Environ. Sci. Technol. 2008, 42, 1705–1710
Characteristics and Reactivity of Rapidly Hydrated Sorbent for Semidry Flue Gas Desulfurization JIE ZHANG, CHANGFU YOU,* SUWEI ZHAO, CHANGHE CHEN, AND HAIYING QI Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing 100084, China
Received September 2, 2007. Revised manuscript received October 12, 2007. Accepted December 27, 2007.
Semidry flue gas desulfurization with a rapidly hydrated sorbent was studied in a pilot-scale circulating fluidized bed (CFB) experimental facility. The desulfurization efficiency was measured for various operating parameters, including the sorbent recirculation rate and the water spray method. The experimental results show that the desulfurization efficiencies of the rapidly hydrated sorbent were 1.5–3.0 times higher than a commonly used industrial sorbent for calcium to sulfur molar ratios from 1.2 to 3.0, mainly due to the higher specific surface area and pore volume. The Ca(OH)2 content in the cyclone separator ash was about 2.9% for the rapidly hydrated sorbent and was about 0.1% for the commonly used industrial sorbent, due to the different adhesion between the fine Ca(OH)2 particles and the fly ash particles, and the low cyclone separation efficiency for the fine Ca(OH)2 particles that fell off the sorbent particles. Therefore the actual recirculation rates of the active sorbent with Ca(OH)2 particles were higher for the rapidly hydrated sorbent, which also contributed to the higher desulfurization efficiency. The high fly ash content in the rapidly hydrated sorbent resulted in good operating stability. The desulfurization efficiency with upstream water spray was 10–15% higher than that with downstream water spray.
1. Introduction SO2emissions from pulverized coal-fired power plants have caused significant environmental and human health effects. Various flue gas desulfurization (FGD) technologies have been developed to remove SO2 using Ca-based sorbents. The wet FGD process is the most widely used due to its high desulfurization efficiency and its wide applicability to various coal and unit types. However, the wet process needs large amounts of water, large facilities, and large investments. The semidry FGD process has the advantages of low water consumption, small space requirements, and low capital costs, which makes this SO2 removal process very attractive in old units which require desulfurization retrofits (1). However, the process has several problems, such as low desulfurization efficiency and poor operating stability, in engineering applications mainly because the calcium conversion ratio (ηCa) is relatively low. Many researchers have tried to develop effective methods to improve ηCa for the semidry FGD process (2–13). Most * Corresponding author tel/fax: +86-10-62781740; e-mail: youcf@ tsinghua.edu.cn. 10.1021/es702208e CCC: $40.75
Published on Web 02/02/2008
2008 American Chemical Society
studies have been conducted in thermogravimetric analysis (TGA) reactors and small fixed-bed reactors using a hydration mixture of lime and fly ash to enhance the sorbent activity (3–11). Sorbent prepared by mixing CaO, CaSO4, and fly ash was found to be very active using hydration at 77 °C for 8 h and drying at 87 °C for a day (4). The high ηCa mainly resulted from the highly active Ca-Si materials formed during the long hydration time and long drying time. However, this preparation method is difficult to apply in actual SO2 removal systems due to the difficult preparation conditions, including high hydration temperatures, long hydration times, and long drying times, which complicate scale-up of the sorbent preparation system in engineering applications. Therefore, improved preparation techniques are needed for commercial applications of high activity Ca(OH)2/fly ash sorbent in real semidry desulfurization systems. A low-cost preparation process for rapidly hydrated sorbents has been demonstrated in pilot-scale tests (2). The process hydrates the lime and the fly ash at ambient temperature for 2 h with drying at 150 °C for 1 h, a greatly reduced hydration temperature and preparation time (2, 12). TGA results showed that the ηCa of the rapidly hydrated sorbent was ten times higher than that of the original lime (12). The improved ηCa for the sorbent was mainly due to the greatly improved particle characteristics including the particle specific surface area and the pore volume. The highly active hydration products were not produced in significant quantities at ambient hydration temperatures (2). CFB reactors not only provide long sorbent residence times but also provide intense gas–solid interactions, which are the optimum reactor conditions for Ca(OH)2/fly ash sorbents to realize high ηCa in real FGD engineering applications (13). The fluidized bed combustion (FBC) residues, mainly composed of the bottom ash and the spent calcium sorbent, were also a kind of Ca(OH)2/fly ash sorbent by water hydration for in situ sulfur removal (14). With hydrationinduced reactivation, the ultimate ηCa increased greatly from about 25% to 80% in a bubbling laboratory-scale reactor at 850 °C (15). In the bubbling reactor, the reactivated FBC residues were characterized by a rather large elutriation rate due to the attrition and fragmentation of sorbent particles (16). Therefore, the effect of the large elutriation rate on the ηCa realization for Ca(OH)2/fly ash sorbent in CFB reactors was important for engineering application. However, few studies have been reported about Ca(OH)2/fly ash sorbent in CFB reactors for semidry FGD processes. The gas–solid flow and reaction conditions in CFB reactors are very different from those in TGA or fixed-bed reactors. The engineering applicability of Ca(OH)2/fly ash sorbents in real semidry FGD processes must be investigated to determine whether the high ηCa measured in TGA or fixed-bed reactors can be realized in CFB reactors. In this investigation, the semidry FGD process at a low temperature of 130 °C was studied in a pilot-scale CFB experimental facility with a rapidly hydrated sorbent. The desulfurization efficiency was measured for various operating parameters, including the sorbent recirculation rate and the water spray method.
2. Experimental Section The pilot-scale CFB reactor system is shown in Figure 1. The main subsystems were the sorbent preparation system, the flue gas generation system and the CFB reactor. The sorbent preparation subsystem included a hydration mixer, a vacuum filter, and an infrared dryer. The flue gas generation VOL. 42, NO. 5, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1705
FIGURE 1. Pilot-scale CFB reactor experimental system diagram. subsystem included a fan, an oil burner, an SO2 mixing chamber, and an air cooler. The CFB reactor included the main bed, a distributor, a water sprayer, a cyclone separator, a bed material circulating facility, a bed material feeder and drain, a bag filter, and a compressor. Flue gas generated by the oil burner was cooled with the air cooler to produce 130 °C simulated flue gas. SO2 was added to the flue gas before the air cooler. The CFB reactor riser was 6 m high with a 0.305 m diameter and a flue gas flow rate of 300 Nm3/h. The flue gas passed through the CFB reactor and reacted with the sorbent, then through the cyclone separator and bag filter before being emitted from the stack. The sorbent particles collected in the cyclone separator and bag filter were fed back into the reactor for recirculation or drained out of the system. The water sprayer was located on the CFB reactor centerline 2.5 m above the distributor, which was 1.0 m higher than the sorbent feeder and 1.3 m higher than the sorbent recirculating inlet. The water sprayer used a Y-type two-fluid nozzle. The droplet spray characteristics were selected according to the spray characteristics of two-fluid nozzles commonly used in semidry FGD reactors (17, 18). The average droplet diameter was about 75 µm and the spray angle was about 20°. The initial droplet injecting velocity was about 100 m/s. The droplet velocity on the axial centerline was reduced to about 23 m/s at a distance of 100 mm from the nozzle exit and about 9 m/s at 300 mm distance, measured by a phase Doppler particle analyzer (PDPA). The rapidly hydrated sorbent was made from lumps of lime and coal fly ash at selected mass ratios, using hydration at ambient temperature for
