Abstract
125 MWe circulating fluidized bed combustion (CFBC) boiler experienced severe fouling in backpass of the boiler leading to obstruction of gas flow passage, while using high sulfur lignite with sorbent, calcium carbonate, to capture sulfur dioxide. Optical microscopy of the hard deposits showed mainly anhydrite (CaSO4) and absence of intermediate phases such as calcium oxide or presence of sulfate rims on decarbonated limestone. It is hypothesized that loose unreacted calcium oxides that settle on tubes are subjected to recarbonation and further extended sulfation resulting in hard deposits. Foul probe tests were conducted in selected locations of backpass for five different compositions of lignite, with varied high sulfur and ash contents supplied from the mines along with necessary rates of sorbent limestone to control SO2, and the deposits build-up rate was determined. The deposit build-up was found increasing, with increase in ash content of lignite, sorbent addition, and percentage of fines in limestone. Remedial measures and field modifications to dislodge deposits on heat transfer surfaces, to handle the deposits in ash conveying system, and to control sorbent fines from the milling circuit are explained.
1. Backdrop
India with growing energy consumption is looking at utilizing all its potential energy resources in the most economic and environmentally sustainable manner. Coal will continue to be the major energy source in India due to its availability. Per capita consumption of electricity and GDP growth has direct relation, and energy intensity in developing countries like India is comparatively more than the developed world, and the gap between supply and demand is ever increasing. The demand for all forms of energy is expected to increase substantially in the foreseeable future and is expected to get doubled by 2030. Although coal would continue to be a major energy source in India due to its availability, lignite is fast emerging as an alternate source of fuel for electricity generation. In India, the total lignite potential is 4177 million tonnes. Indian lignites have a typical analytical range of ash content of 15 to 35%, sulfur content of 1.0 to 7.0%, and moisture content of 10 to 45%. The varieties found in Gujarat and Rajasthan region have moderate to high sulfur (1 to 7%) content. It has become an economic necessity to use these lignites for power generation in view of spurt in energy demand, with SO2 emission controlled. Circulating fluid bed combustion (CFBC) technology is employed considering the impurities, moisture, ash, and sulfur content and wide variations in lignite. Hence, the share of lignite-based pit head thermal projects in Gujarat and Rajasthan is increasing. The size of CFB boilers in India using lignite has reached already over 250 MWe and set to increase above 500 MWe, and that underlines the importance.
Slagging, fouling, and ash deposition are major problems experienced in PF boilers. In contrast, agglomeration of bed particles in fluidized bed combustion system is considered as a primary operational issue. Interaction and coalescence of bed particles and ash (sintering) are considered to be the principal sources of agglomeration in CFB boilers employing bed material and fuel ash as binary system. Choking/blocking in fuel path is another peculiar operational problem experienced worldwide in CFB boilers, firing pet-coke, low rank coals, and biofuels [1].
Lignite mineralogy greatly influences combustion behavior. Agglomeration and clogging/blocking are experienced due to sintering of lignite ash with limestone (sorbent) at lower temperature regime in which CFB boilers operate (640–960°C). At this low temperature range, the extensive knowledge built with respect to slagging, fouling, and corrosion phenomenon occurring at higher temperatures in pulverized fuel combustion may not be applicable. In CFB boilers, ash sintering contributes to deposit formation in cyclone, return leg, and postcyclone flue gas channel (backpass) [1]. In operating units, rapid sintering lead to heavy agglomerate formation, which finally inhibited circulation in dense phase areas (such as seal pot) and in the backpass. Understanding the sintering behavior of fuel is required for resolving such problems.
Over the past decades, designers and operators of fluidized beds have been concentrating on developing the CFBC technology by establishing the optimum operating conditions and troubleshooting associated with refractory and so forth. Due attention has not been paid to understand the limestone characteristics that are important for efficient capture of SO2. Present work describes influence of limestone and its grain size in blocking/clogging of cyclone and hard deposits in second pass of CFB boiler during combustion of high sulfur lignite with high ash content (20 to 30%) in CFB units in Giral, Rajasthan state of India.
2. Operational Issues
High sulfur content lignite, available at Giral, Barmer District, and Rajasthan state, is used as fuel. These lignites had posed several operational issues during initial stage of commissioning and stabilization. High sulfur in the lignite needs high limestone feed rates to control emissions. High limestone feed rates caused huge quantities of backpass deposits, which led to obstruction of gas flow passage. Despite providing steam soot blowers for clearing the deposits obstruction of gas flow increased with increase in limestone feed rate.
2.1. Cyclone Standpipe Blockage
During commissioning, ash holdup occurred in cyclone standpipe at low loads of about 20 to 40 MW. Ash analysis of the hold-up material is carried out.
2.2. Backpass Fouling
Sulfur dioxide emitted during combustion is absorbed in situ by adding limestone of size less than 1.0 mm. The CFB boiler experienced fouling in superheater/reheater (SH/RH) coils while adding required quantity of limestone (Figure 1). Heavy and rapid deposit buildup has been experienced on the flue gas side of the heat transfer tubes. Deposit buildup was most severe at low temperature superheater (LTSH)-SH 1B tube bank. Also, growth of ash deposit in final stage reheater tube bank was observed during the initial period of operation. These deposits increased gas-side pressure drop and in turn increased loading of induced draught (ID) fans, with high current, causing boiler trips.
(a)
(b)
Consequently, CFB boiler was required to be operated with less quantity of limestone which resulted in more sulfur dioxide emissions. The fouling took place mostly in LTSH coils of backpass which is placed between reheater and economizer. Due to fouling in the backpass, fly ash particles collected in hoppers of economizer and in other zones got sintered during intermittent storage. Nonoperation of soot blowers (SB) and water ingress while starting soot blowing caused cakes formation. Dislodgement of such cakes leads to difficulty in ash evacuation. Deashing system pump was chocked often, due to sintered particles (lumps) formed due to water ingress.
3. Experiments: Laboratory and Field
3.1. Lignite
Six samples (sample 1 to sample 6) of high sulfur lignite collected from Giral/Rajasthan/India (covering a range of high sulfur content) are considered for the present study of backpass fouling propensity of the high sulfur fuels in CFB boiler. All the fuel samples are prepared in accordance with ASTM-D 2013. The as-received solid fuels are crushed to pass a number 4 sieve (4.75 mm) and then air dried until the loss in weight is not more than 0.1% per hour. Air dried samples are again crushed to pass a number 72 mesh (212 microns). Samples of sizes less than 72 mesh are used for analyses of proximate, ultimate, and calorific values. Adequate quantity of ash of each fuel is generated using proximate analyses at 750°C for further analyses of chemical composition, ash fusion temperature. The proximate, ultimate, and gross calorific values of the samples were carried out using TGA 701 proximate Analyzer (LECO), Elemental analyzer Vario EL III, and PARR Isoperibol Bomb Calorimeter, respectively. The chemical composition of ashes was carried out by ICP- AES, Perkin Elmer.
3.2. Limestone
The sorbents are characterized based on the CaCO3 content, particle size distribution of the parent sorbent, and a relative sulfation reactivity parameter [3]. Calcium utilization, in general, increases as the sorbents particle size decreases. As the particle size distribution of the feed sorbent changes in a CFB due to attrition, it is taken for granted that the feed size distribution of limestone (input) is not as important as the resultant sorbent size distribution in the boiler. On the contrary, mathematical model results show that sulfur capture efficiency is related to particle attrition/fragmentation of sorbent inventory in addition to input particle size distribution to the performance of circulating fluidized bed CFB combustors [4]. The physical and chemical properties of a sorbent are important when evaluating for use in CFB application. Sorbents although chemically similar, may have different sulfation performance. Extensive literature studies on process of desulfurization in CFBC show that sorbent conversion degree is dependent not only on residence time in combustor but also on its porosity, pore structure and pore size distribution [5]. The detailed analyses of Indian limestones-chemical composition, calcium and magnesium carbonate contents, that are used in CFB were performed using Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP-AES) Perkin Elmer Optima 2000 DU and using Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS) Perkin Elmer. Sulfation of limestones of different size fractions showed that sorbent requirement (g of sorb/g of sulfur) is less for finer size fractions [6].
3.3. Deposit Sampling Using Probes and Field Experiment
3.3.1. Deposit Probes
Field experiment using deposit probes is taken up, as the wide range of characterization of the selected limestones with respect to their potential difference as desulfurisation agents in CFBC boilers yielded no definitive evidence of the fouling and deposition faced in the operating units.
A deposit probe is a good tool for finding out the mechanisms of deposit formation. Air cooled deposit probes of type Figure 2 was used for sampling of deposits, which are equipped with detachable rings [2]. The temperature of the probe can be controlled by varying flow rate of pressurized air. For each test, a new probe/ring is used and the weight of the probe/ring is checked before and after exposure. Taking into account exposure time, a rate of deposit buildup (g/(m2 h)) can be calculated. Deposited probes/rings are stored for analysis.
(a)
(b)
(c)
Deposits were collected from three different locations in the backpass after SH-1B, in between RH-2 bundles, and after RH-2 (Figure 3). Chemical composition analysis of the probe deposits is carried out. The sieve analysis of deposits shows significant share of particles smaller than 50 μm size. It was clear that addition of limestone significantly increased the formation of hard deposits compared to firing only lignite, that is, without any limestone.
3.3.2. Particle Size Distribution of Injected Lime
The sieve analysis of collected deposits showed that these deposits were built up mainly by fine lime particles injected into furnace. Figure 4 shows distribution of the particle size for two samples done by wet sieving. The share of particles smaller than 50 μm size indicated that fine fractions were higher than envisaged during design (0 to 5% less than 50 μm). Earlier researchers have shown that the particle size distribution of sorbent could significantly affect deposit formation rate [2].
4. Results and Discussions
Analyses of proximate, ultimate, and gross calorific value and chemical composition of ashes for the seven lignite samples are listed in Table 1. Analysis of chemical composition of the hold-up material in the cyclone standpipe is furnished in Table 2. Detailed limestone analyses-chemical composition, calcium and magnesium carbonate contents for the Indian limestones that are used in CFB, are furnished in Table 3. Fouling probe test condition/measurement details are furnished and the chemical composition analysis of the probe deposits is furnished in Table 4. Mineralogy of the probe deposits as determined by XRD is furnished in Table 5.
4.1. Correlation with Conventional Ash Deposition Indices
Various conventional indices, based upon ash chemistry, have been calculated as indicators of slagging and fouling propensity [7]. Values for the following indices, for the high sulfur lignite samples 1 to 7 are given in Table 1: Silica ratio = SiO2/(SiO2 + Fe2O3 + CaO+MgO)*100 Base/acid ratio = (Fe2O3 + CaO + MgO + Na2O + K2O)/ (SiO2 + Al2O3 + TiO2) Iron index = Fe2O3*B/A Iron/calcium ratio = Fe2O3/CaO Iron + calcium in ash = Fe2O3 + CaO
The interpretation of such ash deposition indices requires caution, as these have been developed for a particular range or type of coal, and influence of boiler design/operating conditions is not accounted. Ash chemistry indices do not count the mineralogical mode of occurrence of the elements of concern and mineral associations, both of which are equally important as the ash chemistry in determination of slagging and fouling. With the above limitations, it can be seen from Table 1 that the values for most of the common ash deposition indices suggest that the lignite samples would have a high propensity to form ash deposits [8, 9]. The values in bold and italics indicate high propensity for ash deposition. Agglomeration can start well below the ash fusion temperatures in fluidized beds for lignite, and influence of Na2O (AFT decreases) and Al2O3 (AFT increases) on Turkish lignite was studied by earlier researchers [10].
4.2. Sulfation of Free Lime in Backpass of Boiler
The investigations of the deposit hardening phenomenon in the CFB boilers have been widely discussed as the occurrence of three types of deposit consolidation mechanisms [11, 12]. Two out of the three consolidation mechanisms result in increase in volume of free CaO rich zones in deposits. Fine sorbent particles, settled either on the tube surface or in the caverns on the “rough” surface of the old deposits, (Figure 5) are exposed to SO2-containing flue gases. These sorbent particles are fine (i.e., not captured in the cyclone), and the majority of particles are already calcined before entering the second pass of the boiler. During their residence on tube surfaces in the convective section, these particles undergo a continuous sulfation through an exothermic reaction (1). The sulfation process is described by the following overall reaction [2]: Further if the temperature of flue gas in vicinity of the sorbent particle is sufficiently high, then the local temperature of the deposits is likely to exceed the sintering temperature due to exothermic reaction and hence, as a result, the agglomeration could occur.
It had been shown by earlier researchers that the agglomeration can occur between 750 and 950°C via the second mechanism, the extended sulfation process [12]. The temperature for optimum sulfur capture is about 850°C [13]. The issue to be understood is whether there exists an optimum temperature range for extended sulfation (long term) [14]. Sulfation appears to be the dominant agglomeration mechanism in systems that use high sulfur fuel with calcium-based sorbents for low ash fuels like pet-coke [15]. The deposits are shown to be composed predominantly of CaSO4 and in some cases almost pure CaSO4 [16, 17]. Low temperature (down to 750°C), agglomeration mechanism may be via carbonation and then sulfation [18].
Herein the fuel used is lignite having ash content ranging from 15 to 35% and the gas temperature range where the deposits occurred is from to .
In CFBC, sulfation is followed by carbonation of CaO and these reactions can be represented as follows [11]: Carbonation mechanism dominates between temperature range of 650 and 790°C at typical CO2 partial pressures (15 kPa) in a CFB boiler, which is much faster than sulfation and is then followed by sulfation of the deposit.
A third possible mechanism thought to cause agglomeration is hydration followed by carbonation [12]. This type of fouling is not common in FBCs because they are normally operated at temperatures well above at which Ca(OH)2 is stable under atmospheric conditions (≤450°C). The hydration reaction may be represented by the following equation: This must be followed by carbonation at temperatures below 450°C via the following reaction: Traditional fouling mechanism due to presence of elements that are associated with ash softening or melting, in particular K, Na, and V, is not applicable for the fuels studied due to low levels of Na, K, and V present [19].
4.3. Detailed Analysis of Ash Forming Matter in the Giral Lignite
Giral lignite has high ash content, 15 to 35% (Table 1), which makes it unique, with respect to quantum of ash and the rate at which it was deposited at the backpass. The principal ash forming elements that play significant role in the fireside problems of the boiler, as indicated by mineralogy of the lignite (determined by XRD), are aluminum silicate (kaolinite minerals) and iron compounds (pyrite, FeS2).
With no limestone addition, the flue gas was estimated to contain around 6,900 ppm SO2 (with 6.1% sulfur in fuel and 3% O2 in flue gases). With 12 t/h limestone addition, the corresponding emissions measured were 1400 ppm SO2. The tests were conducted at site to study reactions of lime particles in flue gas to understand the formation of deposits containing various calcium compounds. The boiler load was varied by increasing the lignite feed and corresponding increase in the limestone to control the SOx level. The very fine limestone particles were calcined, and less than 50-micron level escaped out of the cyclone to backpass and settled over the superheater and reheater coils. As seen in Table 4 chemical composition analysis indicates that adding limestone changes the whole chemistry of the deposits mainly from silicon-aluminum-iron-based deposits (samples 1 to 3) to calcium-based deposit (samples 4 to 6). The calcium compounds present are mainly CaO, CaCO3, and CaSO4 as seen in XRD (Table 5).
The root cause of the fouling problem is carbonation and then sulfation reactions of the limestone particles. Loose limestone particles deposit sinter on surfaces and form hard deposits, particularly in flue gas temperature range around 500–700°C. As explained earlier, it can be safely concluded, at Rajasthan-Giral, that recarbonation reaction is dominant in range of and the extended sulfation reaction (dominant in range of ) leads to hardened deposits.
Ash formed, due to combustion of high sulfur lignite, does not form (sticky or sintering) deposits without limestone addition. These hard deposits were formed due to fine calcined limestone particles (<50 μm) that leave the cyclone. These particles settle on the superheater surfaces and react with CO2 between 650 and 750°C leading to recarbonation and then with SO2 between 750 and 850°C furthering extended sulfation, forming sintered and hard deposits (Figure 6). The hypothesis is that in CFBC, carbonation takes place as a dominant reaction forming calcium carbonate (at temperature range of 650 to 790°C) and then extended sulfation takes place between 750°C and 850°C. The environment of flue gas and exothermic reactions contributes to the conversion of the deposits already formed as calcium carbonate into calcium sulfate. The particles settle as deposits on the tube surface, continue their reaction journey and form as calcium sulfate.
4.4. Optical Microscopy
Optical microscopy of the deposit samples shows a layered structure (Figure 7) defined mainly by mineralogical variation, principally in anhydrite (CaSO4) and iron oxides. Giral ashes are unusual in the occurrence of complete sulfation of the decarbonated limestone with no evidence of either the occurrence of intermediate phases such as calcium oxide or the presence of sulfate reaction rims (Figure 5) on decarbonated limestone [16, 17]. Reason for this unusual behavior is the high sulfur content of the Giral lignite which might have resulted in complete sulfation of the limestone. Additional factor is the greater proportion of fine particles in the milled Giral limestone which would react completely [6]. This observation is supported by the occurrence of fine anhydrite particles in the Giral backpass sample and a subsequent increase in grain size in the back end of the boiler, suggesting that winnowing of the fine particles has occurred in the hotter sections of the backpass.
(a)
(b)
5. Field Trials/Modifications and Improvement Carried out
5.1. Standpipe Blockage
The chemical compositions of the lignite (Table 1), cyclone ash (Table 2), and limestone (Table 3) were analyzed. During commissioning, cyclone standpipe choking due to clinkers (Figure 8) with low combustor temperature of less than 750°C was noticed. The analysis reveals that the composition does not vary much and contains mostly calcium oxide (CaO). The phenomenon of recarbonation of calcined limestone (CaO + CO2 → CaCO3) unreacted with sulphur dioxide was suspected, as a root cause for loose bonding of material at cyclone standpipe leading to blockage of cyclone [20]. This is reflected in the cyclone ash analysis by the presence of free lime (Table 2). The following steps were taken: (a) limestone feed size was checked with more sampling; (b) excessive limestone feed rate was reduced; (c) the operation procedure was revised to maintain higher combustor temperature before starting limestone addition; and (d) automatic pincing air arrangements at junction of the cyclone and standpipe to disturb the agglomeration were incorporated.
(a)
(b)
After incorporation of changes in operation procedure and with pincing air arrangements, the issue was resolved. The timing of pincing was reduced by maintaining temperature above regime of recarbonation at the cyclone standpipe. Figure 9 shows specific recommendations for avoiding, recarbonation-prone regime for limestone addition [20]. The curve denotes the limit of equilibrium of calcium compounds. As shown in the equilibrium diagram (Figure 8), CaCO3 is stable on the left side of the line, whereas CaO is stable on the right side. In the field, CaO was found abundant because of excess limestone added to the furnace. When the temperature was reduced to recarbonation range, sticky carbonate causing agglomeration blocked (Figure 8) the cyclone standpipe.
5.2. High Pressure Soot Blowing
High pressure soot blowing was introduced in the final superheater (FSH), and reheater (RH) and in low temperature superheater (LTSH). After increase in soot blowing pressure from 10 to 20 kg/cm2g, deposits were completely eliminated. Deposits could be removed easily, nearer to the soot blower location, and deposits located away from lance, accumulated proportional to distance from soot blower. Because continuous soot blowing was needed to keep the boiler surfaces clean, additional soot blowers were introduced at selected locations as shown in Figure 1, and deposits were eliminated completely (Figure 1).
5.3. Limestone Size Distribution
Lignite without limestone addition caused little or no hard deposit buildup in the backpass of CFB boiler. The severity of the fouling (hard deposits) was clearly dependent on the amount of limestone addition. Deposits contained very small fines of less than 50 μm size fractions. It was found that 30–40% of the feed limestone was smaller than 50 μm (Figure 4). Both dry and wet sieving tests indicated fine fractions were higher than envisaged during design. (0 to 5% less than 50 μm). Excess quantity of fines <50 μm generated in the milling process was removed by providing a separate elimination line (Figure 10). In addition, the deashing arrangement was improved by introduction of fluidizing pad at the discharge end and increase in diameter of discharge chute. A screen is provided inside hopper, close to the outlet chute, to separate ash particles below 6 mm into the ash evacuation system (Figure 11).
6. Conclusions
Sorbent limestone is used widely in CFB boilers effectively to control sulfur dioxide emissions. Hard deposits were formed in backpass of CFB boiler while using high sulfur Indian lignite and limestone sorbent to control SO2. In addition large quantum of loose deposits caused severe blocking of the second pass. Unreacted calcium oxides that settled on heat transfer tubes at temperature between 650°C and 750°C were subjected to recarbonation and further extended sulfation which resulted in the hard deposits. Elimination of fines, less than 50 μm, in feed limestone could effectively reduce the hard deposits formation in backpass of CFB boiler. This confirms the finding of the previous studies carried out at other institutions firing high sulfur but low ash fuels. Rate of buildup of deposit and chemistry of deposits in backpass of CFB boiler were studied using special foul probes. The rate of buildup of deposit was proportional to the increase in ash content of lignite and sorbent feed rate. Solution to control the fouling in 125 MWe CFB boiler is to minimize the amount of free lime particles (CaO) in the system formed due to excess addition of fines in feed limestone (less than 50 μm). The fine fractions of limestone feed <50 μm coming out of milling circuit were removed by providing an elimination line.
Other CFB boiler operational issues faced, namely, cyclone standpipe blockage, cleaning the heat transfer surfaces deposited with huge quantum of loose ash, and ash evacuation to separate the large size deposits/particles, were effectively resolved through introduction of pincing air at the junction of cyclone and standpipe, high pressure (20 kg/cm2g) soot blowing in selected locations, and incorporation of fluidizing pads and screens in ash hoppers, respectively.
Frequent soot blowing and provision of soot blowers at additional locations were effective in clearing the huge quantum of loose deposits.
Abbreviations
| AFT: | Ash fusion temperature |
| ASTM: | American Society for Testing Materials |
| Al2O3: | Aluminum oxide |
| CaCO3: | Calcium carbonate |
| CaO: | Calcium oxide |
| CaSO4: | Calcium sulfate |
| CFBC: | Circulating fluidized bed combustion |
| GDP: | Gross domestic product |
| LTSH: | Low temperature superheater |
| LRSB: | Long retract soot blower |
| MWe: | Mega Watt electrical |
| RH: | Reheater |
| SH: | Superheater |
| SiO2: | Silicon dioxide |
| SO2: | Sulfur dioxide |
| TGA: | Thermogravimetric analysis |
| XRD: | X-ray diffraction. |
Acknowledgment
The authors thank the Management of BHEL for the opportunity to present their views through this paper on this important topic. The views expressed in this paper are those of the authors and not necessarily those of BHEL.