Fouling in a 160 MWe FBC boiler firing coal and petroleum coke

Abstract

The 160 MWe fluidized bed combustor (FBC) boiler owned and operated by the Tennessee Valley Authority (TVA) has recently been co-fired with coal and petroleum coke (up to 50%). However, it has suffered some fouling problems. On examination of the deposits it became clear that, in only a few cases could the fouling be partially attributed to alkali metals, and even in those cases the primarily limestone-derived materials were almost quantitatively sulphated to a level which was sufficient to cause strength development by itself. In other cases, it appeared that the fouling mechanism was carbonation of the free lime component of the deposit followed by sulphation. Finally, in a few deposits which were less sulphated than bed materials and fly ash, strength development appeared to have occurred by conversion of the free lime in the deposits to Ca(OH)₂, followed by carbonation. This type of agglomeration has not been reported previously in a FBC.

Introduction

Petroleum coke, an inexpensive by-product of the petroleum refining process, has historically been mixed with coal to provide a lower-cost utility fuel [1]. Due to its high carbon content, coke has a higher calorific value than coal and, when used in combination with the latter, generally improves the overall process efficiency. Coke also has the added benefit of becoming increasingly available on the energy market in North America [2]. However, its high sulphur content presents environmental problems, in terms of sulphur dioxide (SO₂) production. Hence, coke must be combusted in such a way as to minimize SO₂ production. One such method is by employing a fluidized bed combustor (FBC) and adding limestone to the bed material to exploit its inherent ability to capture SO₂ in-situ.

The Tennessee Valley Authority (TVA) 160 MWe FBC boiler has been described in detail elsewhere and will only be described briefly here [3]. It is the largest bubbling FBC boiler in North America. It is designed to operate at a temperature of 790–820°C, a recycle ratio of 2.0 and a flue gas O₂ content of 3.3% or higher. Typically 90% sulphur capture can be achieved with a Ca/S molar ratio of 2.4 providing that the boiler is operated at or near optimum conditions. The unit is 21.3 m by 11 m at the base and rises via a tapered section to an exit with dimensions of 7.5 m by 11 m, at an elevation of 31.3 m above the base. The capacity for the coal and limestone feed systems are 90 and 45 t/h, respectively. However, the actual feed rates will depend on the coal characteristics and the boiler load. At full load conditions a typical coal feed rate is about 60 t/h and the limestone feed is usually between 30 and 50% of that figure, depending on the sulphur content of the fuel. The limestone is fed at a nominal 3mm×0 size range.

TVA experimented with the use of a 50:50 coke–coal blend in their 160 MWe bubbling fluidized bed combustor (BFBC). Aside from start-up difficulties [4], [5], they also discovered significant fouling problems in several regions of the boiler. Initially it was supposed that the fouling was due to vanadium, which can be present at relatively high levels in petroleum coke ash, and/or to the alkali metal content of the coals used for co-firing. However, it appeared that fouling occurred over a wide temperature range, and consequently it was felt there was, almost certainly, a number of fouling mechanisms. Previous work has also shown that in a limestone-derived bed, vanadium was chemically combined in the form of high melting point calcium vanadates [2].

In a previous study, fouling in petroleum coke fired FBC boilers was attributed to reactions between the free lime component of the deposit (CaO) and SO₂ and CO₂ in the system [2]. At atmospheric pressure the sulphation mechanism can be described by the following overall reactions:

$$\text{CaCO}{}_3\text{(s)}\text{→}\text{CaO(s)+CO}{}_2\text{(g)}$$

$$\text{CaO(s)+SO}{}_2\text{(g)+}\text{1}\text{2}\text{O}{}_2\text{(g)}\text{→}\text{CaSO}{}_4\text{(s)}$$

There is no current agreement on the mechanism for sulphation, and a detailed discussion of the mechanism(s) of sulphation is beyond the scope of this paper and the interested reader is referred elsewhere [6]. What is known is that at FBC temperatures a limestone particle first calcines to form a porous particle which then reacts with SO₂ to form an outer shell of CaSO₄ leaving an unreacted CaO core. The reaction rate is almost first order in SO₂ [7], [8], [9], but for conversions to CaSO₄ of over 20–30%, i.e. for reaction of SO₂ through dense CaSO₄ layers, the order falls, and can adopt values in the range of 0.2–0.25 [10]. This suggests a change of mechanism and one possible route, for higher degrees of conversion, such as are observed here, is by solid state diffusion of ionic species, such as S_xO_y⁻, through the CaSO₄ layer that forms around the limestone-derived particles [11]. However, at this point the mechanism and details of the reaction of SO₂ with CaO for high degrees of conversion (>30–40%+) are not well understood.

Parameters thought to affect the bulk sulphation levels are: in-bed retention time, temperature and particle size [12], [13], [14]. For the sulphation levels noted in this work, the SO₂ concentration itself can only be expected to have a second order effect, given that the reaction order is in the range of 0.2–0.25, and this appears to have been confirmed recently [14]. Experimental results also show that if limestone or bed particles are retained long enough in a FBC, essentially all the free lime content will convert to CaSO₄ [12]. Hence, the longer the limestone is subjected to sulphating conditions, the higher the resulting conversion to calcium sulphate. Lastly, it has been proven that an optimal size range exists at which sulphation-associated agglomeration is manifested to the greatest degree, and particles smaller or larger than the optimum do not agglomerate or agglomerate more weakly [12], [14]. However, it is clear that in the absence of the sulphation process or some other chemical change typical of petroleum coke fired boilers, alkali metal salts do not act by themselves to produce agglomeration in petroleum coke fired boilers [13]. Rather they enhance or add to the principal agglomeration mechanism. Furthermore, as opposed to conventional agglomeration, sulphation-associated agglomeration occurs in situations where the fuel contains little or no ash and has a high sulphur content, thus necessitating high limestone use. Therefore, it is a phenomenon of the limestone in the bed, rather than the fuel ash itself [12], [13], [14].

Strength development is thought to occur by means of a process termed “molecular cramming” [12], [13], [14]. In the present context, i.e. for reactants in the form of porous particles, confined within a given volume, “cramming” is understood to be the effect of filling all available space, i.e. pores inside the particles and interparticle spaces, because of the increase in molar volume of the reaction products. If the only available space is that within the pores of a single calcined limestone particle, this limits the maximum degree of Ca utilization realizable under FBC conditions to about 60–75%. If sulphated particles are in contact and expansion into interparticle spaces takes place, as is the case during prolonged sulphation, cramming can increase the number and dimensions of contact between particles eventually leading to chemical hardening. Another process which probably takes place as the highly sulphated shell “peels off” is chemical reaction sintering whereby CaSO₄ bridges form between exposed CaO surfaces. It should be noted that strength development by sulphation has been demonstrated for over ten limestones and several bed materials (including the TVA bed ash) by sulphating them for up to 100 days in a temperature controlled oven [12], [14].

The second mechanism that may contribute to the fouling process is recarbonation followed by sulphation. Carbonation may be described by the following reaction:

$$\text{CaO(s)+CO}{}_2\text{(g)}\text{→}\text{CaCO}{}_3\text{(s)}$$

Literature on the carbonation–sulphation mechanism indicates that, in the temperature range of 650–780°C, carbonation, which is known to be a much faster process than sulphation, most likely initiates strength development in deposits, but the carbonated deposit then sulphates [14]. Hence, the combined effect of carbonation and sulphation seems most likely to cause strength development in this temperature regime.

Finally, the third mechanism is hydration followed by carbonation. The hydration reaction may be expressed as follows:

$$\text{CaO(s)+H}{}_2\text{O(l,g)}\text{↔}\text{Ca(OH)}{}_2\text{(s)}$$

$$\text{Ca(OH)}{}_2\text{(s)+CO}{}_2\text{(g)}\text{→}\text{CaCO}{}_3\text{(s)+H}{}_2\text{O(g)}$$

The Ca(OH)₂ also has a larger molar volume of calcium hydroxide (34 cm³/mol) compared to calcium oxide (17 cm³/mol), and as reaction (4) proceeds to the right, the particle will begin to expand and eventually crack, allowing it to react more rapidly with CO₂, via reaction (5), providing appropriate conditions exist [15]. The main factors influencing the rate and extent of the hydration reaction (using steam), according to Marquis [16], are temperature and partial pressure of water vapour. The rate of hydration decreases with rising temperature. The hydration–carbonation process must occur at temperatures at or below 420–450°C since the partial pressure of H₂O above Ca(OH)₂ ranges from 5 to 12 kPa over this temperature range [17].