Next to the recirculation rate and that way the oxygen concentration in the oxidizer, the stoichiometry offers an option to control flame temperatures in oxyfuel processes. While using both, non-stoichiometric flames and recirculated flue gas as in the concept of controlled staging with non-stoichiometric burners (CSNBs), the recirculation rate could be reduced drastically. Reducing the recirculation rate in the oxyfuel process is a viable option for more efficient steam generator concepts. The auxiliary power demand, the size of the recirculation equipment and the steam generator size is reduced. The investigations in this paper focus on the interaction of two non-stoichiometric operating pulverized fuel oxycoal burners as used in the CSNB concept. Two different staging strategies have been investigated in detail as well as the influence of dry and wet flue gas recirculation. For that purpose species concentrations have been measured along the flue gas path in the combustion chamber. Additionally emissions at the combustion chamber exit as well as mass and energy balances of the process are reported. The species concentration profiles demonstrate the functionality of the CSNB concept. Either way the oxygen excess or the hydrocarbon excess of the first flame level could be balanced by the contrarily operating second flame level. The staged CSNB concept with low recirculation rate shows comparable combustion performance with the unstaged oxyfuel process with high flue gas recirculation. However the influences of the individual burner performance and the flue gas mixing, especially in the second burner level, is emphasized in the CSNB concept. In concerns of NOx emissions shows the arrangement, first sub- then over-stoichiometric combustion lower values because the initial NO formation in the first level is much lower. In case of over- and then sub-stoichiometric arrangement the NO reduction by the second sub-stoichiometric flame is low. In case of wet flue gas recirculation with increased vapor concentration in the flue gas atmosphere the heat transfer in the combustion chamber is slightly higher compared to the dry oxyfuel case.

To compensate the drawback of high flue gas recirculation rates specific for oxyfuel processes, a new concept based on staged combustion, called controlled staging with non-stoichiometric burners (CSNB) was investigated. A combination of over- and sub-stoichiometric burners avoids inadmissible high flame temperatures even with oxygen concentrations up to 40 vol.-% in the oxidant. The non-stoichiometric burners are arranged in such a way that the overall stoichiometry at the combustion chamber outlet is slightly over-stoichiometric similar to conventional combustion processes, so that full burn out is secured. This concept aims at a more efficient oxyfuel process due to the decreasing effort in the recirculation loop and a new designed, more cost effective, steam generator. For further process optimization of the CSNB concept and the adjacent steam generator layout are validated CFD simulations urgently required. This paper shows first steps in validation of the CSNB combustion concept against state of the art CFD codes. The CFD code was optimized for oxyfuel combustion including a new char combustion and gas radiation model. Experimental investigations and CFD modelling are showing good agreement in concerns of temperature, CO and CO2 profiles. However the prediction of the oxygen concentration differs significantly between experiment and simulation.

In oxy-combustion the fuel is burnt in a mixture of oxygen and recirculated flue gas to keep the temperature inside the furnace to levels similar to conventional combustion. This eliminates the atmospheric nitrogen from the process, leading to a flue gas consisting mainly of carbon dioxide and water vapor. Further on, the CO2 can be separated for storage purposes. A major drawback of the conventional oxy-fuel combustion technology consists in the high amount of flue gas that has to be recirculated in order to control the temperature level inside the furnace. A novel oxy-fuel firing concept based on a combination of pulverized coal burners operating under non-stoichiometric conditions is investigated as a solution for lowering the necessary flue gas recirculation rate, while keeping the temperature inside the furnace at feasible levels. This paper presents a numerical analysis of the most relevant aspects for this new firing concept, such as process specifics and limitations, burner design criteria, aerodynamic characterization of the near burner zone, flame ignition and temperature. First the process is defined via thermodynamic calculations which are necessary to establish the operating conditions and to generate sets of parameters for the design phase of the burners. Subsequently the parameters generated in the first phase are used as boundary conditions for the design of the burners via CFD simulations. The CFD code used in this study is updated for oxy-firing conditions with the recent developments in terms of gas phase reactions, char conversion modeling and radiative heat transfer in high temperature atmospheres with elevated CO2 concentration. Additionally, the most relevant aspects regarding the validation of the CFD code against in-flame experimental values are presented and discussed. The simulations show good agreement with the averaged experimental data collected along the flame centerline.

Non-stoichiometric burner operation in oxycoal processes has been proposed as a means to control flame temperature and heat exchange on admissible levels similar to air combustion when reducing the recirculation rate i.e. increasing the oxygen concentration in the oxidizer. Thus, the additional energetic and tangible expenditure of oxyfuel operation compared to conventional air firing can be lowered. Non-stoichiometic burner operation leads to gradual release of combustion enthalpy. On this way lower (adiabatic) flame temperatures are achievable compared to stoichiometric combustion. By combining two or more burners in over- and sub-stoichiometric operation global stoichiometries between lambda = 1.10-1.15 and low oxygen concentrations at the combustion chamber outlet are adjustable.In this work the theoretical foundation of staged combustion is presented. Results from measurements on a single flame are given which indicate the controllability of the flame temperature by varying stoichiometry. Gas species concentration profiles normal to the flame axis have been taken. Images obtained from an optical flame surveillance system allow for the visual interpretation of flame stability and shape. (C) 2012 Elsevier Ltd. All rights reserved.

Oxyfuel combustion is a technology for Carbon Capture & Storage from coal fired power plants. One drawback is the large necessary amount of recirculation of cold flue gases into the combustion chamber to avoid inadmissible high flame temperatures. The new concept of Controlled Staging with Non-stoichiometric Burners (CSNB) makes a reduction of the recirculation rate possible without inadmissible high flame temperatures. This reduction promises more compact boiler designs. We present in this paper experiments with the new combustion concept in a 3×70kW natural gas combustion test rig with dry flue gas recirculation of 50% of the cold flue gases. The new concept was compared to a reference air combustion case and a reference oxyfuel combustion case with recirculation of 70% of the cold flue gases. FTIR emission spectroscopy measurements allowed the estimation of spectral radiative heat fluxes in the 2–5.5μm range. The mixing of the gases in the furnace was good as the burnout and the emissions were comparable to the reference cases. The flame temperatures of the CSNB case could be controlled by the burner operation stoichiometry and were also similar to the reference cases. The heat flux in the furnace through radiation to the wall was higher compared to the oxyfuel reference case. This is an effect of the lowered recirculation rate as the mass flow out of the furnace and therefore the sensible heat leaving the furnace decreases. The higher oxygen consumption with lower recirculation rate could be compensated by a lower furnace stoichiometry. This was possible due to better burnout with increased oxygen concentrations in the burner. The results prove that a reduction of the flue gas recirculation rate in oxyfuel natural gas combustion from 70% down to 50% is possible while avoiding inadmissible high flame temperatures with the concept of Controlled Staging with Non-stoichiometric Burners.

The oxyfuel technology offers the possibility for CO 2 sequestration from coal fired power plants. One drawback is the need for a high external flue gas recirculation to avoid inadmissible high flame temperatures. The concept of controlled staging with non-stoichiometric burners (CSNB) allows a significant reduction of the commonly proposed flue gas recirculation rate while fulfilling all requirements on temperature limitations. The concept aims at a more efficient oxyfuel process with a higher degree of freedom for heat-flux adjustments suitable for a new generation of oxyfuel boilers. The steam generator size could be reduced and in this way a more cost effective steam generator concept is possible. Additionally the energy demand for the flue gas recirculation is lowered. This paper presents the experimental investigations of non-stoichiometric oxycoal flames. Temperature and gas profiles were taken to analyze the combustion behavior of coal with high oxygen concentrations in the oxidizer under oxygen deficiency and accordingly oxygen excess. In addition an optical flame monitoring system allowed a comparison of ignition, flame shape and stability. In the test rig lignite was burned under different stoichiometries ranging from 0.5 to 2.5 and different oxygen concentrations in the oxidant ranging from 30 to 40 vol.%. The thermal input of the burner was 70 kW at a total thermal input of 140 kW and a dry flue gas recirculation was used. The results were compared to a conventional air-blown combustion and showed that similar temperature ranges can be reached even with oxygen concentrations in the oxidizer as high as 40 vol.%. Keywords Oxyfuel Stoichiometry Experiments Flame temperature Species concentrations 1 Introduction Globally coal is the most important fuel for electricity generation. Due to the worldwide increasing energy demand and the missing alternatives for a cheap, secure and reliable electricity supply coal will remain a major part of the global energy mix [1] . Increased power plant efficiency is one way to decrease the greenhouse gas emissions from fossil fuel power plants. For further reductions of CO 2 emissions Carbon Capture and Storage Technologies (CCS) must be adopted. The oxyfuel process, where coal is burned in a mixture of oxygen and recycled flue gases for temperature moderation, is one of the most promising CCS-Technologies. Besides the high energy consumption for the oxygen supply and the CO 2 processing unit, the high temperatures of a combustion with pure oxygen are a drawback of the oxyfuel process. Several options exist to lower these temperatures. 1. Internal flue gas recirculation at the burner (rather low potential). 2. External flue gas recirculation (commonly used). 3. Non-stoichiometric operating burners (CSNB-Concept). The concept of controlled staging with non-stoichiometric burners uses the third option. While leaving the range of stoichiometric burner operation, inadmissible high flame temperatures can be avoided even at high oxygen concentrations in the oxidizer. To secure full burn out in the combustion chamber exit over- and sub-stoichiometric burners are arranged in a way that the global stoichiometry at the combustion chamber exit is slightly over-stoichiometric (see Fig. 1 ). Similar to conventional air-blown coal combustion. In that way the necessary flue gas recirculation for temperature moderation can be reduced drastically. The concept of controlled staging is described in more detail by Becher et al. [2] . Most of the commonly discussed oxyfuel processes are aiming on adjusting the oxyfuel process as good as possible to the known coal combustion process with air. One major concern is to achieve a heat transfer similar to an air-blown steam generator. That implies the need for a high flue gas recirculation rate around 65–75% (based on the mass flow at the combustion chamber exit) [3–6] . Becher et al. showed experimentally with natural gas burners that it is possible with the concept of controlled staging with non-stoichiometric burners to control flame temperature at low recirculation rates down to 50% by varying the stoichiometry [2] . Based on these experiences the concept of controlled staging has been transferred to coal. Because the process of solid fuel combustion is different to gaseous combustion the impact of oxygen deficiency or excess on the combustion behavior (ignition, flame stability and burnout) has to be evaluated. In a first step a single non-stoichiometric operating coal burner is investigated. This paper focuses on flame temperatures and gas species formation of coal flames operated under oxygen deficiency and excess with high oxygen concentrations in the oxidizer. The following questions are the key drivers for the investigations of this single burner arrangement. • Is the stoichiometry a suitable instrument to control flame temperatures even at high oxygen concentrations in the oxidizer? • How is the combustion behavior and the species concentration in non-stoichiometric oxycoal flames with high oxygen concentrations in the oxidizer? • How are the combustion characteristics of the non-stoichiometric flames compared to combustion with air and to combustion with an air-like oxyfuel case? Flames with a stoichiometry range from 0.5 to 2.5 and oxygen contents up to 40 vol.% were analyzed by a suction pyrometer and a gas analysis to be able to answer these questions. A comparison of the high oxygen concentrated non-stoichiometric flames with normal air flames and stoichiometric air-like oxyfuel flames with low oxygen concentration in the oxidizer was done to point out the differences. Firing strategies for conventional air-blown steam generators like over-fire, fuel-staging or special low NO x burners are aiming on emission reduction, burn out improvement and limiting the negative effects of corrosion, slagging and fouling. The oxyfuel process offers an additional degree of freedom in the combustion process. By varying the recirculation rate one can take influence on the oxygen concentration in the oxidizer. That offers additional possibilities in process set up. One of the key questions for retrofit solutions is, how to fit the wall heat flux properties of the oxyfuel combustion to that of the conventional air-blown case. The steam outlet temperature of the evaporator should be in both cases equal. That is important for the economical and technical feasibility of a retrofit solution. Additionally the furnace exit temperature, imposed by the melting behavior of the coal should be equal as well. In consequence the same heat transfer for oxyfuel and air-blown processes has to be reached for retrofit solutions [4] . All of these requirements are interdependent and may lead to a goal conflict for retrofit solutions. However, a newly built oxyfuel boiler could be perfectly adapted to the lowered mass flows and the changed heat transfer properties. Depending on the coal type, recirculation temperature and condition (wet or dry), a recirculation of 65–75% of the mass flow at combustion chamber exit is stated to reach same adiabatic flame temperatures as in air-blown processes. In consequence the oxygen concentration at the burner inlet should be around 28–35 vol.% (dependent on coal quality) if same adiabatic flame temperature is aspired [5,7,4,8,9,3] . Radiation is the dominating effect of heat transfer in boilers. Next to gas radiation, particle and soot radiation contribute to the heat transfer. If particle and soot radiation dominates gas radiation as stated by Kather and Scheffknecht similar temperatures in oxyfuel and air-blown combustion will lead to compareable radiative heat flux characteristics in the boiler [4] . That effect could be observed by Andersson et. al. with lignite fired oxyfuel flames [10] . If the increase of the radiative species CO 2 and H 2 O intensify the radiative heat transfer, lower adiabatic flame temperatures are necessary compared to the air case to achieve same heat flux profiles [5] . Few puplications are available which investigate the oxyfuel process with low recirculation rates. Normann et al. elaborated an oxyfuel firing strategy with higher oxygen concentration in the oxidizer and in consequence elevated temperatures [11] . Focus of their work was, however, the high temperature NO x reduction. Detailed investigations on non-stoichiometric operating oxyfuel burners are solely found with regards to slightly under-stoichiometric operating burners for NO x reduction [12,13] . 2 Experimental setup 2.1 Combustion chamber The experiments were realized in a vertical, cylindrical and air cooled combustion chamber with an inner height of 4 m and an inner diameter of 700 mm as shown in Fig. 2 . The cooling air jacket is divided in eight parts which can be individually supplied with cooling air. On each of the eight levels which are 500 mm apart four ports offer access to the interior of the combustion chamber (see Fig. 3 ). The ports are equipped with ball valves that allow an exchange of the probes even during operation. Up to three horizontally orientated burners can be mounted in the ports with a total thermal power of up to 300 kW. A multi-functional port is installed on each burner level which consists of a camera system and a flame detector. The remaining two ports can be used to insert measuring probes along the flame axis and vertically as shown in Fig. 3 . The flue gas is sucked through the bottom of the combustion chamber. The flue gas is cleaned from particles downstream of the combustion chamber and conditioned to a temperature of 180 °C by a cyclone, air cooled heat exchanger and a fabric bag filter. After the flue gas fan the flue gas can be recycled directly (wet) or after it passed the condenser (dry) where it is cooled to ambient temperature. Each of the burners is supplied with natural gas, primary- and secondary oxidizer and a carrier gas stream which transports the coal. The carrier gas is always dry recycled flue gas and is not enriched with oxygen. Corresponding to the desired thermal input a loss in weight feeder (twin screw from K-Tron) supplies coal into the carrier gas stream. The amount of carrier gas, primary and secondary oxidizer is measured by orifices and regulated by flaps. The natural gas is measured by mass flow controllers. Oxygen is supplied to the mixing line by a mass flow controller for the primary and secondary oxidizer from an external storage tank with a purity of 99.5 vol.%. The facility control and data acquisition is realized by a SPS system from Siemens. Flame temperatures are measured by a suction pyrometer type IFRF with a type B thermocouple protected by three ceramic tubes [14] . A differential pressure sensor is observing the suction process to detect plugging of coal and ash in an early stage. The in-flame gas profiles are obtained by a portable gas sample probe type M& C PSP 4000 H/C/T. After leaving the sample probe the hot flue gas is dried and filtered. CO and CO 2 are measured by a Sidor 700 of SickMaihak (IR-Detector), the oxygen concentration by M& C PMA 30 (paramagnetic) and the NO and NO 2 concentration by a chemiluminescence analyzer of EcoPhysics. In all experiments the fuel was pre dried german lignite. The dry flue gas composition (CO,CO 2 , NO, NO 2 , SO 2 and O 2 ) was measured at the furnace exit with an extractive online ABB Gas-Analyser type AO2000 (URAS 26/ UV, LIMAS 11/ IR, MAGNOS 206/ paramagnetic). The steam content in the flue gas at the furnace exit was measured by an extractive on-line process hygrometer type BARTEC HYGROPHIL H 4230. A finger camera (CT-58SHQXC with a Sony HQ1 Ex-View Color Chip) is mounted in a water cooled jacked with a CO 2 purge gas that allows permanent observation of ignition, flame shape and stability. In addition the in flame measurement process could be supervised. Characteristic flame pictures were taken from recorded flame videos. 2.2 Burner Two different swirl burners have been used in the experiments as shown in Fig. 4 . One was optimized for high oxidizer streams (Burner A) and one with smaller annular clearance area for lower oxidizer streams (Burner B). The principle set up for each burner remains the same. As core tube a standard ignition burner from Hegwein was used. Natural gas as well as the primary oxidizer is supplied through this central duct. Concentrically aligned is the register for the carrier gas and coal supply which is again concentrically surrounded by the swirled secondary stream. The swirl is in both burners generated by an axial swirl generator with an angle of 30°. For a more stable flame and a better ignition close to the burner mouth deflectors have been applied. In all experiments the fuel is german lignite which properties are listed in Table 1 . The initial ash deformation temperature of the coal is 1100 °C and ash flow temperature 1230 °C. Natural gas is provided by the technical infrastructure of the facility. The four major components and the lower heating value (LHV) is given in Table 2 . 2.3 Experimental parameters Results are presented in this paper for experiments in which only two burner were operated. Burner #1 was demounted, burner #2 was operated with coal as fuel and burner #3 was operated with natural gas. The coal flame of burner #2 was the subject of interest and in flame measurements were performed. Burner #3 was used to even out the global stoichiometry of the combustion chamber. To be able to sustain a more stable coal flame 5% of the thermal power input in burner #2 was delivered by natural gas. As listed in Table 3 four cases were investigated in detail, combustion with air and oxyfuel combustion similar to air with an oxygen concentration of 30 vol.%. Both with a stoichiometry of 1.15 (termed air_1.15 and oxy30_1.15). These two cases are the reference for comparing the over-stoichiometric oxyfuel combustion with a stoichiometry of 2.15 and the sub-stoichiometric combustion with a stoichiometry of 0.55 both operated with a low recirculation rate and 40 vol.% oxygen in the oxidizer (termed oxy40_2.15 and oxy40_0.55). The operational procedure was equal in all test cases and divided into three steps. First, the facility was heated up to a certain degree by conventional combustion of natural gas with air. Second, the facility was transferred to oxyfuel operation by exchanging the air stepwise with oxygen enriched recycled flue gas. And finally, after a dry flue gas concentration of over 80 vol.% CO 2 is reached the fuel of burner two is switched from natural gas to coal. The desired stoichiometries are set and the oxygen content of the oxidizer is regulated by the recirculation rate to the desired level. After the system reached stable conditions (flue gas temperature at furnace exit and CO 2 and O 2 concentration in stack) in-flame profile measurements began. Due to unavoidable air leakages in the combustion chamber and the flue gas cleaning, up to 12 vol.% (dry) N 2 depending on combustion chamber under pressure and recycling rate remained in the flue gas. The temperature of the combustion chamber walls were kept in all experiments below 400 °C. In Table 4 the most important process parameters are listed. The parameters of the four test cases were chosen as follows. Stoichiometric combustion with air was set as the reference case. The oxy30_1.15 as reference case for the air-like oxyfuel case with a similar flame temperature as the air case. The oxy40_0.55 was the lower limit where flame stability still allowed reliable measurements. Two aspects influenced the decision for the over-stoichiometric oxy40_2.15 case. On the one hand similar flame temperatures are theoretically expected and practically reached as in the oxy40_0.55 case. On the other hand the balancing of the total stoichiometry at furnace exit was limited by the sub-stoichiometric operating gas burner at the bottom. Due to the different burner geometries and the varying mass flows through the burner the aerodynamics in each of the test cases are different. Focus of the investigation was not a parametric study to investigate single effects, (for example NO x formation) but to show the different combustion behavior of these non-stoichiometric operating flames. 3 Results and discussion 3.1 Influence of stoichiometry and oxygen concentration on flame temperature Using the same fuel and constant surrounding conditions flame temperature is mainly dependent on the stoichiometry and oxygen concentration in the oxidizer [15] . While most articles concerning oxyfuel, investigated the correlation of flame temperatures in dependency of the oxygen concentration in the oxidizer which is controlled by the recirculation rate, the dependency of the stoichiometry is poorly investigated. As stated in the introduction next to fitting the oxyfuel flame temperature to that of air combustion while adapting the oxygen concentration in the oxidizer, stoichiometry offers a second opportunity for temperature control. In Fig. 5 the normalized adiabatic flame temperature is compared to the measured normalized flame temperature. Each of the flame temperatures, measured and calculated are normalized to its own highest measured, accordingly calculated temperature. That way the highest temperature equals one and is close to stoichiometric combustion. The adiabatic flame temperature was calculated with FactSage database [16] . The oxygen content of the oxidizer was in both cases 30 vol.%. The measured temperatures are always the hottest measured temperature on the EW position 150 mm in the flame axis (compare Fig. 3 ). While the oxygen content of the oxidizer was kept constant by adapting the recirculation rate the burner stoichiometry was adjusted by regulating oxidizer mass flow to the burner with the flap system of the mixing line. The global stoichiometry of the furnace was evened out to a slightly over-stoichiometric combustion at combustion chamber exit with the help of a non-stoichiometric natural gas burner. The characteristic shape of the dependency of the adiabatic flame temperature over the stoichiometry is found similar in the measured temperature profile. That means that even at high oxygen concentrations in oxyfuel atmospheres varying the stoichiometry is a suitable instrument to control flame temperatures. The peak temperature in the measured profile is as well close to the stoichiometric combustion and not shifted towards higher or lower stoichiometries. While the stoichiometric combustion and the under-stoichiometric combustion are predicted quite well by the adiabatic flame temperature the over-stoichiometric flame temperatures differ considerably more compared to the profile of the adiabatic flame temperature. The decrease in temperature with increasing stoichiometry is steeper than the theoretically expected one. In Fig. 6 the hottest measured flame temperatures of each profile from different experiments are combined. While the oxygen content of the oxidizer is controlled by the recirculation rate the stoichiometry is adjusted by the amount of oxidizer fed to the burner. Fig. 6 shows the two possibilities in oxyfuel processes to control flame temperatures: Oxygen content in the oxidizer and variation of stoichiometry. With an average oxygen content of 30 vol.% and a stoichiometry of 1.15 the flame temperature of the oxyfuel combustion is pretty similar to the temperature of the air blown flame of around 1280 °C. Distinct higher and lower stoichiometries lead to lower temperatures. Measuring temperatures with the suction pyrometer of near stoichiometric oxyfuel flames with 40 vol.% oxygen in the oxidizer was difficult due to slagging of the pyrometer with molten ash and lead to no reliable data. 3.2 Flame images In Fig. 7 characteristic screen shots from recorded video sequences of the four test cases are shown. The flame images of all three oxyfuel cases were far more luminous than in the air case. The same effect was observed in natural gas experiments and may indicate a higher soot formation rate in oxyfuel flames [17] [10] . The carrier gas mass flow was kept similar in all four cases to secure a proper coal transportation to the burner mouth. However, in the oxyfuel cases the carrier gas is not enriched with oxygen and contained only the excessive oxygen form combustion chamber end compared to the 21 vol.% of the air case carrier gas. This lead to a slightly more stable air flame than in the oxy30_1.15 case. Additional experiments with oxygen enrichment of the carrier gas also in the oxyfuel cases showed that an oxygen enrichment of the coal transporting carrier gas has an significant positive influence on ignition behavior and flame stability. This was as well observed by Hjaertstam et. al. [7] . Generally the most stable flames were achieved if flame ignition was as close as possible to the burner mouth. The oxidizer was distributed between primary and secondary register through the damper system of the facility. While an increased portion through the non swirled primary register leaded to a longer and more jet like flame an increased flow through the swirled secondary stream reduced the flame length due to induced internal recirculation. The highly swirled flames were in consequence more widened and also more turbulent. While all oxyfuel flames ignited directly in the burner mouth, in the air_1.15 case the point of ignition was shifted into the combustion chamber. Close to the burner mouth remained an area with no visible flame development. This effect was found in the temperature and gas profiles as well and will be discussed later. The low oxidizer streams in the oxy40_0.55 case lead to a short and intense flame while the oxy40_2.15 case with its high oxidizer streams developed the longest flame. The air_1.15 and the oxy30_1.15 showed pretty similar flame shapes and length. These flame images helped to evaluate the temperature and gas profiles. 3.3 Comparison of temperature profiles In Fig. 8 the temperatures profiles measured in flame axis (EW-Direction) of the four test cases are presented. The shape of the measured temperature profiles near the combustion chamber wall (from 700 mm inwards) was mainly influenced by the flame lenght. In this case the flame length is mainly influenced by the burner aerodynamics (compare flame pictures in Fig. 7 ). Regardless of flame shape and flame expansion, the maximum temperature and in that way the hottest point of each flame could be fairly good compared in concerns of stoichiometry dependency. As intended the maximum temperatures of the air case and the air-like oxyfuel case oxy30_1.15 were close to each other at around 1280 °C. With the used experimental set up the higher heat capacity of the oxyfuel combustion atmosphere is evened out with an oxygen concentration of around 30 vol.% in the oxidizer which is in the range of results published from other researchers [5] [7,18] . Due to the slightly shifted point of ignition in the air_1.15 case the highest temperature is reached not till 200 mm away from the burner mouth. Close to the burner mouth flame temperatures decreases. That agrees well with the observation made by comparing the flame images. However, as the maximum temperatures the shape of the flame temperature profile of these two flames are very similar after 200 mm. Comparing again the flame images in Fig. 7 one can see that the flame shape and the penetration of the flame into the combustion chamber are nearly equal. The oxy40_0.55 case with an ignition point close to the burner mouth and an very intense combustion, developed the shortest flame due to the low oxidizer streams through the burner. It showed the steepest decrease in the temperature profile and the lowest temperatures close to the combustion chamber wall. On the other hand the oxy40_2.15 showed the longest flame and the smoothest temperature profile. The temperature opposite of the burner mouth is the highest of all four cases. The peak temperatures of the two non-stoichiometric oxyfuel cases with 40 vol.% oxygen in the oxidizer were both on a level close below 1400 °C. The temperature level was roughly 120 °C higher than in the reference cases due to the high oxygen content in the oxidizer but still considered as technical feasible. In flame measurements with the suction pyrometer closer than 150 mm to the burner mouth led to massive disturbance of flame propagation and was not practicable. Due to the video surveillance one could observe that at a distance of 150 mm the physical influence on flame shape and propagation is in a tolerable level. 3.4 Comparison of species concentration profiles Comparing the gas composition of the four flames in the flame axis as shown in Fig. 9 the influence of stoichiometry and combustion atmosphere on the species concentration is becoming more clear. The oxygen concentration of the oxy30_1.15 case was close to the flame root lower than that of the air case. On the one hand the point of ignition in the oxy30_1.15 case was more shifted to the burner mouth as in the air_1.15 case. That can be observed in the flame images of Fig. 7 . On the other hand the higher oxygen concentration in the oxidizer of the oxy30_1.15 case lead to a more intense combustion already in the early stage. Due to the very high oxygen concentrations in the oxy40_0.55 case the oxygen concentration close to the burner mouth was higher than in the air_1.15 and the oxy30_1.15 cases. The oxygen concentration however decreased very fast along the flame axis and remained under 1 vol.%. The high oxygen concentrations in combination with the massive oxygen excess in the oxy40_2.15 case result in very high oxygen concentrations along full length of the flame. The oxygen concentration decreased from the initial measured concentration of 30 vol.% at position 150 mm to 15–20 vol.% (dry) at combustion chamber wall. Regarding the CO concentration the different combustion behavior could be observed as well. The CO concentration of the oxy40_0.55 were rising to the highest values of all four test cases. Peak measured CO concentration reached 7.5 vol.%. While the CO emissions of the oxy30_1.15 are close to the burner mouth the highest among all four test cases and increasing sharply in the following centimeters the CO emissions in the air_1.15 case remain relative low at the burner mouth and increase not until 200 mm after. That implies again the quick and intense ignition process due to the higher oxygen concentrations in the oxyfuel cases. The ignition of the air flame and the CO formation is delayed. This effect is backed by the observation made at the image of the air flame on the one hand and on the other hand at the peak flame temperature. Both show the delayed ignition of the air case compared to the oxyfuel cases. After the different ignition process the CO emissions from the air_1.15 and oxy30_1.15 case were similar to each other as could be observed in the temperature and oxygen profile. Due to the high oxygen availability the CO emissions of the oxy40_2.15 remain at the lower detection level of the CO detector over the full length of the flame. The highest CO concentrations close to the burner mouth in the oxy30_1.15 case create a reducing atmosphere which results in combination with the moderate flame temperature in the lowest NO x concentrations. The rapid combustion of the pre-dried lignite due to the high oxygen concentration in the oxy30_1.15 case at the beginning leads to relative low oxygen concentrations along the flame length and in consequence, in combination with the missing nitrogen to relative low NO x concentrations along the flame axis. The significantly lower level of NO x concentration in oxyfuel atmospheres was also observed by other researchers [18] . Compared to the low NO x concentration of the oxy30_1.15 at the flame root the NO x concentration of the oxy40_0.55 were very high. High temperatures and the oxygen availability shown in Fig. 9 a favors the NO formation close to the burner mouth even in the sub-stoichiometric case. However the increasing CO concentration leads to a significant reduction of the NO x concentration along flame axis. After the middle of the combustion chamber at 350 mm the NO x level decreased to values of the oxy30_1.15 case around 100 ppm. These values are lower than in the air_1.15 case over full flame length. The high availability of oxygen over the full length of the flame and especially directly at the burner mouth resulted in the lowest CO concentrations in the oxy40_2.15 case. The fast increase of the CO concentration which could be observed at the other test cases after flame ignition close to the burner mouth were not visible at all and the CO concentration remained on a constant level. The high oxygen concentrations in combination with the high temperatures lead to a high NO x concentration over full flame length even at the absence of non-fuel-nitrogen. The NO x concentration of the air_1.15 case were at the flame root higher but they decreased after 300 mm due to the increased CO concentration to a level half as high as in the oxy40_2.15 case. 4 Conclusions The focus of the investigations accomplished in this work was to show the feasibility of controlling flame temperature in oxyfuel processes not exclusively by varying the recirculation rate and in consequence the oxygen content in the oxidizer but also by varying the burner stoichiometry. Gas and temperature profile measurements of lignite fired oxyfuel flames at different oxygen concentrations and stoichiometries were presented. Furthermore to show the feasibility of the concept of controlled staging with non-stoichiometric burners a comparison of a sub-stoichiometric and over-stoichiometric flame with low recirculation rates and an high oxygen content of 40 vol.% to the reference cases of an air-blown flame and an oxyfuel flame with high flue gas recirculation is drawn. The stoichiometry is next to the adaption of the recirculation rate a suitable instrument to control flame temperature, to answer the questions raised in the introduction. The temperature dependency of lignite fired oxyfuel flames is similar to that of the calculated adiabatic flame temperatures with highest temperatures close to stoichiometric combustion. Leaving the stoichiometric combustion to either lower or higher stoichiometries leads to a reduction of the maximum flame temperature. The decrease in temperature is similar to the theoretically expected decrease regarding the adiabatic flame temperature. That is valid especially for the sub-stoichiometric cases while the theoretical expected temperatures in the over-stoichiometric cases over predict the measured values. Comparing the flame images of the four investigated cases one can see next to the flame shape, influenced by the burner aerodynamics, the more luminous oxyfuel flames. Furthermore one can state the more intense combustion in all oxyfuel cases compared to the air case. While comparing the four test cases air_1.15, oxy30_1.15, oxy40_0.55 and oxy40_2.15 the maximum temperature of the air_1.15 and the oxy30_1.15 case are as intended similar at around 1280 °C. That temperature level could not just yet be reached with an oxygen content of 40 vol.% in the oxidizer either at a stoichiometry of 0.55 nor 2.15. The maximum temperatures were roughly 120 °C degree higher. Additional experiments showed that by going to broader stoichiometries comparable temperature levels to the reference cases can be achieved. A further spread of the stoichiometries was not possible due to the experimental set up with two burners. With a multi burner set up and more optimized burner especially for the sub-stoichiometric combustion, more extreme stoichiometries can be realizable while remaining an overall stoichiometry at the combustion chamber exit which is slightly over-stoichiometric. That way a further reduction in temperature would be possible. Comparing the species concentration profiles of the four cases the combustion characteristics of the four flames could be reconstructed. As expected the sub-stoichiometric combustion results in high CO concentrations in the flame and in a reducing atmosphere which leads to low NO x concentrations. On the other hand, the oxygen excess in the oxy40_2.15 leads to low CO and high NO x concentrations. How these species concentration characteristics operate in a multi burner arrangement in the concept of controlled staging has to be further clarified. Acknowledgments This work was gratefully supported by the Bayerische Forschungsstiftung, Alstom Power Systems GmbH, EnBW Kraftwerke AG and E.ON Energie AG. References [1] World energy outlook 2010, International Energy Agency, Paris; 2010. [2] Becher V, Bohn J-P, Goanta A, Spliethoff H. A combustion concept for oxyfuel processes with low recirculation rate – experimental validation, Combust Flame. doi:10.1016/j.combustflame.2010.12.029 . [3] M.B. Toftegaard J. Brix P.A. Jensen P. Glarborg A.D. Jensen Oxy-fuel combustion of solid fuels Prog Energy Combust Sci 36 5 2010 581 625 [4] A. Kather G. Scheffknecht The oxycoal process with cryogenic oxygen supply Die Naturwissenschaften 96 9 2009 993 1010 [5] T. Wall Y. Liu L.E. Chris Spero S. Khare R. Rathnam F. Zeenathal An overview on oxyfuel coal combustion-state of the art research and technology development Chem Eng Res Des 87 8 2009 1003 1016 [6] B. Buhre L. Elliott C. Sheng R. Gupta T. Wall Oxy-fuel combustion technology for coal-fired power generation Prog Energy Combust Sci 31 4 2005 283 307 [7] S. Hjärtstam K. Andersson F. Johnsson B. Leckner Combustion characteristics of lignite-fired oxy-fuel flames Fuel 88 11 2009 2216 2224 [8] T. Wall Combustion processes for carbon capture Proc Combust Inst 31 1 2007 31 47 [9] H. Liu R. Zailani B.M. Gibbs Pulverized coal combustion in air and in o 2 /co 2 mixtures with NO x recycle Fuel 84 16 2005 2109 2115 [10] K. Andersson R. Johansson F. Johnsson B. Leckner Radiation intensity of propane-fired oxy-fuel flames: Implications for soot formation Energy Fuels 22 3 2008 1535 1541 [11] F. Normann Emission control of nitrogen oxides in the oxy-fuel process Prog Energy Combust Sci 35 5 2009 385 397 [12] Hannes Stadler Dominik Christ Martin Habermehl Peter Heil Arno Kellermann Andreas Ohliger, Dobrin Toporov, Reinhold Kneer, Experimental investigation of NO x emissions in oxycoal combustion Fuel 90 4 2011 1604 1611 [13] K. Andersson F. Normann F. Johnsson B. Leckner No emission during oxy-fuel combustion of lignite Ind Eng Chem Res 47 6 2008 1835 1845 [14] N. Fricker How do i measure local flame temperatures with the ifrf suction pyrometer? 2001 URL http://www.handbook.ifrf.net/handbook [15] H. Spliethoff Power generation from solid fuels 2010 Springer Berlin Heidelberg [16] C. Bale P. Chartrand S. Degterov G. Eriksson K. Hack R. Ben Mahfoud Factsage thermochemical software and databases Calphad 26 2 2002 189 228 [17] Becher V, Goanta A, Bohn J-P, Gleis S, Spliethoff H. Kontrollierte stufung mit nicht-stöchiometrischen brennern für oxyfuel kraftwerke experimenteller nachweis. In: 40. Kraftwerkstechnisches Kolloquium, Dresden; 2008. [18] Yewen Tan Eric Croiset Mark A. Douglas Kelly V. Thambimuthu Combustion characteristics of coal in a mixture of oxygen and recycled flue gas Fuel 85 4 2006 507 512