46 40.2 0.71 75.2 80.5 <980 990 1100 1190. 2.77 0.22 0.06 0.05 0.24 0.07 0.02 0.02 0.01 0.004. 0.86 0.30 0.03 0.03 0.73 0.06 0.002 0.13 0.11 0.002. 100%. Composition (weight-%). 90% Si Ti P Na Mn Mg K Fe Ca Al. 80% 70% 60% 50% 40% 30% 20% 10% 0% Wood. Bark. Oilseed straw. RCG. Barley straw. Figure 8. Composition of inorganic contents in fuel ash. 13.

(14) Content in dry matter (weight-%). 4 3,5 Si Ti P Na Mn Mg K Fe Ca Al. 3 2,5 2 1,5 1 0,5 0 Wood. Bark. Oilseed straw. RCG. Barley straw. Figure 9. Content of inorganic species in fuel ash as weigh-% of dry matter.. 14.

(15) 2.2.2. Measurement equipment and analyses. Instruments and analyses used during measurements are gathered in Table 3 and Table 4. Table 3. Measurement equipment. Notation refers to the SP quality system.. Instruments Thermocouple type K Dust sampling equipment, STL-Medi Dust sampling equipment, STL-Combi CO/CO2-analyser type Binos (NDIRinstrument) O2-analyser type M&C Model PMA 10 (paramagnetic instrument) O2-analyser type M&C Model PMA 10 (paramagnetic instrument) THC-analyser type JUM (FID-instrument) NO, NO2 EcoPhysics 70 (paramagnetic instrument) Gas watch with pump for wet chemical analysis. Notation ETf-QD Db 2 Inventory no. 200 399 Inventory no. 202 743 Inventory no. 202 045. Measured parameter Flue gas temperature Dust CO2 CO O2 THC propane equivalences NO2. Measurement uncertainties + 3ºC < 10 % at an increase of > 20 mg + 0,3 % CO2 + 450 ppm CO + 0,46 % O2 3-300 A ± 3 ppm ± 31 ppm. Inventory no. 202 589 Inventory no. 202 342 Inventory no. 201 664 Inventory no. 202 106 Inventory no. 200 619. Table 4. Measurement equipment and analyses.. Parameter HCl, hydrogen chloride SO2, sulphur dioxide Particle mass concentration. Unburt in bottom ash Unburt in fly ash. Equipment or analysis Wet chemical absorption followed by IC-TCD analysis Wet chemical absorption followed by IC-TCD analysis DLPI, Dekati Low Pressure Impactor Particle interval: 30 nm – 10 µm ELPI, Electrical Low Pressure Impactor Particle interval: 7 nm – 10 µm Leaching of particles followed by chemical analysis of Cl-, SO42- with the instruments 861 Advanced Compact IC and Metrosep A Supp 5, both from Metrohm Al, Ba, Ca, Fe, K, Mg, Mn, Na, P, Si and Zn was analysed with an ICP-OES-instrument Optima 3000 DV from Perkin Elmer Leaching Oven at 550 ºC. Measured parameter SO2 HCl. Measurement uncertainties Estimated to + 11-20 % Estimated to + 11-20 %. Particle number concentration Inorganic components in dust from DLPI. 15.

(16) 3. RESULTS. 3.1. Used burners and combustion results. The tests were performed using the burners according to the summary in Table 5. Emissions measured on-line are shown in Figure 10 - Figure 17. Time averaged values of emissions together with measured total dust, SO2 and HCl are shown in Figure 18 Figure 33. A table comprehensive table with combustion results is shown in Appendix A. Table 5. Summary of the fuels and the burners used for the tests.. Fuel Wood pellet full load Wood pellet minimum load RCG pellet full load RCG pellet minimum load Bark pellet full load Bark pellet minimum load Oilseed straw pellet full load Oilseed straw pellet minimum load Barley straw pellet. Burner B. Comments Should be possible in all tree burners. B C. Trials in A and B fails, combustion is extinguished by gathering ash. C A. Combustion okay but adjustment of the burner.. after. some. A C. Trials in A and B fails, combustion is extinguished by gathering ash. C Not possible because of slagging. Burner A is developed for high quality wood pellet with low ash content. When wood pellets are combusted the ash is blown away from the grate by the primary air. Nominal load with wood pellet is 20 kW. Every combustion period starts with the scrape moving back and forth several times to clean the grid from any remaining ash or slag. Burner A was, after some adjustments, used for bark pellet with an ash content of 3.4 %. The load had to be lowered (15.7 kW) and the ash gathered on the grid had to be removed manually after about three hours to assure that the glow bed was not covered and extinguished. Time between scrape movements was at full load chosen to 180 minutes. The scrape emptied the grid from glowing char and the burner had to re-start. The scraping and re-starting periods are excluded from the calculation of average values, se Figure 14. At minimum load, 4.2 kW, the periods between cleaning the grate were adjusted to 105 minutes, see Figure 15. The combustion was not optimal. The average CO emissions were at full load 360 and at minimum load 576 mg/Nm3 at 10 % O2 that is higher than is normally accepted during combustion of wood pellet. Average values are presented in Figure 22, Figure 23 (full load 15.7) and Figure 30, Figure 31 (minimum load 4.2).. 16.

(17) 20. Wood pellet 12,8 kW. 3 200. OGC mg/Nm3 at 10 % O2. 20. O2 vol-% in dry gas. CO mg/Nm3 at 10 % O2 in dry gas. CO mg/Nm3 at 10 % O2 in dry gas. 10. 1 600. 5. 10. 800. 0 0. 100. 200. 800. 0 0. 100. Time (minutes). Reed Canary Grass pellet 13,3 kW. 200. 3 200. OGC mg/Nm3 at 10 % O2. Figure 11. Emissions of OGC, NO2 och O2. Wood pellet, minimum load 5.2 kW. 20. RCG pellet 7,8 kW. O2 vol-% in dry gas. O2 vol-% in dry gas CO mg/Nm3 at 10 % O2 in dry gas. 10. 1 600. 5. 10. 50. 100. 150. 200. 250. 300. 350. 800. 0. 0 400. 0. 100. Time (minutes). Bark pellet 15,7 kW. 3 200. OGC mg/Nm3 at 10 % O2. 40. Bark pellet minimun power 4,2 kW. 1 600. 5. 800. 200. 300. NO2 mg/Nm3 at 10 % O2 in dry gas. 30. 20. 2 000. 10. 1 000. 0 400. 0 0. 100. Time (minutes). Oilseed straw pellet 11,9 kW. 3 200. OGC mg/Nm3 at 10 % O2. 20. O2 vol-% in dry gas CO mg/Nm3 at 10 % O2 in dry gas. 5. 800. 0 200. 300. 0 400. Time (minutes). Figure 16. Emissions of OGC, NO2 och O2. Oilseed straw pellet, full load 11.9 kW.. Oilseed straw pellet 8,8 kW. 2 400 CO, NO2. 1 600. NO2 mg/Nm3 at 10 % O2 in dry gas. 15 OGC, O2. 2 400 CO, NO2. OGC, O2. 10. 100. 0 400. 3 200. OGC mg/Nm3 at 10 % O2. O2 vol-% in dry gas NO2 mg/Nm3 at 10 % O2 in dry gas. 0. 300. Figure 15. Emissions of OGC, NO2 och O2. Bark pellet, minimum load 4.2 kW.. CO mg/Nm3 at 10 % O2 in dry gas. 15. 200 Time (minutes). Figure 14. Emissions of OGC, NO2 och O2. Bark pellet, full load 15.7 kW. 20. 3 000 CO, NO2. 2 400 CO, NO2. OGC, O2. 10. 100. O2 vol-% in dry gas. OGC, O2. NO2 mg/Nm3 at 10 % O2 in dry gas. 0. 4 000. OGC mg/Nm3 at 10 % O2 CO mg/Nm3 at 10 % O2 in dry gas. CO mg/Nm3 at 10 % O2 in dry gas. 0. 0 400. 300. Figure 13. Emissions of OGC, NO2 och O2. RCG pellet, minimum load 7.8 kW.. O2 vol-% in dry gas. 15. 200 Time (minutes). Figure 12. Emissions of OGC, NO2 och O2. RCG pellet, full load 13.3 kW. 20. 2 400. 1 600. 5. 800. 0. NO2 mg/Nm3 at 10 % O2 in dry gas. 15. CO, NO2. OGC, O2. CO, NO2. 2 400. OGC, O2. NO2 mg/Nm3 at 10 % O2 in dry gas. 0. 3 200. OGC mg/Nm3 at 10 % O2. CO mg/Nm3 at 10 % O2 in dry gas. 15. 0 400. 300. Time (minutes). Figure 10. Emissions of OGC, NO2 och O2.Wood pellet, full load 12.8 kW. 20. 2 400. 1 600. 5. 0 400. 300. NO2 mg/Nm3 at 10 % O2 in dry gas. 15. CO, NO2. OGC, O2. CO, NO2. 2 400. OGC, O2. NO2 mg/Nm3 at 10 % O2 in dry gas. 15. 3 200. OGC mg/Nm3 at 10 % O2. Wood pellet 5,2 kW. O2 vol-% in dry gas. 10. 1 600. 5. 800. 0 0. 100. 200. 300. 0 400. Time (minutes). Figure 17. Emissions of OGC, NO2 och O2. Oilseed straw pellet, minimum load 8.8 kW.. 17.

(18) Oilseed straw pellets with an ash content of 4.7 % were tried in burner A but, in spite of several trials to adjust air, load and scrape sequences, the glow was slowly extinguished by ash piling up on the grid. With low load and an excess of air the combustion was sustained, but the combustion was poor and after 40 minutes the flame guard shut the burner. Oilseed pellet were then successfully combusted in burner C though the CO emissions were higher than normally accepted during combustion of wood pellet, 940 at full load and at 1168 mg/Nm3 at 10 % O2 at minimum load, see Figure 16, Figure 24, Figure 25 (full load 11.9 kW), and Figure 17, Figure 32, Figure 33 (minimum load 8.8 kW). Minimum load 30 % was not possible to realise because of poor combustion performance. The Reed Canary Grass pellets had an ash content of 7.9 %, and the ash does not fall together but forms “skeletons” that easily covers the glow and prevents the air to reach to the supplied fuel in the burner. Burner A was shut down after a short time and it was considered not possible to combust such an ash rich fuel in this burner. Burner B was tried but combustion was poor. Burner C worked successfully at full load (13.2 kW) with an average CO emission of 153 mg/Nm3 at 10 % O2, see Figure 12. Burner C is equipped with a screw that pushes the bottom ash forward as the pellets are fed into the burner. The pile of “ash skeletons” formed in the burner is shown in Figure 34. It was difficult to adjust the combustion of RCG pellets to a minimum load of 30 % because of poor combustion performance, as the gathered ash was not pushed away but hindered the air to reach the fuel. Finally, minimum load was performed at 7.8 kW with a CO emission of 426 mg/Nm3 at 10 % O2, see Figure 13. The ash piled up in front of the burner and had to be pushed further into the combustion chamber several times during the test. For averaged emission values see Figure 20, Figure 21 (full load 13.3 kW), and Figure 28, Figure 29 (minimum load 7.8). The pellets made of barley straw had an ash melting initial temperature less than 980 ºC. The pellets were first tried in burner C, but combustion was almost not possible to sustain and after some time smoke started to come out through the fuel feeder. A large sintered cake was found at the bottom of the cylindrical burner. Burner B was tried, because the temperature in the glow bed in a burner formed as an open cup has earlier been measured to be between 100 and 200 ºC lower than in a cylindrical burners such as burner C. It was possbilbe to sustain combustion in burner B, but the result was poor - CO emission was “stabilised” around 6000 ppm. The tests with barley straw pellets were concluded. The slag formed in burner B is shown in Figure 35. The wood pellets were successfully combusted in burner B. From earlier experience its know that any of the burners could have been used for the wood pellets; burner B was an easy choice because it was already mounted on the boiler. The average CO- emission at full load was 216 and at minimum load 338 mg/Nm3 at 10 % O2, see Figure 10, Figure 18, Figure 19 (full load 12.8), and Figure 11, Figure 26, Figure 27 (minimum load 5.2 kW). This burner is normally not used on reduced load and therefore no further measures were taken to improve combustion performance on minimum load.. 18.

(19) The following figures show average emission values at full load. Wood pellet 12,8 kW. 800. 600. 10. 400. 5. 200. 0. 0 OGC. HCl. Figure 18. Mean values and stdv of O2 and OGC from 360 minutes. Mean values of HCl and SO2 from 130 minutes. Wood pellet, full load 12.8 kW. Average in mg/Nm3 in dry gas at 10 % O2). 30. CO. SO2. Reed Canary Grass pellet 13,2 kW. 25. Dust. 800. Reed Canary Grass 13,2 kW. 600. 20. 400. 15. 10. 200. 5. 0. 0 O2. OGC. HCl. NO2. 800. Bark pellet 15,7 kW. Dust. Figure 21. Mean values and stdv of CO and NO2 from 251 minutes. Mean values of dust from 120 minutes. RCG pellet, full load 13.2 kW. Bark pellet 15,7 kW. Average in mg/Nm3 in dry gas at 10 % O2). 15. CO. SO2. Figure 20. Mean values and stdv of O2 and OGC from 251 minutes. Mean values of HCl and SO2 from 120 minutes. RCG pellet, full load 13.2 kW. Average in mg/Nm3 in dry gas at 10 % O2). NO2. Figure 19. Mean values and stdv of CO and NO2 from 360 minutes. Mean values of dust from 120 minutes. Wood pellet, full load 12.8 kW. Average in mg/Nm3 in dry gas at 10 % O2 ). O2. 600. 10. 400. 5. 200. 0. 0 OGC. HCl. Figure 22. Mean values and stdv of O2 and OGC from 250 minutes. Mean values of HCl and SO2 from 120 minutes. Bark pellet, full load 15.7 kW. 15. CO. SO2. Oilseed Straw pellet 11,9 kW. NO2. Dust. Figure 23. Mean values and stdv of CO and NO2 from 250 minutes. Mean values of dust from 120 minutes. Bark pellet, full load 15.7 kW. 1400. Rape straw pellet 11,9 kW. Average in mg/Nm3 in dry gas at 10 % O2). O2. Average in mg/Nm3 in dry gas at 10 % O2). Wood pellet 12,8 kW. Average in mg/Nm3 in dry gas at 10 % O2). Average in mg/Nm3 in dry gas at 10 % O2). 15. 1200 1000. 10. 800 600. 5. 400 200 0. 0 O2. OGC. HCl. SO2. Figure 24. Mean values and stdv of O2 and OGC from 360 minutes. Mean values of HCl and SO2 from 122 minutes. Oilseed straw pellet, full load 11.9 kW.. CO. NO2. Dust. Figure 25. Mean values and stdv of CO and NO2 from 360 minutes. Mean values of dust from 120 minutes. Oilseed straw pellet, full load 11.9 kW. 19.

(20) The following figures show average emission values at minimum load. Average in mg/Nm in dry gas at 10 % O2). Wood pellet 5,2 kW. 25. 3. 15 10. 400. 200. 5 0. 0 OGC. Figure 26. Mean values and stdv of O2 and OGC from 360 minutes. Wood pellet, minimum load 5.2 kW. 30 Average in mg/Nm3 in dry gas at 10 % O2 ). Reed Canary Grass pellet 7,8 kW. 25. CO. NO2. Dust. Figure 27. Mean values and stdv of CO and NO2 from 360 minutes. Mean values of dust from 120 minutes. Wood pellet, minimum load 5.2 kW. 800. Reed Canary Grass 7,8 kW. Average in mg/Nm in dry gas at 10 % O2). O2. 600. 20. 3. 15 10. 400. 200. 5. 0. 0 O2. CO. OGC. Figure 28. Mean values and stdv of O2 and OGC from 362 minutes. RCG pellet, minimum load 7.8 kW.. NO2. Bark pellet 4,2 kW. Average in mg/Nm3 in dry gas at 10 % O2). Bark pellet 4,2 kW. Dust. Figure 29. Mean values and stdv of CO and NO2 from 362 minutes. Mean values of dust from 120 minutes. RCG pellet, minimum load 7.8 kW. 1400. 30 Average in mg/Nm3 in dry gas at 10 % O2). Wood pellet 5,2 kW. 600. 20. 3. 800 Average in mg/Nm in dry gas at 10 % O2). 30. 1200. 25. 1000. 20. 800. 15. 600. 10. 400. 5. 200. 0. 0 O2. CO. OGC. Figure 30. Mean values and stdv of O2 and OGC from 239 minutes. Bark pellet, minimum load 4.2 kW.. 1400 Oilseed straw pellet 8,8 kW. Dust. Oilseed straw pellet 8,8 kW. Average in mg/Nm3 in dry gas at 10 % O2). Average in mg/Nm3 in dry gas at 10 % O2). 30. NO2. Figure 31. Mean values and stdv of CO and NO2 from 239 minutes. Mean values of dust from 120 minutes. Bark pellet, minimum load 4.2 kW. 1200. 25. 1000. 20. 800. 15. 600. 10. 400. 5. 200. 0. 0 O2. OGC. Figure 32. Mean values and stdv of O2 and OGC from 381 minutes. Oilseed straw pellet, minimum load 8.8 kW.. CO. NO2. Dust. Figure 33. Mean values and stdv of CO and NO2 from 381 minutes. Mean values of dust from 120 minutes. Oilseed straw pellet, minimum load 8.8 kW. 20.

(21) 3.2. Sintering and fouling. After the combustion test the bottom ash was categorised according to four categories: 1. Only slightly sintered ash that falls apart when touched. 2. Somewhat sintered ash that keeps together when touched but can be broken apart. Granules are easily distinguished in the material. 3. Sintered ash still possible to brake into pieces. Granules are still possible to distinguish, but melted material/parts can be seen by eye. 4. Totally sintered ash, not possible to break apart by hand. The ash has melted and formed larger blocks. No individual granules are possible to distinguish by eye. Bottom ash from bark pellets was categorised as 2 and from barley as 4. The other fuels did not sinter. Results are summarised in Table 6. Slag formed in the burner after combustion of barley straw pellets is shown in Figure 35. Table 6. Summary of the bottom ash categories and fouling from the fuels.. Wood pellet RCG pellet Bark pellet Oilseed straw pellet Barley straw pellet. Ash category 1 1 2 1 4. Fouling Very little, grey More than wood, less than bark, black Substantial, black Substantial, white Combustion tests were not possible. After each combustion test the combustion chamber was inspected for fouling. Combustion of wood pellet resulted in very little fouling with grey colour. Combustion of Reed Canary Grass resulted in more, and black, fouling. Combustion of bark pellet resulted in even more, and black, fouling. Combustion of oilseed straw pellet resulted in a thick layer of white fouling, see Figure 36.. Figure 34. “Ash skeletons” formed from combustion of RCG pellets in burner C.. 21.

(22) Figure 35. Slag in burner B after combustion of barley straw pellets.. Figure 36. Ash and fouling formed during combustion of oilseed straw pellets. Burner C.. 22.

(23) 3.3. Results of SO2, HCl, NOx and total dust. During combustion fuel sulphur may form compounds with high melting temperature that stay in the bottom ash, such as calcium sulphate, CaSO4, or form aerosols that ends up in the fly ash, such as potassium sulphate, K2SO4, or form gaseous compounds, mainly sulphur dioxide, SO2. Fuel chlorine may stay in the bottom ash, or form fly ash, for example as potassium chlorine, KCl, or in gaseous form as hydrogen chloride, HCl. Flue gas was absorbed during 2 hours of the full load tests and analyzed for sulphate and chlorine. The corresponding emissions of SO2 and HCl are shown in Figure 37 and Figure 39 together with maximum possible values calculated from fuel content. Results from earlier measurements from oat grain combustion are also shown. Measured emission of SO2 from wood and bark are low, 1 and 2 mg/Nm3 at 10 % O2 respectively. The emissions correspond to between 3 % (wood pellet) and 4 % (bark pellet) of maximum possible conversion of sulphur to sulphur dioxide, see Figure 38. Emission values from RCG pellets and oilseed straw pellets are 25 and 9 mg/Nm3 at 10 % O2 respectively. The emissions correspond to between 9 % (RCG pellet) and 3 % (oilseed straw pellet) of maximum possible conversion of fuel sulphur to sulphur dioxide. This can be compared to combustion of oat grain that exhibits a conversion in these cases to about 100 %.. SO2 mg/Nm3 in dry gas at 10 % O2. Measured values of HCl in this project are generally low, between 1 and 6 mg/Nm3 at 10 % O2. The emissions correspond to between 1 % (oilseed straw pellet) and 17 % (wood pellet) of maximum possible conversion of chlorine to hydrogen chloride, see Figure 40. This can be compared to combustion of oat grain that exhibits a conversion in these gases of H to HCl of 70-98 %.. 360. Measured SO2 Max SO2. 270. 180. 90. 0 Wood full load. RCG full load. Bark full load. Oilseed full load. Oat in PelLing. Oat in Agrotec. Figure 37. Measured SO2 and maximum possible SO2. Full load.. 23.

(24) SO2 mg/Nm3 in dry gas at 10 % O2. 30 9%. 25. Measured SO2. 20 15 3%. 10 5. 3%. 4% 0. Wood RCG full Bark full Oilseed full load load load full load Figure 38. Measured SO2 and percentage of possible SO2. Full load.. HCl mg/Nm3 in dry gas at 10 % O2. 80. Measured HCl Max HCl. 220. 60. 40. 20. 0 Wood full load. RCG full load. Bark full load. Oilseed full load. Oat in PelLing. Oat in Agrotec. Figure 39. Measured HCl and maximum possible HCl. Full load.. 24.

(25) HCl mg/Nm3 in dry gas at 10 % O2. 8 Measured HCl 12 %. 6. 4 17 % 2 9%. 1%. 0 Wood RCG Bark Oilseed full load full load full load full load. Figure 40. Measured HCl and percentage of possible HCl. Full load.. Measured emissions of nitrogen oxides, calculated as NO2, are shown in Figure 41 together with results from combustion of oat grain. The percentage of maximum possible emissions of NOx depends on the fuel content and is indicated for each fuel.. NO2 mg/Nm3 in dry gas at 10 % O2. 11 % 800. Measured NO2. 19 %. 8%. 600 24 % 30 %. 400. 200. 37 %. 0 Wood Wood RCG full RCG Bark full Bark Oilseed Oilseed Oat in Oat in full load min load load min load load min load full load min load PelLing Agrotec. Figure 41. Measured NO2 and percentage of possible NO2. Full and minimum load.. NOx emissions depend on combustion conditions as well as on fuel nitrogen content. In theory, formed NOx are reduced to N2 when oxygen is not available. Therefore, high CO emissions are usually connected to lower NOx emissions. Emissions of CO and NO2 are shown together in Figure 42. At minimum load, CO emission is higher and NOx emission lower, though the difference is marginal in all cases but bark.. 25.

(26) Carbon monoxide, CO Nitrogen oxide, NO2. 1000 800 600. 3. CO and NO2 mg/Nm in dry gas at 10 % O2. 1200. 400 200 0 Wood Wood RCG full RCG Bark full Bark Oilseed Oilseed Oat in Oat in full load min load load min load load min load full load min load PelLing Agrotec. Figure 42. Measured CO and NO2 during full and minimum load.. Measured emissions of total dust are shown in Figure 43 together with results from combustion of oat grain. Of the fuels in this project only oilseed straw pellets exhibits large dust emission, 639 mg/Nm3 at 10 % O2 at full load and 359 mg/Nm3 at 10 % O2 at minimum load. Emissions from bark are 126 and 116 mg/Nm3, from RCG 31 and 12 mg/Nm3 and from wood 25 and 69 mg/Nm3 respectively. Emissions from oat grain are 322 and 236 mg/Nm3. Emissions from full load are higher than emissions from minimum load (except for wood). This may be a result of lower temperatures in the glow bed during minimum load that reduces the forming of aerosols.. 700 Dust mg/Nm3 in dry gas at 10 % O2. Total dust. 600 500 400 300 200 100 0 Wood Wood RCG full RCG Bark full Bark Oilseed Oilseed Oat in Oat in full load min load load min load load min load full load min load PelLing Agrotec. Figure 43. Measured total dust during full and minimum load.. 26.

(27) 3.4. Particle mass- and number concentration and composition of inorganic contents. During small-scale combustion, dust (fly ash) is formed mainly from unburned hydrocarbons and from volatile species that leave the fuel as gas and form particles < 1 µm when the temperature in the flue gas is lowered. Particles larger than 1 µm generally consist of inorganic species with higher melting temperature than are pulled from the glow bed by the gas flow. Large scale combustion generally exhibits a larger fraction of particles > 1 µm, resulting from a higher flue gas velocity in larger appliances. Particle mass concentrations as function of aerodynamic particle size measured with a DLPI (Decati Low Pressure Impactor) are shown in Figure 44. The particles are mainly between 100 nm and 1 µm. A peak in mass concentration is found between 130 nm (wood pellets) and 350 nm (oilseed pellets). Maximum is shifted towards larger diameters with higher concentrations. This effect is caused by more particles coming together and forming larger particles when the concentration is increased. The shift is also indicated in Figure 45 where the number concentrations are shown, measured with ELPI (Electrical Low Pressure Impactor).. Particle mass concentration Δm/Δlog(Dp) (mg/mn3 in 10 % O2 dry gas). 600 500 400. RCG pellet Wood pellet Oilseed pellet Bark pellet. 300 200 100 0 0,01. 0,10 1,00 Aerodynamic particle size (µm). 10,00. 100,00. Figure 44. Comparison of mass concentration of particles from the four fuels.. The combustible parts of the fly ash particles, measured as weight-% of dust after filter sampling of total dust, were: wood 13 %, RCG 9 %, bark 3 %, and oilseed straw 1 %. This reflects a generally good combustion performance with low amount of unburned hydrocarbons and soot in the flue gas. This also reflects the OGC emissions that were less than 5 mg/Nm3 at 10 % O2 at full load. Composition of the inorganic contents in the particles gathered on the different steps of the impactor is shown in Figure 46 (wood), Figure 47 (RCG) and Figure 48 (bark). Unfortunately, results from oilseed straw pellets were corrupt. The submicron particles formed during wood pellet combustion are dominated by potassium, sulphur and chlorine. Normally, potassium chloride KCl, and potassium sulphate, K2SO3, are found in dust from wood combustion, and this is probably the case also here.. 27.

(28) Number concentration of particles 3 dN/dlog(Dp) 10 % O2, (#/cm ). 1,E+09. Wood pellet Reed canary grass pellet Bark pellet Oilseed straw pellet. 1,E+08 1,E+07 1,E+06 1,E+05 1,E+04 1,E+03 1,E+02 0,01. 0,1. 1. 10. Aerodynamic diameter (µm) Figure 45. Particle number concentration from four fuels. All cases full load.. Submicron particles formed during combustion of Reed Canary Grass consists mainly of phosphate, potassium, sulphur and chlorine. Silica could not be analysed and is therefore not included. Potassium and phosphate may form K2HPO4, KH2PO4 or K3PO4. Probably potassium chloride KCl, and potassium sulphate, K2SO3 are found also here.. 100%. Composition (mole-%). 90%. Ti Zn P Na Mn Mg K Fe Ca Ba Al S Cl. 80% 70% 60% 50% 40% 30% 20% 10% 0% 0,05. 0,17. 0,44. 1,11. 2,84. 7,35. Particle size (µm) Figure 46. Composition of main inorganic contents in from DLPI. Combustion of wood pellet, full load 12.8 kW.. 28.

(29) 100%. Composition (mole-%). 90%. Ti Zn P Na Mn Mg K Fe Ca Ba Al S Cl. 80% 70% 60% 50% 40% 30% 20% 10% 0% 0,05. 0,17. 0,44. 1,11. 2,84. 7,35. Particle size (µm). Figure 47. Composition of main inorganic contents in dust from DLPI. Combustion of RCG pellet, full load 13.3 kW.. Submicron particles formed during combustion of bark consists mainly of silica, sodium, potassium, sulphur and chlorine. The number of analysed species was fewer in this sample – Ti, Zn, Mn, Fe and Al are not included. Silica and sodium is often a result of contamination, sodium may be in the form of NaCl. Probably potassium chloride KCl, and potassium sulphate, K2SO3 are found also here. Particles larger than 1 µm generally consist of inorganic ash fragments that follow the flue gas. Therefore, particles > 1 µm consists of a wide range of species, for example for wood pellets aluminium, calcium, iron, magnesium and titan, and the composition is generally not related to combustion conditions.. 100%. Composition (mole-%). 90% P Si Na Mg K Ca Ba S Cl. 80% 70% 60% 50% 40% 30% 20% 10% 0% 0,05. 0,17. 0,44. 1,11. 2,84. 7,35. Particle size (µm). Figure 48. Composition of main inorganic contents in dust from DLPI. Combustion of bark pellet, full load 15.7 kW.. 29.

(30) 3.5. Contents in bottom ash. Composition of inorganic species in bottom ash from the fuels in this project is shown in Figure 49. The amount of produced bottom ash was not measured, and therefore no absolute values can be given. The bottom ashes consist of species and compounds with high melting temperature, typically including calcium and silica. Also potassium and aluminium are found in the bottom ash. 100% Si Ti Zn P Na Mn Mg K Fe Ca Ba Al S Cl. Composition (mole-%). 80%. 60%. 40%. 20%. 0% Wood. RCG. Bark. Oilseed straw. Figure 49. Composition of main inorganic contents in bottom ash. Full load.. 30.

(31) 4. DISCUSSION. 4.1. Implications on combustion by high ash content. The conversion of a solid fuel can be divided into three more or less consecutive steps: drying, devolatilization and char combustion. The combustible volatile compounds leave the fuel as CO, CH4 and different hydrocarbons, which are combusted in the gas phase downstream of the solid fuel. The fuels in this project are fairly similar in size, form, moisture and volatile content. To evaluate influence of ash content on drying and volatilization of the separate fuel pieces a different experimental approach is needed, and these effects are not considered here. From general knowledge, the influence from high ash content on drying and volatilization are considered to be quite small. Here, the influence on char combustion is considered. From the theory of coal combustion, see among others [2], high ash content may have the following effects: 1. Thermal effects Ash consumes energy as it is heated to high temperatures and changes phase. 2. Irradiative properties Irradiative properties of ash differ from those of char. 3. Particle size Particle break up toward end of burnout. 4. Catalytic effects Minerals may change char reactivity. 5. Hindrance effects Ash forms a barrier for oxygen to reach the char surface. Softening and melting may worsen the effect. From biomass combustion theory the following implication should be considered: 6. Char reactivity may differ in different biomasses. To evaluate the implications of (1) – (4) and (6) detailed experiments are needed. In this project it is assumed that thermal, irradiative and catalytic effects are marginal because of the, compared to some coals, quite low ash contents. Char particle break up has not been evaluated. Reactivity has not been evaluated, but when the combustion is diffusion controlled (see below) differences in reactivity are not crucial. The remaining parameter is hindrance effects (5) which are considered an important parameter for the results in this project. Combustion of ash rich pellets can be described using the constant density, shrinking core model, see Figure 50. This model applies to diffusion controlled combustion of a fairly large particle in a hot environment. In this model, oxygen is consumed as soon as it reaches the surface of the char particle. Therefore the particle shrinks until all char is combusted, and the density of shrinking core keeps constant. The inorganic parts (the ash) stay as a “shell” covering the burning char particle. This shell acts as a hindrance for oxygen diffusion and the char combustion time is prolonged. From the combustion tests in this project it was concluded that an increase of ash content may lead to a poorer combustion and even extinction of the glow bed. Poorer combustion is manifested by high CO emissions and (in one case) low flue gas temperature.. 31.

(32) Figure 50. Shrinking core model of a large particle in a hot environment. [3]. From (5) it is concluded that ash may hinder diffusion of oxygen. Ash can act as a hindrance in two ways: 1. Ash is not transported away from the grate but piled up above the supplied fuel. 2. Ash forms a shell around each fuel particle and decrease the oxygen diffusion velocity. When oxygen is hindered from reaching the char, the char combustion time is extended and the risk for increasing amount of unburnt char in the bottom ash is obvious, though this was not the case for any of the fuels here. Another effect of lower diffusion velocity may also be lower power. As all fuels were not tried in all burners no conclusions about power can be drawn in this project.. 4.2. The burners. The three burners used in the project fulfil high quality requirements. Burners A and B are designed and P-marked for wood pellets. Burner C is designed for oat grain. They represent three different designs: burner A has no mechanical mean for moving bottom ash on the grate during combustion, and the glow bed is covered by a flame director, see Figure 1. Burner B pushes fuel and ash upwards and the glow bed is exposed to the surrounding combustion department. Burner C pushes fuel and ash forwards, inside a cylinder. In burner A it was possible to combust bark pellets with an ash content of 3.4 %, though the load had to be reduced and the grate had to be cleaned after 3 hours (full load) and 2 hours (minimum load). The combustion was not optimal with averaged CO emissions higher and the flue gas temperature lower than is normally expected from combustion of wood pellets. This indicates insufficient combustion of the char with low temperature in the glow bed, leading to low temperature also in the gas phase which resulted in incomplete combustion of CO. The ash piling up on the glow bed acts as a hindrance for oxygen to reach the char surface. With fuels with even higher ash content this effects was even more accentuated, and combustion was not possible to sustain. Normally, when burner A is used in a house, the combustion periods are shorter and cleaning of the grate with the scrape is done with shorter intervals and a fuel such as bark should not be a problem, provided that the load and the cleaning periods are adjusted to the fuel. Still, to achieve really good gas combustion and thorough burnout of char, actions should to be taken to enhance oxygen diffusion and/or temperature in the flue bed. Burner B continuously pushes the char and ash upwards and, finally, above the rim of the burner. Depending on the volume and structure of the ash it piles up to a certain height that covers the char before it falls off. The piled up ash acts at a “hat” that effectively covers the char. The oxygen is hindered from reaching the char and the temperature is lowered. Neither Reed Canary Grass pellets nor oilseed straw pellets were possible to combust in this burner. In this design, the glow bed is exposed to the surrounding 32.

(33) combustion compartment, and therefore the temperature in the glow bed is lower compared to the glow bed in burner C. Nonetheless, the barley straw pellets reached a temperature where they sintered into hard pieces that suffocated the combustion. Burner C continuously pushes the char and ash forward. In this process, fresh fuel and glowing char is not covered by ash and can more easily be reached by oxygen compared to the glow bed in burner B. Also, the closed design of the burner keeps the temperature in the glow bed higher compared to a design where the glow bed is exposed to the surrounding.. 4.3. Emissions from the different fuels. CO emissions from the ash rich fuels were higher that should be expected from good combustion of wood pellets. The combustible parts of the fly ash particles, measured as weight-% of dust after filter sampling of total dust, were low, showing a generally good combustion performance with low amount of unburned hydrocarbons and soot in the flue gas. This also reflects the OGC emissions that were less than 5 mg/Nm3 at 10 % O2 at full load. Unburnt in bottom ash were for the ash rich fuels 2 – 11 weigh-% which shows a good burn out of the char. Unburnt in bottom ash from wood pellets were even higher, reflecting the not-so-good performance of the wood combustion test. Measured emissions of SO2 and HCl in this project are low. Only combustion of RCG resulted in emission values exceeding the order of combustion of wood pellets, but they are still low compared to the emissions measured during combustion of oat grain. From combustion of grains, rapid corrosion in flue gas channels has been reported. Especially chlorine corrosion that appears in the form of pot holes, are reported to destroy flue gas channels in a short time. From the use of grain in combustion devices in Sweden and Denmark strategies for combustion has developed that are generally accepted and aim to avoid corrosion problem. These strategies are shortly: 1) Avoid condensation of moisture in the heat exchanger and flue gas channel by keeping the temperature high. Recommended temperatures in the boiler are often above 70 ˚C, and in the flue gas channel top above 80 ˚C, but these recommendations can vary between appliances. 2) Keep the flue channel dry by the use of a pressure regulator. 3) Use materials in the flue gas channel that are not sensitive to corrosion. 4) Avoid low loads that may cause condensation on cold areas. 5) Use a burner with sustained glow that keeps the heat exchanger and flue gas channel warm also between combustion periods. If these strategies are applied in wise combinations, corrosion problems are avoided. Because of the low emissions of SO2 and HCl measured from the fuels in this project, no urgent measures are to be taken. Nevertheless, it is important to in the future survey these corrosive emissions, because variations in fuel content and ash composition may influence the formation of corrosive gases. Let out of the chimney, these gases are also acidifying and harmful for the environment. Therefore, if there will be a massive expansion of combustion of fuels containing sulphur and chlorine in the future, it might be necessary to reduce these emissions. Fuel nitrogen content in the fuels was: wood 0.1 %, bark 0.4 %, oilseed straw 0.5 %, RCG 0.9 %. Emissions of nitrogen oxide follow the fuel content with: wood 137, bark 410, oilseed straw 478 and RCG 940 mg/Nm3 at 10 % O2 (full load). At minimum load, CO emissions are higher and NO2 emission lower than at full load, though the difference in all cases but bark is quite small. The conversion of fuel nitrogen to NOx follows the nitrogen content with the highest conversion (37 % of fuel nitrogen resulting in NO2). 33.

(34) during wood combustion and the lowest (8 and 11 %) during oat grain combustion. Oat grain has a fuel nitrogen content of 2.2 %. Methods to reduce NOx are well known from large scale combustion where emissions are limited. A widely used primary measure is the use of stages air supply to create a reducing zone with adequate residence time and temperature. In small-scale appliances, these measures are not easily realized because of the small size of the equipment. Secondary means to reduce NOx are considered to expensive and also to complicated to apply on small-scale furnaces. To reduce NOx emissions from fuels with high nitrogen content, there is still a need to transfer and adapt knowledge and methods from large scale combustion. Emissions of total dust are for RCG pellets in the same order as for wood pellets (29 – 75 mg/Nm3 at 10 % O2), and for bark pellets slightly above 100 mg/Nm3 at 10 % O2. For oilseed straw pellet the dust emissions are considerably higher, 639 at full load and 359 mg/Nm3 at 10 % O2 at minimum load. Emission of total dust don not directly correspond to fuel ash content, which is shown in Figure 51. Dust mg/Nm in dry gas at 10 % O2. 700 Total dust. 600 500 400. 3. 300 200 100 0 0,3 0,3 (Wood (Wood full load) min load). 3,1 (Oat 3,4 (Bark 3,4 (Bark 3,6 (Oat 4,7 4,7 7,9 (RCG 7,9 (RCG grain) full load) min grain) (Oilseed (Oilseed full load) min load) full load) min load) load). Ash (weight-% dm) Figure 51. Measured total dust as function of ash content.. Measured particle mass concentrations showed that the fly ash consisted mainly of submicron particles. Chemical analyses showed that submicron particles from wood combustion were dominated by potassium as positive ion, and sulphur and chlorine as negative ions, which is typical for wood fly ash. Particles from bark combustion were quite similar to particles from wood combustion, but potassium was now accompanied by sodium, and the share of chlorine was higher. Particles from Reed Canary Grass pellets exhibited phosphor as positive ion together with potassium. Also here, sulphur and chlorine were main negative ions. Unfortunately, the chemical analysis of fly ash from oilseed straw was corrupt. The comparably high fuel chlorine content of oilseed straw, the low content of chlorine in bottom ash and the low emission of HCl indicates that most of the chlorine were found in the fly ash.. 34.

(35) 5. CONCLUSIONS AND RECOMMENDATIONS. 5.1. Combustion performance. From the experiments it was concluded that appliances optimized fro wood pellets will have to be further adapted to and optimized for ash rich pellets. To succeed with combustion of ash rich pellets the following has to be ensured: –. The ash needs to be hindered from piling up and cover the glow bed.. –. The ash has to be removed from the grate.. –. High ash content acts as a hindrance for oxygen diffusion and thus – char residence time has to be long enough for complete combustion of the char, – the temperature in the glow bed has to be high enough for complete char combustion.. –. Combustion of the char bed is crucial for complete combustion of CO in the gas phase. Therefore temperature and mixing in the gas phase have to be secured.. In this project, the horizontal burner that pushes the fuel and ash forwards, showed best performance. By pushing the glow bed forwards, ash did not pile up as and form a top lid, which more easily happens in a burner that burns upwards. Also, the temperature in the horizontal burner is higher than in the upwards burning burner, and temperature is crucial when oxygen diffusion velocity is slowed down. To provide long enough char residence time the grate has to be large (long) enough. The temperature in the glow bed can be kept high if it is protected from irradiative heat exchange from colder surfaces. This means that ash melting temperature can not be low, and/or that the ash sinters in a way that does not disturb the glow bed, for example in the form of small pebbles that are easily moved away from the grate. Different fuels may have different ash properties and therefore react differently. For example, in this project Reed Canary Grass pellets had highest content of ash but better combustion result than oilseed straw pellets that had slightly lower ash content. Residence time for high temperature in the gas phase has to be long enough and mixing good enough to secure thorough combustion of CO.. 5.2. Fuel specific emissions. The risk for corrosion problems is not directly connected to fuel content of sulphur and chlorine. Still, it is important to in the future keep an eye on these emissions, because variations in fuel content and ash composition may influence the formation of these corrosive gases. SO2 and HCl are also acidifying and harmful for the environment, which should be paid attention to if combustion of fuels containing sulphur and chlorine increase a lot in the future. Emissions of nitrogen oxides may be fairly high from these kinds of fuels. To reduce this emission from small-scale appliances, existing knowledge and means of reduction during large scale combustion has to be transferred and adapted to small-scale appliances. Secondary means of flue gas cleaning may be needed.. 35.

(36) Emissions of total dust from small-scale combustion consist, during good combustion, mainly of submicron, inorganic particles. Emission levels are not linearly connected to ash content. The amount and composition depends on the content and composition of specific species, for example sulphur, chlorine, alkali and phosphor. Emission limits below 100 mg/Nm3 at 10 % O2 was reached by one of the fuels in this study (besides the wood fuel). For the other fuels actions such as adapting low particle emission combustion designs, additives or secondary cleaning devices must be used to reach such low emissions.. 36.

(37) REFERENCES [1] M. Rönnbäck, H. Persson and K. Segerdahl, Spannmålsbrännare - funktion, säkerhet och emissioner, SP Sveriges Provnings- och Forskningsinstitut: Borås. (2005) [2] G.L. Borman and K.W. Ragland, Combustion Engineering. (1998), U.S.A.: WCB/McGraw-Hill [3] H. Thunman, Biofuel Combustion, Department of Energy Conversion, Chalmers University, Göteborg: Göteborg. (2004). ACKNOWLEDGEMENT The Swedish Energy Agency has supported this project which is gratefully acknowledged.. 37.

(38) APPENDIX A Table 7 Results from measurements and analyses.. Wood Wood pellet, full pellet, min Time for calculation of average values (min). RCG, full. RCG, min 362. 360. 362. 360. 12,8 1 Very little, grey. 5,2 1 Very little, grey. Flue gas temperature (ºC). 157. 109. 13,2 1 More than wood, black 167. Unburt in bottom ash (weight-% db). 79. Unburt in fly ash (weight-% db). 13. Power (kW) Ash category Fly ash deposits. Bark Bark Oilseed straw Oilseed straw pellet, full pellet, min pellet, full pellet, min 250. 239. 351. 381. 7,8 15,6 4,2 1 2 2 More than Substantial, Substantial, wood, black black black 110 116 66. 11,9 1 Substantial, white. 8,8 1 Substantial, white. 182. 123. 2. 11. 2. 9. 3. 11. O2 (vol-%). 10,8. 12,6. 11,2. 11,9. 9,1. 12,2. 11,5. 12,4. CO2 (vol-%). 9,5. 7,6. 9,1. 8,3. 11,0. 8,0. 8,7. 7,8. OGC (mg/Nm³ d.g. at 10 % O2). 3. 6. 5. 16. 3. 17. 4. 8. CO (mg/Nm³ d.g. at 10 % O2). 201. 388. 153. 426. 360. 576. 940. 1168. NO2 (mg/Nm³ d.g. at 10 % O2). 137. 134. 664. 641. 410. 343. 478. 475. 0,016. 0,031. 0,012. 0,034. 0,03. 0,05. 0,075. 0,093. Dust (mg/Nm³ d.g. at 10 % O2). 38. 69. 75. 29. 126. 116. 639. 359. HCl (mg/Nm³ d.g. at 10 % O2) SO2 (mg/Nm³ d.g. at 10 % O2). 2 1. CO (%). 6 25. 1 2. 1 9.

(39) 39 1002 ISO/IEC 17025. Rapport 1 i projektet: Utvärdering av utvecklingsstatus för småskalig förbränning av pellets från nya askrika råvaror askor och filter. Föremål Två askor och 2 stoftfilter insända av uppdragsgivaren. Provmärkning: 1) Aska träpellets nominell (bottenaska) 2) Aska rörflen nominell 070821 (bottenaska) 3) Stoftfilter, träpellets nominell 4) Stoftfilter, rörflen nominell Provmängd: Askor 80 g /prov. Förpackning: Plastpåsar Ankom KM: 2008-02-18 Provningsdatum: Vecka 8-10, 2008.. Uppdrag Askor Bestämning av fukt, aska, klor, svavel, huvudelement (Al,Si,Fe,Mn,Ti,Mg,Ca,Ba,Na,K,P) samt zink. Filterprov Bestämning av oförbränt som glödförlust.. Metoder Askor Aska: mod. SS 18 71 87 Klor: SS 18 71 85 Svavel: SS 18 71 86 Huvudelement: - Al, Si, Fe, Mn, Ti, Mg, Ca, Ba, Na, K, P mod. ASTM D 3682 Zink: mod. ASTM D 3683 Filter Oförbränt som glödförlust:. Inaskning vid 550ºC*. * Ej ackrediterad metod.

(40) 40. Resultat - Askor På torrt prov 1. 2. Aska, vikt-% Svavel, S, vikt-% Klor, Cl, vikt-%. 21,3 0,05 0,04. 97,9 0,06 0,01. Aluminium, Al, vikt-% Kisel, Si, vikt-% Järn, Fe, vikt-% 0,21 Mangan, Mn, vikt-% Titan, Ti, vikt-% Magnesium, Mg, vikt-% Kalcium, Ca, vikt-% Barium, Ba, vikt-% Natrium, Na, vikt-% Kalium, K, vikt-% Fosfor, P, vikt-%. 0,23 0,88 0,67 0,69 0,01 1,04 5,56 0,05 0,07 2,29 0,33. 0,79 39,2 0,29 0,07 1,07 3,56 0,04 0,28 3,00 1,29. Zink, Zn, mg/kg. 49. 61. 3. 4. 9. 12. Resultat på torrt filterstoft. Oförbränt som glödförlust, mg. SP Sveriges Tekniska Forskningsinstitut Kemi och Materialteknik - Oorganisk analytisk kemi. Conny Haraldsson. Mathias Johansson. Tekniskt ansvarig. Teknisk handläggare. Bilaga Mätosäkerhet.

(41) 41 1002 ISO/IEC 17025. Rapport 2 i projektet: Utvärdering av utvecklingsstatus för småskalig förbränning av pellets från nya askrika råvaror askor och filter. Föremål Två askor insända av uppdragsgivaren. Provmärkning: 1. Aska Barkpellets nominell effekt (bottenaska) 2. Aska Rapspellets nominell effekt (bottenaska) 3. Stoftfilter Barkpellets nominell 07-06-05 4. Stoftfilter Rapspellets nominell Provmängd: Prov 1 ca 150g Prov 2 ca 80g Förpackning: Plastpåsar Ankom KM: 2007-06-13 Provningsdatum: Vecka 25-34, 2007. Uppdrag Askor: Bestämning av fukt, aska, klor, huvudelement (Al,Si,Fe,Mn,Ti,Mg,Ca,Ba,Na,K,P) samt zink. Filterprov: Bestämning av oförbränt som glödförlust.. Metod Askor Aska: Klor:. mod. SS 18 71 71 Lakning- kvantifiering med jonkromatografi. Huvudelement: - Al, Si, Fe, Mn, Ti, Mg, Ca, Ba, Na, K, P - Zink: Filter Oförbränt som glödförlust:. mod. ASTM D 3682 mod. ASTM D 3683. Inaskning vid 550ºC.

(42) 42. Resultat- Askor På torrt prov. 1. 2. Aska, vikt-% Svavel, S, vikt-% Klor, Cl, vikt-%. 89,1 0,25 0,08. 98,2 1,47 0,97. Aluminium, Al, vikt-% Kisel, Si, vikt-% Järn, Fe, vikt-% 1,30 Mangan, Mn, vikt-% Titan, Ti, vikt-% Magnesium, Mg, vikt-% Kalcium, Ca, vikt-% Barium, Ba, vikt-% Natrium, Na, vikt-% Kalium, K, vikt-% Fosfor, P, vikt-% Zink, Zn, mg/kg. 2,44 11,5 0,22 1,55 0,11 2,35 26,7 0,39 0,76 3,72 1,33 800. 0,23 4,91. På inaskat prov vid 550ºC Aluminium, Al, vikt-% Kisel, Si, vikt-% Järn, Fe, vikt-% 1,45 Mangan, Mn, vikt-% Titan, Ti, vikt-% Magnesium, Mg, vikt-% Kalcium, Ca, vikt-% Barium, Ba, vikt-% Natrium, Na, vikt-% Kalium, K, vikt-% Fosfor, P, vikt-% Zink, Zn, mg/kg. 2,73 12,8 0,22 1,74 0,12 2,62 29,9 0,43 0,86 4,16 1,49 900. Resultat på torrt filterstoft Oförbränt som glödförlust (mg). 0,07 0,02 1,57 29,4 0,04 0,90 8,90 1,36 36 0,23 4,98 0,07 0,02 1,59 29,8 0,04 0,91 9,02 1,38 37. 3. 4. 23. 11. SP Sveriges Tekniska Forskningsinstitut Kemi och Materialteknik - Oorganisk analytisk kemi Conny Haraldsson. Mathias Johansson. Tekniskt ansvarig. Teknisk handläggare. Bilaga Mätosäkerhet.

(43) SP Technical Research Institute of Sweden develops and transfers technology for improving competitiveness and quality in industry, and for safety, conservation of resources and good environment in society as a whole. With Sweden’s widest and most sophisticated range of equipment and expertise for technical investigation, measurement, testing and certification, we perform research and development in close liaison with universities, institutes of technology and international partners. SP is a EU-notified body and accredited test laboratory. Our headquarters are in Borås, in the west part of Sweden.. SP Technical Research Institute of Sweden. Energy Technology. Box 857, SE-501 15 BORÅS, SWEDEN. SP Report 2008:31. Telephone: +46 10 516 50 00, Telefax: +46 33 13 55 02. ISBN 978-91-85829-48-4. E-mail: info@sp.se, Internet: www.sp.se. ISSN 0284-5172. www.sp.se.

(44)

No results found