Article pubs.acs.org/EF
Techno-Economic Assessment of Integrated Chemical Looping Air Separation for Oxy-Fuel Combustion: An Australian Case Study Cheng Zhou, Kalpit Shah, and Behdad Moghtaderi* Priority Research Centre for Frontier Energy Technologies & Utilisation Chemical Engineering, School of Engineering, Faculty of Engineering and Built Environment, The University of Newcastle, Callaghan, NSW 2308, Australia S Supporting Information *
ABSTRACT: A techno-economic analysis was carried out to assess the oxy-fuel conversion of eight major coal-fired power plants in the state of NSW, Australia. For this purpose, several alternative retrofit configurations, differing only in the air separation unit (ASU) but otherwise identical, were considered. More specifically, three types of oxygen plants were studied: a cryogenic-based air separation unit and integrated chemical looping air separation units using steam (ICLAS[S]) and recycled flue gas (ICLAS[FG]) as the reduction medium. The main objective of the techno-economic analysis was to determine if the economic viability of oxy-fuel operations could be enhanced by incorporating ICLAS technology. The results show that the normalized oxygen demand for the NSW fleet of coal-fired power plants was about 450−550 m3/MWh, with Bayswater having the lowest normalized oxygen demand and Munmorah having the highest one. Moreover, it was found that by replacing a cryogenic-based ASU with an ICLAS unit, the average reduction in the ASU power demand was up to 47% and 76%, respectively, for ICLAS[S] and ICLAS[FG]. Similarly, the average thermal efficiency penalty associated with the cryogenic and the ICLAS[S] and ICLAS[FG] units was found to be about 9.5%, 7.5%, and 5%, respectively, indicating that the ICLAS[FG] unit is the most energy efficient option for oxy-fuel plants. Economic analyses suggest that a retrofit cost reduction of about 32% can be achieved by incorporating an ICLAS[FG] unit. On average, the levelized cost of electricity associated with the cryogenic and the ICLAS[S] and ICLAS[FG] units for the NSW fleet of coal-fired power plants was found to be about $118/MWh, $105/ MWh, and $95/MWh, respectively.
1. INTRODUCTION Oxygen is the second largest-volume chemical produced in the world with a 30% share of the global industrial gas market. The global demand for oxygen in 2011 is forecast to be 950 billion cubic meters with an annual growth rate of about 6%. It has major commercial applications in metallurgical industry, chemical synthesis, glass manufacturing, pulp and paper industry, petroleum recovery/refining, and health services. Emerging markets for oxygen include advanced power generation systems, such as integrated gasification combined cycle, oxy-fuel combustion, and solid oxide fuel cells. Among these, oxy-fuel combustion is particularly an attractive lowemission technology because of its inherent ability for in situ separation of CO2. However, oxy-fuel combustion requires oxygen and, thereby, an air separation unit (ASU) to function effectively. Conventional ASUs (e.g., cryogenic systems) may consume between 10% and 40% of the gross power output of a typical oxy-fuel plant1−8 and constitute 40% of the total equipment cost (about 14% of the total plant cost).7,8 Oxygen is commonly produced at industrial scales by air separation using cryogenic distillation and adsorption-based technologies [pressure swing adsorption (PSA) and vacuumPSA (VPSA)]. Advanced technologies such as membrane separation (e.g., ion-transport membrane) and in situ air separation are also being developed for small-volume point-ofuse oxygen generation. Cryogenic processes are generally expensive owing to the energy intensity of their air compression sub-process. Similar to the cryogenic methods, air compression is a key step in the adsorption-based air separation methods, © 2015 American Chemical Society
and, as such, the specific power consumptions of PSA and VPSA plants are not much lower than their cryogenic counterparts. Membranes have been in commercial use for several decades, but much of their past applications have been in liquid−liquid and liquid−solid separation. The use of membranes for large volumetric gas flow rates, such as those in air separation, has not been demonstrated yet. Membrane systems also suffer from high cost of manufacture. There is a need for a more simple and cost-effective air separation technology with much smaller energy footprint and lower capital cost than conventional and emerging air separation methods. In recognition of this need and relying on the track record of our group in chemical looping9−20 and development of novel energy technology concepts,21,22 we embarked on a journey to develop an innovative chemical looping-based method of air separation for oxy-fuel applications in NSW, Australia. The relevant research work was supported by generous funding from the State government. The method known as integrated chemical looping air separation (ICLAS) is used for tonnage production of high-purity oxygen, and is specifically tailored for ease of integration with oxy-fuel type power plants running on organic (e.g., biomass) or fossil-based (e.g., coal, gas, oil, etc.) fuels. Received: October 1, 2014 Revised: March 23, 2015 Published: March 23, 2015 2074
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
exiting from the reduction reactor is passed onto the boiler for oxy-fuel combustion process. In case of using steam as the reduction medium, the mixture of steam and oxygen need to enter a condenser unit for the removal of water before it being sent to the boiler. The above processes are then retrofitted to an existing coal-fired power plant, where coal, oxygen (from the ASU), and recycled flue gas are co-fed into the boiler and the mixture is combusted at high temperatures. The heat generated from the combustion process runs a steam cycle which in turn converts the thermal energy into electricity (see Figure 2). The use of recycled flue gas here is an important and integral part of the oxy-fuel combustion process because firing pure oxygen in a boiler would result in excessively high flame temperatures which may damage the boiler. Therefore, the mixture must be diluted by mixing with recycled flue gas before it can be fed into the boiler. This is particular convenient for the overall retrofit system if the ICLAS unit also uses recycled flue gas, instead of steam, as the reduction medium.
By incorporating the concept of oxygen decoupling into the two-step redox reaction, ICLAS is able to separate oxygen from normal air. As Figure 1 illustrates, the ICLAS process works in a
Figure 1. Schematic of the ICLAS process.
MexOy − 2 (s) + O2 (g) → MexOy (s) Oxidation or O2 coupling
cyclic fashion by continuous recirculation of metal oxide particles between a set of two interconnected reactors, where oxidation (O2 coupling, see R1) and reduction (O2 decoupling, R2) of carrier particles take place, respectively. During this cyclic process, oxygen is taken from air in one reactor, carried by particles then released in a second reactor. The system therefore consists of two reactors linked together through a loop-seal to prevent gas leakage from one reactor to another. Air is fed into the oxidation reactor so that the incoming reduced carrier particles can be regenerated to a higher oxidation state. The regenerated carrier particles, in turn, are transported back to the reduction reactor, where oxygen decoupling occurs in the presence of recycled flue gas, or alternatively steam. The mixture of recycled flue gas and oxygen
(R1)
MexOy (s) → MexOy − 2 (s) + O2 (g) Reduction or O2 uncoupling
(R2)
From energy efficiency point of view the ICLAS process is quite efficient because of its low energy demands. This is partly due to the fact that the theoretical net heat released over reactions R1 and R2 is zero. Therefore, in theory the heat transported by the incoming carrier particles into the reduction reactor must be sufficient to support the endothermic reaction R2. In practice, though, some heat must be supplied to the reduction reactor to compensate for heat losses to the surrounding. Furthermore, under steady state operation much
Figure 2. Schematic of an oxy-fuel coal-fired power plant retrofitted with an ICLAS unit (adapted in part from CCSD online publications). 2075
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
CO2 from the power plant to the sequestration site is feasible from both technical and economic points of view. It is also important to assume that for any given power plant the oxy-fuel retrofit is not limited by any space and layout limitations. Obviously these assumptions which are quite site-specific may not always be valid. However, recent studies of fleet retrofit potential suggest that a considerable number of existing coalfired power plants around the world can be retrofitted for oxyfiring. Besides, it is essential to make such assumptions in a benchmarking exercise like that described here. The details of the methodology and key findings from the present technoeconomic assessment are provided in the proceeding sections.
of the heat required for pre-heating of air is offset by the heat contents of the hot stream leaving the reduction reactor and the reduced air stream exiting from the oxidizer. As Figure 1 shows, this is achieved by exchanging the sensible heat between various streams in a series of heat exchangers. The additional thermal energy required to carry out the ICLAS process can be provided by electrical power. Our mass and energy balance calculations carried out using the HYSYS process simulation package suggest that the heat/power demand for the ICLAS process is much lower than that required in cryogenic systems. The heat demand for the ICLAS process is 0.03 kWh per cubic meters of oxygen produced (i.e., 0.03 kWh/m3n), which is about 7.5% of the specific power of a conventional cryogenic system (i.e., 0.4 kWh/m3n). More advanced cryogenic systems due to enter the market by 2012, however, are expected to approximately reach a specific power of 0.3 kWh/m3n. Such specific powers are still 10 times greater than the average specific power for the ICLAS process. The successful execution of the ICLAS process also largely depends on oxygen carriers capable of reacting reversibly with gaseous oxygen at high temperatures. This additional thermodynamic constraint is a means of differentiating oxygen carriers feasible for the ICLAS process from those only suitable for common redox applications. This part of study has been carried out extensively in other published works of the research group9,13−20 and thus will not be covered in this study. The specific scope of this paper, however, was concerned with a techno-economic assessment of ICLAS retrofitting to NSW fleet of coal-fired power plants. As shown below in Table 1, there are eight major coal-fired power plants (capacity >1 MW) in NSW with a combined capacity of 11.75 GW. The most recent literature on oxy-fuel combustion1 indicate that as a technology option for mitigation of GHG emissions in short to medium terms, oxy-fuel is particularly suited to older coal-fired power plants (>20 years), where the capital cost is typically written off after 20 years and, as such, the retrofit costs can be covered over a longer period. Of those eight NSW major coal-fired power plants, six plants (Bayswater, Eraring, Liddell, Munmorah, Vales Point, and Wallerawang) with a combined capacity of 10.2 GW have been in service for over 20 years and, as such, are particularly attractive for oxy-fuel option. Given the above, there are two key objectives of this study:
2. METHODOLOGY 2.1. Overview. The techno-economic study summarized here was concerned with retrofitting the eight major coal-fired power plants listed below in Table 1 for oxy-firing operation. A number of relevant publications were used to establish an accurate database of technical information and cost components.1−8 The most important principle followed in the technoeconomic assessment was to use a consistent and correct baseline. For this purpose, several alternative retrofit configurations, differing only in ASU but otherwise identical, were considered for each plant: Base Case: “business as usual” operation scenario for the existing plant (air-fired) with no carbon capture and storage (CCS) Cryogenic Case: oxygen-fired retrofit with a cryogenictype ASU and CCS ICLAS[S] Case: oxygen-fired retrofit with an ICLAS unit (steam reduction) and CCS ICLAS[FG] Case: oxygen-fired retrofit with an ICLAS unit (recycled flue gas reduction) and CCS The technical assessment for each plant comprised an investigation of the impact of both ASU and CPU (CO2 processing unit) on the plant’s performance characteristics. The impact was evaluated and expressed in terms of the following: • plant thermal efficiency • plant thermal efficiency penalty • electricity output penalty (EOP) • power requirement for oxygen production • power requirement for CPU Similarly, in the economic assessment of technology options for the NSW fleet of coal-fired power plants, we considered the base case, case 1, case 2, and case 3 scenarios outlined above. The economic assessment was evaluated and expressed in terms of the following: • power plant retrofit costs (i.e., capital cost) • operating costs • levelized cost of electricity • incremental cost of electricity relative to the base case • CO2 abatement costs relative to the base case In addition to the above comparisons among various oxy-fuel alternatives, a series of bench-marking comparisons were also made between chemical looping-based oxy-fuel options and a suite of other low-emission technologies for renewable energy sources and fossil fuels. 2.2. Technical Assessment Methodology. Figure 3 illustrates the technical assessment sub-model (TAS) developed by the research team to determine the technical aspects of oxy-
1. Carry out a techno-economic assessment of the NSW fleet of coal-fired power plants for oxy-fuel conversion. 2. Determine the extent to which the economic viability of oxy-fuel operations is enhanced by integration of the chemical looping-based air separation processes into such operations. In this work the results of a techno-economic assessment for conversion of the above eight power plants to oxy-firing are presented. The assessment has been carried out at an elementary level mainly to evaluate the relative merits of various air separation processes when incorporated into oxyfuel operations. As such the current techno-economic study focuses on the key question of “Is oxy-f iring worth doing in NSW?” rather than “Can oxy-f iring be done in NSW?”. The answer to the latter question requires a much more detailed feasibility study which is beyond the scope of the present investigation. Furthermore, for an elementary study such as that outlined here it is necessary to assume that the transport and storage of 2076
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
The associated flue gas recirculation rate was determined by dividing the plant’s oxygen demand (calculated separately as part of the analysis in m3/h) by the oxygen concentration (i.e., 20%). The choice of copper was due to the fact that copper was found in one of our previous studies12 to be the most expensive oxygen carrier, and, as such, it did allow us to consider the most expensive scenario (worst case scenario) in terms of cost. As a result, there was no need to consider lower cost figures associated with other types of oxygen carriers. The associated solid circulation rate of copper oxide for each power plant was calculated by dividing the plant’s oxygen demand by experimentally determined rate of oxygen transport (ROT) of copper oxide at a conversion rate of 80% and a temperature of 1000 °C. The ROT used in our study was about 3%/s. Plant Output Calculations.
Figure 3. Flowchart of TAS.
SOE × 1000 OC
TPO =
fuel retrofit to the NSW coal-fired power plant fleet. The basic plant data for the NSW coal-fired power plants of interest were obtained from ref 23, and are summarized in Table 1. The properties of the fuel (ultimate analysis) used in each power plant are summarized in Table 2. Moreover, the equations used in TAS to calculate various outputs and/or parameters are listed in this section to provide the reader with a basic understanding of the science underlying TAS. The data used in calculations of the power demand for ASU were those obtained from the experimental components of the present study, whereas the data for CPU power demand calculation were adopted from the literature for systems combining highest efficiencies and best practices.1−8 The project data used as the specific power requirement of the ICLAS process was related to CuO/Cu2O metal oxide system on an alumina substrate at a reduction temperature of 1000 °C and oxidation temperature of 900 °C, resulting in a 20% O2 concentration in the product gas stream from the ASU (i.e., 20% O2 and 80% flue gas). The specific power also accounted for the energy required for the transport of particles between the redox reactors.
(1) (2)
OC = LF × 365 × 24
where TPO is typical plant output (MW), SOE is send-out electricity (GWh), OC is operating capacity (h/year), and LF is load or capacity factor (assumed to be 85% in this study). CO2 Captured Calculations. CE × (CO2 ER × 10−3)(SOE × 1000)
CCO2 =
106
(3)
where CCO2 is captured CO2 per year (Mt/yr), CE is capture efficiency (assumed to be 90% in this study), and CO2ER is emission rate (kg/MWh). O2 Demand Calculations. SOE × 1000 (OC × Eff)
Q in =
FCR =
(4)
Q in (5)
FCV
Table 1. Basic Plant Data for Eight Major Coal-Fired Power Plants in NSW23 Bayswater operator commencement year boiler typea maximum capacity (MW) send-out electricity (GWh) no. of units unit capacity (MW) condenser cooling cooling medium
annual average thermal efficiencyb (%) fuelc annual fuel consumption (Mt) CO2 emission rate (kg/MWh) a
Eraring Eraring Energy
Liddell
Macquarie Generation 1984 PF 2640
1982 PF 2640
Macquarie Generation 1971 PF 2000
15955
13859
8070
4 660 natural draft cooling towers fresh water (Hunter River) 36.5
4 660 natural draft cooling towers salt water (Lake Macquarie)
Munmorah
Redbank
Vales Point B
Delta Electricity 1992 PF 1320
Mt Piper
Delta Electricity
National Power
Delta Electricity
1969 PF 600
2001 CFBC 150
1978 PF 1320
Delta Electricity 1976 PF 1000
Wallerawang
7921
226
1027
6671
5458
2 660 evap. cooling towers salt water (Lake Macquarie)
2 500 evap. cooling towers fresh water (Cox River)
35.6
33.2
2 660 evap. cooling towers fresh water (Cox River)
2 300 evap. cooling towers salt water (Lake Mun mora)
36.4
4 500 custom built lake fresh water (Hunter River) 32.7
37.2
32.0
1 150 evap. cooling towers fresh water (Hunter River) 32.0
coal 7.23
coal 5.63
coal 4.16
coal 3.1
coal 0.1
coal tailing 0.53
coal 2.82
coal 2.27
879
870
949
843
984
978
896
893
PF, pulverised fuel; CFBC, circulating fluidized bed combustor. bHHV basis. cSee Table 2 for details. 2077
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels Table 2. Fuel Properties for Eight Major Coal-Fired Power Plants in NSW23 Bayswater
Eraring
Liddell
Mt Piper
Munmorah
Redbank
Vales Point B
Wallerawang
53.5 3.5 1.2 0.5 6.4 24.7 10.1 22.4
59.0 3.7 1.2 0.4 6.6 21.0 8.2 24.3
49.7 3.3 1.1 0.5 6.3 30.4 8.8 20.9
59.3 3.7 1.4 0.5 6.1 21.2 8.0 24.7
63.6 4.0 1.2 0.4 7.3 19.0 4.5 26.3
51.8 3.3 2.2 0.3 5.4 8.0 30.0 21.6
58.2 3.7 1.2 0.4 6.3 22.2 8.0 23.8
59.1 3.7 1.3 0.5 5.8 21.7 8.0 26.0
C (%) H (%) N (%) S (%) O (%) ash (%) moisture (%) calorific Value (MJ/kg)a a
HHV basis.
CRR = FCR × (C%)
(6)
ORR = [FCR × (O%)]/2
(7)
⎡ (MWO2) ⎤ ⎥ − ORR ODTm = ⎢CRR × ⎢⎣ MWC ⎥⎦
(8)
ODA m = ODTm × (1 + EX air)
(9)
⎛ ODA ⎞ ⎛ R ⎞ ⎛ 1 ⎞ ̅ m⎟ ODA V = ⎜⎜ ⎟ × ⎜ T × P ⎟ × ⎝⎜ 3600 ⎟⎠ MW ⎝ ⎝ O2 ⎠ a a⎠
SDB is bulk solid density of CuO (∼6400 kg/m3), and ROT* is rate of oxygen transport expressed on the basis of the weight loss of CuO corresponding to 80% conversion (∼3%/s for the system under investigation). ASU Power Demand Calculations. ASUW =
where CRR is rate of carbon release from fuel (kg/s), Eff is plant’s thermal efficiency, EXair is excess air (assumed to be 1.5% in this study; see ref 4), FCR is fuel consumption rate (kg/s), FCV is fuel calorific value (MJ/kg), MWC is molecular weight of carbon (12 kg/kmol), MWO2 is molecular weight of oxygen (32 kg/kmol), ODAm is actual oxygen demand on mass basis (kg/s), ODAV is actual oxygen demand on volumetric basis (m3/h; assuming ideal gas behavior), ODTm is theoretical oxygen demand on mass basis (kg/s; assuming 100% combustion efficiency), ORR is rate of oxygen release from fuel (kg/s), Pa is ambient pressure (assumed to be 101.3 kPa in this study), Qin is total input thermal energy to the power plant (MW), R̅ is the universal gas constant (8.314 kJ/(kmol·K)), and Ta is ambient temperature (assumed to be 22 °C in this study). ICLAS Redox Calculations. (11)
ODA m ROT*
(12)
Binv =
SCR × 60 1000
(13)
Bvol =
Binv × 1000 SDB
(14)
SCR =
AMOD = Binv × NBR
(16)
⎧ 0.34 kW/(m 3· h) cryogenic; see ref 4 ⎪ ⎪ SPA = ⎨ 0.18 kW/(m 3· h) ICLAS[S]; see ref 24 ⎪ ⎪ 0.08 kW/(m 3· h) ICLAS[FG]; see ref 24 ⎩
(10)
ODA V GRR = OCPS
ODA V × SP 1000
(17)
where ASUW is the air separation unit power demand (MW), and SPA is the specific power of the ASU. CPU Power Demand Calculations. CPUW =
⎛ SPC ⎞ ⎛ CCO2 × 106 ⎞ ⎜ ⎟ × ⎜ ⎟ ⎝ 1000 ⎠ ⎝ OC ⎠
(18)
where CPUW is the CO2 processing unit power demand (MW), and SPC is the specific power of the CPU (assumed to be 135 kWh/t CO2 captured; see ref 5). Efficiency Penalty Calculations. ⎛ TPO − ASU ⎞ W ⎟⎟ EPA = Eff − ⎜⎜ Q in ⎝ ⎠
(19)
⎛ TPO − CPU ⎞ W ⎟⎟ EPC = Eff − ⎜⎜ Q in ⎠ ⎝
(20)
EPT = EPA + EPC
(21)
where EPA is efficiency penalty due to ASU (%), EPC is efficiency penalty due to CPU (%), and EPT is total efficiency penalty. Electricity Output Penalty Calculations. %EOP =
⎛ ASUW + CPUW ⎞ ⎜ ⎟ ⎝ ⎠ TPO
(22)
⎛ [ASU + CPU ] × 1000 × OC ⎞ W W ⎟ EOP = ⎜ CCO2 × 106 ⎝ ⎠
(15)
where AMOD is quantity of metal oxide needed per year (tonnes), Binv is bed inventory (tonnes), Bvol is ved volume (m3), GRR is gas (either steam or flue gas) recirculation rate (m3/s), NBR is bumber of bed replacements per year (typically eight times per year, given our experimental data suggest that CuO shows significant loss of reactivity after 1000 h of operation), OCPS is oxygen concentration in product gas stream (∼20% for ICLAS), SCR is solid circulation rate (kg/s),
(23)
where %EOP is electricity penalty expressed as percentage of typical plant output (%), and EOP is electricity penalty per tonne of CO2 captured (kWh/t CO2). 2.3. Economic Assessment Methodology. Figure 4 illustrates the cost assessment sub-model (CAS) developed by the research team to determine the economic aspects of oxy2078
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
Table 4. Properties of Reference Coal Used in Determining Unit Costs Listed in Table 3 C (%) H (%) N (%) S (%) O (%) ash (%) moisture (%) calorific value (MJ/kg)a a
49.83 2.91 0.71 0.22 13.87 32.46 3.88 20.23
HHV basis.
Table 5. Components of the Fixed Operating and Maintenance Costs item
2009 $/ kWyr
2014 $/ kWyr
2009 $/ MWh
2014 $/ MWh
operator labor maintenance cost admin and support labor total
11.8 50.3 0.96 63.1
11.1 47.3 0.9 59.3
1.6 6.8 0.1 8.5
1.5 6.4 0.1 8.0
Figure 4. Flowchart of CAS.
fuel retrofit to the NSW coal-fired power plant fleet. The basic cost data were obtained from refs 1−8, and are summarized in Tables 3−6 for the reference case (i.e., the oxy-fuel retrofit reference case). Moreover, the equations used in CAS to calculate various outputs and/or parameters are listed in this section to provide the reader with a basic understanding of the underlying methodology particularly in respect to determination of the levelized cost of electricity. Table 3 summarizes the unit cost (expressed as Australian dollars per kilowatt of plant output unless otherwise specified) for various component of the retrofit work. Unit costs (see the second column from left) were expressed in terms of the value of dollar in 2009 and on the basis LHV of a reference coal (Table 4) used in ref 7.These figures have been adjusted in this study based on the 2014 dollar and HHV of the reference coal (see columns 3−5 of Table 3). Table 5 shows the unit costs for fixed operating and maintenance cost items. The unit cost for operator labor was obtained assuming a labor rate of $60/h and 13 operators per shift. The original unit costs from refs 2, 3, and 7 listed in the second column are in $/kWyr for the 2009 dollar. These unit costs have been adjusted for the 2014 dollar and have been also
calculated on the basis of $/MWh of plant output. For conversion from $/kWyr to $/MWh, it was assumed that the capacity factor was 85%, which according to eq 2 equates to an operating capacity of 7446 h/yr. Table 6 summarizes the variable operating and maintenance cost items, primarily adopted from information provided in refs Table 6. Components of the Variable Operating and Maintenance Cost unit cost consumable
value
unit
water water treatment chemicals fabric filter bags and cages fly and bottom ash disposal oxygen carriers for ICLAS
1.1 0.8 0.2 4.3 5000−5500
$/MWh $/MWh $/MWh $/MWh $/tonne
Table 3. Components of Capital Cost
a
unit cost based on LHV ($/kW)
unit cost based on HHV ($/kW)
item
2009 $a
2014 $
2009 $
2014 $
coal handling system coal prep and feed systems feed water and miscellaneous balance of plant systems PC boiler and accessories ASU (cryogenic) ASU (cryogenic) ASU (ICLAS[S])b ASU (ICLAS[FG])b flue gas cleanup purification and compression HRSG, ducting, and stack steam turbine generator cooling water systems ASH/spent sorbent handling access electric plant I&C
129.8 44.7 159.2 865.3 652.8
122.1 42.1 149.8 814.2 614.2 4917.2 4300 3980 237.4 270.5 38.0 447.6 107.7 43.6 224.0 45.6
115.3 39.7 159.2 865.3 652.8
108.5 37.4 149.8 814.2 614.2 4917.2 4300 3980 237.4 270.5 38.0 447.6 107.7 43.6 224.0 45.6
252.3 287.5 40.4 475.7 114.5 46.3 238.1 48.5
252.3 287.5 40.4 475.7 114.5 46.3 238.1 48.5
notes
unit costs per kW of ASU power demand unit costs per kW of ASU power demand unit costs per kW of ASU power demand S&L models
unit costs adjusted based on coal ash and sulfur content
From refs 2,3,5−8. bFrom ref 25, based on figures for catalytic cracking units in petrochemical plants and refineries. 2079
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels 2, 3, and 7, With the exception of oxygen carriers cost, all other cost items are assumed to be the same for all alternative air separation options. The cost of oxygen carriers is based on information received from several vendors. Fuel Cost Calculations. ⎛ FCV ⎞ UFC = UFCR × ⎜ ⎟ ⎝ FCVR ⎠
(24)
⎛ SOE ⎞ TFC = (UFC × 1000) × ⎜ 6 ⎟ ⎝ 10 ⎠
(25)
FC = (TFC × 106) × (FCV × AFC × 109)
(26)
Figure 5. Bar chart plots of gross oxygen demand for coal-fired power plants in NSW.
where AFC is annual fuel consumption (Mt/yr), FC is fuel cost per unit of energy content ($/MJ), FCV is fuel calorific value (MJ/kg), FC is fuel cost per unit of energy content ($/MJ), TFC is total fuel cost per year ($/yr), UFC is unit fuel cost in dollar per MWh of send-out electricity ($/MWh), and subscript “R” refers to the reference fuel described in Table 4. Levelized Cost of Electricity Calculations. If, for simplicity, it is assumed that the net output of the power plant and the operating, maintenance, and fuel costs are constant over the life of the plant, then the levelized cost of electricity can be calculated from the following set of equations:26
is directly proportional to the plant capacity, it is not surprising that Bayswater requires about 1 million cubic meters of oxygen per hour, whereas Munmorah only needs about 16 000 m3/h. However, the gross value of oxygen demand does not provide the whole picture because in addition to the plant capacity the oxygen demand is also related to the type of fuel consumed in the plant (note that the fuel carbon and oxygen contents as well as its calorific value influence the overall oxygen demand). For that reason and to have a better understanding of the oxygen demand, a normalized version of the information provided in Figure 5 is illustrated in Figure 6.
⎛ (TCR × FCF) + FOM ⎞ LCOE = ⎜ ⎟ + VOM + CS ⎝ ⎠ NPO + (HR × FC)
FCF =
r(1 + r )t (1 + r )t − 1
(27)
(28)
NPO = TPO − ASUW − CPUW
(29)
AFC × FCV × 109 NPO × OC
(30)
HR =
where CS is cost of CO2 storage ($/t CO2, assumed to be $72/ t CO2 for NSW),6 FCF is fixed charged factor, FOM is fixed operating and maintenance cost ($/yr, see Table 5), HR is net power plant heat rate (MJ/MWh), LCOE is levelized cost of electricity ($/MWh), NPO is net power plant output (MW), r is rate of capital discharge (%; assumed to be 7% in this study),2,3,7 t is power plant life (yr; assumed to be 25 yr in this study), TCR is total capital requirement ($; see Table 3), and VOM is variable operating and maintenance cost ($/MWh; see Table 6). Cost of Abatement Calculations. ⎛ LCOE − LCOE ⎞ ref ⎟ COA = ⎜⎜ ⎟ CO − CO ⎝ 2ref 2cap ⎠
Figure 6. Bar chart plots of normalized oxygen demand for coal-fired power plants in NSW.
This time though the oxygen demand is expressed in cubic meters of oxygen per MWh of the power plant output. As can be seen the normalized oxygen demand varies between 450 and 550 m3/MWh for the coal-fired power plants in NSW. Interestingly, the normalized values shown in Figure 6 indicate that Bayswater has one the lowest normalized oxygen demand, while Munmorah exhibits a very high demand. This confirms that the gross value of oxygen demand cannot be the sole factor in determining the oxygen demand for a given oxy-fuel retrofit. The normalized oxygen demand appears to be a better and more accurate indicator as it considers both the plant capacity and fuel type. 3.1.2. CO2 Avoided (Captured). Figures 7 and 8 respectively present the gross and normalized values of CO2 avoided (i.e., captured) by each NSW coal-fired power plant if modified and retrofitted for oxy-firing operation. The gross values (Figure 7) are expressed in million tonnes of CO2 captured per year, while the normalized values are expressed in tonnes of CO2 avoided per unit of the original send out electricity (i.e., before retrofit). As Figure 7 shows, the CO2 avoided in gross terms vary from 0.2 to 13 Mt/yr for the NSW power plants under investigation.
(31)
where COA is cost of CO2 abatement (expressed as $/t captured CO2), CO2Cap is CO2 captured/avoided (t CO2/ MWh), and the subscript “ref” refers to business as usual case with no CCS.
3. RESULTS AND DISCUSSION 3.1. Plant Technical Assessment. 3.1.1. Oxygen Demand. Figure 5 shows the gross value of the oxygen demand (assuming 100% oxy-firing) for the eight coal-fired power plants under investigation. Given that the gross oxygen demand 2080
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
Figure 10. Bar chart plots of CPU power demand for coal-fired power plants in NSW.
Figure 7. Bar chart plots of gross CO2 avoided for coal-fired power plants in NSW.
depicted in Figure 5. Similarly, the CPU power demand for each plant has been determined assuming that the CO2 quantities shown in Figure 7 ought to be avoided (captured). Also, as noted earlier it is assumed in the present analysis that all oxy-fuel options utilize the same CPU design and hence differ only in terms of the ASU design. For that reason, Figure 10 shows only one option for the CPU component of each power plant, while the cryogenic, ICLAS[S] and ICLAS[FG] options are considered in Figure 9 for the air separation component of the retrofit. As shown in Figure 9 the difference in ASU power demands for cryogenic and ICLAS-based designs is quite significant for each power plant studied here. In fact, the ratios of ICLAS[S]/ cryogenic and ICLAS[FG]/cryogenic power demands for the ASU component are 0.53 and 0.24, respectively. Thus, on average the ICLAS[S] process leads to 47% reduction in the ASU power demand when compared with a cryogenic-based system, while for the same operating conditions ICLAS[FG] lowers the ASU power demand by as much as 76% if it replaces a cryogenic-based unit. Figure 11 illustrates the bar chart plots of the combined parasitic load (i.e., sum of ASU and CPU power demands) for
Figure 8. Bar chart plots of normalized CO2 avoided for coal-fired power plants in NSW.
The combined value of CO2 avoided for these power plants is about 47 Mt/yr which is equivalent to the annual GHG emissions from 9.5 million average cars (note that the total number of cars in Australia is about 13.5 million). Similarly the normalized values of the CO2 avoided vary between 0.78 and 0.88 t CO2/MWh for the power plants considered in this study (see Figure 8). The Bayswater and Eraring power plants are standout cases here because they have relatively modest values of normalized CO2 avoided (∼0.79 t CO2/MWh),, while in gross terms these two power plants can potentially achieve the highest levels of CO2 captured in the state if retrofitted for oxy-firing operation (Figure 7). 3.1.3. Parasitic Power Losses Due to Oxy-Firing Retrofit. Figures 9 and 10 show the power demands of the ASU and CPU components of the oxy-firing retrofit for the eight NSW coal-fired power plants. The ASU power demand for each plant has been calculated with the underlying assumption that the ASU is capable of delivering the relevant oxygen demand
Figure 11. Bar chart plots of combined parasitic load for coal-fired power plants in NSW.
the NSW coal-fired power plants. As shown, there are significant losses of output power when a conventional plant is converted to oxy-firing operation. These losses are clearly much higher when a cryogenic-based air separation unitis employed. For benchmarking, the combined parasitic load for each power plant was normalized against its typical output and the results were summarized in Figure 12 and reported as “plant capacity loss”. The formula used for normalization was PCL =
Figure 9. Bar chart plots of ASU power demand for coal-fired power plants in NSW. 2081
⎛ CPL ⎞ ⎜ ⎟ ⎝ TPO ⎠
(32) DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels where PCL is plant capacity loss (%), and CPL is combined parasitic load (MW).
Figure 14. Bar chart plots of thermal efficiency penalty for the NSW fleet of coal-fired power plants.
compared with the conventional cryogenic air separation option. The efficiency penalties shown in Figure 14 are primarily due to power demands of the ASU and CPU components of the oxy-fuel retrofit. The breakdown of the efficiency penalty in terms of its respective components is shown in Figure 15.
Figure 12. Bar chart plots of plant capacity loss for coal-fired power plants in NSW.
It can be seen from this figure that when cryogenic air separation is employed the capacity loss can be as high as 30% of the plant output, whereas the capacity loss can be progressively reduced to figures as low as 15% (i.e., 50% reduction in capacity loss) if an ICLAS[FG] type design is implemented. 3.1.4. Thermal Efficiency. Figure 13 depicts a comparison of power plant thermal efficiency for the four cases under
Figure 13. Bar chart plots of thermal efficiency for the NSW fleet of coal-fired power plants. Figure 15. Bar chart plots of thermal efficiency penalty for the NSW fleet of coal-fired power plants as a function of efficiency penalty due to ASU and CPU power demands.
investigation; that is the base case; oxy-fuel with cryogenic air separation; oxy-fuel with ICLAS with steam reduction (ICLAS[S]) and finally oxy-fuel with flue gas reduction (ICLAS[FG]). As shown all three oxy-fuel options lower the plant thermal efficiency when compared with the business as usual case (i.e., base case). However, both chemical looping-based air separation processes considered here and in particular the ICLAS[FG] consistently lead to lower levels of efficiency loss. This is more evident in Figure 14, where the efficiency penalty (rather than the actual thermal efficiency) associated with the three different air separation options has been shown for the oxy-fuel conversion of the NSW coal-fired power plants (assuming the same CPU for all cases). As can be seen, the average efficiency penalty due to the implementation of the cryogenic option in NSW is about 10 percentage point which is quite consistent with figures reported in a number of recent studies.1 The efficiency penalty of the power plant, however, can be significantly reduced by the introduction of chemical looping-based air separation processes. As evident from Figure 14, the average efficiency penalty associated with ICLAS[S] is about 7.5%, while that of the ICLAS[FG] is approximately 5%; representing a 50% reduction in efficiency losses when
As this figure indicates, in the case of the cryogenic air separation the efficiency penalty due to the ASU approximately represents 61% of the total efficiency penalty, whereas for ICLAS[S] and ICLAS[FG] the efficiency penalties associated with ASU respectively represent about 45% and 27% of the overall efficiency penalty. In other words, while for the cryogenic case the limiting unit in terms of the design of the retrofit plant is ASU, in the case of chemical looping-based air separation processes and in particular ICLAS[FG], it is the CPU which underpins the efficiency losses rather than the ASU. This, as shown in the proceeding sections, leads to much lower operating and maintenance costs and results in significantly higher economic gains as more electricity can be exported compared with that associated with a cryogenic air separation component. 3.1.5. Electricity Output Penalty. Electricity output penalty (EOP) is essentially another way of expressing the plant losses due to parasitic load. However, EOP is usually normalized by the amount of CO2 avoided and, as such, is a measure of the effectiveness of the CO2 capture technology. Figure 16 shows 2082
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
varies between $70 million to about $6 billion. The highest capital cost is associated with the Bayswater plant and the lowest one corresponds to Munmorah. The bar chart plots of the normalized capital cost (expressed in $/kW of send out electricity) shown in Figure 18 reveal that the average cost of oxy-firing retrofit with a cryogenic type air separator is approximately $3,900/kW, whereas the corresponding retrofit costs with ICLAS[S] and ICLAS[FG] options are $3100/kW and $2600/kW, respectively. Therefore, an ICLAS[FG] type design can potentially lead to 32% reduction in capital cost and as noted earlier to 50% reduction in efficiency penalties when compared with a cryogenic-based design. Interestingly, Bayswater which in gross terms represents the highest capital cost, in normalized terms features a close to average capital cost. The same trend by and large is exhibited by other large power plants such as Eraring and Wallerawang, while small plants such as Munmorah and Redbank post very high values for the normalized capital cost. This suggests that regardless of the air separation design, the oxy-firing retrofit is more economical and cost-effective if implemented in coal-fired power plants with relatively high plant capacity (i.e., high MW output). To gain a better insight into the impact of the type of ASU on the capital cost, the breakdown of capital cost items for the Bayswater power plant is shown in Figures 19−22. Figure 19 illustrates the pie chart plot of the capital cost items for a cryogenic-based design, while Figures 20 and 21 show similar breakdowns of the capital cost for the ICLAS[S] and ICLAS[FG] designs. As can be seen in these figures in the case of the cryogenic design the ASU capital cost represents 26% of the total cost and is as high as that corresponding to the boiler modifications. For this configuration, the capital cost of the CPU is as low as 9% of the overall cost. However, for an ICLAS[S] type design (see Figure 20) the capital cost of the ASU: (i) only constitutes 15% of the overall cost, (ii) is comparable with the capital cost of the CPU, and (iii) is only one-third of the capital cost associated with boiler modification. For an ICLAS[FG] type design (see Figure 21), the capital cost of the ASU reduces even further to levels as low as 7% of the overall capital cost. In comparison, the capital cost of the ICLAS[FG] component of the retrofit becomes one-fifth of the boiler modification cost, and, as such, for an ICLAS[FG]-type design, the capital cost of the ASU is no longer the limiting factor. This is better shown in Figure 22, where the relative importance of the ASU and CPU capital costs are compared with respect to the combined cost of these items. For a cryogenic design (Figure 22a), the ASU costs almost 3 times more than the CPU, whereas in the case of an ICLAS[S] design the relative cost of ASU to CPU becomes 1.5 (note: 60.4/39.6 − 1.5; see Figure 22b). This downward trend continues to the extent that for an ICLAS[FG] design the relative cost of ASU to CPU drops to 0.64 (note: 39/61 − 0.64; see Figure 22c). To put the capital cost estimates presented in this section into perspective, a benchmarking comparison was undertaken to compare the normalized capital cost for 22 different lowemission technologies (both renewable and fossil fuel-based) with the current results for oxy-firing retrofit of the NSW coalfired power assets with a configuration involving ICLAS[FG] and CCS. The majority of data for the above-mentioned lowemission technologies were adopted from ref 6. The results of
Figure 16. Bar chart plots of EOP for the NSW fleet of coal-fired power plants.
EOP for the eight coal-fired power plants that are being examined in this study. Yet again there is a clear difference between the results corresponding to the cryogenic and those related to chemical looping-based air separation with ICLAS[FG] showing the lowest EOP figures for each power plant. Interestingly, the results appear to be independent of the power plant showing little variation from plant to plant. For instance, in the case of cryogenic air separation the EOP varies between 342.8 and 344.6 kWh/t CO2 captured. Similarly, for the ICLAS[S] the EOP ranges from 245 to 246 kWh/t CO2 captured, while ICLAS[FG] varies over a much tighter range of EOP between 184 and 184.3 kWh/t CO2 captured. The corresponding averages for the cryogenic, ICLAS[S] and ICLAS[FG] are 344.0, 245.6, and 184.2 kWh/t CO2 captured, respectively. 3.2. Plant Economic Assessment. 3.2.1. Capital Cost. Figures 17 and 18 show the gross and normalized version of the
Figure 17. Bar chart plots of the gross capital cost for the retrofit of NSW fleet of coal-fired power plants.
Figure 18. Bar chart plots of the normalized capital cost for the retrofit of NSW fleet of coal-fired power plants.
capital cost for the oxy-firing retrofit of NSW coal-fired power plants. The data set used in the calculation of the capital cost was presented earlier in Table 3. As Figure 17 indicates, depending on the power plant specifications/features and the type of air separation technology selected, the gross capital cost 2083
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
Figure 19. Breakdown of capital cost items for the Bayswater power plant (cryogenic option).
Figure 20. Breakdown of capital cost items for the Bayswater power plant (ICLAS[S] option).
the benchmarking comparison are summarized in Figure 23. As can be seen, the capital cost for an ICLAS[FG]-based oxy-fuel retrofit in NSW, which ranges between $2300/kW and $3400/ kW) is much cheaper than many of the technology options shown in Figure 23 (except open and combined gas turbine cycles) and is quite comparable with the more cost-effective technologies such supercritical pulverized black coal, combined cycle gas turbine with CCS and wind (onshore). Notably, the capital cost estimates from this study are also much lower than the values presented in Figure 23 for oxy-fuel combustion of black coal. This is partly due to the lower capital and operating costs of the ICLAS[FG] (note previous studies only considered oxy-fuel with cryogenic ASU) and partly due to the fact that there is no need for costly flue gas desulphurization (FGD)
units in NSW, given the low sulfur content of the coal consumed in NSW. 3.2.2. Levelized Cost of Electricity. levelized cost of electricity (LCOE) is considered the most convenient measure of the overall competiveness of alternative power generation technologies. It represents the cost in real dollars (per kWh basis) of building and operating a power plant over an assumed life cycle. In simple terms, LCOE represents the minimum price of electricity at which a power plant generates enough revenue to pay for production cost and provide sufficient return to investors. Key inputs in calculating LCOE as noted in section 2.3, include costs such as capital costs, fixed and variable operations and maintenance costs, fuel cost, financing costs, and an assumed utilization rate for the power plant. 2084
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
Figure 21. Breakdown of capital cost items for the Bayswater power plant (ICLAS[FG] option).
Figure 22. Relative costs of ASU and CPU with respect to their combined capital cost. Figure 23. Benchmarking of the normalized capital cost for the NSW coal-fired power plants retrofitted for oxy-firing (with ICLAS[FG] and CCS) against the normalized capital cost for 22 alternative lowemission technologies.
The LCOE for the coal-fired power plants under investigation in this study are summarized in Figure 24. According to this figure the most expensive power plants as far as electricity generation is concerned are Munmorah and Redbank. The main reason for the observed trend is the small capacity of these plants and their relatively high values of normalized capital cost. In contrast, large capacity power plants such as Bayswater or Eraring post relatively modest LCOE (see Figure 24). It is evident from Figure 24 that yet again retrofit configurations with chemical looping-based air separators and in particular ICLAS[FG] outperform those retrofits which incorporate cryogenic air separation. Based on Figure 24, the average value of LCOE for a cryogenic-based oxy-fuel plant in
NSW is $118/MWh, whereas the average figures for the ICLAS[S] and ICLAS[FG] are $105/MWh and $95/MWh, respectively. Similar to the benchmarking exercise discussed in the previous section and to put LCOE estimates presented in this section into perspective, a benchmarking comparison was undertaken to compare the LCOE for 23 different low-emission technologies (both renewable and fossil fuel-based technologies). The results are presented in Figure 25, where most of the 2085
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels
Figure 24. Bar charts of LCOE for NSW coal-fired power plants retrofitted for oxy-firing operation.
Figure 26. Bar chart plot of rises in the cost of generating electricity (no carbon tax).
Figure 27. Bar chart plot of rises in the cost of generating electricity (carbon tax included).
without carbon tax (a fixed value of $23/t CO2 was assumed). Clearly, regardless of the ASU design or the carbon price, significant increases in the generation cost of electricity are to be expected if oxy-fuel technology deployed across NSW. Although under the scenario of a carbon tax regime (Figure 27), the expected increases in the generation cost are more modest. Also, by and large, ICLAS[FG] results in much smaller levels of cost increase and, as such, is the most attractive option. 3.2.3. Cost of Abatement. The cost of abatement (COA) refers to the cost associated with the voluntary or compulsory removal of CO2 from a generating plant. As described in section 2.3, COA is expressed is expressed in $/t CO2 captured. The COA for NSW coal-fired power plants if retrofitted for oxy-firing operation are shown in Figures 28 and 29 as a
Figure 25. Benchmarking of the LCOE for the NSW coal-fired power plants retrofitted for oxy-firing (with ICLAS[FG] and CCS) against the LCOE for 22 alternative low-emission technologies6
data were adopted from ref 6. In this figure the results designated as “this study” correspond to oxy-fuel retrofits with ICLAS[FG] and CCS only. As can be seen, the LCOE for an ICLAS[FG]-based oxy-fuel retrofit in NSW, which is about $100/MWh, is much cheaper than many of the technology options shown in Figure 23 and is quite comparable with the more cost-effective technologies such as supercritical pulverized fuel (black coal) and combined cycle gas turbine. Notably, the LCOE from this study is also much lower than the values presented in Figure 23 for oxy-fuel combustion of black coal. This is partly due to the lower capital and operating costs of the ICLAS[FG] and partly due to the fact that there is no need for costly FGD units in NSW, given the low sulfur content of the coal used in NSW. It should be highlighted that the average cost of electricity generation in NSW is currently about $38/MWh2. Therefore, the rollout of oxy-fuel retrofits in the State will inevitably lead to higher generation cost. To understand the level and magnitude of such cost blowouts, the increase in the cost of electricity (Δ Cost of Electricity) due to oxy-fuel retrofit has been examined for the eight power plants under investigation. The results are summarized in Figures 26 and 27, where the increase in the production cost has been examined with and
Figure 28. Bar chart plot of cost of abatement for NSW coal-fired power plants (no carbon tax).
function of ASU design under two scenarios of “no carbon tax” and “with carbon tax”. Under the “no carbon tax” scenario the cost of abatement ranges between $61/t CO2 and $85/t CO2 with averages of 81, 74, and 67 $/t CO2 for the cryogenic, ICLAS[S] and ICLAS[FG] designs, respectively. However, when carbon tax is included the cost of abatement drops to a range between $37/t CO2 and $64/t CO2 with averages of 60, 2086
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Energy & Fuels
■
Article
AUTHOR INFORMATION
Corresponding Author
*Phone: +61-2-4044-9062. Fax: +61-2-4033-9383. E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge financial support from the University of Newcastle, the NSW Coal Innovation, and Glencore (formerly Xstrata) for this project.
Figure 29. Bar chart plot of cost of abatement for NSW coal-fired power plants (carbon tax included).
■
51, and 43 $/t CO2 for the cryogenic, ICLAS[S], and ICLAS[FG] designs, respectively.
(1) Wall, T.; Liu, Y.; Spero, C.; Elliott, L.; Khare, S.; Rathnam, R.; Zeenathal, F.; Moghtaderi, B.; Buhre, B.; Sheng, C.; Gupta, R.; Yamada, T.; Makino, K.; Yu, J. An overview on oxyfuel coal combustion: State of the art research and technology development. Chem. Eng. Res. Des. 2009, 87, 1003−1016. (2) EPRI Australian electricity generation technology costsReference case 2010, Electric Power Research Institute, 2010. (3) Pulverized coal oxycombustion power plants, DOE/NETL-2007/ 1291, National Energy Technology Laboratory, 2007. (4) Andersson, K.; Johnsson, F. Process evaluation of an 865 MWe lignite fired O2/CO2 power plant. Energy Convers. Manage 2006, 47 (20), 3487−3498. (5) Retrofitting CO2 Capture to Existing Power Plants, IEAGHG Report 2011/02, 2011. (6) Australian Energy-Technology Assessment 2012 Report, Australian Government, Bureau of Resources and Energy Economics, 2012. (7) Shah, K.; Elliott, L.; Spörl, R.; Belo, L.; Wall, T. Coal quality impacts and gas quality control in oxy-fuel technology for carbon capture and storagecost impacts and coal value, a milestone report provided to Glencore as part of a research contract with the Xstrata Coal Low Emissions Research, 2014. (8) Rubin, E. S.; Rao, A. B.; Berkenpas, M. B. Development and Application of Optimal Design Capability for Coal Gasification Systems; Technical Documentation: Oxygen-based Combustion Systems (Oxyfuels) with Carbon Capture and Storage (CCS). Final Report to U.S. Department of Energy, DE-AC21-92MC29094, May 2007. (9) Moghtaderi, B. Effects of particle cloud extinction on synthesis gas reduction of metal oxides in chemical looping reactors. Fuel 2012, 99, 262−270. (10) Moghtaderi, B.; Doroodchi, E. Performance characteristics of a miniaturised chemical looping steam reformer for hydrogen enrichment of fuels. Int. J. Hydrogen Energy 2012, 37 (20), 15164−15169. (11) Shah, K.; Moghtaderi, B.; Wall, T. Effect of flue gas impurities on the performance of a chemical looping based air separation process for oxy-fuel combustion. Fuel 2013, 103, 932−942. (12) Shah, K.; Moghtaderi, B.; Zanganeh, J.; Wall, T. Integration options for novel chemical looping air separation (ICLAS) process for oxygen production in oxy-fuel coal fired power plants. Fuel 2013, 107, 356−370. (13) Moghtaderi, B.; Song, H. Reduction properties of physically mixed metallic oxide oxygen carriers in chemical looping combustion. Energy Fuels 2010, 24 (10), 5359−5368. (14) Song, H.; Doroodchi, E.; Moghtaderi, B. Redox characteristics of Fe-Ni/SiO 2 bimetallic oxygen carriers in CO under conditions pertinent to chemical looping combustion. Energy Fuels 2012, 26 (1), 75−84. (15) Song, H.; Shah, K.; Doroodchi, E.; Moghtaderi, B. Development of a Cu-Mg-based oxygen carrier with SiO2 as a support for chemical looping air separation. Energy Fuels 2014, 28 (1), 163−172. (16) Song, H.; Shah, K.; Doroodchi, E.; Wall, T.; Moghtaderi, B. Reactivity of Al2O3- or SiO2-Supported Cu-, Mn-, and Co-based
4. CONCLUSIONS A techno-economic assessment was carried out to evaluate the technical and economic performances of the NSW fleet of coalfired power plants converting to oxy-fuel operations. The assessment was performed using a spreadsheet-based analysis toolkit developed in-house specifically for this study. The results indicate that the normalized oxygen demand for coalfired power plants in NSW varies between 450 and 550 m3/ MWh, while their normalized CO2 emission avoided varies between 0.78 and 0.88 tonne per MWh. On average, the ICLAS[FG] was found to be the best ASU option which consistently leads to lower levels of efficiency loss, capital cost, and electricity cost. Specifically, the capital cost (or retrofit cost) associated with the cryogenic-based ASU, ICLAS[S]based ASU, and ICLAS[FG]-based ASU was found to be about $3900/kW, $3100/kW, and $2600/kW, respectively. In contrast, the cost of electricity associated with the cryogenicbased ASU, ICLAS[S], and ICLAS[FG] cases in NSW was found to be about $118/MWh, $105/MWh, and $95/MWh, respectively. The following key conclusions can thus be drawn on the basis of the above results of the techno-economic analyses: • Broadly speaking and subject to the simplifying assumptions made in the present study, oxy-fuel conversion of the coal-fired power generation assets in NSW appears to be one of the most viable low-emission options for the State, provided that a chemical loopingbased design for air separation is employed. • This, however, requires a detailed feasibility study for each power plant and, more importantly, further studies to demonstrate the effectiveness of the chemical loopingbased air separation at scale. • Also, a range of suitable options ought to be identified and/or established for the storage and sequestration of CO2 in NSW. Without a viable storage option, the rollout of oxy-fuel technology across NSW will not reach its full potential and will unlikely lead to any tangible outcome for the State or companies involved.
■
REFERENCES
ASSOCIATED CONTENT
S Supporting Information *
Sample calculations for the oxy-fuel retrofit reference case and all coal-fired power plants studied; data generated by the techno-economic model for the figures presented in this paper. This material is available free of charge via the Internet at http://pubs.acs.org. 2087
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
Article
Energy & Fuels oxygen carriers for chemical looping air separation. Energy Fuels 2014, 28 (2), 1284−1294. (17) Song, H.; Shah, K.; Doroodchi, E.; Wall, T.; Moghtaderi, B. Analysis on chemical reaction kinetics of CuO/SiO2 oxygen carriers for chemical looping air separation. Energy Fuels 2014, 28 (1), 173− 182. (18) Zhang, Y.; Doroodchi, E.; Moghtaderi, B. Thermodynamic assessment of a novel concept for integrated gasification chemical looping combustion of solid fuels. Energy Fuels 2012, 26 (1), 287−295. (19) Zhang, Y.; Doroodchi, E.; Moghtaderi, B. Chemical looping combustion of ultra low concentration of methane with Fe2O3/Al2O3 and CuO/SiO2. Appl. Energy 2014, 113, 1916−1923. (20) Zhang, Y.; Doroodchi, E.; Moghtaderi, B. Utilization of ventilation air methane as an oxidizing agent in chemical looping combustion. Energy Convers. Manage. 2014, 85, 839−847. (21) Zhou, C.; Doroodchi, E.; Moghtaderi, B. An in-depth assessment of hybrid solar−geothermal power generation. Energy Convers. Manage. 2013, 74, 88−101. (22) Zhou, C.; Doroodchi, E.; Moghtaderi, B. Assessment of geothermal assisted coal-fired power generation using an Australian case study. Energy Convers. Manage. 2014, 82, 283−300. (23) A Life Cycle Assessment of the NSW Electricity Grid, CCSD Technology Assessment Report 58, 2003. (24) Moghtaderi, B. Application of chemical looping concept for air separation at high temperatures. Energy Fuels 2009, 24 (1), 190−198. (25) Perry, R. H.; Green, D. W. Perry’s Chemical Engineers’ Handbook, 7th ed.; McGraw-Hill, 1997. (26) Toward A Common Method Of Cost Estimation For CO2 Capture And Storage At Fossil Fuel Power Plants; Global CCS Institute, 2013.
2088
DOI: 10.1021/ef5022076 Energy Fuels 2015, 29, 2074−2088
