Update of Emissions Control for Circulating FluidizedBed Combustion Power Plants

1010231

Effective December 6, 2006, this report has been made publicly available in accordance with Section 734.3(b)(3) and published in accordance with Section 734.7 of the U.S. Export Administration Regulations. As a result of this publication, this report is subject to only copyright protection and does not require any license agreement from EPRI. This notice supersedes the export control restrictions and any proprietary licensed material notices embedded in the document prior to publication.

Update of Emissions Control for Circulating Fluidized-Bed Combustion Power Plants

1010231 Technical Update, December 2005

EPRI Project Manager J.M. Wheeldon

ELECTRIC POWER RESEARCH INSTITUTE 3420 Hillview Avenue, Palo Alto, California 94304-1395 PO Box 10412, Palo Alto, California 94303-0813 USA 800.313.3774 650.855.2121 askepri@epri.com www.epri.com

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES

THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT. ORGANIZATION(S) THAT PREPARED THIS DOCUMENT Electric Power Research Institute (EPRI)

This is an EPRI Technical Update report. A Technical Update report is intended as an informal report of continuing research, a meeting, or a topical study. It is not a final EPRI technical report.

NOTE
For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or e-mail askepri@epri.com. Electric Power Research Institute and EPRI are registered service marks of the Electric Power Research Institute, Inc. Copyright 2005 Electric Power Research Institute, Inc. All rights reserved.

CITATIONS
This document was prepared by Electric Power Research Institute (EPRI) 3420 Hillview Avenue Palo Alto, CA 94304 Principal Investigator J.M. Wheeldon This document describes research sponsored by the Electric Power Research Institute (EPRI). This publication is a corporate document that should be cited in the literature in the following manner: Update of Emissions Control for Circulating Fluidized-Bed Combustion Power Plants. EPRI, Palo Alto, CA 2005 1010231.

iii

ABSTRACT
This short report reviews the environmental performance of three state-of-the-art circulating fluidized-bed (CFB) combustors that have recently entered commercial service. These units (on a net basis) are Jacksonville Energy Authoritys (JEA) 2 x 265-MW, Reliant Energys 2 x 260MW, and East Kentucky Power Cooperatives 268 MW. These units are operating with high availability and in full environmental compliance. The environmental performance is compared to CURC/DOE/EPRI Roadmap Goals for 2010 and 2020. These targets were proposed as a means of developing a time line for the progression to near-zero emissions from coal-based power plants. As one of the JEA units was funded as part of the Clean Coal Technology program, there are a lot more data available than from the others. The Roadmap capture efficiency goals for 2020 for sulfur and mercury are 99 and 95 percent, respectively, and the JEA plant achieved these targets in 2004. It also achieved the 2020 PM10 dust emission target of 0.00086 kg/GJ (0.002 lb/MBtu) on some occasions, but performance improvements are required to enable the target to be met consistently. Although NOX emission were low, at or below 0.034 kg/GJ (0.08 lb/MBtu), the 2020 Roadmap goal is 0.0043 kg/GJ (0.001 lb/MBtu) so marked improvements are required. The report discusses a number of approaches being investigated to improve further the environmental performance of coal-based CFB power plants. Some additional approaches are identified that are achieving good success on pulverized coal units and may be applied successfully to CFB plants. Also discussed are ways in which CFB technology is being developed to facilitate a reduction in CO2 emissions. These include deployment of supercritical CFB plants, using oxygen in place of air for combustion, and concepts employing chemical looping.

ACRONYMS AND ABBREVIATIONS

ACF CCPC CCT CFB CURC DOE EKPC EPA ESP FBC FD FF FGD ICR ID IGCC JEA LNB PC SCR SDA SEA SNCR USC actual cubic feet Canadian Clean Power Coalition Clean Coal Technology circulating fluidized bed Coal Utilization Research Council US Department of Energy East Kentucky Power Cooperative US Environmental Protection Agency electrostatic precipitator fluidized-bed combustion forced draft fabric filters flue gas desulfurization Information Collection Request induced draft integrated gasification combined cycle Jacksonville Energy Authority low-NOX burners pulverized coal selective catalytic reduction spray dryer absorber sorbent-enhancement additive selective non-catalytic reduction ultra-supercritical

CONTENTS
1 INTRODUCTION ....................................................................................................................1-1 Study Objective ....................................................................................................................1-2 Report Organization .............................................................................................................1-2 2 ENVIRONMENTAL PERFORMANCE OF CURRENT CFB DESIGNS .................................2-1 JEA CFB Demonstration Project....................................................................................2-1 EKPCs E. A. Gilbert Unit 3 ............................................................................................2-8 Reliant Energys Seward Units 1 and 2..........................................................................2-9 EPAs ICR Data on Mercury Capture in FBC Units........................................................2-9 3 POSSIBLE MEANS OF IMPROVING ENVIRONMENTAL PERFORMANCE OF CFBS ......3-1 Sulfur Capture ......................................................................................................................3-1 NOX Emissions .....................................................................................................................3-4 Dust Emissions ..............................................................................................................3-5 Mercury Capture ............................................................................................................3-5 4 CARBON DIOXIDE CAPTURE PROCESSES INVOLVING CFB TECHNOLOGY ...............4-1 Oxygen-Fired Combustion ...................................................................................................4-1 Oxy-Combustion in a PC boiler......................................................................................4-1 Oxy-Combustion in a CFB Boiler ...................................................................................4-3 Chemical Looping ..........................................................................................................4-7 Generating Efficiency and CO2 Reduction ...........................................................................4-9

vii

1
INTRODUCTION
Coal-based power plants are expected to advance towards near-zero emission designs in response to increasingly stringent environmental permit limits. The rate of progress towards near-zero emission designs is indicated by the DOE/CURC/EPRI Roadmap Goals that establish a time line to achieve defined environmental and economic targets. The environmental targets are tabulated below.
2010 SO2 removal efficiency, % NOX emissions, kg/kJ (lb/MBtu) PM10, kg/kJ (lb/MBtu) Mercury removal efficiency, % By-product utilization, %
(1)

2020 > 99 <0.0043 (< 0.01) 0.00086 (0.002) 95 Near to 100

99 0.021 (0.05) 0.0021 (0.005) 90 50

(1)

Technically achievable value.

The values present the goals as specified in 2001: they are currently being updated. At present it is unclear whether 99-percent SO2 capture will be required with low-sulfur coal or just with highsulfur coals. In preparing the goals it was recognized that there was a need to conserve fresh water supplies by progressing towards cooling systems with near-zero water usage. However, further study was considered necessary before specifying targets. The need to reduce CO2 emissions was also recognized and although a frame work for future development was established, defining performance targets was deferred. An example of more stringent requirements is presented by two PC plants that received their permits only a few years apart. Both plants are designed to fire low-sulfur sub-bituminous coal.
Council Bluffs Unit 4 Location Permit date Plant output, MW NOX controls NOX emission limits, kg/kJ (lb/MBtu) SO2 controls SO2 emission limits, kg/kJ (lb/MBtu) SO2 removal efficiency, % PM10 limits, kg/kJ (lb/MBtu)
(1)

J. K. Spruce Unit 2 San Antonio, Texas Expected in 2006 750 (subcritical) LNB + SCR (0.05) FF + wet FGD (0.06) 95 (1)

Council Bluffs, Iowa 2002 790 (supercritical) LNB + SCR (0.07) ESP + SDA + FF (0.10) 92 (1) (0.025) Estimated by current study

(0.015)

1-1

An interesting feature of the J. K. Spruce design is the switch from SDA to wet FGD. This is considered by many to be necessary to meet low SO2 emission limits even for low-sulfur coals. Both designs require SCR units to meet NOX emission limits as even with sub-bituminous coal low-NOX burners have not achieved emissions below 0.043 kg/GJ (0.10 lb/MBtu). Study Objective This short report will review the environmental performance being achieved by CFB units currently coming into service. The values will be compared to those proposed by the Roadmap Goals and the prospects of improving CFB performance still further will be discussed. Report Organization Chapter 2 presents the environmental performance of three state-of-the-art CFB power plants that have recently entered commercial service. Chapter 3 discusses how the environmental performance of future CFB power plants might be improved. Chapter 4 discusses how CFB technology might be used to address CO2 reduction requirements.

1-2

2
ENVIRONMENTAL PERFORMANCE OF CURRENT CFB DESIGNS
Three state-of-the art CFB power plants have recently entered commercial service. This Chapter presents a brief description of the three projects along with published performance data. Each plant has these common features: Limestone sorbent is used for in-bed SO2 capture. Anhydrous ammonia is injected ahead of the cyclones to promote selective non-catalytic reduction (SNCR) of NOX. A back-end polishing scrubber is used to capture additional SO2 from the flue gas allowing +98-percent capture to be achieved overall with reduced sorbent demand. The scrubber also removes SO3, chlorine, fluorine, and heavy metals including mercury. The scrubber designs used are described in Reference 2-1. Baghouses are used for particulate control. Almost all the coal and limestone-based ash is sent to landfill. Very little is used in byproduct applications although opportunities are being sought. In addition, mercury capture data from six fluidized-bed combustors (FBC) collected as part of the Environmental Protection Agencys Information Collection Request (ICR) are discussed. JEA CFB Demonstration Project JEA (Jacksonville Energy Authority) has constructed 2 x 265-MW(net) (2 x 298-MW(gross)) CFB units at its Northside site. Both units are supplied by Foster Wheeler and one (Unit 2) was partly funded under the US-DOEs Clean Coal Technology program. Features of the design include the following: The CFB is designed to feed 100-percent delayed petroleum coke with 6.7-percent sulfur and 100-percent bituminous coal with up to 4.8-percent sulfur. Fuel properties are presented in Table 2-1. The steam conditions are 172.4 bar/538C/538C (2,515 psia/1,000F/1,000F) and the design heat rate is 10,500 kJ/kWh (9,950 Btu/kWh) on an HHV basis. The boiler has three cyclones and each fines recycle return line includes an external heat exchanger that provides part of the final reheat and superheat duty. The polishing scrubber is a spray dryer absorber that uses lime. The baghouse has a spare compartment to allow for on-line maintenance.

2-1

Table 2-1 Properties of Fuels Burned by the JEA CFB Unit Pittsburgh #8 Composition, wt % Carbon Hydrogen Nitrogen Oxygen Sulfur Moisture Ash Chlorine HHV, kJ/kg (Btu/lb) 72.6 4.8 1.4 2.1 4.8 7.3 6.9 0.09 29,960 (12,880) 64.8 4.3 1.3 7.1 3.2 12.1 7.1 0.15 27,100 (11,650) 79.0 3.6 1.0 0.3 6.7 9.0 0.4 32,600 (14,000) Illinois #6 Petroleum coke

The unit is in commercial operation firing an 80:20 blend of petroleum coke and coal. Feeding 100-percent coke resulted in the formation of agglomerates in the external heat exchanger that resulted in poor fluidization. Using the blend reduced the agglomeration problem to manageable proportions. All projects in the CCT program are required to complete a test program investigating performance. Consequently, there are more data available from this project than the other two (see Reference 2-2). The environmental performance data are presented in Tables 2-2 through 25. The variation of emissions with load for the two coals is plotted in Figures 2-1 and 2-2. Fullload emissions performance for the two coals is very similar and although the part-load values are also similar they show different trends. The plant operating conditions for the tests are not reported and differences in these conditions may be responsible for some of the differences seen in the data.
Performance with Pittsburgh #8

Table 2-2 presents data obtained for full-load and three part-load operating conditions Overall SO2 capture is 98.6 percent with the lime fed to the SDA capturing 56.9 percent of the SO2 leaving the furnace with the flue gas. This is close to the 2010 Roadmap Goal of 99percent capture, which could presumably be met by feeding more limestone to the bed and/or more lime to the SDA. Capture efficiency was not measured at the part-load conditions but as the SO2 concentration remained reasonably constant presumably so too did capture efficiency.

2-2

Table 2-2 Environmental Performance of JEA CFB Fired with Pittsburgh #8 Coal Operating point, percent of MCR 100 NOX, (1) CO,
(1)

80 0.034 (0.080) 0.019 (0.044) 1.0

60 0.031 (0.072) 0.051 (0.118) 1.5

40 0.035 (0.082) 0.023 (0.053) 1.4

0.032 (0.074) 0.011 (0.026) 0.0017 (0.0040) 1.1

PM10, (1) Opacity, percent Sulfur capture data SO2 leaving boiler (1) SO2 leaving SDA In-bed capture, % SDA capture, % Combined capture, % In-bed Ca/S ratio Limestone/coal ratio, (1)

0.102 (0.237) 0.044 (0.102) 96.8 56.9 98.6 1.77 0.28

(1).

0.035 (0.082) -

0.035 (0.081) -

0.046 (0.108) -

Emissions in kg/GJ (lb/MBtu)

Table 2-3 Environmental Performance of JEA CFB Fired with a 50:50 Blend of Petroleum Coke and Pittsburgh #8 Coal Operating point, percent of MCR 100 NOX, CO,
(1)

80 0.023 (0.040) 0.010 (0.024) 1.4

60 0.018 (0.043) 0.012 (0.028) 1.1

40 0.014 (0.033) 0.034 (0.080) 0.8

0.030 (0.070) 0.0064 (0.015) 0.0017 (0.0041) 1.0

(1)

PM10, (1) Opacity, percent Sulfur capture data SO2 leaving boiler (1) SO2 leaving SDA In-bed capture, % SDA capture, % Combined capture, % In-bed Ca/S ratio Limestone/coal ratio, (1)

0.0868 (0.202) 0.040 (0.093) 97.5 54.0 98.8 1.70 0.34

(1)

0.034 (0.080) -

0.029 (0.067) -

0.047 (0.109) -

Emissions in kg/GJ (lb/MBtu)

2-3

Table 2-4 Environmental Performance of JEA CFB Fired with a 80:20 Blend of Petroleum Coke and Pittsburgh #8 Coal Operating point, percent of MCR 100 NOX, (1) CO,
(1)

80 0.012 (0.027) 0.0064 (0.015) 1.0

60 0.0077 (0.018) 0.0095 (0.022) 1.0

0.0056 (0.013) 0.00546 (0.0127) 0.0010 (0.0024) 0.7

PM10, (1) Opacity, percent Sulfur capture data SO2 leaving boiler (1) SO2 leaving SDA In-bed capture, % SDA capture, % Combined capture, % In-bed Ca/S ratio Limestone/coal ratio, Mercury capture Leaving boiler (2) Capture, %
(1).
(1)

0.0494 (0.115) 0.025 (0.058) 97.8 49.5 98.9 2.29 0.27

0.023 (0.054) -

0.035 (0.058) -

<0.0317 (<0.0738) 98
(2)

Emissions in kg/GJ (lb/MBtu)

Emissions in kg/PJ (lb/TBtu)

2-4

Table 2-5 Environmental Performance of JEA CFB Fired with Illinois #6 Coal Operating point, percent of MCR 100 NOX, CO, PM10,
(1)

80 0.028 (0.064) 0.013 (0.031) 1.3

60 0.023 (0.053) 0.015 (0.034) 1.6

40 0.034 (0.078) 0.059 (0.138) 1.4

0.037 (0.086) 0.0086 (0.020) 0.00082 (0.0019) 1.5

(1)

(1)

Opacity, percent Sulfur capture data SO2 leaving boiler (1) SO2 leaving SDA In-bed capture, % SDA capture, % Combined capture, % In-bed Ca/S ratio Limestone/coal ratio, Mercury capture Leaving boiler (2) Capture, %
(1). (1)

0.119 (0.276) 0.0460 (0.107) 94.9 61.0 98.0 2.68 0.27

0.035 (0.079) -

0.035 (0.144) -

0.046 (0.108) -

0.148 (0.345) 95

(2)

Emissions in kg/GJ (lb/MBtu)

Emissions in kg/PJ (lb/TBtu)

2-5

0.14

0.12

0.10

Emissions, lb/MBtu

0.08

0.06

0.04 SO2 0.02 NOx CO

0.00 0 20 40 60 80 100 120

Load as percent of MCR

Figure 2-1 Variation in Emissions with Part Load for JEA CFB Fired on Pittsburgh #8
0.16

0.14

0.12

Emissions, lb/MBtu

0.10

0.08

0.06

0.04 SO2 0.02 NOx CO

0.00 0 20 40 60 80 100 120

Load as percent of MCR

Figure 2-2 Variation in Emissions with Part Load for JEA CFB Fired on Illinois #6

At full-load NOX emissions are 0.032 kg/GJ ((0.074 lb/MBtu) and these remain fairly constant as load falls. These levels are above the 2010 Roadmap Goal of 0.02 kg/GJ (0.05 lb/MBtu). To increase SNCR reactions more ammonia could be injected but this would

2-6

increase the ammonia slip, which may then exceed permitted values. It could also result in ammonium sulfate formation producing a sticky, corrosive ash that could give rise to problems in the baghouse. At full-load CO emissions are 0.011 kg/GJ ((0.026 lb/MBtu) and increase as load falls. This is an intrinsic feature of CFB, for as load falls so too does furnace temperature. Emissions of SO2 and NOX from the furnace also increase but overall are kept low by the SDA and SNCR at the back-end, respectively. At full-load PM10 emissions are 0.0017 kg/GJ ((0.0040 lb/MBtu), which is below the 2010 Roadmap Goal but greater than that proposed for 2020. Part-load emissions were not measured but presumably did not increase.

Performance with a 50:50 Blend of Petroleum Coke and Pittsburgh #8 Coal

Table 2-3 presents data obtained for full-load and three part-load operating conditions Overall SO2 capture is 98.8 percent with the lime fed to the SDA capturing 54.0 percent of the SO2 leaving the furnace with the flue gas. At full-load NOX emissions are 0.030 kg/GJ ((0.070 lb/MBtu), which is similar to 100percent Pittsburgh #8. With the blend as load falls emissions fall by around 50 percent whereas for 100-percent Pittsburgh #8 the emissions remained constant,. The part-load levels are below the 2010 Roadmap Goal of 0.02 kg/GJ (0.05 lb/MBtu). At full-load CO emissions are 0.0064 kg/GJ ((0.015 lb/MBtu) and increase as load falls. The levels are approximately half those measured for Pittsburgh #8, which is surprising as petroleum coke is less reactive and might be expected to produce more CO. At full-load PM10 emissions are 0.0017 kg/GJ ((0.0041 lb/MBtu), which is below the 2010 Roadmap Goal and similar to that achieved with Pittsburgh #8.

Performance with an 80:20 Blend of Petroleum Coke and Pittsburgh #8 Coal

Table 2-4 presents data obtained for full-load and two part-load operating conditions Overall SO2 capture is 98.9 percent with the lime fed to the SDA capturing 49.5 percent of the SO2 leaving the furnace with the flue gas. The amount of limestone fed is greater than the previous two Pittsburgh #8 data sets and the SDA is capturing less of the total. At full-load NOX emissions are 0.0056 kg/GJ ((0.013 lb/MBtu), which is almost a factor of six lower than that measured for 100-percent Pittsburgh #8. With the 80:20 blend, as load falls emissions increase whereas for the 50:50 blend emissions decreased. At full-load CO emissions are 0.00546 kg/GJ ((0.0127 lb/MBtu) and increase as load falls. The levels are approximately 20-percent lower than with the 50:50 blend. As CO trends are usually the opposite of NOX trends, these results are also surprising. At full-load PM10 emissions are 0.0010 kg/GJ ((0.0024 lb/MBtu), which is below the 2010 Roadmap Goal and lower than that achieved with Pittsburgh #8. Mercury measurements were made for the full load condition and capture efficiency was determined to be 98 percent. This is greater than that required by the 2020 Roadmap Goals.

2-7

Performance with Illinois #6 Coal

Table 2-5 presents data obtained for full-load and three part-load cooperating conditions. Overall SO2 capture is 98.0 percent with the lime fed to the SDA capturing 61.0 percent of the SO2 leaving the furnace with the flue gas. At full-load NOX emissions are 0.037 kg/GJ ((0.086 lb/MBtu), which is slightly greater than that measured for Pittsburgh #8. As load falls emissions decrease slightly and are slightly lower than for Pittsburgh #8 At full-load CO emissions are 0.0086 kg/GJ ((0.020 lb/MBtu) and increase as load falls. Levels are lower than measured for Pittsburgh #8. At full-load PM10 emissions are 0.00082 kg/GJ ((0.0019 lb/MBtu), which is below the 2020 Roadmap Goal and lower than that achieved with any of the four fuels fed. Mercury measurements were made for the full-load condition and capture efficiency was determined to be 95 percent. This is equal to the value required by the 2020 Roadmap Goals.

EKPCs E. A. Gilbert Unit 3 EKPC (East Kentucky Power Cooperative) has constructed a 268-MW(net) (295-MW(gross)) ALSTOM Power CFB unit at its Spurlock site. Features of the design include the following: The CFB is designed to feed a range of bituminous coals with sulfur contents from 1.50 to 4.5 percent. See Table 2-6. EKPC believe that this flexibility will allow it to negotiate more competitive fuel prices. The unit is also designed to burn petroleum coke, biomass, and up to five-million tires per year. The steam conditions are 173.7 bar/538C/538C (2,535 psia/1,000F/1,000F). The boiler has three cyclones but does not have any external heat exchangers, all heat transfer surface being in the furnace or the convection pass. The polishing scrubber is a Flash Dryer Absorber (FDA) that uses unreacted limestone in the fly ash leaving the furnace with the flue gas. Part of the baghouse catch is reactivated by hydration, which causes it to swell and expose the CaO at the core of the particles. This avoids the use of lime which is 4 to 7 times more expensive than limestone. Commissioning of the unit started in December 2004 and entered commercial service in March 2005. ALSTOM report that at full load the unit has achieved 98 percent SO2 capture with an inbed Ca/S molar ratio of 1.9. This is slightly higher than the in-bed Ca/S molar ratio reported for Pittsburgh #8 at JEA, but the EKPC unit does not require additional lime to be fed to the SDA. NOX emissions were reported at 0.034 kg/GJ (0.08 lb/MBtu) (see Reference 2-3).

2-8

Table 2-6 Properties of Coals used at EKPCs Gilbert CFB Unit Design fuel Composition, wt % Carbon Hydrogen Nitrogen Oxygen Sulfur Moisture Ash HHV, kJ/kg (Btu/lb) 24,190 (10,400) 53.5 4.4 0.9 7.2 4.5 9.5 20.0 23,760 (10,210) 61.7 2.4 1.3 5.0 1.5 8.1 20.0 29,080 (12,500) 70.8 3.9 1.4 4.0 4.0 7.9 8.0 Kentucky Pittsburgh #8

Reliant Energys Seward Units 1 and 2 Reliant energy has constructed 2 x 260-MW(net) (2 x 292-MW(gross)) ALSTOM Power CFB units at its Seward site. The CFB designs are very similar to that of EKPCs Gilbert Plant, the most notable difference being the fuel used. The Reliant plants are designed to handle bituminous gob with up to 50-percent ash, blended to achieve a heating value of 12,800 kJ/kg (5,500 Btu/lb). Locally mined coal may be added to the coal waste to achieve this heating value. Commissioning of the unit started in mid-2004 and entered commercial service in March 2005. ALSTOM report that at full load the unit has achieved 98 percent SO2 capture with an in-bed Ca/S molar ratio of 1.9, and NOX emissions of 0.07 lb/MBtu (see Reference 2-4). EPAs ICR Data on Mercury Capture in FBC Units The mercury content of coal varies by coal type and the location of the mine. The majority of the mercury exists as compounds associated with the organic and sulfur fractions of the coal. When the coal is burned the mercury is vaporized and forms elemental mercury. Through interaction with other flue gas constituents some of the elemental mercury is converted to oxidized mercury, also termed ionic mercury, or condenses onto dust becoming particulate-bound mercury. This mercury transformation process is known as speciation. Although not fully understood, speciation is believed to be affected by the following factors. Properties of the coal and boiler fly ash. Combustion conditions in the furnace. Flue gas temperature and composition. Post-combustion flue gas clean-up technologies used.

2-9

A major mechanism for the formation of oxidized mercury is the reaction of elemental mercury with chlorine. This is supported by the observation that coals with high chlorine contents (more than 0.05 percent) produce more oxidized mercury than coals with low chlorine contents (less than 0.05 percent). In general bituminous coals have higher chlorine contents than subbituminous ones. Iron present in fly ash is also believed to promote the formation of oxidized mercury. Chemical equilibrium suggests that these gas-phase reactions start at around 680C (1,260F) and are essentially complete by 430C (800F). The catalysts used in SCR units also catalyze the reaction and these units normally operate in the range 430C (800F) to 300C (570F) so extending the operating temperature range for the oxidation reactions. Speciation is important as the different forms of mercury require different capture technologies. Oxidized mercury dissolves readily in alkaline mixtures such as the limestone slurries present in wet-FGD units and can also be collected by suitable adsorbents. Elemental mercury does not dissolve and can only be removed by adsorption. Particulate-bound mercury is removed in the particulate control device, but this is usually only a small proportion of the total mercury present. As fluidized-bed combustors do not include wet FGD units, the elemental and oxidized mercury present in the flue gas will both have to be removed by adsorption. CFBs operate at lower temperatures than PC units and as a consequence combustion efficiencies are slightly lower. This results in more unburnt carbon present in the flue gas and promotes intrinsic mercury capture, the carbon being more adsorptive than coal ash. The US EPAs Information Collection Request issued in 1999 included six FBC units. A summary of the data is presented in Table 2-7. Other parameters such as fuel ash and sulfur contents also varied but chlorine content is considered the most significant variable. The capture efficiencies achieved are intrinsic values and no additives or measures were used to enhance capture. Table 2-8 presents the measurements showing the mercury speciation. The data are taken from Reference 2-5. This small amount of data with the variability in equipment design and fuel types is considered insufficient to develop complete understanding of the capture mechanisms. However the following observations are made. The three fuels with the highest total capture efficiency were bituminous coals with the highest chlorine contents. Two of the three lowest total capture efficiencies were with low-rank fuels that are more reactive than bituminous coal, resulting in higher combustion efficiencies and less unburnt carbon in the flue gas. These observations appear to be in keeping with the measurement results from JEA for the Illinois bituminous coal has a high chlorine content of 0.15 percent and 95 percent of the mercury was captured. However the 80:20 petroleum coke with Pittsburgh #8 blend achieved 98 percent capture and the chlorine content will have been only around 0.02 percent. None of the FBC units tested by the EPA included SDA units, and the extent to which the SDA affected the results from JEA is not known.

2-10

Table 2-7 Selected Data from FBC Units Sampled in Support of EPAs ICR (1999) Plant and power output Heskett Station (1) TNP Scrubgrass Barbers Point
(2) (2)

Fuel fired Dakota lignite Texas lignite Bituminous waste Indonesian bituminous Anthracite/bituminous Bituminous/pet. coke

Fuel chlorine, wt. % 0.02 0.017 0.06 0.005 0.023 0.042

Total mercury capture efficiency, % 48 58 99 57 99 98

Kline Township Stockton

(1)

Bubbling bed using ESP for final particulate collection. Others are all circulating beds using baghouses. (2) Includes SNCR for NOX control
Table 2-8 Mercury Speciation Data from FBC Units Sampled in Support of EPAs ICR (1999) Plant Heskett Station TNP Scrubgrass Barbers Point Kline Township Stockton Mercury species, % Particulate 48.2 56.0 99.6 16.4 99.1 100 0 Oxidized 22.6 25.0 0.3 7.0 0.2 0 Elemental 30.2 19.0 0.1 76.6 0.7 Capture, % 48 58 99 57 99 98

From the data available it appears that for bituminous coals with chlorine contents above 0.02 weight percent, a CFB unit will achieve high mercury removal efficiencies intrinsically and so meet the Roadmap Goals. Bituminous coals with lower chlorine contents and low-rank fuels achieve lower intrinsic capture efficiencies and mercury reduction measures will be required to reach the required capture efficiency levels.
Summary of CFB Performance

The range of environmental performance achieved on JEAs CFB unit operating at full-load is presented in Table 2-9 and compared to the Roadmap Goals for 2010 and 2020. As coal is the primary focus of the Roadmap, test data collected when petroleum coke was being fired is not included.

2-11

Table 2-9 Environmental Performance Compared with Roadmap Goals for JEAs CFB Operating at Full Load and Fired with Bituminous Coal Performance range SO2 removal efficiency, % NOX emissions, kg/kJ (lb/MBtu) PM10, kg/kJ (lb/MBtu) Mercury removal efficiency, % By-product utilization, %
(1)

2010 99 0.021 (0.05) > 99

2020

98.0 to 98.6 0.032 to 0.037 (0.074 to 0.086) 0.00082 to 0.0017 (0.0019 to 0.0040) 95 to 98 NA
(1)

<0.0043 (< 0.01)

0.0021 (0.005)

0.00086 (0.002)

90 50

95 Near to 100

Technically achievable value.

Based on the data presented in Table 2-9, the following comments are made: SO2 capture efficiency is close to meeting the Roadmap Goals for 2010 and 2020 and with additional sorbent the goals will do so albeit at increased operating cost. The EKPC and Reliant CFB plants appear to be achieving similar levels of capture performance. The improvement required is achieving this performance with reduced sorbent demand. NOX emissions are around 50 percent higher than the 2010 Roadmap Goal and up to 9 times higher than that for 2020. The EKPC and Reliant CFB plants appear to be achieving similar levels of NOX performance. Improvements in SNCR might allow the 2010 Goal to be reached but a more significant performance improvement is required to meet the 2020 Goal. Dust loadings are met for the 2010 Roadmap Goal. The Illinois #6 meets the 2020 Roadmap Goal but the Pittsburgh #8 does not indicating that improvements are required to be certain of achieving performance requirements with all coals. Mercury capture efficiency is achieved intrinsically for both the 2010 and 2020 Roadmap Goals. This may not be the case for fuels other than bituminous coal, but as will be discussed in Section 3 appropriate control measures are available if required. Most ash is sent to landfill and very little is utilized at present. Suitable applications for which 50 percent could be used on a consistent basis have not been identified. Based on these comments there appears to be a need to improve CFB performance if the technology is to progress towards near-zero emissions operation. Possible ways of improving performance are discussed in Chapter 3.

2-12

References

2-1
2-2

Evaluation of Processes for Post-Combustion Control of Emissions from CFB Power Plants, EPRI Report 1004883, December 2003.
Final Technical Report for the JEA Large-Scale CFB Combustion Demonstration Project DOE Report, June 2005. http://www.netl.doe.gov/cctc/resources/pdfs/jacks/Test%204%20DOE%20Issue%20Rev%20 1.pdf Treff, P. and Johnson, C., Clean Power from Coal: Design and Status of East Kentucky Power Cooperatives E. A. Gilbert Unit, 18th International FBC Conference, Toronto, May 22nd to 25th, 2005. Skowyra, R., ALSTOM CFB Technology, ibid.

2-3

2-4

2-5

An Assessment of Mercury Emissions from U.S. Coal-Fired Power Plants, EPRI Report 1000608, October 2000.

2-13

3
POSSIBLE MEANS OF IMPROVING ENVIRONMENTAL PERFORMANCE OF CFBS
This chapter discusses ways in which the environmental performance of CFBs for currently regulated species, might be improved. Sulfur Capture As discussed in Chapter 2, CFBs are achieving capture efficiencies in line with Roadmap Goals. The improvement required is to achieve the same, or better performance, using less limestone. It appears that 99 percent capture can be achieved with a Ca/S molar ratio of around 1.9. In contrast, a wet FGD unit can achieve similar capture efficiency with a Ca/S molar ratio of less than 1.1. Using the properties of Pittsburgh #8 tabulated in Chapter 2, containing 4.8-percent sulfur, and limestone containing 95-percent CaCO3, the effect of the two Ca/S molar ratios on the rates of limestone feed and ash discharge can be calculated. The values are shown in Table 3-1, which shows that the CFB requires 22,500 kg/hr (49,500 lb/hr) more limestone feed and produces 13,400 kg/hr (29,500 lb/hr) more ash for disposal. For a low-sulfur coal the effect is less pronounced. For the same coal but containing only 0.5percent sulfur the mass flow rates are presented in Table 3-2. For this calculation the Ca/S molar ratio for the CFB is increased to 2.5 to allow for the lower SO2 concentration in the flue gas. All other variables remain the same. For this coal the CFB requires 3,890 kg/hr (8,560 lb/hr) more limestone feed and produces 2,990 kg/hr (6,690 lb/hr) more ash for disposal. These calculations show that fuel sulfur content has a marked impact upon the rates of limestone feed and ash discharge and that for low-sulfur content fuels the ash discharge rate is not increased as markedly as it is for fuels with high-sulfur contents. However, as CFBs tend to be used for opportunity fuels with high-sulfur contents, such as petroleum coke, measures that increase limestone utilization efficiency will be economically advantageous. One means of improving utilization efficiency is being pursued by ALSTOM at their 340-MW Sulcis unit in Sardinia (see Reference 3-1). This includes steam reactivation of a portion of the fines being recycled from the cyclones back into the furnace. The Sulcis unit includes an external heat exchanger and it is understood that a portion of the solids flow is cooled to a specific temperature in a separate compartment before the steam is injected. As of November 2005, the CFB unit was in the early stages of commissioning and the effectiveness of steam reactivation had not been investigated.

3-1

Table 3-1 Flow Rates of Limestone and Ash for 500 MW CFB and PC Plants Fired on Pittsburgh #8 with 4.8Percent Sulfur CFB Heat rate, kJ/kWh (Btu/kWh) Coal feed rate, kg/hr (lb/hr) Ca/S molar ratio, Limestone feed rate, kg/hr (lb/hr) Limestone/coal mass ratio, Coal ash produced, kg/hr (lb/hr) Spent limestone, kg/hr (lb/hr) (1) Total ash discharged, kg/hr (lb/hr)
(1)

PC + FGD 10,400 (9,850) 173,400 (382,400) 1.1 30,100 (66,400) 0.17 12,000 (26,400) 38,100 (84,100) 50,100 (110,500)

10,500 (9,950) 175,200 (386,300) 1.9 52,600 (115,900) 0.30 12,100 (26,700) 51,400 (113,300) 63,500 (140,000)
(1)

Dry basis

Table 3-2 Flow Rates of Limestone and Ash for 500 MW CFB and PC Plants Fired on Pittsburgh #8 with 0.5Percent Sulfur CFB Heat rate, kJ/kWh (Btu/kWh) Coal feed rate, kg/hr (lb/hr) Ca/S molar ratio, Limestone feed rate, kg/hr (lb/hr) Limestone/coal mass ratio, Coal ash produced, kg/hr (lb/hr) Spent limestone, kg/hr (lb/hr)
(1)

PC + FGD 10,400 (9,850) 173,400 (382,400) 1.1 3,010 (6,640) 0.017 12,000 (26,400) 3,810 (8,410) 15,810 (34,810)

10,500 (9,950) 175,200 (386,300) 2.5 6,900 (15,200) 0.039 12,100 (26,700) 6,700 (14,800) 18,800 (41,500)
(1)

Total ash discharged, kg/hr (lb/hr) (1)

Dry basis

Foster Wheeler proposed including steam reactivation stages for the baghouse catch and recycled fines in their CCPI-1 project with Colorado Springs Utilities for a 150-MW CFB (see Reference 3-2). This was estimated to achieve 98-percent sulfur capture with a Ca/S molar ratio of around 1.6. It was not considered cost effective to also reactivate the material from the bed ash cooler. From this estimate, even with steam reactivation of the cyclone and baghouse material, CFB units still require more limestone for a given sulfur capture than does a PC unit. The combination of a SDA with baghouse for a PC plant costs around $180/kW and the combination of ESP + wet FGD costs around $210/kW. These costs vary with local circumstances and coal type being used. There are no comparable cost estimates available for CFB plants, but assuming that the same costs apply, replacing the spray dryer/baghouse used in
3-2

current designs with ESP/wet FGD might only increase the capital cost of the CFB by $30/kW. Other advantages discussed later, including reduced limestone demand, could reduce this differential or even eliminate it. Before proceeding further, this idea is entirely speculative and has not been subjected to detailed analysis. The CFB will still require bed material to circulate and dissipate the heat, but without the need to capture a high proportion of the sulfur the amount of solids could be reduced. A material more resistant to breakdown than limestone could be used and no sulfur captured. This material would be continually circulated with a make up rate far less than the limestone normally fed to the bed. A possible draw back of this approach is material released by the fuel during combustion could coat the particles and alter their characteristics unfavorably. They could become sticky and form agglomerates or increase in size and defluidize in the furnace. Because of possible operating problems with an alternative bed material, it is suggested limestone make up the bed material but that the feed rate is limited to that necessary for good heat transfer performance. Hence, there will still be some in-bed SO2 capture. This material when removed from the furnace could be utilized further in the wet FGD lowering limestone demand further. Following this approach would offer the following advantages: Limiting the sorbent feed rate to the CFB lowers the capital cost of feeders. Bed ash cooling duty is reduced lowering the capital cost of the cooler. Reducing furnace solids loading reduces in-furnace pressure drops and lowers FD and ID fan duty. Using an FGD, possibly followed by a wet ESP, allows the baghouse to be replaced by an ESP. This will lower capital and operating costs and, as it has a lower pressure drop, also ID fan duty. Because of its higher electrical resistivity, an ESP collects limestone-based ash less efficiently than coal ash. This could be addressed by adding collection fields or introducing SO3 ahead of the ESP to lower resistivity. Furnace control strategy could be simplified by removing sulfur retention considerations. Reduced equipment and solids circulation rates will lower maintenance costs. Removing in-situ sulfur capture considerations simplifies the design and may allow for the development of a more standardized CFB unit. Overall limestone demand and the amount of ash sent to landfill are reduced with capital and operating cost savings. If advantages were established for operating the CFB with a reduction of in-situ sulfur capture, there are alternative approaches to removing the SO2 from the flue gas that result in a by-product more saleable than gypsum. This will also contribute to the Roadmap Goal of increasing byproduct utilization. Powerspans ECO process oxidizes the SO2 to SO3, which then reacts with ammonia to form ammonium sulfate that is sold as a fertilizer (see Reference 3-3). The Airborne process reacts the SO2 with sodium bicarbonate which is regenerated with ammonia also forming ammonium sulfate for sale as fertilizer (see Reference 3-4).

3-3

NOX Emissions In the CFB the ammonia for SNCR is introduced ahead of the cyclone to take advantage of its swirling motion to mix the ammonia (or urea) with the flue gas. An alternative approach developed for use in PC units might be equally effective. Mobotec (see Reference 3-5) has developed a modified over-fired air system termed ROFA (rotating opposed-fired air) that creates high turbulence in the furnace. A small proportion of the combustion air is injected into the furnace at high velocity through nozzles placed asymmetrically in the boiler walls to create a rotational component that eliminates dead spots. The enhanced mixing evens out the temperature profile reducing thermal NOX formation, and by contacting the air more effectively with the fuel increases carbon burnout and reduces carbon monoxide. As CFBs operate at lower temperatures, only around 30 percent of the NOX formed is thermal NOX so using ROFA may not result in a large reduction. But as seen from Chapter 2, CFBs do produce appreciable amounts of CO that ROFA may lower. From data presented in Chapter 2, CFBs fired on bituminous coal emit around 0.011 kg/GJ (0.025 lb/MBtu) of CO and this increases as load falls. These emissions are four to five times higher than for a PC unit, which could cause problems in permitting a CFB unit in a non-attainment area. Mobotec report using their ROFA system on a 130-MW CFB burning waste coal but do not discuss any results. A variant of ROFA is Rotamix, which adds ammonia to the injection air. If applied to a CFB this approach could achieve good mixing of the flue gas and ammonia ahead of the usual cyclone entry point, extending the gas residence time and increasing the degree of NOX reduction achieved. Additional NOX reduction can be achieved by adding an SCR unit. This SNCR-SCR combination approach has been used in Japan in the flue gas stream from a coal-fired bubbling PFBC unit and in Sweden on a biomass-fired unit to eliminate ammonia slip. A pulverized coal unit with lowNOX burners fired on bituminous coal will achieve around 0.064 kg/GJ (0.15 lb/MBtu) NOX requiring a 93-percent efficient SCR unit to achieve the 0.004 kg/GJ (0.01 lb/MBtu) target. For a CFB fired with bituminous coal using a standard SNCR arrangement and with 0.03 kg/GJ (0.07 lb/MBtu) of NOX exiting the boiler, the SCR would only have to be 70-percent efficient to achieve the 0.004 kg/GJ (0.01 lb/MBtu) target. This would offer a capital and operating cost to the CFB design. If using Rotamix to improve the effectiveness of SNCR reduced NOX emissions to say 0.02 kg/GJ (0.05 lb/MBtu) or lower, the SCR efficiency required would be even lower with still greater savings. An advantage of using the SCR is that it can be incorporated with a separate catalyst bed to reduce CO emissions. If ROFA is effective in reducing CO this catalyst bed will not be required. The SCR catalyst can also oxidize mercury making it more easily captured by the backend pollution control equipment. In PC applications where the flue gas passing through the SCR contains all the SO2, some is oxidized to SO3. The presence of SO3 increases the visibility of the stack plume, giving it either a blue or brown tint depending upon prevailing weather conditions. However, in the CFB application, as most of the SO2 is removed in the bed there is less in the flue gas and so less potential for the SCR to produce SO3.

3-4

Dust Emissions As discussed in Chapter 2 current baghouse performance comes close to meeting the 2020 Roadmap Goal of PM10 dust emission limits of 0.00086 kg/GJ (0.002 lb/MBtu) but improvements are required to do this consistently for all fuel types. One approach would be to lower the face velocity still further to reduce particle penetration through the bags. All baghouses would include redundant compartments to allow for on-line maintenance in the event of a bag rupture. In a PC unit the main PM2.5 constituent is SO3, but such emissions from CFBs are expected to be lower as the operating temperatures are lower and the sulfur is captured in the bed. Consequently PM2.5 from a CFB may be inherently low although this has not been established by measurement. Nevertheless as environmental controls become tighter it may need to be controlled. Wet ESPs are beginning to be installed on new PC plants to remove PM2.5 material as most of it is not removed by a baghouse. If such a measure where required for CFBs then the flue gas would have to be cooled or if a wet FGD were installed, the wet ESP would be installed in place of the demister. Mercury Capture DOE-NETL is funding an extensive program investigating mercury capture from coal-derived flue gases. This covers laboratory-scale experiments, field trials on flue-gas slip streams at operating power plants, and full-scale demonstration plants. The program is not yet complete but has already identified technologies capable of capturing over 90 percent of the mercury in flue gases generated by the three major coal types, bituminous, sub-bituminous, and lignite. Further, as understanding has progressed so the costs of capture have decreased. The field-trial program started in 2001 and is currently scheduled to be completed in 2006 although it is highly likely that it will be continued. The amount of data produce are extensive (see Reference 3-6) so for brevity only the two approaches considered most applicable to CFB technology are discussed here. AmerenUEs 140-MW Meramec Unit 2 burns sub-bituminous coal using low-NOX burners. Intrinsic mercury capture was less than 10 percent. Conventional activated carbon injected ahead 3 of the cold-side ESP at a rate of 32 kg/million am (2 lb/million ACF) removed around 65 percent of the mercury. See Figure 3-1. Injecting a brominated activated carbon at the same rate removed 90 percent of the mercury.

3-5

100

80

Mercury removal, %

60

40

20

Meramec brominated AC Meramec untreated AC Laramie River brominated AC Laramie River untreated AC

0 0 1 2 3 4 5 6 7 8

Sorbent injection concentration, lb/million ACF

Figure 3-1 Mercury Capture Improvements using Brominated Activiated Carbon

Similar results were obtained from Missouri Basins 550-MW Laramie River Unit 2. This burns sub-bituminous coal and is supplied with a spray dryer absorber and cold-side ESP. The degree of improvement for the Laramie River unit was greater but a higher feed rate of activated carbon was required for a given capture efficiency. These differences may arise because of the spray dryer, emphasizing the complexity of deriving a universally applicable set of design criteria. As the lowest intrinsic mercury capture efficiency for the FBC units in EPAs ICR was 48 percent, adding even conventional activated carbon would appear to be capable of raising collection efficiency to over 90 percent. Of course the feed rate of brominated activated carbon may be lower and so a more economic choice. Basin Electric Power Cooperatives 440-MW Antelope Valley Unit 1 burns lignite and is supplied with a spray dryer and fabric filter. Brominated activated carbon injected ahead of the 3 spray dryer at a rate of 32 kg/million am (2 lb/million ACF) removed around 70 percent of the mercury. To enhance capture efficiency a proprietary, halogenated sorbent-enhancement additive (SEA-2) was fed to the boiler. When SEA-2 and brominated activated carbon were fed at 1.2 and 32 kg/million am3 (0.078 and 2 lb/million ACF), respectively, mercury capture was 90 percent. Either of the two approaches discussed appear to be capable of elevating mercury capture efficiency from the intrinsic level of around 50 percent up to the 95-percent level required by the 2020 Roadmap Goals.

3-6

References 3-1 3-2 3-3 3-4 3-5 3-6 Skowyra, R., ALSTOM CFB Technology, 18th International FBC Conference, Toronto, May 22nd to 25th, 2005. Foster Wheelers CCPI-1 CFB Project with Colorado Springs Utilities http://www.netl.doe.gov/coal/CCPI/pubs/withdrawnProj218.pdf Powerspan Web Page: http://www.powerspancorp.com/home/index.shtml Airborne Technologies Inc. Web Page: http://www.airbornepollutioncontrol.com/ghg.html Mobotec USA Web Page: http://www.mobotecusa.com/ DOE/NETL's Mercury Control Technology R&D Program Review Conference Pittsburgh, July 12th to14th, 2005. http://www.netl.doe.gov/publications/proceedings/05/Mercury/HgConf05.html

3-7

4
CARBON DIOXIDE CAPTURE PROCESSES INVOLVING CFB TECHNOLOGY
Worldwide concern over global warming is promoting considerable research effort into reducing emissions of carbon dioxide. There is unlikely to be a universal technology applicable to all situations, rather a variety of technologies each suited to particular technical and economic circumstances. Two techniques under development for which CFB technology seems well suited are combustion with oxygen instead of air, and indirect combustion using chemical looping. Increasing generating efficiency also lowers CO2 emissions and a 460-MW supercritical CFB is to be built in Poland. This is an important development and is expected to open up the opportunity for further large-scale, high-efficiency CFB power plants. These three technology developments are discussed briefly in this Section. There are other ways in which CFBs can contribute to reduced CO2 emissions but these are adaptations of existing designs and are not discussed in this report. The outstanding fuelflexibility characteristic of CFBs makes them well suited to co-fire biomass fuels and displace coal-derived CO2. This can be achieved by modifying existing CFB units or designing new ones with the ability to fire biomass as a high percentage of its base fuel. There are also a number of post-combustion CO2 removal processes that can be used with any power plant technology including CFBs. Oxygen-Fired Combustion This approach has been investigated for PC plants by the Canadian Clean Power Coalition (CCPC) (see Reference 4-1) and is being investigated for CFB application by ALSTOM Power (see Reference 4-2). Oxy-Combustion in a PC boiler In a conventional power plant, coal is burned in air to release heat and generate steam to be converted to electricity. The resulting flue gas is diluted greatly by the air nitrogen, a typical composition being 75-percent nitrogen, 15-percent CO2, 7-percent moisture, and 3-percent oxygen (by volume on a wet basis). Separating the low concentration of CO2 from the other species using post-combustion removal processes is energy intensive and lowers plant output significantly. To achieve a higher CO2 concentration, instead of burning the coal in air, oxygen from an on-site oxygen plant can be used. The resulting flue gas is more than 90-percent CO2 with the balance being moisture. However if oxygen alone is introduced into a boiler, the flame temperature reached would be far too high for the equipment to tolerate.

4-1

Air Nitrogen Flue Gas (~97% CO 2 ) Recycle ~75%

Air Separation

Oxygen

Oxygen
Coal

Boiler Flue gas Boiler + ~97% CO 2 controls

Drier Compressor W ater

G Generator Turbine

CO2 to be sequestered

Feed Pump

Taken from Reference 4-1.

Figure 4-1 Oxygen Fired PC Boiler

The risk of over heating is prevented by recirculating over 75 percent of the flue gas (mainly CO2) to control the flame temperature. An arrangement investigated by the CCPC is presented in Figure 4-1. Recirculating such a high volume of gas close to its dew point and distributing it to the coal burners is a complicated and costly procedure. The flue gas leaving the plant is cooled, condensing out water and removing other pollutants (NOX, SO2, and HCL), and the CO2 is compressed prior to sequestration. As there is no flue gas discharge flow, there is no need for a stack. The flue gas stream contains around one-volume percent oxygen (see Reference 4-3). If the CO2 is to be transported by pipeline to the sequestration site, the oxygen must be reduced to below 0.01 percent to satisfy current pipeline insurance requirements. The concern is that the reactive nature of oxygen promotes corrosion and may accelerate pipeline failure. Suitable means of reducing the oxygen have not yet been identified. If sequestration is at the power plant site and no pipeline is involved then the oxygen content may be of no concern.

4-2

Oxy-Combustion in a CFB Boiler The overheating issue associated with firing a PC boiler with oxygen, is more readily overcome in a CFB unit. An arrangement proposed by ALSTOM Power is presented in Figure 4-2. An intrinsic feature of CFB technology is that solids are circulated within the combustor for various process reasons. In some designs these solids are cooled in an external heat exchanger to raise steam, and re-introduced back into the combustor. By cooling the solids more extensively they can be used to control the potentially higher combustion temperature arising from using oxygen. This eliminates the need to recirculate flue gas for cooling duty although some recirculation is still required to achieve adequate fluidization in the furnace. This simplifies boiler design and lowers capital and operating costs appreciably. ALSTOM has carried out a program of oxygen combustion testing funded by the US-DOE (see References 4-4 and 4-5). The bench-scale bubbling FBC test reactor was an electrically-heated tube 1.8-m (6-ft) tall with an inner diameter of 0.1 m (4 inches). The bed material was alumina and the unit operated without fines recycle from the cyclones. The combustion gas was a mixture of oxygen and CO2 with the oxygen content varying from 30 to 70 percent. The fluidizing velocity was varied from 0.55 to 1.0 m/s (1. 8 to 3.3 fps). The oxygen-to-fuel ratio varied from 2.0 to 4.7 resulting in flue gas oxygen contents of 13 to 51 volume percent (dry basis). It was accepted that test conditions were not representative of those in a commercial CFB but the results were to be used to evaluate how the variables influenced performance parameters on a comparative basis. A CFB unit would operate with fines recycle from the cyclones and operate at a fluidizing velocity around 5.5 m/s (18 fps). Even with a combustion gas containing 70-percent oxygen the flue gas would not contain more than 3-percent oxygen. The properties of the three fuels tested, two bituminous coals and a delayed petroleum coke, are presented in Table 4-1. Limestone was fed during the tests to investigate sulfur capture performance. To capture sulfur the limestone must first be calcined to form CaO and it is this that reacts with SO2. CaCO3 = CaO + CO2 CaO + SO2 + [O] = CaSO4 The calcination temperature varies with the partial pressure of CO2.
Partial Pressure of CO2, bar (psia). Calcination temperature, C (F)

0.3 (4.4) 0.5 (7.3) 0.7 (10.2)

820 (1,510) 850 (1,570) 880 (1,610)

4-3

Taken from Reference 4-2.

Figure 4-2 Oxygen-Fired CFB Boiler

4-5

Table 4-1 Properties of Fuels Used by ALSTOM in the Oxygen Combustion Tests Base coal Composition, % a.r. Carbon Hydrogen Nitrogen Sulfur Oxygen Ash Moisture HHV, kJ/kg (Btu/lb) 65.2 3.3 1.3 1.8 3.0 21.5 3.9 26,390 (11,340) 59.2 4.1 1.1 3.7 6.9 8.6 16.4 24,730 (10,630) 81.6 3.6 3.0 1.1 1.0 0.3 9.4 32,680 (14,050) Illinois #6 Petroleum coke

To ensure that calcination occurs at the highest CO2 partial pressure the operating temperature has to be above 880C (1,610F). To aid comparison of test data, it was decided to carry out all the tests at the same temperature of 900C (1,650F). This is higher than the optimum for sulfur capture, normally around 850C (1,570F), so a high Ca/S molar ratio of 3.5 was selected to compensate. The following conclusions were reported. Increasing the oxygen content of the combustion gas increased the reactor temperature by up to 120C (250F) but did not result in any slagging or de-fluidization. Combustion efficiency increased from around 90 to 95 percent with increasing oxygen content resulting in lower unburnt carbon losses. A commercial CFB operating on air would expected to achieve greater than 98 percent combustion efficiency, so any increase with increasing oxygen content is expected to be less pronounced. CO emissions were higher than for air but decreased with increasing oxygen content. It is speculated that the higher partial pressure of CO2 inhibited CO oxidation at the particle-gas boundary layer. The test rig had a gas residence time of 1.5 seconds but a commercial unit will have over 5 seconds and this is expected to lower CO emissions. For tests with high oxygen content combustion gas in which the bed temperature increased markedly, SO2 capture for a given Ca/S molar ratio decreased. This reduction with increasing temperature is an established phenomena in air-fired FBCs. When the temperature remained reasonably constant SO2 capture increased slightly suggesting that the capture mechanism was enhanced by the increased oxygen content. NOX emissions were reduced by around 30 percent presumably because the nitrogen-free combustion gas eliminated the formation of thermal NOX. These test results were used in support of economic studies reported in Reference 4-4. Two CFB technologies with CO2 capture were included, oxygen fired and chemical looping (discussed next). Table 4-2 presents some performance data compared to IGCC.

4-6

Table 4-2 Performance Parameters Relative to those for IGCC for Two CFB-Based CO2 Capture Technologies Oxygen-fired CFB Heat rate TPC COE 1.09 1.19 1.15 Chemical looping 0.78 0.67 0.73

A CFB with oxygen firing is an expensive option for capturing CO2 compared to IGCC. Both required an air separation unit but the effect upon the thermal and economic performance of the CFB is greater because combustion requires approximately 2.5 times more oxygen than gasification. The development of alternative means of producing oxygen at lower costs appears essential. To this end, ionic-transport membrane technologies are under development in US-DOE funded projects and in Europe. Because of the higher oxygen demand any cost reduction would show the greatest benefit for combustion. ALSTOMs analysis indicates clearly that chemical looping offers the potential for capture CO2 at the lowest cost. However, this is a long-term option and will require considerable development effort. In contrast oxygen-fired CFB could be deployed in the short term using components currently in commercial service, and of course IGCC is already in service and being offered commercially. Recent EPRI studies re-evaluated amine-based, post-combustion CO2 capture technology and identified a number of thermal- and economic-performance improvements (see Reference 4-6). One aspect of the study was that the relative performance of technologies was affected by coal type. ALSTOMs studies were all for bituminous coal, and conclusions may be different for lowrank fuels. Chemical Looping This is an indirect combustion process that uses an oxygen-carrier medium to provide the oxygen for combustion so eliminating the need for air separation. There are several variants under investigation, and this section describes one of the simpler schemes. First coal is gasified in oxygen to produce a syngas containing carbon monoxide and hydrogen. This still requires air separation but the amount of oxygen required is only 40 percent of that required for combustion. These gaseous constituents pass into a syngas reactor where they mix with a hot oxygen-carrier medium that provides the oxygen to oxidize them to carbon dioxide and water. As this reaction is endothermic the amount of heat carried into the reactor by the syngas and oxygen carrier controls the extent of the reaction. The hot reduced carrier medium is transferred into an oxidizer to be reoxidized in an air stream. The heat from the exothermic reaction is carried away by the vitiated air leaving the oxidizer and transferred to the steam-water circuit to raise steam for power generation. The oxidized carrier medium is then transferred back into the syngas reactor. The net heat released over the two reactors is the same as for direct combustion in a combustor, with the important difference that the carbon dioxide is released as a separate stream containing moisture but no nitrogen. Hence, the nitrogen and the carbon dioxide are separated without extra energy being expended. Further, unlike combustion in oxygen there is no free oxygen present with the CO2, so the gas can be passed to the pipeline without a scavenging stage
4-7

As solids have to flow between the two reactors, a fluidized-bed system is well suited for the chemical looping equipment. A proposed atmospheric-pressure system is presented in Figure 4-3 and is discussed in more detail in Reference 4-7. The oxidizing reactions take place at around 950C (1,740F) in what is normally the furnace of the CFB. The oxygen-carrier particles are separated in the cyclone and flow into a bubbling-bed syngas reactor. The reduced carrier particles then flow through the seal pot and back into the oxidizer. The high temperature flue gas leaving the cyclone passes to a waste heat boiler and the CO2 leaving the syngas reactor is cooled to remove the moisture, and then compressed ready for sequestration. Unreacted syngas does not condense with the CO2 and is returned to the syngas reactor for further processing. Some is also bled to the oxidizer should the operating temperature need to be increased. This step results in some CO2 being discharged to atmosphere with the flue gas. The flue gas will carry some of the oxygen carrier away from the process and this will have to be replaced with make up material. The extent of material lost in this manner will affect the economics of the process. A number of metals have oxides suitable for use as oxygen carriers: aluminium, cobalt, copper, iron, nickel, and manganese, either individually or in combination. The oxide carrier will experience thermal cycling and physical expansion and contraction as it is oxidized and reduced. It must also be able to resist the mechanical forces applied by the fluidizing action of the reactors. As mentioned previously as the material breaks down and becomes finer it leaves the system with the flue gas exiting the cyclone. These same forces are the ones that have made it difficult to develop a hot gas desulfurization sorbent with an economic life. Finding a suitable oxygen-carrier medium with an economic life may pose a similar challenge. If the carrier is a metal oxide, it is likely to also react with trace contaminants present in the syngas such as H2S and HCL, and these may combine with the oxygen rendering the carrier inactive. To avoid this economic penalty the syngas will have to be efficiently cleaned prior to being introduced into the reactor. Although the chemical-looping concept is new and investigative work has only recently commenced, ALSTOM Powers preliminary assessment shows that it has tremendous potential. It is, however, a long-term option facing multiple challenges and requiring significant development effort.

4-8

flue gas to waste heat boiler

syngas recycle metal-oxide carrier make up syngas CO2 to sequestration

Modified version of Figure 2 from Reference 4-7.

Figure 4-3 Chemical-Looping Combustion Concept Incorporating a CFB

Generating Efficiency and CO2 Reduction Replacing existing low-efficiency boilers with higher efficiency designs can reduce carbon dioxide emissions significantly. Table 4-3 present the CO2 emissions for four steam cycles. Supercritical boilers are widely used and can reduce CO2 emissions by almost 12 percent compared to using a subcritical steam cycle. Ultra supercritical (USC) boilers, expected to come into service shortly in Europe and Japan, can reduce CO2 emissions by over 16-percent. The reduction can be increased to almost 26 percent by the Advanced USC boilers once they become available in a few years time. See Reference 4-8. The percent reductions of CO2 achieved will be much greater for a new unit replacing an inefficient existing unit. The advantage of using this approach is that except for the steam circuit and the necessary different materials of construction, the rest of the boiler remains unchanged. This offers continuity in equipment design and plant operation based on many years previous experience.

4-9

Table 4-3 Reductions in CO2 Achieved by Increasing Steam Conditions Steam Conditions Subcritical 165 bar/538C/538C (2,410 psia/1,000F/1,000F) Supercritical 241 bar/566C/566C/566C (3,410 psia/1,050F/1,050F/1,050F) Ultra-supercritical 344 bar/649C/649C/649C (5,010 psia/1,200F/1,200F/1,200F) Advanced ultrasupercritical
(1)
(2)

Heat Rate (1) 10,070 (9,550) 8,880 (8,420) 8,420 (7,980) 7,550 (7,150)

CO2 Emitted (2) 0.950 (2.07) 0.830 (1.83) 0.785 (1.73) 0.704 (1.55)

CO2 Reduction (3) NA

11.6

16.4

374 bar/760C/760C/760C (5,460 psia/1,400F/1,400F/1,400F)

25.9

Heat rate (HHV) kJ/kWh (Btu/kWh) CO2 emitted kg/kWh (lb/kWh) for a Pittsburgh #8 bituminous coal with HHV 29.2 MJ/kg (12,500 Btu/lb) and containing 75-weight percent carbon. (3) Reduction as a percent of subcritical CO2 emissions.

Increasing the steam conditions increases the capital cost of the plant because of the more exotic materials used. However, the improved efficiency results in fuel cost savings that offset this increase such that the cost of electricity is almost the same for all four steam conditions. Previously this was seen as a disincentive to adopt the more efficient designs, especially as more capital was required. However, this view is expected to change in a carbon-constrained world. Recent EPRI studies (see Reference 4-6) show that as more efficient plants produce less CO2 the cost of post-combustion capture is reduced so allowing the Advanced USC cycle to produce the lowest cost of electricity. The improvements in steam conditions have previously been applied only to PC boiler designs not to those of CFBs. However, this is changing and the first supercritical CFB boiler has been ordered. In March 2003 Poland's Poludniowy Koncern Energetyczny (PKE) announced that it had signed a contract with Foster Wheeler Energia Oy to design and build the worlds first supercritical CFB unit. The 460-MW power plant will replace two units within the existing Lagisza Plant located in southern Poland. The contract is valued at approximately $146 million and includes engineering calculations, proposal drawings, cost estimates, and services to define the coal handling system. Design work was completed in the end of 2003 but construction did not proceed as planned because of difficulties in securing the necessary financing. In November 2005 funding was secured and equipment purchasing commenced. Commercial operation was originally expected to commence in summer 2006 but a new date is yet to be announced. A most probable date is late 2008 or early 2009. The request for bids included designs for both CFB and PC supercritical units, and after a rigorous evaluation PKE determined that the CFB design was the most cost-effective option. It would appear that the flexibility of the CFB to accept a wider range of coal properties was an important factor in the selection. The ability to also burn coal slurry was also required. Some of
4-10

the coals had a high chlorine content and the lower metal temperatures in a CFB boiler will help avoid superheater tube corrosion. The stated characteristics for the bituminous coal are tabulated below on an as-received basis.
Design coal Sulfur, wt. % Ash, wt. % Moisture, wt. % LHV, MJ/kg (Btu/lb) 1.2 23 12 20 (8,600) Coal range 0.6 to 1.4 10 to 25 6 to 23 18 to 23 (7,740 to 9,890) Slurry range 0.6 to 1.6 28 to 65 27 to 45 7 to 17 (3,010 to 7,310)

The design incorporates Siemens Benson vertical once-through unit technology, which is ideally suited to the uniform temperature distribution experienced in a fluidized-bed combustor. The initial superheat and reheat circuits are located in the back pass while the final circuits are located in the INTREX unit. This is an external heat transfer unit integrated with the boiler through which the fines collected by the cyclones are returned. As it is fluidized by air not flue gas, this serves further to reduce corrosion potential. The boiler produces almost 1,040,000 kg/hr (2,860,000 lb/hr) of steam at 27.5 MPa/560C/580C (3,910 psia/1,040F/1,080F) and the design heat rate is approaching 9,150 kJ/kWh (8,700 Btu/kWh) on a higher heating value basis. The plant emissions performance has not been stated other than that they will meet the requirements of the new European Union Large Combustion Plant Directive. For SO2 and NOx these are 200 mg/Nm3 (0.165 lb/MBtu) and for dust 50 mg/Nm3 (0.041 lb/MBtu). Currently the largest CFB boilers are around 300 MW, so the design for the Lagisza unit will require some scale up. It is believed that the design will be based on Foster Wheelers Compact Boiler design that is being supplied at 232 MW (net) to the Turow Power Station, also in Poland. The bed depth (boiler front-to-back dimension) is not increased so the ability to distribute air and solids in that plane will not change. The boiler height will be 10-percent greater, which is not considered to affect performance at all. The length will be increased by approximately 20 percent but this is accommodated by increasing the number of coal feed points. ALSTOM Power also has supercritical designs prepared for CFB plants up to 600 MW. It is understood that these designs have been included in at least three responses to RFPs within the USA and Europe.

4-11

References 4-1 Stobbs, B. Canadian Clean Power Coalition: Clean Coal-Fired Power Plant Technology to Address Climate Change Concerns, paper presented at Gasification Technologies Conference: 2003, San Francisco, October 13th to 15th 2003. G.N. Stamatelopoulos, G. N. et al, CO2 Control Technologies: ALSTOM Power Approach, International Conference on Clean Coal Technologies for our Future, Sardinia, 21st to 23rd October 2002. Vogel, D. et al, Engineering Feasibility of CO2 Capture on and Existing US Coal-Fired Power Plant, First National Conference on Carbon Sequestration, Washington DC, May 15th to 17th, 2001. Greenhouse Gas Emissions Control by Oxygen Firing in Circulating Fluidized-Bed Boilers: Phase 1 A Preliminary Systems Evaluation, Volume 1 of a report prepared by ALSTOM Power for the US-DOE on Cooperative Agreement DE-FC26-01NT41146, May 2003 Volume 2 of the report in Reference 4-4 Evaluation of Amine-Based, Post-Combustion CO2 Capture Plants, EPRI Report 1011402, November 2005. Mattisson, T. and Lyngfelt, A., Capture of CO2 using Chemical-Looping Combustion paper presented at First Biennial Meeting of the Scandinavian-Nordic Section of the Combustion Institute, Gteborg, Sweden, 18th to 20th April 2001 Viswanathan, R. et al, Materials for Ultra-Supercritical, Coal-Fired Power Plants, Materials for Advanced Power Engineering, 7th Lige Conference, Belgium, 30th September to 2nd October 2002.

4-2

4-3

4-4

4-5 4-6 4-7

4-8

4-12

Export Control Restrictions Access to and use of EPRI Intellectual Property is granted with the specific understanding and requirement that responsibility for ensuring full compliance with all applicable U.S. and foreign export laws and regulations is being undertaken by you and your company. This includes an obligation to ensure that any individual receiving access hereunder who is not a U.S. citizen or permanent U.S. resident is permitted access under applicable U.S. and foreign export laws and regulations. In the event you are uncertain whether you or your company may lawfully obtain access to this EPRI Intellectual Property, you acknowledge that it is your obligation to consult with your companys legal counsel to determine whether this access is lawful. Although EPRI may make available on a case-by-case basis an informal assessment of the applicable U.S. export classification for specific EPRI Intellectual Property, you and your company acknowledge that this assessment is solely for informational purposes and not for reliance purposes. You and your company acknowledge that it is still the obligation of you and your company to make your own assessment of the applicable U.S. export classification and ensure compliance accordingly. You and your company understand and acknowledge your obligations to make a prompt report to EPRI and the appropriate authorities regarding any access to or use of EPRI Intellectual Property hereunder that may be in violation of applicable U.S. or foreign export laws or regulations.

The Electric Power Research Institute (EPRI)

The Electric Power Research Institute (EPRI), with major locations in Palo Alto, California, and Charlotte, North Carolina, was established in 1973 as an independent, nonprofit center for public interest energy and environmental research. EPRI brings together members, participants, the Institutes scientists and engineers, and other leading experts to work collaboratively on solutions to the challenges of electric power. These solutions span nearly every area of electricity generation, delivery, and use, including health, safety, and environment. EPRIs members represent over 90% of the electricity generated in the United States. International participation represents nearly 15% of EPRIs total research, development, and demonstration program. TogetherShaping the Future of Electricity

2005 Electric Power Research Institute (EPRI), Inc. All rights reserved. Electric Power Research Institute and EPRI are registered service marks of the Electric Power Research Institute, Inc. Printed on recycled paper in the United States of America 1010231

ELECTRIC POWER RESEARCH INSTITUTE 3420 Hillview Avenue, Palo Alto, California 94304-1395 PO Box 10412, Palo Alto, California 94303-0813 USA 800.313.3774 650.855.2121 askepri@epri.com www.epri.com