Experimental investigations on combustion and emission behaviour during oxy-coal combustion

Abstract

As the most abundant non-renewable energy source available, coal has traditionally played a major role in ensuring the security of energy, and will continue to play a key role in the world energy mix. The burning of coal has however always been a subject of environmental concern. In recent years, the emission of green house gases and global climate change has emerged as the largest environmental challenge. As coal fired power plants are categorised among the least carbon efficient energy producer in terms of CO2 emission per unit of electricity generated, an immediate technological response is anticipated. Although, improving the efficiency of coal fired plants can decrease CO2 emissions to some extent, fossil fuel based carbon capture and storage technologies will have to bear the significant share in power generation, if reduction or even stabilisation of CO2 emission is to be envisaged. One promising technology with a potential for near complete capture of CO2 is the oxy-coal combustion process with flue gas recirculation. This technology is however still at an early stage of development, and until now there are no full-scale commercial plants based on this technology. The combustion of coal in this process takes place with almost pure oxygen and recycled flue gas resulting in a flue gas stream of almost pure CO2, ready for geo-sequestration. The flue gas is recycled back into the furnace to control the temperature and maintain the heat flux profiles within the furnace. Research efforts are in progress for a better understanding of the oxy-coal combustion process. However, many issues still need to be addressed in order to obtain improved fundamental understandings. The primary objective of this study is therefore, a comprehensive and well-planned experimental investigation to further understand the emission and combustion behaviour during O2/CO2 combustion using a series of fuels, and is sub-divided into:

I. Determination of combustion and emission behaviour during un-staged combustion. II. Determination of NOx reduction potential during oxidant staged combustion. III. Determination of the fate of recycled NO. IV. Determination of the fate of SO2 and its impact on oxy-coal combustion.

A 20 kW electrically heated, once through furnace was used for this investigation, which enabled highly flexible parametric studies as well as reliable and repeatable measurements. To simulate an oxy-coal combustion environment, almost pure O2 and CO2 supplied from tanks were mixed in a highly flexible mixing station and were supplied to the furnace via three different streams of the burner. Parametric studies were also performed for air-blown combustion to compare the combustion and emission behaviour in an O2/N2 and an O2/CO2 mixture. In addition, measurements during simulated O2/RFG (dry recirculation) environment was also conducted to investigate the behaviour of recycled flue gas species i.e. NO and SO2 by injecting known concentrations of NO and SO2 into the furnace along with the O2/CO2 mixture or air. Four coals (ranging from medium volatile bituminous coal to pre-dried brown coal), natural gas and char were used for various parametric studies. As the initial application of oxy-coal combustion technology will most likely be a retrofit in existing pulverised coal furnaces, modifications are required to match the flame and heat transfer characteristics of air-blown furnaces. An investigation was therefore commenced with detailed in-flame and furnace exit measurements during un-staged combustion, using two bituminous coals (Klein Kopje and Ensdorf) and two pre-dried brown coals (Lausitz and Rhenish). Emission and combustion behaviour were determined at different O2/CO2 volumetric concentrations i.e. 21% O2/ 79% CO2 (OF21), 27% O2/ 73% CO2 (OF27) and 35% O2/ 65% CO2 (OF35). To further establish the emission and combustion behaviour during oxy-coal combustion, investigation in explicitly gas-phase using natural gas was also performed. This investigation was able to develop a further understanding of the combustion behaviour of volatiles in O2/CO2 environment, whilst avoiding the influence of particles. This is thought to aid in improved prediction of homogeneous reactions taking place during oxy-coal combustion for scale-up and modelling purpose. Similar flame temperature profile and gaseous concentration profile along the length of the reactor was observed for coals as well as natural gas firing during OF27 and air-blown combustion. The ash/char composition collected along the reactor length during coal firing was also observed to be similar for OF27 and air-blown combustion. This indicates that approximately 27 vol. % oxygen will be required during O2/CO2 combustion (representing oxy-coal combustion with dry flue gas recirculation) to achieve similar combustion performance of air-blown pulverised coal furnaces. The result further reinforces the findings from previous authors that oxygen concentration if properly adjusted gives the heat transfer and flame character similar to air-fired furnaces. It was further found that, the conversion of fuel-N to NO for all 4 coals tested is lower in a CO2 environment, and is in agreement with most historical studies. When a comparison of the NO emission rate and the fuel nitrogen conversion rate between different coals during OF27 combustion is made, a trend typical to a conventional air-fired, fuel lean combustion scenario was observed, i.e. the higher the fuel-N content (for coals with comparable volatile content), the greater the NO emission, and the higher the volatile content (for coals with comparable fuel-N content), the greater the conversion of fuel-N to NO. It is however noted that this investigation was carried out in a once through furnace. As the purity requirement of CO2 for storage is still uncertain, it is rational to minimise impurities in the furnace when possible. The next focus of this investigation was therefore on the reduction of nitric oxide by oxidant staging during oxy-coal combustion. Detailed investigation of NOx formation mechanisms during oxidant staged combustion in CO2 (27% O2/73% CO2 by volume) and N2 (air-blown) environment was conducted by firing two pre-dried brown coals (Lausitz and Rhenish) and two bituminous coals (Klein Kopje and Ensdorf). Investigations of oxidant staged combustion in a CO2 environment of 27% O2/73% CO2 was carried out as the initial results during un-staged combustion indicated that the combustion and emission behaviour for OF27 combustion was comparable to air-blown combustion (it is however noted that as the technology matures, future applications will be with reduced flue gas recycle resulting in higher combustion temperature). Gaseous concentration profile measurements for different burner oxygen ratios and residence times in the reduction zone were conducted. Measurement of in-flame HCN and NH3, using a FTIR was also performed to determine the NOx formation mechanisms during O2/CO2 combustion. This investigation not only aided in understanding the NOx reduction potential during O2/CO2 combustion but also provided design features required for a low NOx oxy-coal burner. Fuel NOx formation mechanisms in a fuel rich environment during O2/CO2 combustion, as demonstrated by the formation and destruction of HCN and NH3 is similar to air-blown combustion. NO formed in the mixing zone reacts with hydrocarbon radicals to produce HCN or NH3, which are converted to N2 in the oxygen deficient reduction zone for both cases. As with conventional air-blown combustion, the formation of HCN and NH3 is also dependant on the coal rank, as only HCN was detected for a medium volatile bituminous coal, Klein Kopje while both HCN and NH3 were detected for the Lausitz brown coal. Furthermore, results indicate that oxidant staging for NOx reduction is equally or even more effective for O2/CO2 combustion in terms of NOx reduction potential, as the conversion of fuel-N to NO and NOx emission rate is lower than corresponding air-blown combustion. However, as mentioned earlier, this investigation was carried out in a once through furnace. As with conventional air-blown combustion, the reduction of NO was observed to be proportional to the partitioning of fuel-N into the gas-phase. This is seen as an encouraging result from oxy-coal combustion perspective, as the amount of fuel-N partitioned into the gas-phase is a function of temperature, and manipulation of local temperature is considered to be simpler during oxy-coal combustion. Direct injection of O2 into the flame in order to increase the devolatilisation rate without affecting the overall heat transfer performance could be one viable option that can take advantage of existing infrastructure of an oxy-fuel plant for further reduction of NO. Oxy-coal combustion also requires recirculation of the flue gas to moderate the furnace temperature. Apart from CO2 and water vapour, the recycled flue gas also contains pollutants from coal combustion and the impact of such pollutants recycled back into the furnace needs to be evaluated for successful application of oxy-fuel process. The next focus of this investigation was therefore the determination of the fate of nitrogen oxide recycled back into the furnace. Fate of NO recycle back into the furnace was determined by injecting a known concentration of pure NO via the burner or the over-fire port, along with air or O2/CO2 mixture, depending upon the investigation being carried out. A medium volatile bituminous coal (Klein Kopje) and a brown coal (Lausitz) were used for this investigation. Homogeneous and heterogeneous reduction of recycled NO was also determined by firing natural gas and char of a brown coal. Investigations were carried out during OF27 and air-blown combustion, with and without oxidant staging. This investigation has not only enhanced the understanding of NOx re-burning mechanism during oxy-coal combustion, but may also provide the design features of low NOx oxy-coal burner and assess the requirements of flue gas clean-up devices. Reduction of recycled NO by heterogeneous reactions with active carbon sites as indicated by measurements with char of brown coal is prominent when oxygen is in excess. However, as oxygen availability decreases with a decrease in burner oxygen ratio during staged combustion; homogeneous reactions (reactions with active hydrocarbon radicals) become prominent. Similarly, concentrations of recycled NO within the range of 1200 ppm showed no influence on recycled NO reduction efficiency during char and coal combustion, indicating that the reduction efficiency will remain unaffected with an increase in the concentration of recycled NO. Also, the reduction of recycled NO during coal combustion for both air-blown and oxy-coal combustion was seen to be entirely dependent on combustion conditions i.e. the burner oxygen ratio and residence time in the reduction zone. Recycled NO reduction of almost 100% was achieved during staged combustion with a burner oxygen ratio of 0.75, and approximately 50% reduction was achieved during un-staged combustion. Similarly, a longer residence time in the reduction zone during staged combustion had a positive influence on recycled NO reduction when NO was introduced via the burner. The reduction of recycled NO, at a burner oxygen ratio of 0.75 is almost similar for air-blown and oxy-coal combustion, indicating that the combustion media has little influence on recycled NO reduction when oxygen availability is low, or when homogeneous reactions are prominent. However, as oxygen availability increases with an increase in burner oxygen ratio (or when heterogeneous reactions are dominant), reduction of recycled NO is higher in a CO2 media. This is most probably due to the presence of higher in-flame CO concentrations during oxy-coal combustion. Reduction efficiency of recycled NO is also dependent on the location of NO injection, as the reduction of recycled NO when injected via the over-fire port is much lower than when introduced via the burner. By considering the overall reduction of recycled NO and stable NO concentration thus achieved, the NOx emission rate is between 0.24 to 0.37 times lower during oxy-coal combustion with 73 vol.% flue gas recycle when compared to corresponding air-blown combustion. This is in agreement with historical investigations carried out during oxy-coal combustion with flue gas recycle, further confirming that the reduction of NO recycled back into the furnace is the most important factor resulting in lower NOx emissions rate during oxy-coal combustion. Besides the emission of CO2, NOx, CO, etc., the combustion of coal is also associated with the conversion of fuel sulphur into compounds such as SO2, SO3, H2S, etc. The presence of sulphur components, especially H2S and SO3 are associated with high temperature corrosion at the furnace walls and super heater sections, and low temperature corrosion through the condensation of sulphuric acid in the economiser or air heater. Primary emphasis of the next investigation was therefore to assess the impact of combustion media (N2 or CO2) and recycled SO2 on the formation of gaseous sulphur components in the furnace (SO2 and H2S). The SO2 retention capacities of coals (ash) in the radiative and convective section of the furnace were also investigated. Investigations were carried out during air-blown and OF27 combustion, with and without oxidant staging. The impact of recycled SO2 was investigated by injecting a known concentration of synthetic SO2 through the secondary stream of the burner. This investigation not only assisted in understanding the risk of corrosion during oxy-coal combustion, but also aided in understanding the SO2 retention capacity of different coals in the low temperature convective section of the furnace and the flue gas recycle path. Lower temperature window in the convective and recirculation path before the ESP offers a potential for SO2 retention by sulphate forming elements in the ash. Regardless of a combustion media (CO2 or N2), if high concentration of active sulphate forming elements are present in the ash, the SO2 retained by the ash increases with an increase in SO2 concentration up to a certain concentration of SO2 in the flue gas. As SO2 concentration is much higher during O2/RFG combustion, the SO2 retained by the sulphate forming elements in the ash will also be higher. The most important information from these results is that, if the coal being used has high concentration of active sulphate forming elements, the opportunity of retaining maximum possible SO2 in the convective section and flue gas recirculation path leading to decreased SO2 accumulation needs be considered, when designing a cost effective oxy-coal plant. Similarly, the H2S/SO2 formation mechanism in a CO2 environment is seen to be basically similar to a N2 environment. However, the fraction of H2S with respect to SO2 is lower in a CO2 environment as well as in an O2/RFG environment, when compared to air-blown combustion. This may be due to the enhanced consumption of H2 to produce CO by the water shift reaction resulting in a lower reaction rate of sulphur and hydrogen molecules to form H2S. The concentration of H2S in an O2/RFG scenario on the other hand is much higher than air-blown combustion due to high in furnace SO2 concentrations. Therefore, from corrosion perspective, special attention needs to be given to oxy-coal operated furnaces. Low rank (high volatile), high sulphur coals might require even more attention, as both H2S fraction and the concentrations are seen to be much higher than medium volatile bituminous coal.

Fossile Energieträger haben aufgrund ihrer weltweiten Verbreitung und wirtschaftlichen Verfügbarkeit einen großen und verlässlichen Beitrag an der Energieversorgung. Die Technologien wurden ständig weiterentwickelt um zum Einen den spezifischen Verbrauch zu reduzieren und zum Anderen die Umwelteinflüsse wie Staub-, Schwefel- und Stickstoffoxidemissionen zu minimieren. Moderne Kohlekraftwerke emittieren ca. 750g/kWh CO2 bei einem Wirkungsgrad von rd. 46%. Diese hohen spezifischen CO2-Emissionen und die Tatsache, dass Kohle weiterhin eine weltweit hohe wirtschaftliche Verfügbarkeit besitzt, legen nahe für diese Kraftwerke Technologien zur CO2-Abscheidung und langfristigen CO2-Speicherung zu entwickeln. Die in der Entwicklung befindlichen Technologien firmieren unter der Bezeichnung „Carbon Capture and Storage“ kurz CCS-Technologien; diese stellen hierbei eine mittelfristig realisierbare Möglichkeit dar, die Kohle weiterhin klimaverträglich zur Energieversorgung zu nutzen. Die vorliegende Arbeit beschäftigt sich mit der so genannten „Oxyfuel-Technologie“, hierbei wird Sauerstoff anstelle von Luft für die Verbrennung verwendet. Dies führt zu hohen CO2-Konzentrationen im Abgas und begünstigt somit die Wirtschaftlichkeit des Gesamtprozesses. Um die Verbrennungstemperaturen im Feuerraum und die Wärmeübertragung zu kontrollieren werden Rauchgase in den Feuerraum zurückgeführt. Eine zentrale Aufgabenstellung der Arbeit bestand darin, die Auswirkungen der Rauchgasrezirkulation auf das Verbrennungsverhalten und auf die Schadstoffbildung zu untersuchen. Die Schwerpunkte der Untersuchungen bilden die Abbrandcharakteristik und die Bildung und Reduktion von Stickstoff- und Schwefeloxiden in einer CO2-reichen Atmosphäre. Neben einer Variation der Verbrennungsatmosphäre (Luft bzw. CO2-reich) wurden Kohlen mit Unterschieden im Inkohlungsgrad, des Stickstoff- und Schwefelgehalts, der Aschemenge und der Aschezusammensetzung verwendet. Durch die gezielte Auswahl der Einsatzstoffe können aus den Ergebnissen für ein breites Brennstoffband umfangreiche Erkenntnisse abgeleitet werden. Um die geplanten Parametervariationen erfolgreich durchzuführen, wurde ein für Luftverbrennung am Institut verfügbarer Reaktor um die Erfordernisse der Oxyfuel-Verbrennung erweitert. Adaptionen waren im Wesentlichen bei der Sauerstoff- und CO2-Zuführung am Brenner sowie für eine Sauerstoffstufung im Feuerraum erforderlich. Darüber hinaus wurden Möglichkeiten geschaffen, um NO und SO2 gezielt den Verbrennungsgasen beizumischen, um so die Auswirkungen einer Rezirkulation dieser Komponenten auf Feuerraum- und Abgaskonzentrationen zu untersuchen. Neben Standardkomponenten wie O2, CO2, CO, NO, NOx, SO2 wurden insbesondere im Brennernahfeld die Komponenten NH3, HCN und H2S gemessen. In einem ersten Schritt wurden im Vergleich zu Verbrennungsversuchen mit Luft drei Feuerungseinstellungen mit unterschiedlichen Sauerstoffkonzentrationen (21, 27 und 35 Vol.%) im Verbrennungsgasgemisch (O2/CO2) durchgeführt (Kapitel 5). In Übereinstimmung mit anderen Publikationen konnte gezeigt werden, dass die Abbrandcharakteristik der 27% Oxyfuel-Einstellung große Übereinstimmungen mit der konventionellen Luftverbrennung aufweist. Vor diesem Hintergrund konzentrierte sich ein Großteil der vergleichenden Parameterstudien in den folgenden Abschnitten auf diese beiden Einstellungen (Luft bzw. 27%O2,73%CO2). Die Ergebnisse dieser Untersuchungen unterstrichen, dass eine belastbare vergleichende Diskussion und Darstellung der Luft- und Oxyfuel-Einstellungen mit der bisher verwendeten Maßeinheit mg/m³ i.N nicht zielführend ist. Für vergleichende Betrachtungen wird somit die Schadstofffracht in mg/MJ am Feuerraumende herangezogen.

In den weiteren Kapiteln 6 und 7 wird gezielt auf die Bildung und Reduktion der Stickoxide eingegangen. Umfangreiche Profilmessungen der Komponenten NO, CO, NH3 und HCN bei variierenden Verweilzeiten (1s, 2s, 3s) und Sauerstoff- zu Brennstoff-Verhältnissen (0,75, 0,85, 0,95, 1,15) in einer Primärreaktionszone bilden hier die Basis der Untersuchungen. Darüber hinaus wurde eine mittelflüchtige und hochflüchtige Steinkohle (Klein Kopje, Ensdorf) sowie zwei Braunkohlen aus dem Rheinischen und Lausitzer Revier eingesetzt. Anhand der umfangreichen Datenbasis konnte gezeigt werden, dass auch in einer CO2-reichen Oxyfuel-Atmosphäre die sauerstoffgestufte Verbrennung ein probates Mittel darstellt, um die NO-Bildung aus Brennstoff-Stickstoff zu minimieren. Für alle eingesetzten Kohlequalitäten konnten bei Anwendung der Sauerstoffstufung Umwandlungsraten zu NO von weniger als 10% erzielt werden. In Kapitel 7 wurde untersucht, inwieweit mit den abgekühlten Rauchgasen rezirkuliertes NO in der Primär- und Ausbrandzone einer Feuerung bei ungestufter und gestufter Feuerungsführung reduziert wird. Es konnte gezeigt werden, dass sowohl für die mittel- wie auch hochflüchtigen Kohlen bereits bei ungestufter Verbrennung eine Reduktion von ca. 50% und bei Sauerstoff- zu Brennstoff-Verhältnissen von kleiner 0,95 in der Primärzone Reduktionsraten von 80 bis 100% erzielt werden können. Die im Rahmen von Kapitel 6 und 7 durchgeführten Untersuchungen verdeutlichen, dass die in einzelnen Versuchsanordnungen mit realer Rauchgasrezirkulation ermittelten verringerten NO-Frachten im Wesentlichen auf die Reduktion von rezirkuliertem NO in der Flamme zurückzuführen sind. Dieser Sachverhalt erklärt auch die starke Schwankungsbreite der bisher veröffentlichten Datenbasis, da das Reduktionspotential sowohl von der rezirkulierten Rauchgasmenge wie auch stark von den lokalen Sauerstoff- zu Brennstoff-Verhältnissen im Brennernahfeld abhängig ist. Eine Vielzahl der in der Arbeit erzielten Ergebnisse unterstreichen, dass eine primärseitige NO-Reduktion durch gestufte Verbrennung unter Oxyfuelbedingungen ein großes Potential hat und für ein breiteres Brennstoffband effizient angewendet werden kann. Das Verhalten von Schwefel im Hochtemperaturbereich der Feuerung und entlang des Rauchgasweges wird in Kapitel 8 betrachtet. Grundsätzlich konnte gezeigt werden, dass unabhängig von der Verbrennungsatmosphäre der im Brennstoff enthaltene Schwefel weitgehend zu SO2 umgewandelt wird. Es lässt sich auch kein nennenswerter Zusammenhang zwischen einer CO2-reichen Atmosphäre und der Bildung von H2S im Brennernahbereich ableiten. Es werden vielmehr aufgrund der deutlich erhöhten SO2-Konzentrationen unter Oxyfuel-Bedingungen insbesondere für hochflüchtige Brennstoffe höhere H2S-Konzentrationen im Brennernahfeld gemessen. Neben dem Sauerstoff- zu Brennstoff-Verhältnis im Brennernahfeld hängt die erzielte H2S-Konzentration vor allem von der SO2-Konzentration im Feuerraum und der Kohlequalität ab. So wurde z.B. für eine hochflüchtige Steinkohle (Ensdorf) eine Umwandlungsrate zu H2S von 44% bestimmt, während für die mittelflüchtige Steinkohle (Klein Kopje) nur eine Umwandlungsrate von 11% ermittelt wurde. Um das Potential der Eigeneinbindung von Schwefel in die Flugasche der jeweiligen Kohle zu untersuchen wurde die SO2-Konzentration im Rauchgas über einen Bereich von 500 vppm bis 4000 vppm variiert. Das Ca/S-Verhältnis der Brennstoffe lag zwischen 1,03 und 3,96. Die Eigeneinbindung des Schwefels steigt unter Oxyfuel-Bedingungen deutlich an und ist verursacht durch die prozessbedingten höheren SO2-Konzentrationen. Die Untersuchungsergebnisse unterstreichen, dass für eine Betrachtung der Schwefelbilanz die Eigeneinbindung verstärkt zu berücksichtigen ist. Die erhöhte Eigeneinbindung und der damit verbundene erhöhte SO3-Anteil in der Asche sollte auch im Hinblick auf die weiteren Verwertungswege berücksichtigt werden. Dieser Sachverhalt gilt insbesondere für die Steinkohleaschen, die üblicherweise unter Berücksichtigung der EN 450 der Zementherstellung zugeführt werden. Abschließend werden die Untersuchungsergebnisse kapitelübergreifend zusammengefasst und diskutiert sowie Empfehlungen für ein weiteres Vorgehen abgeleitet.