Experimental 2021 Materials

Two limestones originating from Megalopolis (Southern Greece) and Florina (Northern Greece) areas were investigated for their CO2 and SO2 capture capability. Grinding and sieving were applied to provide a baseline particle size fraction between 150 and 250 ^m and two other particle size fractions of 38-53 ^m and

500-800 jam, for the investigation of the effect of granulometry on the cycling performance of the limestones.

20.2.2 Apparatus and procedure

The sorption and cyclic tests were conducted in a TA instruments Q600 thermo-gravimetric analyzer. The accuracy of this instrument is ±1 pg. Temperature is measured by a thermocouple located immediately below the sample holder. Each sample of ~20 mg is placed in an open alumina pan, which in turn is placed on a horizontal beam holder. The temperature and weight of the sample are recorded continuously. The same temperature was applied for both the sorption and the calcination stages. A schematic layout of the reactor is shown in Fig. 20.1. The initial calcination of the limestones, however, was performed under non-isothermal conditions, raising the temperature from 30°C to the sorp-tion/calcination temperature with a heating rate of 20°C/min. The samples were then maintained at this temperature for 5 min to ensure complete calcination. The flow rate remained constant at 800 ml/min for both the sorption and calcination steps. Pure nitrogen was used for the calcination process in all cases, whereas carbonation of the samples was performed either with pure CO2 or with a mixture containing 80% CO2 and 20% N2 by volume, respectively. A gas mixture composed of 80% CO2, 2900 ppm SO2, 3% O2, and N2 as the balance was used during the co-capture tests, whereas a mixture containing 2900 ppm SO2 and 3% O2 in N2 was utilized for all the sulfation cycle tests.

Thermocouple

Fig. 20.1 Schematic diagram of the thermogravimetric analyzer at atmospheric pressure.

Thermocouple

Fig. 20.1 Schematic diagram of the thermogravimetric analyzer at atmospheric pressure.

The sorption time plays an important role in determining the capture capability and performance of limestones. A fast carbonation stage is normally completed within 8 min. Therefore, a sorption time of 8 min was defined as the baseline for the tests performed. The effect of time on the sorption capability and performance of the samples was also investigated by tests with two alternative sorption times, 3 and 30 min. In all cases, the time for the calcination stage was maintained at 5 min. Two additional particle size fractions, 38-53 jam and 500-800 jam, were used to investigate the influence of sorbent granulometry on the carbonation behavior of the limestones. Moreover, the effect of the reaction temperature on the capture capability of the sorbents was studied by runs at two additional capture temperatures, 650°C and 750°C.

Chemical analyses of the limestone samples were conducted using inductively coupled plasma-atomic emission spectroscopy on an ICP-AES multielement spectrophotometer (Perkin-Elmer). Various acid solutions (hydrochloric, hydrofluoric, etc.) were used to dissolve the samples and then to determine their elements. More details on the process are given elsewhere (Skodras et. al., 2007). A complete data set referring to the chemical analysis of both samples appears in Table 20.1.

Table 20.1 Chemical properties of the sorbents.

Samples

CaO (%wt)

MgO (%wt)

Al2O3 (%wt)

Na2O (%wt)

Fe2O3 (%wt)

(%wt)

TÍO2 (%wt)

Rest (%wt)

Megalopolis

45

0.48

0.03

0.01

0.02

0.00

0.00

54.46

Florina

58.1

0.36

1.33

1.42

0.22

0.55

0.06

37.96

20.3 Results and Discussion 20.3.1 Limestone properties

As shown in Table 20.1 both limestones mainly consist of calcite and small proportions of other impurities. Haji-Sulaiman and Scaroni (1990) considered any substances other than CaO as impurities and reported that natural limestones contain ~1-20% substances which are non-active during sorption. These additional substances include silica, quartz, and trace elements. The presence in high proportions of these impurities results in shrinkage of the limestones during calcination. ICP analysis revealed that Florina limestone had the highest percentage of CaO and at the same time the highest percentage of other substances, mainly Al and Na. As previously shown (Laursen et al., 2001), sodium can assist in the sorption of sulfur. Hence Florina limestone is the more promising candidate for the absorption of SO2. On the other hand, these impurities result in lower surface areas for the limestones and their calcines. As indicated in a previous study (Haji-Sulaiman and Scaroni, 1990), the impurities content play a significant role in causing sintering during CO2 capture and calcinations. It has been reported that concentrations of Fe2O3 beyond 0.5 wt.% contribute to the sintering of sorbents. In both Greek limestones, ferrous ions are present at low percentages, of the order of 0.02 wt.% and 0.22 wt.% for Megalopolis and Florina, respectively, so that its effect on sintering is expected to be limited. Na, which is also considered to favor sintering, is also present at a very low percentage in the Megalopolis sorbent.

According to Haji-Sulaiman and Scaroni (1990), impurities could be present in the limestone in one of two forms: (a) homogeneous, with the impurities well dispersed inside the sample particles and (b) heterogeneous, with the impurities concentrated between the crevices or strata. Cases where the impurities are not uniformly dispersed are also included in the second option.

Fig. 20.2 X-ray mapping of epoxy-embedded as-received Megalopolis limestone.

SEI6BE,255

|MgHal„ 8

|AIKa, 10

S^ilfefir fflftl

•n.

|5iKa. 26

tar,a. 24

IFeKa. 7

#

Fig. 20.3 X-ray mapping of epoxy-embedded as-received Florina limestone.

Fig. 20.3 X-ray mapping of epoxy-embedded as-received Florina limestone.

As can been seen from the SEM-X-ray images (Figs. 20.2 and 20.3), in the case of Florina limestone the silica content is concentrated in high proportions into crevices and is not dispersed in other places inside the particle. Aluminum, which together with silica constitutes the main impurities in Florina limestone, follows similar dispersion behavior. On the other hand, most of the Megalopolis limestone impurities are well dispersed inside the particles. Magnesium, the major impurity, is so well dispersed that it exhibits a similar profile as calcium. Four types of voids are presented in Ca-based sorbents - micro-pores, micro-fractures, macro-pores, and macro-fractures (Laursen, 2002; Cheng et al., 2004). These structural characteristics are also clearly visible in the SEM images. Megalopolis limestone exhibits a higher proportion of micro-pores and micro-structures, whereas Florina limestone presents a more stable structure with a higher number of macro-pores and macrostructures, as shown in Figs. 20.4 and 20.5.

20.3.2 Carbonation of the sorbents

The cyclic CO2 capture capability for a pure CO2 gas stream for both limestone samples is presented in Fig. 20.6. At the beginning of each carbonation cycle, the reaction between CaO and CO2 is rapid, as indicated by the slope of the mass profile. This stage is governed by the gas-solid chemical reaction taking place on the surface of the particles. This reaction results in the formation of CaCO3, which slowly obstructs the micro-pores and renders the internal CaO inaccessible to CO2. This first stage is followed by a second slower stage, which is controlled by the diffusion of CO2 in the CaCO3 layer (Abanades, 2002; Abanades and Alvarez, 2003; Barker, 1973; Chrissafis et al., 2005; Grasa and Abanades, 2006; Mess et al., 1999; Silaban and Harrison, 1995; Sun et al., 2005). As can be seen from Fig. 20.6, the chemically controlled stage is completed in the first 2-3 min of the carbona-tion, and this time remained approximately constant for all the cycles performed. This indicates that the time for the chemically controlled stage is unaffected by the number of carbonation/calcination cycles. On the other hand, the conversion during the fast carbonation stage decreased as the number of cycles increased. For example, the time needed for calcination during the first three cycles was almost 5 min, whereas at elevated number cycle the calcination time was significantly reduced. Therefore, the calcination time seems to be affected by the cyclic process, in agreement with observations of previous studies (Grasa and Abanades, 2006; Ryu et al., 2006).

As depicted in Figs. 20.6 and 20.7, the CO2 capture decreased during repeated cycles, meaning that part of the calcium was not able to be utilized as the cycle number increased. Even from the first cycle the carbonation capacity is limited to 89 wt.% and 86 wt.% for Megalopolis and Florina limestones, respectively. Similar values for the first carbonation of the limestones have also been reported in previous studies (Silaban et al., 1996).

Despite the fact that complete calcination was achieved in each cycle, the ability of CaO to react with CO2 continuously declined. In all cases the reduction in CO2 capture ability is considered to have resulted from textural transformations occurring after at least one carbonation/calcination cycle.

Fig. 20.4 High-resolution SEM image of as-received Megalopolis limestone.
Fig. 20.5 High-resolution SEM image of as-received Florina limestone.

This reduction was highest in the first two cycles of the process. Munoz-Guillena et al. (1995) explained that the textural properties differ, even between the original limestone and the calcined CaO. Sun et al. (2005) attributed the continuously increasing failure of calcium carbonate to capture CO2 to sintering effects during carbonation and to the lack of re-crystallization during calcination. During the successive carbonation/calcination cycles the voids between the grains are reduced due to formation of necks. These necks between the grains result in a decrease in the surface area and to the loss of porosity (Stanmore and Gilot, 2005). More specifically, sintering leads to the closure of micro-pores and simultaneously to the widening of the larger pores. As a result, the available CaO for CO2 capture decreases continuously as the number of carbonation/calcination cycles increases.

■ Megalopolis

50 100 150

Time (min)

Fig. 20.6 Weight change during carbonation/calcination cycles, Carbonation-calcination at 850°C, 8 min, pure CO2, 15 cycles.

The CO2 capture capability of Megalopolis limestone was higher for the first two cycles than for the Florina limestone, even though Megalopolis limestone has a lower Ca content than the Florina one. The most plausible explanation is that the calcium available for sorption is higher in the case of Megalopolis limestone, despite lower calcium concentration. Notwithstanding its initial higher capture, the decay for the Megalopolis limestone in the next cycles was higher from the 3rd cycle onward so that its capture capability became significantly lower than for the Florina sorbent. As noted above, the Megalopolis limestone structure has a greater number of micro-pores. Sintering causes closure of the micro-pores and therefore a rapid decrease in the capture capability. The number of micro-pores for Florina limestone is lower and thus its loss of capture ability was lesser than for the Megalopolis limestone. Hence the reduced capture capability of Florina limestone during the three initial cycles can be explained by the enhanced number of macro-pores during the early stages.

Abanades (2002) suggested an empirical equation to predict the decay of the limestone samples when subjected to continuous carbonation/calcination cycles:

where XcN represents the CO2 capture capacity, N is the number of cycles, while f and b are empirical constants with values f = 0.782 and b = 0.174. A schematic diagram of the empirical equation is depicted in Figs. 20.7 and 20.8 in comparison with the experimental results of the cyclic tests for both Megalopolis and Florina limestones. As can be seen from Fig. 20.7, there was excellent agreement between the correlation and the baseline test results. During the first five cycles the decay of Megalopolis limestone is higher than predicted by the correlation. From the 7th cycle onward the decay was below that indicated by Eq. (20.1). In both cases the deviation between the baseline and the correlation is minimal. On the other hand the deviation between the baseline corresponding to the capture capability of Florina limestone and the theoretical curve is higher both before the first five cycles and after the 7th cycle. Nevertheless, these deviations are relatively small.

a) Particle size effect

The conversion data for the Megalopolis and Florina limestones are compared in Figs. 20.7 and 20.8. For all the tests the particle size used is between 150 and 250 pm except where specified. The general trend is the same for all particle size ranges tested. The loss in capture capacity of the limestone is rapid during the first four cycles, especially for the first cycle. The rate of decay decreased during the following cycles. For the Megalopolis limestone, a reduction in the particle size resulted in increased capture capability and less decay for all cycles tested. Smaller particles result in higher external surface area and shorter penetration distances and thus lower decay during at least the first few cycles. However, the decay of the limestone was almost at the same level as that for the baseline during the following cycles. The capture capability during the successive cycles seems to have been unaffected by the increased particle size for Megalopolis limestone. The conversions for the 150-250 and 500-800 pm particle fractions were very similar in all cases.

As depicted in Fig. 20.8, during the first five cycles the capture capability of Florina limestone was almost identical for all the three particle sizes tested. From this cycle onward, the 150-250 pm particle fraction shows the best CO2 capture ability, followed by the smallest particles (38-53 pm). The lowest CO2 capture capability was for the largest particles (500-800 pm) for this limestone. The conversion deviations due to different particle sizes are of the order of 0.25. It is also notable that in all cases the conversion curves for all three particle size fractions investigated are higher than for Eq. (20.1).

10 1

(0

0.9 -

o

(A 0

0.8 -

13

b

0.7 -

TO

13

0.6 -

t

T

U

0.5' -

3

W

0.4 -

u

m

o

03-

o

<u

0.2 -

13

0.1 -

Baseline, 150-250 |im, 8min, 850°C 38-53 |im, 8min, 850°C 500-800 |im, 8min, 850°C 150-250 |im, 3min, 850°C 150-250 |im, 30min, 850°C 150-250 |im, 8min, 750°C Empirical

6 8 10 Number of reaction cycle Fig. 20.7 Cyclic CO2 capture ability of Megalopolis limestone in 80% v/v CO2-20% v/v N2. Calcination in N2.

1.0 1

w

0.9 -

o

(A <U

0.8 -

13

b

0.7 -

<B

13

0.6 -

t

T

(U

0.5 -

3

m

0.4 -

u

O

03-

O

<u

0.2 -

13

s

0.1 -

0.0 -

Baseline, 150-250 ^m, 8min, 850°C 38-53 ^m, 8min, 850°C 500-800 ^m, 8min, 850°C 150-250 ^m, 3min, 850°C 150-250 ^m, 30min, 850°C 150-250 ^m, 8min, 750°C Empirical

Number of reaction cycle Fig. 20.8 Cyclic CO2 capture ability of Florina limestone in 80% v/v CO2-20% v/v N2. Calcination in N2.

b) Effect of reaction time and temperature

As previously reported, the baseline tests were performed at a temperature 850°C and for a time of 8 min. The conversion profiles corresponding to the carbonation capacity of Megalopolis and Florina limestones at varying times (3 min and 30 min) are plotted in Figs. 20.7 and 20.8, respectively. It is seen that the reaction time affected the CO2 capture capability quite differently for the two limestones. Megalopolis limestone seems to have been less affected by the variation in carbonation times than the Florina limestone. Note that the 3-min time interval was long enough for the carbonation reaction. In previous studies (Grasa and Abanades, 2006; Sun et al., 2005) a reaction time of 3 min was considered to be too short to complete the reaction between CaO and CO2. These differences in results probably reflect different experimental systems in each situation. In the present study, the experimental conditions such as the weight, the gas flow rate, and size of the furnace probably allow the reaction between the gas CO2 and the solid sorbent to take place more quickly.

As shown in Fig. 20.7, Megalopolis carbonation was unaffected by a reduced reaction time, since the conversion profiles for the baseline (8 min) and the 3-min test almost coincide for every cycle. This behavior may be because the sintering of the limestone occurs to the same extent for 8 min as for 3 min of car-bonation. The 30-min test showed that until the 3rd cycle, the carbonation capacity was not affected by the longer reaction times. As depicted in Fig. 20.7, for cycle number >3, the 30-min carbonation resulted in less CO2 capture capacity for the Megalopolis limestone. The decay of the specific limestone also appeared to be greater for this elevated time interval, suggesting that the sintering effect was stronger for longer carbonation times. As reported previously (Stanmore and Gilot, 2005), various factors such as temperature, time, and the presence of CO2 alter the sintering effect.

Different trends were observed for Florina limestone. As Fig. 20.8 illustrates, the carbonation efficiency in the 3-min tests was less than for the baseline. In fact, the conversion of Florina limestone was limited, even in the first cycle, with more than 3 min needed for calcium utilization of this limestone. The calcium utilization for the 30-min tests was similar to that of the baseline. However, the decay in the 3-min tests was unexpectedly lower, probably due to the morphological characteristics and the pore structure of the Florina limestone. Notwithstanding the sintering effect, the crevices and the larger pores in this sorbent provide pathways that allow CO2 to reach the calcium available in the interior. This, in conjunction with the high reaction time, leads to the absorption of higher quantities of CO2.

Carbonation temperature constitutes another parameter that affects the CO2 capture capability of the Ca-based sorbents. Carbonation is an exothermic reaction, while calcination is endothermic. Therefore, the latter is favored by higher temperatures, whereas the carbonation reaction is favored by lower temperatures around 750°C at atmospheric pressure (Li et al., 2005). In the present study a temperature of 850°C was used as the baseline for both the carbonation and calcination stages. As shown in Figs. 20.7 and 20.8, the CO2 capture capacity was significantly lower for the tests performed at 750°C and a reaction time of 8 min for both limestones. In the case of Megalopolis sorbent, the calcium utilized for sorption was only 25.3% of the calcium utilized at 850°C. The corresponding value for Florina limestone was somewhat higher (35%). Note that the decay of the limestones varies throughout the entire cycling test. In effect, only part of the calcium was utilized and underwent sintering, and therefore most of it was available for the following carbonation cycles. The tests performed at calcination and carbonation temperature of 650°C showed even worse CO2 capture profiles for both the limestones tested.

20.3.3 Sulfation of the sorbents

Figure 20.9 shows results of sulfation tests on 150-250 ^m particles performed with 2900 ppm SO2 without CO2 being present for 2 h. It is seen that the Florina limestone had a higher sulfation efficiency than Megalopolis, with the calcium utilization in the latter case being almost 13%, whereas the corresponding value for Florina limestone was almost 26% after 2 h of continuous operation. As depicted in Fig. 20.9, the lime utilization could continue after 2 h. According to previous research (Stanmore and Gilot, 2005), the total calcium utilization for SO2 capture has the potential to continue to increase, reaching conversions up to 95% after extended reaction times. Fundamental studies on the sulfation of Ca-based sorbents (Laursen, et al., 2002; Ryu et al., 2006) have reported three different sulfation modes: unreacted core, uniform, and network sulfation. The mode in a given case depends on the morphological characteristics of the limestones after sulfation. These studies found that the unreacted core sulfated particles exhibited higher SO2 sorption capacity than the uniform and the network sulfated particles. The uniformly sulfated particles should contain a large fraction of micro-pores, and their sulfation should occur in a homogeneous manner. Although the quantity of SO2 absorbed by unreacted core particles is higher, the CaSO4 layer produced at the rim of the particles is thicker than for the uniformly sulfated particles. Taking into account these general characteristics, Florina limestone seems to follow an unreacted core mode, while Megalopolis is subject to the uniform sulfation mode.

Another important finding is that the sulfation of the samples took place in two clearly distinguished stages. In the first stage, the sulfation rate increased almost linearly with time, whereas in the second, the sulfation proceeded at a significantly slower rate. These observations are consistent with the findings of Ye et al. (1995), who claimed that sulfation is divided into two different stages, the first accompanied by pore filling by gaseous SO2 considered to be the rate-limiting step. As illustrated in Fig. 20.8, this step lasted for ~50 min for Florina limestone. For the Megalopolis limestone, this step required a shorter time interval (30 min). As previously mentioned, the reaction between CaO and SO2, which mainly occurs on the outer surface and inside the pores of the sorbent particles, results in the formation of an outer CaSO4 layer, which obstructs SO2 from reaching the available CaO. However, some SO2 permeates through this layer to gain access to the interior calcium. Therefore, sulfation of the limestone continues slowly in the second reaction stage. In the case of Megalopolis limestone, the contribution of the second stage to the overall sulfation of the limestone is low, accounting for only 15% for a reaction time of 2 h. The second stage contributes more to the sulfation of Florina limestone, accounting for almost 26% of the overall sulfation.

The effect of calcination on the sulfation of the limestones was investigated with experiments, which include a number of sulfation/calcination cycles at 850°C and a constant reaction time of 8 min. The results are also presented in Fig. 20.9. It is well known that the SO2 absorbed by CaO cannot be desorbed upon calcination.

It is clear that the calcination affected the sulfation of both limestones, but to a different extent. The sulfation of Megalopolis limestone followed behavior similar to the continuous sulfation pattern, until the first reaction stage was complete. During the second sulfation stage, SO2 absorption was higher for the cyclic sul-fation/calcination process. The sulfation utilization of Florina limestone also increased with implementation of the sulfation/calcination cycles. This enhancement in sulfation capacity accounts for 1.5% of the 2 h continuous sulfation for both limestones. As in previous research (Laursen et al., 2002), the unreacted core sul-fated particles tended to be regenerated, whereas the uniformly sulfated particles showed very little regeneration. The different effects of calcination on the two limestones are probably related to the different sulfation modes. It is also notable that for the Florina limestone, the times for each of the reaction stages were transformed. As shown in Fig. 20.9, the reaction-limited stage (first stage) extended to 80 min, while the second stage shortened to 40 min. A plausible explanation could be that the calcination cycles prevented the CaSO4 layer from forming, resulting in longer reaction-limited and shortened diffusion-limited stages. Consequently, calcination cycles during the sulfation are beneficial to achieve higher sulfation efficiencies.

O 0.3013

o tn

0 20 40 60 80 100 120 140

Time (min)

Fig. 20.9 Sulfation profiles for Megalopolis and Florina limestones in no CO2 - 2900 ppm SO2 - 3% v/v O2 - balance N2. Calcination in N2.

20.3.4 Co-capture capability of the sorbents

The sorption capacity and performance of the limestone samples when both CO2 and SO2 are present is of significant importance in fluidized bed combustion. Therefore, simultaneous carbonation and sulfation behavior of the Greek limestones was investigated with a gas mixture containing 80% CO2, 2900 ppm SO2, 3% O2, and the balance N2. In every cycle, the quantities of CO2 and SO2 captured were calculated from the thermogravimetric curves. The main consideration is that

Transition from 131 to 2nd stage

□ Megal - 2h sulfation ■ Megal - sulf/calc cycles o Flor - 2h sulfation • Flor - sulf/calc cycles

Transition from 131 to 2nd stage

□ Megal - 2h sulfation ■ Megal - sulf/calc cycles o Flor - 2h sulfation • Flor - sulf/calc cycles

0 20 40 60 80 100 120 140

Time (min)

Fig. 20.9 Sulfation profiles for Megalopolis and Florina limestones in no CO2 - 2900 ppm SO2 - 3% v/v O2 - balance N2. Calcination in N2.

the entire quantity of SO2 captured in each cycle remains absorbed in the sorbent after calcination. A schematic diagram of the proportion of the CO2 captured, in conjunction with the total available calcium in each cycle, is depicted in Fig. 20.10. During simultaneous CO2/SO2 capture a portion of the available CaO is utilized to absorb CO2, while another part participates in the absorption of SO2. This results in reduced CO2 capture capacity compared to carbonation alone. The reactions between the sorbent and the gases along with a detailed analysis are presented by Sun et al. (2005).

As shown in Fig. 20.10, the falloff in CO2 capture capability, during the 10-12 cycles is similar to that for the carbonation tests. This observation creates the perception that for the first 10-12 cycles the reduced availability of CaO during co-capture results in the same deterioration in CO2 capture as due to the sintering effect during carbonation. Nevertheless, the falloff in CO2 capture capability of both the Greek limestones increased after the 12th cycle. As the number of cycles increased, both sulfation and direct sulfation resulted in formation of a thick CaSO4 layer on the surface of the sorbent particles. From the 12th cycle onward, a major portion of the available CaO was utilized for SO2 absorption, and together with the sintering effect caused by the CO2 absorption, resulted in a rapid decay in the CO2 capture capability.

As for the carbonation of the sorbents, each cycle of the simultaneous CO2/SO2 capture is characterized by a rapid sorption stage, followed by a second much slower stage. For the first cycle, the fast sorption stage lasted for about 4 min. As the number of cycles increased, the time needed for the fast sorption stage continuously decreased and the second stage became dominant. During the final sorption cycle, the first stage lasted for less than 1 min. These results are consistent with the observations of Ryu et al. (2006), who ascribed this behavior to the reduced capture capability of the limestones. It is obvious that the first effective sorption stage involves carbonation of the available calcium. As previously mentioned, sulfation is also present during this stage. Therefore, after a sufficient period of sorption, the products of both reactions prevent CO2 from reaching the available calcium, causing a reduction of reaction rate. The slow reaction stage which follows is characterized by the diffusion of CO2 through the product layers. As the number of cycles increases, the thickness of the CaSO4 layer also increases. As a result, the fast reaction time is limited and the slow reaction stage becomes dominant.

Another important aspect of the results is the effect of CO2 in co-capture tests on the sulfation of the limestones. The presence of CO2 in the gas stream initially reduces the available calcium for SO2 absorption and consequently the SO2 sorption is expected to be reduced. In fact as the number of cycles increases the SO2 capture is also increased at a slow but constant rate. Taking into account that SO2 captured in each cycle remains absorbed during calcination, the sulfation of the limestone is expected to increase. Therefore, sulfation could be extended to higher conversions than carbonation. In the meantime, the CaCO3 produced during each cycle could be converted to CaSO4 due to direct sulfation reaction (Stanmore and Gilot, 2005; Sun et al. 2007).

□ Megal - 2h sulfation ■ Megal - sulf/calc cycles o Flor - 2h sulfation • Flor - sulf/calc cycles

□ Megal - 2h sulfation ■ Megal - sulf/calc cycles o Flor - 2h sulfation • Flor - sulf/calc cycles

Fig. 20.10 Cyclic CO2/SO2 sorption ability of Megalopolis and Fiorina limestones in 80% v/v CO2 - 2900 ppm SO2 - 3% v/v O2 - balance N2. Calcination in N2.

Fig. 20.10 Cyclic CO2/SO2 sorption ability of Megalopolis and Fiorina limestones in 80% v/v CO2 - 2900 ppm SO2 - 3% v/v O2 - balance N2. Calcination in N2.

20.3.5 Sequential carbonation and sulfation

Two sets of consecutive carbonation/sulfation experiments were performed in an effort to elucidate the partial effects of sulfation and carbonation on the performance of the limestones. During the first experiments, the samples were initially calcined and subjected to a single carbonation cycle. The second cycle included sulfation of the product calcine in a gas mixture containing 2900 ppm SO2, 3% O2 in N2, and then calcination. The sample was then subjected to 15 carbona-tion/calcination cycles. The second experimental set included 21 cycles, with one sulfation for every four carbonation/calcination cycles. Each carbonation, sulfation, or calcination step lasted for 8 min. The results appear in Fig. 20.11 and 20.12 for the Megalopolis and Florina limestones, respectively. The baseline curves are also included for comparison.

As expected, the CO2 capture capability of both Megalopolis and Florina limestones was reduced after one sulfation cycle is presented. However, the decay of both limestones after the sulfation cycle followed the same trend as for the baseline test. This behavior indicates that the formation of a CaSO4 layer constrains the available CaO, but does not affect the decay mechanism of the sorbents during carbonation/calcination cycles. The slope of the thermogravimetric curves during carbonation cycles, which follow the sulfation, remains the same as for the carbonation/calcination tests. On the other hand, the sulfation rate was slightly affected by the carbonation/calcination cycles. Li et al. (2005) affirmed that the sulfation rate increased when a carbonation/calcination cycle was inserted between the initial calcination and the sulfation of the limestone.

Another important aspect of the results is the increased decay in CO2 capture for the Megalopolis limestone exposed to one sulfation, compared to the

corresponding behavior of the Florina limestone. As already noted, Florina limestone follows an unreacted core pattern when exposed to sulfation conditions, and therefore a CaSO4 layer formed on the surface of the sorbent particles. CO2 molecules are able to penetrate this layer and reach the CaO available on the interior. CaCO3 results in the formation of crevices on the CaSO4 layer. Therefore, the decrease in the ability to capture CO2 occurs due to sintering during carbonation/calcination cycles. The same is also true for the uniformly sulfated Megalopolis limestone. As previously mentioned, sintering affects the CO2 capture capability of Florina limestone to a lesser extent, and this could be the reason for the greater decay of Megalopolis limestone when subjected to calcination and sulfation followed by sequential carbonation/sulfation cycles.

The four-sulfation tests showed that the decay of each limestone strongly depended on the number of sulfation cycles interspersed with the carbonation cycles. After each sulfation cycle, a sharp decay was observed in the sorbent performance, indicating loss of an additional portion of the available CaO. However, the previously noted independence between the sulfation and the rate of decay is again relevant. This finding suggests that the decay mechanism remains unaffected, whether the number of sulfation stages is one or much higher. It is also worth emphasizing the decreased effect of each sulfation stage on the Florina limestone compared to the Megalopolis sorbent. In the one-sulfation tests, the decay in CO2 capture capability of the Florina limestone was less than for the limestone from the Megalopolis one. The sulfation and re-carbonation mechanisms previously described seem to have also applied to the four-sulfation tests. Consequently, Florina limestone presents increased sulfation tolerance compared to the Megalopolis limestone.

8 10 12 14 Number of reaction cycle

Fig. 20.11 Sequential CO2/SO2 capture capability of Megalopolis limestone. Carbonation in 80% v/v CO2 - 20% v/v N2, sulfation in no CO2 - 2900 ppm SO2 - 3% v/v O2 - balance N2. Calcination in N2.

8 10 12 14 Number of reaction cycle

Fig. 20.11 Sequential CO2/SO2 capture capability of Megalopolis limestone. Carbonation in 80% v/v CO2 - 20% v/v N2, sulfation in no CO2 - 2900 ppm SO2 - 3% v/v O2 - balance N2. Calcination in N2.

Baseline

o

1 sulfation cycle

&

A

4 sulfation cycles

v

^ 2 ° ° A A Ä &

A A A A

A

A

8 10 12 14 Number of reaction cycle

8 10 12 14 Number of reaction cycle

Fig. 20.12 Sequential CO2/SO2 capture capability of Florina limestone. Carbonation in 80% v/v CO2 - 20% v/v N2, sulfation in no CO2 - 2900 ppm SO2 - 3% v/v O2 - balance N2. Calcination in N2.

Was this article helpful?

0 0
Solar Panel Basics

Solar Panel Basics

Global warming is a huge problem which will significantly affect every country in the world. Many people all over the world are trying to do whatever they can to help combat the effects of global warming. One of the ways that people can fight global warming is to reduce their dependence on non-renewable energy sources like oil and petroleum based products.

Get My Free Ebook


Post a comment