Back to homepage To references: sulfur capture

Sulphur capture in fluidized bed boilers

- research at Chalmers University

Based on paper presented at the Sulphation Workshop of the 35th IEA Fluidized Bed Conversion Meeting,Vienna, November 7-9, 1997

A later and more comprehensive paper giving a more general overview of sulphur capture research in FBC is:
Lyngfelt, A. and Leckner, B., "Sulphur capture in circulating fluidized bed boilers - can the efficiency be predicted?",
Chem. Eng. Sci., Vol. 54 (1999) pp. 5573-5584.

(no figures available in this version)

Abstract

Since the middle of the eighties sulphur capture in fluidized bed boilers (FBBs) of commercial size has been studied at Chalmers. The boilers used in the studies include a 16 MW stationary FBB and a 12 MW circulating FBB located at Chalmers, a 40 MW circulating FBB in Nyköping and a 160 MW circulating FBB in Örebro. The following areas have been subject for detailed investigations:

The temperature dependence of sulphur capture performance. A result of this study was the discovery of the important effect of reducing conditions on the sulphur capture process.

The effect of local reducing conditions on sulphur capture. Besides temperature dependence, this includes the effect of air staging in circulating FBBs, the formation of calcium sulfide in stationary and circulating FBBs, and the effect boiler type, i.e. circulating vs. stationary FBB. The sulphur capture performance was shown to be dependent on the occurrence of reducing conditions inside the combustion chamber as measured with zirconia cell oxygen probes. Furthermore the reductive decomposition has been studied directly inside the combustion chamber with extraction of gas.

The knowledge about the effect of reducing reducing conditions was applied in the development of a method, reversed air staging, where nitrous oxide emissions are reduced by 80-90% without increased emissions of SO2.

From chemical analysis of sieved fractions of the exiting solid flows from the boilers it has been possible to determine the residence times for the various particle sizes. The sorbent particle size decrease in the boilers has also been studied. From the chemical analyses the degree of sulphation has been obtained as a function of particle size.

Laboratory studies of the reactivity of up to 11 particle sizes have been made. The results were generalized in the form of a reaction constant, which is a function of degree of conversion and particle size.

The reactivity data derived in laboratory and the residence time data were used in a model to predict conversion versus particle size. The model was validated with data from the three circulating fluidized bed boilers.

Introduction

The start of the research activities regarding SO2 capture in fluidized bed boilers (FBBs) at the Dept. of Energy Conversion, Chalmers University of Technology, is connected to a 16 MW stationary FBB which was put into operation in 1982. The boiler supplied the university with heat, but an additional purpose was to serve as a plant for demonstration and research. In addition to this boiler, comprehensive test series have been carried out in a 40 MW circulating FBB in Nyköping, a 160 MW circulating FBB in Örebro. The last years the research has been concentrated to a 12 MW circulating FBB at Chalmers, which was put into operation in 1991.

The boiler research has been persued in close co-operation with the Dept. of Inorganic Chemistry, Chalmers, where laboratory experiments have been made. In addition close contacts have been kept with two other Nordic groups working with sulphur capture research, Dept. of Chemical Engineering, Technical University of Denmark, and The Combustion Chemistry Research Group at Åbo Akademi University, Finland.

The purpose of the paper is to give a brief overview of the SO2 capture research activities at Chalmers.

Temperature dependence and reducing conditions

The sorbent used in most of the studies, Ignaberga, is porous limestone from the Cretaceous period and was chosen as one of the most reactive of Swedish limestones commercially available.[1] The first results from the 16 MW stationary FBB at Chalmers showed a very strong temperature dependence of the sulphur capture performance.[2] (Fig. 1.) The results from the stationary FBB was compared to results from a comprehensive test series in a 40 MW circulating FBB where the same limestone, Ignaberga, was used.[3] This comparison showed that the temperature dependence was much more significant in the stationary FBB. In addition the sulphur capture performance was much more efficient in the circulating FBB.

It had previously been assumed that the temperature dependence of sulphur capture was a "property" associated with the limestone. A literature study, however, showed that the marked fall in sulphur capture performance with temperature observed in FBBs was not observed in laboratory experiments.[4] (Fig. 2.) Instead the temperature dependence was explained as an effect of reducing conditions in the bottom bed. The effect of reducing conditions in the bottom bed was also advanced to explain the more efficient sulphur capture of the circulating FBB.

In an additional study of the temperature dependence the appearance of reducing conditions in the bottom bed of the stationary FBB was explained by the by-pass of air according to the two-phase model.[5] Measurements in the bottom bed with a zirconia-cell oxygen probe had shown that the conditions were reducing 80-90% of the time, despite an air ratio of 1.4. (Since no secondary air was used the in-bed air ratio was the same as the total air ratio.) Evidence was also put forward that the temperature dependence is explained by reductive decomposition of CaSO4:

                                                                                                          (1)

Subsequently a boiler test was designed to provide evidence that a reductive decomposition of CaSO4 actually takes place in the boiler.[6,7] This test showed that at temperatures above 880-890° C the sulphur retention, defined as fraction of added S captured, became negative, indicating a net release of sulphur from the sorbent in the bed. At a temperature of 930° C the amount of sulphur leaving the boiler was more than twice the amount of sulphur added to the boiler in the form of fuel sulphur. Thus conclusive evidence of the decomposition of CaSO4 was at hand.

It was possible to describe the results obtained with a model where sulphur is captured when the sorbent is exposed to oxidizing or reducing conditions and decomposition occurs under intermediate conditions.[8]

The decomposition of CaSO4 was studied in laboratory.[9] Later, laboratory experiments where limestone was sulphated under conditions alternating between oxidizing and reducing were made in Denmarks University of Technology.[10,11,12] These tests showed that SO2 is released during the shifts between oxidizing and reducing conditions.

Conversion versus size, particle size decrease and limestone reactivity

In order to obtain a better understanding of the effect of particle size, exiting solid flows from the boilers were sieved and the various particle sizes were analysed for Ca and S. Knowing the exiting flows of both the fly-ash and the bottom bed ash the fate of the added limestone particles could be determined. A comparison of the 16 MW SFBB and the 40 MW CFBB was made and showed that in both cases the added sorbent particles, Ignaberga, with an initial mass median size of about 0.6 mm, decreased to about 0.12 mm.[13] (Fig. 3.) Also a comparison of the conversion, i.e. degree of sulphation, versus size was made and showed surprisingly that the conversion was rather independent on size. (Fig. 4.) It also showed a considerably higher conversion for all particle sizes in the circulating FBB. The latter cannot be explained in terms of residence time. Although the residence time for the small particles is longer in the circulating FBB, the large particles have a much longer residence time in the stationary FBB. The lower conversion of the large particles in the stationary FBB was attributed to the negative effect of reducing conditions.

In a subsequent study two methods of estimating the residence time distribution of the sorbent particles were investigated and compared for the 40 MW CFBB.[14]

In a second test series in the 40 MW Nyköping CFBB, Ignaberga limestone was compared to a much less reactive limestone called Köping.[15] Laboratory tests had shown that the final conversion under similar conditions was four times larger for Ignaberga, 36%, compared to Köping, 9%. Surprisingly enough, the difference in sulphur capture performance was small. The results could in part be explained by the substantial size decrease, both limestones decreased from a mass median size of about 0.6 mm to 0.12 mm.

A subsequent test series with these two limestones in the 165 MW circulating FBB in Örebro, showed a somewhat larger difference between the two limestones, and a somewhat lower sulphur capture performance.[16]

CaS formation

In connection with one of the previously mentioned tests in the 16 MW stationary FBB, high concentrations of CaS were found in the bed ash.[17] As much as half of the sulphur present in the bed material was in the form of sulphide. However, sulphide was only present after stopping of limestone addition, i.e. in connection with high SO2 concentrations.

A subsequent study of CaS formation in the 12 MW circulating FBB at Chalmers revealed that only small amounts of sulphide, a few per cent of total sulphur, was formed independent on the extent of air staging.[18] However, when limestone addition was stopped and the SO2 concentration increased, substantial amounts of CaS was formed.

Air staging and reducing conditions

The effect of air staging on sulphur capture was studied in the 12 MW CFBB at Chalmers. Three cases of staging was studied: no staging, normal staging and intensified staging.[19] The sulphur retention dropped to 40% under intensified staging, compared to about 90% under normal and no staging. (Fig. 5.) When the air staging cases were compared at a higher temperature, 930° C, the sulphur retention was negative both for normal and intensified staging, indicating a net release of sulphur from the sorbent.

During these tests the fraction of time under reducing conditions was measured with a zirconia cell oxygen probe.[20] The fraction of time under reducing conditions in the bottom bed was small for no staging, 70-80% for normal staging and 100% for intensified staging. (Fig. 6. ) Later oxygen probe measurements show the occurrence of reducing conditions in various parts of the combustion chamber in more detail.[21] Similar to the stationary FBB there is a through-flow of air which by-passes the bottom bed. The presence of reducing conditions in the bottom bed is a combined effect of this by-pass and of air-staging . This by-pass was examed in the 12 MW stationary FBB with the extraction of gas and solid samples from the combustion chamber.[22]

The reductive decomposition of CaSO4 was studied directly by the extraction of gas from the combustion chamber at various heights.[23] The study shows that the decomposition takes place in the bottom bed, below about 0.6 m, and that much of the sulphur released from the sorbent is recaptured in the splash zone above the bed.

In connection with the development of a patented method, "reversed air staging", for obtaining low N2O emissions from fluidized bed combustion the effect of air staging on sulphur capture was further studied.[24,25,26,27,28] While the emission of N2O decreases with raised temperature or lowered air ratio, the emission of SO2 increases. This conflict between the desire to reduce N2O emission with maintained efficient sulphur capture was circumvented by addressing the conditions in the lower and the upper part of the combustion chamber. In contrast to the emission of SO2, the emission of N2O is not much affected by the conditions in the bottom part. Thus the emission of N2O could be decreased by lowering the air ratio in the upper part if the air ratio in the lower part is simultaneously raised.

Modelling and limestone reactivity

In order to obtain accurate data for modelling of sulphur capture the reactivity of the two limestones Ignaberga and Köping was studied in detail in the laboratory.[29] Up to 11 particle sizes were studied and an effective first order rate constant was derived as function of conversion. It was found that this rate parameter could be approximated as function of conversion with an exponential decay, and that the constants in this decay function could be expressed as function of particle size. (Fig. 7.)

These reactivity data and data on residence time as function of particle size were then applied in a model predicting the sulphur capture performance of the 12, 40 and 165 MW circulating FBBs.[30] Furthermore the degree of conversion as function of particle size was derived and compared to results from analysis of sieved ash samples. (Fig. 8.) The model gave a reasonable agreement and the difference in sulphur capture performance between the small 12 MW CFBB and the two larger could be explained in terms of residence time and particle size distribution. (Fig. 9)

The model was only adopted for Ignaberga limestone. For Köping limestone this was not possible because of the large differerence in conversion obtained in laboratory and in the boiler. (Fig. 10.) This is presently investigated to see if this difference can be explained in terms of an effect of alternating conditions.[31]

References

1) Andersson, E., Blid, C., Lindqvist, O. and Nielsen, B., 1982, Svenska karbonatstenars kapacitet att absorbera svaveldioxid vid atmosfärstryck, Report A8201, Dept. of Inorganic Chemistry, Chalmers University of Technology, Göteborg

2) Åmand, L-E., Johansson, S., Karlsson, M., and Leckner, B., 1986, "Emissions from a fluidized bed boiler," Report A86-156, Dept. of Energy Conversion, Chalmers University of Technology, Göteborg

3) Leckner, B., and Åmand, L.-E. , 1987, "Emissions from a circulating and a stationary fluidized bed boiler: A comparison," Proc. Int. Conf. Fluid. Bed Combustion, Vol. 9, pp. 891-897.

4) Lyngfelt, A., and Leckner, B., 1989, "Sulphur capture in fluidised-bed combustors: temperature dependence and line conversion," J. Inst. Energy, Vol. 62, pp. 62-72.

5) Lyngfelt, A., and Leckner, B., 1989, "Sulphur capture in fluidized bed boilers: the effect of reductive decomposition of CaSO4," Chem. Eng. J., Vol. 40, pp. 59-69.

6) Lyngfelt, A., Åmand, L.-E., and Leckner, B., 1988, "The effect of reducing conditions on sulphur capture - a comparison of three boilers," Proc. of the Intitute of Energy's Fourth International Fluidised Combustion Conference, London, pp. II/10/1-II/10/11.

7) Lyngfelt, A., and Leckner, B., 1989, "SO2 capture in fluidised-bed boilers: re-emission of SO2 due to reduction of CaSO4," Chem. Eng. Sci. Vol. 44, pp. 207-213.

8) Lyngfelt, A., and Leckner, B., 1993, "Model of sulphur capture in fluidised-bed boilers under conditions changing between oxidising and reducing," Chem. Eng. Sci. Vol. 48, pp. 1131-1141.

9) Ghardashkhani, S., Ljungström, E., and Lindqvist, O. , 1989, "Release of sulfur dioxide from calcium sulfate under reducing conditions," Proc. Int. Conf. Fluid. Bed Combustion, Vol. 10, pp. 611-615.

10) Hansen, P., Dam-Johansen, K., Bank, L., and Östergaard, K. , 1991, "Sulphur retention on limestone under fluidized bed combustion conditions - an experimental study," Proc. Int. Conf. Fluid. Bed Combustion, Vol. 11, pp. 73-82.

11) Hansen, P., Dam-Johansen, K., and Östergaard, K. , 1993, "High-temperature reaction between sulphur dioxide and limestone -V. The effect of periodically changing oxidizing and reducing conditions," Chem. Eng. Sci. Vol. 48, pp. 1325-1341.

12) Hansen, P., 1991, Sulphur Capture in Fludized Bed Combustors, PhD Thesis, Dept. of Chemical Engineering, Technical University of Denmark, Lyngby.

13) Lyngfelt, A. and Leckner, B., 1991, "Sorbent size reduction and conversion versus particle size in fluidized bed boilers," Proc. of the Intitute of Energy's Fifth International Fluidised Combustion Conference, London, pp. 179-190.

14) Lyngfelt, A., and Leckner, B., 1992, "Residence time distribution of sorbent particles in a circulating fluidised bed boiler," Powder Technol., Vol. 70, pp. 285-292.

15) Mjörnell, M., Leckner, B., Karlsson, M., and Lyngfelt, A., 1991, "Emission control with additives in CFB coal combustion," Proc. Int. Conf. Fluid. Bed Combustion, Vol. 11, pp. 655-664.

16) Leckner, B., Karlsson, M., Mjörnell, M., and Hagman, U. , 1992, "Emissions from a 165 MWth circulating fluidised-bed boiler," J. Inst. Energy, Vol. 65, pp. 122-130.

17) Lyngfelt, A., Langer, V., Steenari, B.-M., and Puromäki, K., 1995, "Calcium sulphide formation in fluidized bed boilers," Can. J. Chem. Eng., Vol. 73, pp. 228-233.

18) Mattisson, T., and Lyngfelt, A., 1995, "The presence of CaS in the combustion chamber of a 12 MW circulating fluidized bed boiler," Proc. Int. Conf. Fluid. Bed Combustion, Vol. 13, pp. 819-829.

19) Lyngfelt, A., and Leckner, B., 1993, "SO2 capture and N2O reduction in a circulating fluidized-bed boiler: influence of temperature and air staging," Fuel, Vol. 72, pp. 1553-1561.

20) Lyngfelt, A., Bergqvist, K., Johnsson, F., Åmand, L.-E., and Leckner, B., 1993, "Dependence of sulphur capture performance on air staging in a 12 MW circulating fluidised bed boiler," Gas Cleaning at High Temperatures. Eds. Clift, R., and Seville, J.P.K., Blackie Academic & Professional, Glasgow, pp. 470-491.

21) Lyngfelt, A., Åmand, L.-E., Müller, E., and Leckner, B., 1996, "Reversed air staging - a method to reduce nitrous oxide emissions from circulating fluidized bed boilers," in 7th International Workshop on Nitrous Oxide Emissions, Cologne, April 21-23, 1997

22) Lyngfelt, A., Åmand, L.-E., and Leckner, B., 1996, "Progress of combustion in the furnace of a circulating fluidised bed boiler," Twenty-sixth Symposium (International) on Combustion. pp. 3253-3259.

23) Lyngfelt, A., and Leckner, B., 1997, "Sulphur capture in circulating fluidized bed boilers - decomposition of CaSO4 under local reducing conditions" accepted for publication in Fuel

24) Lyngfelt, A., Åmand, L.-E., and Leckner, B., 1994, "Procedure for two-stage combustion of solid fuels in a circulating fluidized bed," SE patent application 9402789-3, document 502292/C/, Kvaerner EnviroPower AB, Goeteborg (Sweden)

25) Lyngfelt, A., Åmand, L.-E., and Leckner, B., 1995, "Low N2O, NO and SO2 emissions from circulating fluidized bed boilers," Proc. Int. Conf. Fluid. Bed Combustion, Vol. 13, pp. 1049-1057.

26) Lyngfelt, A., Åmand, L.-E., Karlsson, M., and Leckner, B., 1995, "Reduction of N2O emissions from fluidized bed combustion by reversed air staging Combustion and emissions control," Second International Conference on Combustion and Emissions Control, The Institute of Energy, London, pp. 89-100.

27) Lyngfelt, A., Åmand, L.-E., Gustavsson, L., and Leckner, B., 1996, "Methods for reducing the emission of nitrous oxide from circulating fluidized bed combustion," Energy Conversion and Management, Vol. 37, pp. 1297-1302.

28) Lyngfelt, A., Åmand, L.-E., and B. Leckner, 1997, "Reversed air staging - a method for reduction of N2O emissions from fluidized bed combustion of coal," accepted for publication in J. Inst. Energy

29) Mattisson, T. and Lyngfelt, A., 1997, "A method of evaluating limestone reactivity with SO2 under fluidized bed combustion conditions," submitted for publication

30) Mattisson, T. and Lyngfelt, A., 1997, "A sulphur capture model for circulating fluidized bed boilers," accepted for publication in Chem. Eng. Sci.

31) Mattisson, T. and Lyngfelt, A., 1997, "The reaction between sulfur dioxide and limestone under periodically changing oxidizing and reducing conditions - the effect of cycle time", submitted for publication


Last update: Sept. 3, 1997

Please e-mail any comments and suggestions regarding this page to :
Anders Lyngfelt
anly@entek.chalmers.se