Scholarly article on topic 'Sulphur Oxide Emissions from Dust-fired Oxy-fuel Combustion of Coal'

Sulphur Oxide Emissions from Dust-fired Oxy-fuel Combustion of Coal Academic research paper on "Chemical engineering"

Share paper
Academic journal
Energy Procedia
OECD Field of science
{"sulphur oxides" / "sulphur dioxide" / "sulphur trioxide" / "SO2 " / "SO3 " / desulphurisation / "oxy-fuel operation"}

Abstract of research paper on Chemical engineering, author of scientific article — Reinhold Spörl, Jörg Maier, Günter Scheffknecht

Abstract This article summarizes scientific knowledge and practical experiences on the release and capture of SO2 and SO3 with a special focus on the implications of oxy-fuel combustion conditions on sulphur emission behaviour. Results, obtained at the experimental, 500 kWth atmospheric, pulverized fuel combustion rig (KSVA) of the IFK regarding sulphur oxide (SO2/SO3) emissions and capture behaviour are presented. The experimental plant was operated with pre-dried sulphur rich lignite in air and oxy-fuel combustion mode. The following issues are highlighted in particular: • General and transient behaviour of SO2 emissions under air and oxy-fuel combustion conditions SO3 • separation behavior of the ESP under oxy-fuel conditions • Differences of ESP ash qualities from air and oxy-fuel operation

Academic research paper on topic "Sulphur Oxide Emissions from Dust-fired Oxy-fuel Combustion of Coal"

Available online at

SciVerse ScienceDirect

Energy Procedia 37 (2013) 1435 - 1447


Sulphur Oxide Emissions from Dust-Fired Oxy-Fuel

Combustion of Coal

Dipl.-Ing. Reinhold Spörla*, Dipl.-Ing. Jörg Maiera, Univ.-Prof. Dr. techn.

Günter Scheffknechta

aInstitute of Combustion and Power Plant Technology (IFK), Department of Firing Systems, University of Stuttgart, _Pfaffenwaldring 23, D-70569 Stuttgart, Germany_


This article summarizes scientific knowledge and practical experiences on the release and capture of SO2 and SO3 with a special focus on the implications of oxy-fuel combustion conditions on sulphur emission behaviour. Results, obtained at the experimental, 500 kWth atmospheric, pulverized fuel combustion rig (KSVA) of the IFK regarding sulphur oxide (SO2/SO3) emissions and capture behaviour are presented. The experimental plant was operated with pre-dried sulphur rich lignite in air and oxy-fuel combustion mode. The following issues are highlighted in particular:

• General and transient behaviour of SO2 emissions under air and oxy-fuel combustion conditions

• SO3 separation behavior of the ESP under oxy-fuel conditions

• Differences of ESP ash qualities from air and oxy-fuel operation

© 2013 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of GHGT

Keywords: sulphur oxides, sulphur dioxide, sulphur trioxide, SO2, SO3, desulphurisation, oxy-fuel operation

1. Background

1.1. Oxy-Fuel combustion

In oxy-fuel operation, coal is burned with a mixture of oxygen and recirculated flue gas, instead of air. The mixing of O2 with recirculated flue gas is, among other issues, necessary to lower the temperature in the furnace which otherwise would exceed the limits of construction materials of the boiler. Due to the lack of dilution of flue gasses by airborne nitrogen in oxy-fuel operation, the concentrations of flue gas components such as CO2, SO2 and H2O generally increase considerably by a factor up to 4 [2, 3].

* Corresponding author. Tel.: +49 (0) 711-68563748; fax: +49 (0) 711-68563491. E-mail address:

1876-6102 © 2013 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of GHGT doi:10.1016/j.egypro.2013.06.019

1.2. Formation and capture of SO2

Fossil fuels contain sulphur in various forms, such as sulphides (especially pyrite: FeS2), sulphates (e.g., gypsum), as organically bound sulphur and to a small portion as elemental sulphur. The sulphur content of coals varies significantly. Thus, central European lignite contains between 0.2 up to about 5% of sulphur [4]. In power plants, the sulphur from fuel is oxidised almost completely, mainly to sulphur dioxide following equation (1) [5].

S + O2->SO2 (1)

The most important parameters influencing the concentration of SO2 in the flue gas are:

• stoichiometric ratio/oxygen partial pressure

• sulphur content of coal

• alkaline/earth-alkaline content of the ash

• temperature

• residence time

In combustion tests using coal under N2/O2 and CO2/O2 atmospheres, Hu et al. [1] found a significant influence of the stoichiometric ratio X on the S02 concentrations/emissions (see Fig. 1): starting from fuel-lean mixtures, the SO2 concentration rises with decreasing X, until it reaches a maximum at ^-values around 0.85. When decreasing X further, the SO2 concentration falls again. The effect is more pronounced with high O2 contents in the oxidant than at O2-concentrations of 20% and is found in N2 as well as in CO2 atmospheres. The mentioned rise in SO2 concentrations in CO2 atmospheres starts at lower X-values than in N2 atmospheres, then runs, however, steeper.

Since SO2 is formed by combustion of fuel-sulphur, the SO2-concentration rises with rising sulphur content of the coal. Alkaline and earth-alkaline compounds can take up sulphur oxides from the flue gas to form solid sulphates. Therefore, the content of those compounds in the coal ash, primarily the calcium





SO2 ppm


0H-i-i-1-1 0

0,7 1 1,7 5 0,7 1 1,7 5

A O2-20 % « O2-50 % ■ O2-80 % ♦ pure O2

Fig. 1: Influence of X on S02-concentrations in coal combustion, for various O2-contents in CO2- and N2-based oxidants at 1000°C; the marked regions correspond to Oxy30 (on the left) and air combustion (on the right)

Adapted from Fuel Vol. 7 (15), Hu, Y.; Naito, S.; Kobayashi, N.; Hasatani, M., CO2, NOx and SO2 emissions from the combustion of coal with high oxygen concentration gases, pp 1925—1932, Copyright (2000), with permission from Elsevier. (graphic adapted: equivalence ratio replaced by stoichiometric ratio)

System desulphurisation efficiency, [%]

O2/CO2, S = 2 wt.-

100-1 System desulphurisation efficiency, [%]

1.400 Temperature [K]

2 4 6 8

Residence time of particles [s]

Fig. 3: Calculated dependency of the desulphurisation efficiency on residence time at 1400K, X =1.2, Ca/S-ratio = 5 and 10|im particle diameter for coal combustion in

O2/CO2 atmosphere with a recirculation rate of 0.84 and for air combustion

Fig. 2: Calculated temperature dependency of the desulphurisation efficiency at X = 1.2, 8s residence time (once-through), Ca/S-ratio = 5 and lOpm particle

diameter for coal combustion in O2/CO2 atmosphere with a recirculation rate of 0.84 and for air combustion

Adapted with permission from Energy Fuels, Vol. 15 (2), Liu, H.; Katagiri, S.; Okazaki, K., Drastic SOx Removal and Influences of Various Factors in O2/CO2 Pulverized Coal Combustion System, pp 403—412. Copyright (2001) American Chemical Society.

content, influences the SO2 emissions of firing systems substantially. The calcination and desulphurisation reactions with CaCO3, which is the most important desulphurisation reaction chain is shown exemplarily in equations (2) and (3) [6]:

CaCO3 CaO + CO2 CaO + SO2 + Y2O2^ CaSO4

(2) (3)

The SO2 capture in the ash depends besides the alkaline/earth-alkaline content of the ash, on the level of the SO2 in the flue gas (and therefore on the sulphur content of the coal), but also on the temperature (see Fig. 2) and on the residence time of flue gases in the temperature range relevant for desulphurisation (see Fig. 3). It can be seen from figure 2 that there is a increase in the efficiency of desulphurisation with increasing sulphur content of the coal and that there is - dependent on temperature - a desulphurisation maximum (in air-firing at approx. 1300-1350°C, in oxy-fuel combustion at approx. 1400°C). Such a maximum can be explained as follows: Initially, the desulphurisation reactions run more and more efficiently with rising temperatures. However, CaSO4 starts to decompose at high temperatures, whereby SO2 is released. Increasing the temperature, the decomposition reactions reduce the efficiency of the flue gas desulphurisation, until the CaSO4 decomposition outbalances its formation. From 2 and 3 another effect can be seen: The flue gas desulphurisation efficiency in O2/CO2 atmosphere compared to air-firing is increased by a factor of about 4-6 and obviously runs over a wider temperature range, with higher maximum temperatures. This rise can be explained by a - compared to air combustion - increased residence time of the flue gases at high temperatures, due to the decreased flue gas volume flow through the furnace in oxy-fuel operation and due to the recycling of flue gas. Moreover, the substantial increase in SO2 concentrations under oxy-fuel conditions suppresses the decomposition of CaSO4 and shifts it to higher temperatures [6]. Besides that, the elevated CO2 partial pressure has a positive impact on the desulphurisation efficiency, since it improves the porosity of calcium-rich particles by which the diffusion

resistance for SO2 towards Ca is decreased [7, 8]. Liu et al. [6] state however that the decomposition of CaSO4 is accelerated in the absence of O2 in CO2-rich atmospheres. To sum up, the conversion of fuel-sulphur to SO2 in a power station depends on a variety of individual factors, which differ between fuels, plants and combustion settings.

Sheng et al. [9] indicate a sulphur capture in the ash of 20-25% for conventional air-fired pulverised coal combustion. This level, which is low in comparison to fluidized bed combustion systems and to lab ashing experiments, ca be explained by higher temperatures (1300-1600°C) and shorter residence times (1-2s) in pulverised coal combustion [9, 10]. Croiset et al. [11] report a fuel-sulphur to SO2 conversion of 92% at a 210 kW^ experimental plant during air-firing of hard coal. Conversion rates of 75% and 64% are reported for combustion in CO2/O2 atmosphere and for oxy-fuel combustion with flue gas recirculation respectively. Fleig et al. [12] found in experiments at a 100 kW^ pilot plant with lignite a 67% fuel-sulphur to SO2 conversion rate for air combustion and between 41-46% for oxy-fuel combustion. Lignite combustion experiments at the 500 kWth KSVA plant of the IFK by Monckert et al. [13] provided a conversion of fuel-sulphur to SO2 of 50.4% for air-firing and 39.1% for oxy-fuel-firing. Generally, in oxy-fuel operation SO2 concentrations that are elevated by a factor of 2-4 [14] compared to conventional air combustion can be found. Extreme SO2 concentrations up to above 20.000 ppm were measured during oxy-fuel combustion of sulphur rich coal [15]. In contrast to the SO2-concentrations, due to the improved sulphur capture in the ash, the SO2 emissions (SO2 mass flow related to energy supply by fuel) in oxy-fuel operation are lower than in air-firings [16].

1.3. Formation and capture of SO3

During combustion and along the flue gas path, SO3 is formed from SO2 in a homogeneous (bi-or trimolecular) gas phase reaction (Eq. (4)) or in a heterogeneous solids catalysed reaction for example on iron oxides [5, 17]. The ratios between homo- and heterogeneous SO3-formation cannot be given in general.

SO2+O (+M)->SO3(+M) (4)

The following parameters have a significant influence on the formation of SO3:

• oxygen partial pressure

• sulphur content of coal (SO2-partial pressure)

• alkaline/earth-alkaline content of the ash

• content of catalytic active compounds in the ash (e.g. Fe2O3)

• temperature

• residence time

• application of an SCR-Reactor

The O2 concentration in the flue gas has a direct influence on the SO3 formation rate, which increases with increasing oxygen content [18, 19]. Likewise, the sulphur content of the coal, which affects directly the SO2 partial pressure, significantly influences the formation of SO3. Higher sulphur contents cause higher SO2 and therefore higher SO3 concentrations [19]. The SO3 concentration in the flue gas is also highly affected by the content of alkaline and earth-alkaline compounds in the ash, since these compounds can capture SO3 from flue gas, forming sulphates. The underlying reactions are temperature-dependent and proceed fast at high temperatures, while being relatively slow in colder sections of the plant. SO3 capture by alkaline and earth-alkaline compounds is most significant in the temperature range between 800°C and 300°C [20]. If NH3 is present in the flue gas (e.g.: at/after a SCR unit), efficient SO3 capture in form of ammonia sulphates takes place even at lower temperatures. SO3/H2SO4 can also be separated from gas phase effectively by adsorption or by condensation on fly ash particles at temperatures

near or below the sulphuric acid dew point temperature [20]. Another important factor influencing the SO3 formation is the content of catalytically active compounds in the ash (e.g.: Fe2O3). In this context, an increased iron content of the ash leads to a rise of the SO3 formation [17]. Catalytic SO3 formation is highly dependent on temperature, with a maximum of the SO2/SO3 conversion at about 700°C [21]. This catalytic formation takes place on entrained fly ash particles, as well as on surfaces of ash deposits. The catalytic SO3 formation on ash deposits is influenced by temperatures and the amount of deposits. In the past, it was possible to observe the influence of cyclic soot blowing activity on measured SO3 concentrations [22]. Iron oxides of the pipes' and walls' construction materials only play a minor role for the catalytic SO3 formation, since they are normally not in direct contact with the flue gas due to ash deposits [17]. In addition to the above mentioned SO3 formation on catalytic compounds of the ash, SO2 can also be catalytically oxidized to SO3 on vanadium pentoxide from SCR DeNOx catalysts. Therefore, the application, the type, as well as the operational parameters of a SCR catalyst have an influence on the SO3 concentrations in the flue gas. In modern, commercial SCR catalysts generally SO2/SO3 conversion rates lower than 0.5 % are reached [20]. In the late 90s, much higher conversion rates in the range of 1 to 2% were reported [23]. By simulation, Fleig et al [19] investigated the homogeneous SO3 formation in greater detail, focussing on its dependency on operational parameters and flue gas composition. It was found that besides the aforementioned factors the SO3 formation is also dependent on the NO and H2O content of the flue gas, as well as on the concentrations of products of incomplete combustion and therefore on the combustion stability. The homogeneous, as well as the catalytic SO3 formation clearly show a temperature-dependent behaviour [19, 21] and are relatively slow [17]. This is the reason why in a technical firing systems, no equilibrium concentrations can be observed. Therefore, the temperature-residence-time-profile of the flue gas in a plant can be seen as one of the main influencing factors on the SO3 formation [5, 19]. For example, the (maximum achievable) SO3 concentration in thermodynamic equilibrium lies at a SOx content of 1170 ppm and temperatures 1205°C at 10 ppm, while it is about 1000ppm at 538°C [17]. In practice, SO3 concentrations up to about 40 ppm can be found in air-fired and over 180 ppm in oxy-fuel facilities [15, 24]. Wall et al. state that the SO3 concentration in oxy-fuel operation is increased by a factor of about 2.5 to 3 compared to air combustion [3]. Reported SO2/SO3 conversion rates under various conditions lay between 1 and 5%, depending on fuel and operational parameters [25]. A comprehensive study of Fleig et al. [19] in published experimental data, regarding changes of this conversion rate between air and oxy-fuel combustion, shows no clear trend. Possibly, a combination of factors with partially opposite impacts that are influenced by the specific plant layout, the fuel used and the operating parameters, cause that the conversion rate can increase as well as decrease when changing from air to oxy-fuel firing.

Depending on temperature, gaseous sulphuric acid H2SO4 is formed from SO3 and water vapour:

SO3+H2O->H2SO4 (5)

This transformation starts at about 400°C and is nearly complete at 200°C [26, 27]. The dew point of the formed H2SO4 depends on the concentrations of H2O and SO3/H2SO4. It can be calculated according to Verhoff and Banchero [28] or more exactly according to ZareNehzhad [30]. In power plants, H2SO4 dew points typically range between 95 and 160°C. As a result of increased H2O and SO3/H2SO4 concentrations in oxy-fuel operation, considerably higher dew points can be found. When temperature falls below the H2SO4 dew point temperature, for instance locally on cold surfaces, H2SO4 from gas phase starts to condense. In power plants, relevant temperatures are found in the region of air pre-heaters, where substantial problems due to low temperature corrosion can occur. H2SO4 condensates attack not only metallic components but also concrete and plastic parts of a plant and corrode those within relatively short time [29]. Also fouling by sulphate deposits can be a severe problem at cold parts of a power plant.

2. Methodology

2.1. 500 kWth experimental combustion rig KSVA

The 500 kWth atmospheric, pulverised fuel combustion rig KSVA (Fig. 4) applies a conventional pulverized fuel firing concept and can be operated not only in air, but also in oxy-fuel combustion mode. It is perfectly suitable to study combustion of pulverised fuels such as hard coal, lignite, sewage sludge, biomass, SRF and others. Similarly to a power station, the plant is equipped with an air/oxidant pre-heater and flue gas cleaning equipment, such as a high dust SCR unit as well as an ESP. A fan conveys fresh air or recirculated flue gas in air and oxy-fuel operation, respectively. The furnace has a total length of 7 m and an internal diameter of 0.8 m. A water jacket integrated into the refractory lined walls of the furnace allows for heat discharge. The milled fuel is fed to the top mounted burner pneumatically by air or CO2 from tank in air or oxy-fuel operation, respectively and mixed there with the oxidant gas (air/oxygen enriched recirculated flue gas). There are measuring ports all along the furnace, at all flue gas cleaning units and along the ducts. Using dedicated sampling probes, concentration and temperature measurements can be performed in the furnace and bottom and fly ash samples can be taken. Flue gas measuring systems are installed at the end of the furnace for continuous measurement of O2, CO, CO2, SO2, and NOx. The H2O is continuously measured at the ESP outlet. In oxy-fuel operation, flue gas is recirculated wet after the ESP, preheated in the oxidant preheater, mixed with O2 from tank and supplied to the burner.

2.2. Operation Details/Analysis of the Fuel

Table 1 presents the analysis of the lignite used for air and oxy-fuel combustion experiments. The lignite's ash had a CaO content of 8.43 wt.-%, equivalent to a Ca/S molar ratio of the coal of around 0.56. The operational parameters from air and oxy-fuel experiments are shown in the table 2. An overview of the average flue gas composition during operation is given in the table 3.

hopper \_/

Fig. 4: 500 kWth experimental air and oxy-fuel combustion rig KSVA

Tab. 1: Lignite analysis [% = wt.-%] (complying with DIN51729-1, DIN51729-11, DIN51900-1, DIN51900-3) ~H w A F cfX C H N S o

[J/g] [%, raw] [%, wf] [%, waf] [%, waf] [%, waf] [%, waf] [%, waf] [%, waf] [%, waf]

18078 9.10 18.82 59.52 40.48 65.97 7.09 0.66 1.98 24.29

Tab. 2: Operational parameters at KSVA: air and oxy-fuel operation

Unit Air Oxy-Fuel

Fuel feed [kg/h] 60.6 61.1

Thermal power [kW] 304.3 306.8

Carrier gas (CO2/air) [m3 (STP)] 30 24

Secondary oxidant gases [m3 (STP)] 325 204

O2 supply [m3 (STP)] - 70

Purge gases (CO2/air) [m3 (STP)] 9 7

Flue gas flow (wet) [m3 (STP)] 396 304

Oxidant O2 concentration [Vol.-%, dry] 20.9 36.1

X (in oxy-fuel: furnace-A, according to [31]) [-] 1.15 1.12

Tab. 3: Measured and averaged flue gas composition in air and oxy-fuel operation

Unit Air Oxy-Fuel

O2 [Vol.-%, dry] 3.2 4.7

CO2 [Vol.-%, dry] 16.3 89

CO [mg/m3 (STP), dry] 32 31

SO2 [mg/m3 (STP), dry] 4608 14562

NOx [mg/m3 (STP), dry] 296 573

H2O [Vol.-%, wet] 7 28

2.3. Measurement Techniques SO2/SO3

SO2 concentrations were measured online by IR-photometry and collected ashes were analysed by ICP-OES. Measurements of SO3 concentrations were performed applying the controlled condensation method in correlation to VDI guideline 2462. Problematic with this measurement method is the possibility of a negative bias of the measurement, due to a capture of SO3 on the sample gas filter in form of condensed H2SO4 or by reaction with ash particles. To minimize this bias, the filter must be reliably kept above the H2SO4 dew point temperature. However, there is also the possibility of a positive bias if a filter is heated above the flue gas temperature, since SO3 can be formed by decomposing sulphates of the ash [20]. To minimize both effects, a heated glass probe with in-stack quartz wool plug was used for SO3 sampling.

3. Experimental results - Discussion

3.1. Measured SO2 concentrations

To evaluate the release of fuel sulphur to the flue gas, theoretically achievable maximum SO2 concentrations for air and oxy-fuel firing (cSO2,max,air or cS02,max,oxy) were calculated according to equations (6) and (7), assuming a complete conversion of the fuel-sulphur to SO2. In equation (7), the yleackage accounts for dilution of the flue gas by air ingress or contaminations the O2 feed. It was estimated based on the measured gas composition at KSVA to be 5%.

Based on the measured SO2 concentrations in the flue gas, a conversion rate of fuel-sulphur to SO2 of 93.3% and 78.7% can be calculated for air and oxy-fuel operation, respectively. These relatively high conversion rates can be explained by the relatively low Ca/S ratio of the lignite. Nonetheless, one

ys * mm,so2

0,79 0,21

Yo 2Mm

21 y o2

M,SO2 *106


100 yO2 yleackage

observes significantly lowered SO2 release under oxy-fuel conditions, which can be explained by the altered firing conditions (most importantly by the increased SO2 concentration).

The air and oxy-fuel experiments, allowed also to investigate the effect of discontinuities of plant operation on SO2 concentrations. It was observed that during oxy-fuel operation at low O2 concentrations substantial fluctuations of the SO2 concentrations occurred. Two recorded sequences of measured concentrations are shown in figure Fig. 5. While in the upper diagram, at an O2 level above 4 Vol.-%, the SO2 concentration remains fairly stable between 14666 and 16160 mg/m3 (STP), in the lower diagram at a lower O2 level down to 1 Vol.-%, the SO2 concentration is fluctuating between 13718 und 19931 mg/m3 (STP), which is equivalent to a fluctuation of 38% in relation to the mean value over time. Figures 6, 7 and 8 provide an explanation of this behaviour. In Figure 7, the SO2 axes represent a concentration range

of 65 to 105% of the calculated, theoretical maximum SO2 concentrations cSO2r

and cso2

. for air

and oxy-fuel operation, respectively. It can be seen that the fluctuations of the SO2 concentration relative

20000 16000 12000 8000 4000 0

9:09 9:10 9:11 9:12 9:13 9:14 9:15 9:16 9:17 9:18

-SO2 [mg/Nm3 (STP)] (dry normalized to 6% O2) -O2 [Vol.-%, dry]

Fig. 5: Transient behaviour of O2 and SO2 concentrations in oxy-fuel operation at different O2 levels

SO2 concentration [mg/Nm3 (STP), dry] 18500


14500 -

10500 -

6500 -

0 2 4 6 8

O2 concentration [Vol.-%, dry]

Fig. 6: Excess O2 dependency of SO2 formation in air and oxy-fuel operation (same scale for air and oxy-fuel); values in Fig. 6, 7 and 8 represent the same 2-h time frame; SO2 concentrations normalized to 6% O2

SO2 concentration [mg/Nm3 (STP), dry] 19150 t-

SO2 concentration [mg/Nm3 (STP), dry] 4370

O2 concentration [Vol.-%, dry]

Fig. 7: Excess O2 dependency of SO2 formation in air and oxy-fuel operation (mind different scales for air and oxy-fuel: both axes represent 65-105% of the cSO2,maXiair and cSO2,max,oxy concentrations, calculated at 6% excess O2; values in Fig. 6, 7 and 8 represent the same 2-h time frame; SO2 concentrations normalized to 6% O2


to the theoretical maximum values are in air-fired less pronounced than in oxy-fuel operation. Besides that, it can be noticed from figures 6 and 7 that SO2 concentration maxima are only reached, when the O2 concentration falls below a certain level. Below that level, a distinct rise of SO2 concentrations can be observed. The threshold level lies at around 5 Vol.-% O2 in oxy-fuel operation. In air-fired operation the effect is less pronounced and the threshold is sifted to somewhat lower O2 concentrations. This observation is also related to the higher conversion rate of fuel sulphur to SO2 in air-firing. Since in the air-fired operation on average above 93% of the fuel-sulphur can be found in the flue gas as SO2, a rising sulphur release from fuel can only lead to relatively low additional SO2 emissions. The predominance of the observed effect in oxy-fuel operation can be explained with help of figure 1. In an oxy-fuel firing, with, compared to air-firing, increased O2 concentrations in the oxidant, the same excess O2 concentration corresponds to a lower ^,-value. At lower ^.-values and therefore under oxygen lean conditions the reaction rates of SO2 capture reactions with compounds from the ash are reduced. Therefore, in oxy-fuel operation, a reduced SO2 capture can appear at higher excess O2 concentrations than in air-fired systems. In addition, according to figure 1 the effect of rising SO2 concentrations at decreasing ^-values is more pronounced in oxy-fuel operation. This might also influence the observed sulphur release behaviour. Besides the sensitivity of SO2 formation at low excess O2 concentrations, in figure 8 a similar trend can be observed for the formation of CO. It can be noticed that in oxy-fuel combustion, compared to air-firing, CO concentration peaks are higher and occur already at higher excess O2 concentrations. The CO peaks at excess O2 concentrations lower than about 4 Vol.-% indicate local, short termed, near or sub-stoichiometric combustion conditions. The oxy-fuel-firing obviously responds more sensitive to short-term fluctuations in operating conditions. Presumably, the most dominant parameter in this respect is a discontinuity in the coal-dosing. Due to the lower available amount of excess oxygen at similar excess O2 concentrations, in comparison to air firing, in oxy-fuel operation, the same small, short-term fluctuations of the coal feed can easier lead to short-term oxygen depletion in the furnace.

The reduced SO2 capture at low excess O2 concentrations should be considered for an optimal oxy-fuel process operation. A low O2 excess that is desired for economic reasons, may lead to a significant increase

CO concentration [mg/Nm3 (STP), dry] 1000 1-

O2 concentration [Vol.-%, dry]

Fig. 8: O2 dependency of CO formation in air and oxy-fuel operation; values in Fig. 6, 7 and 8 represent the same 2-h time frame; CO concentrations normalized to 6% O2

E1-oxy E1-air E2-oxy E1-air E3-oxy E1-air

Fig. 9: S, Ca and Fe content in fly ashes from ESP compartments 1, 2 and 3 from air and oxy-fuel operation

in short-term SO2 and potentially SO3 concentration peaks, with associated consequences such as increased corrosion, fouling and higher costs for desulfurization. Anyway, it should be possible to operate an oxy-fuel power plant even at low excess O2 concentrations, without formation of SO2 concentration peaks, by application of an optimized oxy-fuel combustion system. In some instances it may, however, be beneficial, to slightly increase the excess O2 concentration compared to air-firing to achieve an optimal SO2 emission behaviour.

3.2. Measured SO2 concentrations

During the experiments at the KSVA, SO3 measurements were performed at the ESP inlet and outlet, for a characterization of the SO3-separation performance at this unit. To avoid premature SO3/H2SO4 condensation and thereby a negative bias of the measurement, the ESP was operated at an outlet temperature of about 180-200°C. Table 4 shows the results of the SO3 measurements. The single SO3 measurements performed at each measurement point establish a high repeatability. At the ESP inlet, significantly higher SO3 concentrations were found in oxy-fuel compared to air- fired operation. It can be seen that in the lignite-fired oxy-fuel combustion tests at KSVA, the ESP worked as a highly efficient SO3 sink, capturing about 62% of the inlet SO3 amount.

3.3. Analysis of E-filter ashes

Besides the study of gaseous sulphur emissions, ash samples were taken from the ESP and analyzed in the

Tab. 4: Results of SO3 measurements at ESP in oxy-fuel and air-fired operation at KSVA (concentrations normalized to 6% O2)

Operational Measurement Measured SO3 concentrations [mg/m3 i . N., dry]

mode point Sample 1 Sample 2 Sample 3 Mean

Air ESP inlet 42.8 45.4 43.4 43.9

Oxy-fuel ESP inlet 280.6 290.8 296.5 289.3

Oxy-fuel ESP outlet 118.2 100.4 109.5 109.4

SO3 capture at ESP ~ 62%

laboratory. The ESP at KSVA has 3 compartments (E1, E2 and E3) from which fly ash samples can be taken. In the first compartment, most of the ash is separated by mass-inertia of larger particles, whereas in the following chambers, more fine particles are separated by the electric field. Sulphur, calcium and iron contents of sampled ashes from all 3 ESP compartments from air and oxy-fuel operation are summarised in figure 9. In air, as well as in oxy-fuel operation, one finds considerably higher sulphur contents in the E2 and E3 ashes compared to the E1 ash, which goes together with higher contents of calcium, but also of iron. A reason for the connection between Ca and S contents is that most of the sulphur in the ash is present in from of calcium sulphate. Possibly also iron can be associated with sulphur capture in the ash. An obvious increase of the sulphur content of all ESP ashes under oxy-fuel conditions, compared to air-firing, is in line with the reduced oxy-fuel SO2 release. Moreover, it can be assumed that the observed SO3 capture in the fly ash of the ESP in oxy-fuel operation is responsible for a small part of this increase.

4. Summary

In this article, a comprehensive compilation and explanation of key parameters, influencing the SO2 and SO3 concentrations in the flue gas in air-fired and oxy-fuel operation is provided. In addition, conversion rates of fuel sulphur to SO2 for the combustion of lignite under air and oxy-fuel conditions are determined, based on experimental data and compared. An important finding, with relevance for commercial application of the oxy-fuel technology is that excess O2 concentration levels that are unproblematic in air-firing, can lead to a significant increase of SO2 concentrations in an oxy-fuel application. Moreover, in oxy-fuel operation at excess O2 concentrations below about 5% an increase of substantial, short-term SO2 emission peaks can occur. In the illustrated case, a fluctuation of the SO2 concentration of 38% related to the mean concentration over time was observed within only about 5 minutes. This illustrates the importance of an optimization of oxy-fuel combustion systems to minimize discontinuities of operational parameters. Such an approach is especially important, to operate an oxy-fuel plant successfully at low excess O2 concentrations.

As part of investigations at IFK, the SO3 separation characteristic of the ESP at KSVA was investigated and a SO3 capture rate of around 62% was found for oxy-fuel-firing conditions. Moreover, an increased sulphur content in the ESP ashes was detected und oxy-fuel combustion conditions.


X stoichiometric ratio

c, Mass concentration of component i in the flue gas [mg/m3 (STP), dry]

Yi Weight fraction of component i in the fuel [wt -%, raw]

Mm,, Molecular weight of component i [g/mol]

yi Volume fraction of component i in the flue gas [Vol. -%, dry]

yLeckage Volume fraction of estimated contaminant gases/ingressed air in the flue gas [Vol. -%, dry]

-M-coal Mass flow of coal feed [kg/h]

VCO2 Flow rate of CO2 for fuel conveying and purge [m3 (STP)/h]

Vmol Molar volume: 22.414 l (STP)/mol


[1] Y. Hu, S. Naito, N. Kobayashi, M. Hasatani, CO2, NOx and SO2 emissions from the combustion of coal with high oxygen concentration gases. Fuel 2000; 79:1925ff.

[2] A. Kather, G. Scheffknecht, Review: The oxycoal process with cryogenic oxygen supply. Naturwissenschaften 2009, 96:993ff.

[3] T. Wall, Y. Liu, C. Spero, L. Elliott, S. Khare, R. Rathnam, F. Zeenathal et al., An overview on oxyfuel coal combustion-State of the art research and technology development. Chemical Engineering Research and Design 2009, 87:1003ff.

[4] P. Adolphi, M. Storr, P.G. Mahlberg, H.H. Murray, E.M. Ripley, Sulfur sources and sulfur bonding of some central European attrital brown coals. International Journal of Coal Geology 1990, 16:185ff.

[5] R.K. Srivastava, C.A. Miller, C. Erickson, R. Jambhekar, Emissions of sulfur trioxide from coal-fired power plants. Journal of the Air & Waste Management Association 2004, 54:750ff.

[6] H. Liu, S. Katagiri, K. Okazaki, Drastic SOx Removal and Influences of Various Factors in O2/CO2 Pulverized Coal Combustion System. Energy Fuels 2001, 15(2):403ff.

[7] C. Chen, C. Zhao, C. Liang, K. Pang, Calcination and sintering characteristics of limestone under O2/CO2 combustion atmosphere. Fuel Processing Technology 2007, 88(2):171ff.

[8] C. Chen, C. Zhao, Mechanism of Highly Efficient In-Furnace Desulfurization by Limestone under O2/CO2 Coal Combustion Atmosphere. Ind. Eng. Chem. Res 2006, 4(14):5078ff.

[9] C. Sheng, M. Xu, J. Zhang, Y. Xu, Comparison of sulphur retention by coal ash in different types of combustors. Fuel Processing Technology 2000, 64:1ff.

[10] J. Cheng, J. Zhou, J. Liu, Z. Zhou, Z. Huang, X. Cao, X. Zhao et al., Sulfur removal at high temperature during coal combustion in furnaces: a review. Progress in Energy and Combustion Science 2003, 29:381ff.

[11] E. Croiset, K.V. Thambimuthu, NOx and SO2 emissions from O2/CO2 recycle coal combustion. Fuel 2001, 80:2117ff.

[12] D. Fleig, K. Andersson, F. Johnsson, B. Leckner, Conversion of Sulfur during Pulverized Oxy-coal Combustion. Energy Fuels 2011, 25(2):647ff.

[13] P. Monckert, D. Reber, J. Maier, G. Scheffknecht, Operation of a retrofitted 0.5 MWth PF combustion facility under oxyfuel conditions - an experience report, Presented at: 32nd International Technical Conference on Coal Utilization & Fuel Systems, Clearwater 2007.

[14] R. Stanger, T. Wall, Sulphur impacts during pulverised coal combustion in oxy-fuel technology. Progress in Energy and Combustion Science 2011, 37:69ff.

[15] J. Ahn, R. Okerlund, A. Fry, E.G Eddings, Sulfur trioxide formation during oxy-coal combustion. International Journal of Greenhouse Gas Control, 2011, 5:127ff.

[16] G. Scheffknecht, L. Al-Makhadmeh, U. Schnell, J. Maier, Oxy-fuel coal combustion - A review of the current state-of-the-art. International Journal of Greenhouse Gas Control 2011, 5:16ff.

[17] U. S. EPA, Evaluation and Mitigation of Visible Acidic Aerosol Plumes from Coal Fired Power Boilers September 2006 November 14, 2006.

[18] L. Zheng, E. Furimsky, Assessment of coal combustion in O2+CO2 by equilibrium calculations. Fuel Processing Technology 2003, 81(1):23.

[19] D. Fleig, K. Andersson, F. Normann, F. Johnsson, SO3 Formation under Oxy-Fuel Combustion Conditions. Ind. Eng. Chem. Res 2011, 50(14):8505ff.

[20] Y. Cao, H. Zhou, W. Jiang, C.-W. Chen, W.-P. Pan, Studies of the Fate of Sulfur Trioxide in Coal-Fired Utility Boilers Based on Modified Selected Condensation Methods. Environ. Sci. Technol 2010, 44(9):3429ff.

[21] P. Marier, H.P. Dibbs, The catalytic conversion of SO2 to SO3 by fly ash and the capture of SO2 and SO3 by CaO and MgO. ThermochimicaActa, 1974, 8:155ff.

[22] W. Ellison, H. Weiler (Eds.), Experience with Twelve-Months-a-Year European Operation of SCR as Applied to System Design in U.S., High-Sulfur, Coal Service, Presented at: 22nd annual international Pittsburgh Coal Conference 2005.

[23] E. Tronconi, A. Cavanna, C. Orsenigo, P. Forzatti, Transient Kinetics of SO2 Oxidation Over SCR-DeNOx Monolith Catalysts. Ind. Eng. Chem. Res 1999, 7:2593ff.

[24] J.R. Kenney, M.M. Clark; A.A Levasseur; S.G Kang, SO3 Emissions From a Tangentially Fired Pilot Scale Boiler Operating Under Oxy-Combustion Conditions, Presented at: IEAGHG Special Workshop on Oxyfuel Combustion, London, January 26 th 2011.

[25] Y. Tan, E. Croiset, M. A. Douglas, K. V. Thambimuthu, Combustion characteristics of coal in a mixture of oxygen and recycled flue gas, Fuel 2006, 85:507ff.

[26] M. P. Perujo, Dissertation: Condensation of Water Vapor and Acid Mixtures from Exhaust Gases 2004.

[27] R. Hardman, R. Stacy, E. Dismukes, Estimating Sulfuric Acid Aerosol Emissions from Coal-Fired Power Plants. Presented

at: DOE-FETC Conference on Formation, Distribution, Impact and Fate of Sulfur Trioxide in Utility Flue Gas Streams, Pittsburgh, PA, March 1998.

[28] F.H. Verhoff, J.T. Banchero, Predicting dewpoints of flue gases. Chemical Engineering Progress 1974, 70: 71ff.

[30] B. ZareNezhad, New correlation predicts flue gas sulfuric acid dewpoints. Oil & Gas Journal 2009, 107:60ff.

[31] T. Wild, S. Lysk, F. Kluger, H. Bischoff, Besonderheiten von Feuerungen in Oxyfuel-Dampferzeugern und Auswirkungen auf Auslegung und Betrieb.VGBPowerTech 2010, 11:32ff.

Appendix A.

A.1. Normalization of concentrations

In the following the equations used to normalize measured concentrations (normalization to 6 Vol.-% O2) in this study are introduced. Equation (8) applies for air-firing conditions, equation (9) for oxy-fuel-firing. The exponent 'ref denotes reference conditions (e.g.: 6 Vol.-% O2), while 'o' denotes operating conditions.