Techno-economic evaluation of oxy-combustion coal-fired power plants

Increasing attention is being paid to the oxy-combustion technique of coal-fired power plants because CO2 produced from fossil fuel combustion can be captured and sequestrated by it. However, there are many questions about the economic properties of the oxy-combustion technique. In this paper, a detailed techno-economic evaluation study was performed on three typical power plants (2 × 300 MW subcritical, 2 × 600 MW supercritical, 2 × 1000 MW ultra supercritical), as conventional air fired and oxycombustion options in China, by utilizing the authoritative data published in 2010 for the design of coal-fired power plants. Techno-economic evaluation models were set up and costs of electricity generation, CO2 avoidance costs as well as CO2 capture costs, were calculated. Moreover, the effects of CO2 tax and CO2 sale price on the economic characteristics of oxy-combustion power plants were also considered. Finally, a sensitivity analysis for parameters such as coal sample, coal price, air separation unit price, flue gas treatment unit price, CO2 capture efficiency, as well as the air excess factor was conducted. The results revealed that: (1) because the oxy-combustion technique has advantages in thermal efficiency, desulfurization efficiency and denitration efficiency, oxy-combustion power plants will reach the economic properties of conventional air fired power plants if, (a) the CO2 emission is taxed and the high purity CO2 product can be sold, or (b) there are some policy preferences in financing and coal price for oxy-combustion power plants, or (c) the power consumption and cost of air separation units and flue gas treatment units can be reduced; (2) from subcritical plants to supercritical and finally ultra-supercritical plants, the economics are improving, regardless of whether they are conventional air fired power plants or oxy-combustion power plants.

As a branch of practical economics, techno-economics is widely used to research the economic benefits of technology application activities, achieve the best combination of technology and economy, seek ways to enhance economy benefit, and provide a decision basis for investment decision makers. Because coal-fired power plants are technologyintensive and capital-intensive processes, a techno-economic evaluation is particularly important. Many techno-economic evaluation studies have been conducted on the desulfurization (De-SO x ) and denitration (De-NO x ) processes in conventional coal-fired power plants. CO 2 emission control has become a global issue [1], and actions to minimize emis-sions are a priority [2]. At present, CO 2 capture and sequestration from power plants is a feasible and effective choice. And as CO 2 emission control technologies, such as oxycombustion technology, integrated gasification combined cycle (IGCC) technology, monoethanolamine (MEA) and MEA/MDEA (activated methyldiethanolamine) scrubbing technology, have reached the commercialization phase, greater attention has been paid to the economic costs of these new technologies. Techno-economic analysis of the emission control technologies is one of the key problems that must be solved. Oxy-combustion is a new technology that adds a cryogenic air separation process (ASU) and a flue gas clean and purification process (CPU) to a conventional combustion process. High purity oxygen product from the ASU, instead of air, is used in the oxy-combustion process, and about 70%-80% of the flue gas is recycled into the furnace, keeping the combustion temperature inside the furnace within the conventional range. A schematic diagram of the oxy-combustion technology is shown in Figure 1. Because there is no nitrogen dilution, the CO 2 concentration in the oxy-combustion flue gas is high, and a high purity CO 2 product (95%-99%) can be obtained through purification, compression and separation. Moreover, efficient De-SO x and De-NO x can be achieved in such a system and consequently oxy-combustion has become one of the most competitive coal combustion technologies of this century. At present, oxy-combustion technology has reached the demonstration stage in many countries, and there were eight demonstration power station projects operating worldwide in 2008-2010. In this paper, techno-economic evaluations of oxy-combustion and also conventional coal-fired power plants are performed. The results of these two evaluations are compared and presented. In conventional coal-fired power plants, coal is combusted with air in the furnace and the flue gas containing about 15 mol% CO 2 is emitted directly into the atmosphere. IHI in Japan [3], Chalmers University of Technology in Sweden [4], ALSTOM in America [5], Argonne National Laboratory in America [6], CANMET in Canada [7] and EDF in France [8] have all carried out techno-economic evaluations of the oxy-combustion technology. The results of IHI [3] show that the efficiency of the oxy-combustion power plant (1000 MW) decreases 10.5%; the results from Chalmers University of Technology [4] show that the efficiency of the oxy-combustion power plant (865 MW) decreases 9.1%, the CO 2 avoidance cost is $26/t and the cost of electricity is $64.3/kW; the results of ALSTOM [5] show that the CO 2 avoidance cost of the oxy-combustion power plant (450 MW) is $42/t and the unit investment cost is $823/kW; the results of Argonne National Laboratory [6] show that the CO 2 avoidance cost is $34/t; the results of CANMET [7] show that the CO 2 avoidance cost of the oxy-combustion power plant (400 MW) is $35/t, the cost of electricity increases 20%-30% and the unit investment cost is $791/kW; the results of EDF [8] show that the efficiency of the oxycombustion power plant (1200 MW) decreases 10%, the investment cost increases 69%, the cost of electricity increases 48% and the CO 2 avoidance cost of the oxy-combustion system is 29% lower than that of the MEA scrubbing system. These results can be summarized as: if conventional coal-fired power plants are retrofitted to be oxy-combustion power plants, the net power output will decrease by about 25%, the cost of electricity will increase by 30%-50%, the CO 2 avoidance cost is about $30/t and about 85% CO 2 can be captured. However, the techno-economic characteristics of CO 2 emission control systems are complicated. They depend on the energy efficiency of the system, technology maturity level, pollutants (including SO x , NO x , PM10 and CO 2 ) emission policies in the country or the local region, and even financial policies (such as the loan interest rate and inflation rate). Since there are large differences among the evaluating system sizes and combustion conditions from various academic institutions, and the tax policies and financial policies between Western countries are usually adopted from country-specific data, the published research results are not transferable to the Chinese situation. Therefore, to provide the basis of policy decisions, it is very important to perform techno-economic evaluations for different CO 2 emission control systems based on Chinese conditions and data, for energy and power systems, by comparing various electricity costs, CO 2 avoidance costs and CO 2 capture costs for these CO 2 emission control systems.
The authors have previously performed a techno-economic evaluation of oxy-combustion coal-fired power plants retrofitted from conventional coal-fired power plants, by using a thermo-economic cost model [9] and practical investigation data [10]. However, some internal cost items (such as depreciation cost, amortization expense, material cost, personnel wages and other expenses) were ignored in the previous models. Cost models for De-SO x and De-NO x technologies were very simple, and also detailed comparisons among several typical coal-fired power plants were not carried out. In this paper, a more systematic and comprehensive techno-economic evaluation of the oxy-combustion technology was thus conducted. Each factor during the electricity cost formation and detailed investment and operating costs of De-SO x and De-NO x devices, was considered. Moreover, three typical coal-fired power plants (2 × 300 MW subcritical, 2 × 600 MW supercritical and 2 × 1000 MW ultra-supercritical) in China were chosen to calculate the electricity costs in oxy-combustion power plants and conventional power plants, and CO 2 avoidance costs and CO 2 capture costs in oxy-combustion power plants. The effects of a CO 2 tax, and CO 2 sale price, on the cost results are also discussed. Finally, a sensitivity analysis of some important parameters in oxy-combustion systems, such as the coal price, ASU cost, CPU cost and CO 2 capture efficiency, were performed to study their influences on the economics of the oxy-combustion technology.

Basic methods
Because there are no demonstration or commercially operated oxy-combustion coal-fired power plants larger than 30 MW, the techno-economic evaluation of an oxy-combustion plant was performed based on its corresponding conventional coal-fired power plant. Keeping the gross power outputs of the oxy-combustion plant and its corresponding conventional plant equivalent, the differences in the oxy-combustion plant from the conventional plant mainly lie in: retrofitting the burner, heat exchange surface and flue gas recycle in the boiler island; an ASU and a CPU are added. Consequently, the techno-economic evaluation process of an oxy-combustion plant is as follows: (1) Collect basic thermodynamic parameters (such as coal consumption rate, power generation load, and boiler efficiency), operational conditions (such as annual operation hours, maintenance factor, amortization rate, depreciation rate, and personnel wages), and investment and operational costs of De-SO x and De-NO x devices, in the conventional plant system, that can be obtained from a system process simulation, or investigation. In this paper, data were adopted mainly from the book "Reference cost indexes in quota design for coal-fired projects (2009 levels)" [11] published by the China Power Engineering Consulting Group Corporation in 2010. The boiler retrofit cost, investment cost and power consumption of CPU could be estimated and adjusted by referring to published papers [12,13]. The investment cost and power consumption of ASU can be obtained from oxygen production companies and by simulating the ASU system.
(2) Generally, commercial loans exist for the construc-tion of a power plant, so it is necessary to know the market economy policies, such as interest rate, fuel price, water price, steam price, limestone price and gypsum price.
(3) From the data mentioned above, each basic cost item (such as fuel cost and investment cost) relating to the oxycombustion and conventional plants can be calculated. Then the CO 2 avoidance costs and CO 2 capture costs of the oxycombustion plants can be further calculated. Finally, a sensitivity analysis may be performed.

Cost calculation for power plants
The total cost of a power plant includes the power generation cost, period cost, and by-products revenue (C 10 ). The power generation cost includes fuel cost (C 1 ), operation and maintenance (O&M) cost (C 3 ), depreciation cost (C 4 ), amortization cost (C 5 ), pollutants' emission tax (C 6 ), personnel wages (C 7 ), material cost (C 8 ) and other costs (C 9 ). The period cost includes a management expense and financial expense (including loan interest (C 2 )). Because the management expense and financial expense involve complicated financial accounting theory and industry rules, only some "hard" costs (annualized cost C T ) were considered in this paper, which can be described as (i) Cost calculation for conventional power plants. Conventional power plant costs can be calculated as follows: (1) Fuel cost in which, m F,0 is the unit standard coal consumption rate for power generation (315, 299 and 275 g/(kW h) for the subcritical, supercritical and ultra-supercritical power plant, respectively in this paper) [11], c F is the unit standard coal price (680 ¥/t with tax [11], ¥ is the symbol of Chinese Yuan (CNY (2) Loan interest cost in which, C IT,0 is the total investment cost of the conventional power plant and C IT,0 = C IT,base,0 + C IT,S,0 + C IT,N,0 . The C IT,base,0 for the three kinds of power plant (excluding De-SO x and De-NO x devices) can be estimated by using 4412, 3675 and 3591 ¥/kW [11]. The device costs of the De-SO x devices (considering the wet flue gas desulfurization (FGD) technology with a 95% desulfurization efficiency ( S,0 ) ) in the three plants are 111.43, 185.45, 247.09 M¥, respectively [11]. The device costs of the denitration devices (considering the selective catalytic reduction (SCR) denitration technology with a 80% denitration efficiency ( N,0 ) ) in the three plants are 72.99, 108 and 140 M¥, respectively [11]. In addition, the costs of De-SO x and De-NO x devices are set to be 80% of their investment costs (C IT,S,0 and C IT,N,0 ) [14,15] and other costs, such as construction, installation and technical service, account for the remaining 20%; p loan is the loan percentage (80% [11]), and the "average capital method" was chosen to payback the load, the average interest rate can be calculated by   = i × (1 + 1/P)/2, P is the loan period (15, 18 and 18 years, respectively [11]), i is the loan interest rate for a period longer than 5 years (5.94% [11]).
(3) Operation and maintenance cost in which, p OM,base,0 is the O&M coefficient (2.5% [7], including the major maintenance expense) for the conventional power plants (excluding De-SO x and De-NO x devices); C OM,S,0 is the O&M cost for the FGD device, including limestone expense (C OMS0,1 ), process water expense (C OMS0,2 ), effluent processing expense (C OMS0, 3 ) and equipment maintenance expense (C OMS0, 4 ). Personnel wages, depreciation cost, amortization cost and electricity consumption cost for the De-SO x device and the following De-NO x device are considered from the viewpoint of the whole power plant. Furthermore, C OMS0,1 = c CaCO 3 × S ar × M F,0 × H n /H i × W × H × 100/32 × r Ca 2 S /P CaCO 3 , in which, H n is the lower heating value of the standard coal, 29270 kJ/kg, c CaCO 3 is the unit price of limestone (60 ¥/t [11]), r CaCO 3 is the mole ratio of Ca to S (1.03 [14]), P CaCO 3 is the purity of limestone (90% [14]); C OMS0,2 =c pw ×M pw,0 ×H, where c pw is the unit price of process water (1.54¥/t [14]), M pw,0 is the process water consumption rate (10 t/h [14] for the 2 × 300 MW power plant); C OMS0,3 = c ef × M ef,0 × H, where c ef is the unit effluent processing cost (1.6 ¥/t [14]), M ef,0 is the effluent discharge rate (120 t/h [14] for the 2 × 300 MW power plant); C OMS0,4 = C IT,S,0 × p OM,S,0 , p OM,S,0 is the O&M coefficient (1.5% [14], including the major maintenance expense) for the De-SO x device. And for the 2 × 600 MW supercritical and 2 × 1000 MW ultra-supercritical power plants, the C OMS0,2 and C OMS0,3 are proportional to the limestone consumption rate in each power plant, respectively. C OM,S,0 is the O&M cost for the SCR device, including ammonia expense, catalyst expense, steam expense and equipment maintenance expense [15,16]. Adjusted for the annual operation hours, the ammonia expense, catalyst expense and steam expense for the 2 × 300 MW power plant considered in this paper are 4.62, 13.34 and 0.11 M¥/y, respectively [16]. The corresponding data for the 2 × 600 MW power plant are 9.15, 26.43 and 0.22 M¥/y, respectively [15]. However, because 2 × 1000 MW ultra-supercritical power plants with SCR devices are very limited in China, data for this size of SCR device is very difficult to obtain. In this paper, the corresponding data for the 2 × 1000 MW power plant (14,40.42 and 0.35 M¥/y, respectively) were proportional to those of the 2 × 600 MW power plant. The O&M coefficient of the SCR device used in this paper is 1.5%.
in which, p fa is the fixed assets formation percentage (95% [11]), p lv is the residual value percentage (5% [11]) and the Y d is the depreciation period (15 years).
in which p ia is the percentage of intangible and deferred assets (5%) [17] and Y a is the amortization period (5 years). (6) Pollutants emission tax in which E S,0 is the SO 2 emission amount in the conventional power plant, which can be estimated by referring to [18]. to that of N element [18],  N,0 is the transforming rate (25% [18]) of fuel N, m n,0 is the percentage of NO x coming from fuel N to total NO x (80% [18]), T S and T N are the unit pollutant emission tax (0.6 ¥/0.95 kg) for SO 2 and NO x , respectively. In addition, pollutant emission taxes for CO and particles were not considered in this paper and tax differences from different regions and environment functions were also not considered. If the emission tax of CO 2 is considered, then eq. (8) should be modified to be , and T CO 2 is the unit CO 2 emission tax (¥/t), t C is the ratio of C ar transformed to be CO 2 after coal combustion (usually 100%),  C,0 is the CO 2 capture ratio (for conventional plants,  C,0 =0; and for oxy-combustion plants,  C,1 =90%).
in which, N base,0 , N S,0 , N N,0 are personnel numbers for the base power plant, the FGD system and the SCR system, respectively. For the three kinds of plant, N base,0 is 234, 247 and 300 [11], respectively; N S,0 is 15, 18, 21 (three groups, and each of 5, 6 and 7 persons), respectively; N N,0 is 15, 18, 21 (three groups, and each of 5, 6 and 7 persons), respectively. c pay is the annual wage for each person (50000 ¥/y), and r w is the welfare and labor insurance coefficient (60% [11]).
(ii) Cost calculation for oxy-combustion power plants. We can calculate the C T in oxy-combustion plants similarly to that of the conventional plants, and the differences lie in the boiler retrofit, ASU and CPU additions. Also, the De-SO x and De-NO x devices can be simplified significantly in the oxy-combustion plants. Because of the N 2 -lean combustion environment and flue gas recycle, a lower cost De-SO x technology (such as limestone injection into the furnace and the activation of unreacted calcium, LIFAC) could be adopted to reach a satisfactory De-SO x result. In addition, SO x in the flue gas can also be removed in the CPU, thus a total 95% De-SO x efficiency was used in this paper. On the other hand, because of the N 2 -lean environment, it can be considered that there is only fuel NO x generated (viz. m n,1 = 100%) and at the same time, the flue gas recycle, low air excess factor (tiny positive pressure combustion, air excess factor  1 = 1.05) and adopting low NO x air staging burners can effectively suppress the fuel NO x generation (considering the fuel N transforming efficiency  n,1 is 15%). Also, NO x in the flue gas can be co-removed in the CPU (assuming the De-NO x efficiency  N,1 = 30%), so an additional SCR is not needed. In general, costs for the oxy-combustion plants can be calculated as follows: (1) Because the flue gas recycle can effectively reduce the heat loss from the flue gas, the efficiency increase ratio  e =  b /( b + 0.02) is applicable, and this reduces coal consumption. The unit standard coal consumption rate in the oxy-combustion plant is m F,1 = m F,0 ×  e , and its fuel cost C 1,1 = C 1,0 ×  e . The boiler efficiencies ( b ) for the three kinds of plant are set to be 92%, 94% and 95%, respectively.
(2) The total investment cost (C IT,1 ) for oxy-combustion plants can be calculated as IT,1 IT,base,0 I,bioler,0 IT,S,0 ASU IT,base,0 7% 3 2.5%, in which, the second item on the right side of the equation is the boiler retrofit cost, which can be estimated to be 7% [12] of the boiler cost (C I,bioler,0 ), and the C I,bioler,0 for the three sizes of boilers are 652.75, 1299.9 and 2800 M¥ [11], respectively; the third item on the right side is the cost of the LIFAC De-SO x device, which is assumed to be 1/3 of that of the FGD; while the fourth item is the cost of the ASU. According to the investigation data from some oxygen production companies (such as Hangzhou Oxygen Production and the Sichuan Air Separation), the investment cost of large-scale oxygen production machines (60000 N m 3 /h) satisfying the oxygen concentration demand of oxy-combustion technology is 120 M¥, and the actual oxygen con- Therefore, the C ASU = V O,1 /60000 × 120 M¥; and the fifth item on the right side is the cost of the CPU, which is about 2.5% of the total investment cost of the whole base power plant [13]. Similar to that of the base plant, and the loan interest cost, depreciation cost and amortization cost can be calculated based on the C IT,1 .
in which, p OM,base,1 is the O&M coefficient of the oxycombustion base plant (also 2.5%, including the major maintenance expense); the O&M cost of the De-SO x device (LIFAC) is set to be 1/3 of that of FGD; p OM,ASU is the O&M coefficient of ASU (1.5%) and the p OM,CPU is the O&M coefficient of CPU (1.5%).
(4) Each pollutant emission amount and corresponding emission tax can be estimated by using methods introduced for conventional power plants. (5) The personnel wages for an oxy-combustion base plant (including LIFAC) are considered to be equivalent to that of the conventional plant. (6) The material cost ratio and other cost ratios in oxycombustion plants are equivalent to that of conventional plants. (7) There is no gypsum revenue in oxy-combustion plants, but the high purity CO 2 may be considered as a product. So in that case, the by-products revenue could be C 10,1 = M CO 2 × c CO 2 , in which CO 2 capture amount M CO 2 = C ar × m F,1 × H n /H i × H × W ×  C × 44/12, and c CO 2 is the unit price of CO 2 product.

Cost of electricity
The cost of electricity (c COE ) for coal-fired power plants can be calculated as in which, W net is the net power output. For conventional power plants, W net,0 = W × (1r pe,0 )W S,0 W N,0 , r pe,0 is the auxiliary power ratio (5.5%, 5.2% and 4.5% [11] for the three sizes of plant, respectively), W S,0 is the power consumption of the De-SO x device (1.5%, 1.1% and 0.7% [11] of the total load, respectively), W N,0 is the power consumption of the De-NO x device (1.3, 1.6 and 2.0 MW [15,16], respectively). For oxy-combustion power plants, W net,1 = W × (1r pe,1 )W S,1 W ASU W CPU , r pe,1 is equivalent to r pe,0 , the power consumption of the De-SO x device is W S,1 = W S,0 /3, the power consumption of ASU is W ASU = V O,1 /60000 × 21 MW (the power consumption of the 60000 Nm 3 /h ASU is 21 MW) and the power consumption of CPU, W CPU , is estimated to be 8% [13] of the gross power output.
The c COE values of the conventional (four cases: without De-SO x or De-NO x device; with De-SO x device; with De-NO x device; with De-SO x and De-NO x devices) and oxycombustion plants (two cases: with LIFAC and without De-SO x device, the CO 2 tax and the CO 2 sale price are not considered) under the three different loads are listed in Table 2. Figure 2 gives a comparison of the c COE in different cases.
The results in Table 2 and Figure 2 show that (the descriptions in the following paragraph all correspond to the 2 × 300 MW subcritical, 2 × 600 MW supercritical and 2 × 1000 MW ultra-supercritical plants sequentially): (   for about 12%-14% [11] of the total c COE . If these effects are considered, the c COE of conventional power plants are approximately the same according to the results presented in [11], which indicates that the techno-economic analysis performed in this paper is in reasonable agreement. (2) The static investment cost increases by 8.7%, 8.32% and 5.88% if the De-SO x and De-NO x devices are added in the conventional power plants; in comparison to the conventional power plants with De-SO x and De-NO x devices, the static investment costs for oxy-combustion plants (with LIFAC) increase by 19.45%, 23.28% and 24.68%, respectively. From the subcritical to the supercritical and finally the ultra-supercritical, the material upgrade and some special imported parts make the boiler cost increase rapidly.
(3) Even if the De-SO x and De-NO x devices are not included in the oxy-combustion power plants, a low SO x and NO x emission level can still be achieved. However, if the LIFAC system is installed, the static investment costs of the oxy-combustion plants increase by only about 1%, the annualized total costs remain nearly unchanged, power outputs decrease about 0.5% and c COE increases no more than 1%, and a De-SO x efficiency similar to the FGD technology can be realized.
(4) The static investment costs for oxy-combustion plants increase mainly because of the high commercial price of ASU, and the investment in the CPU system. Further developments to the oxygen production technology and increasing the scale of the ASU market should decrease the costs of ASU systems significantly, and then the economic characteristics of the oxy-combustion technology will improve significantly.
(5) In comparison to the conventional power plants with De-SO x and De-NO x devices, the annualized total costs for oxy-combustion plants (with LIFAC) increase by 1.51%, 1.95% and 2.55%, respectively. The increases are slight because the De-SO x and De-NO x devices with high O&M costs are removed and coal consumption decreases because of the enhanced boiler efficiency in oxy-combustion plants. However, the net power outputs for oxy-combustion plants decrease substantially in comparison to conventional plants because of the high power consumptions of ASU and CPU systems, which also increase the c COE of oxy-combustion plants substantially. Therefore, developing low cost and low power consumption ASU and CPU systems is the key to enhance the economic characteristics of the oxy-combustion technology. The components and corresponding proportions of annualized total costs for three different load plants under conventional combustion and oxy-combustion are shown in Figure 3. The results show that fuel costs, the depreciation and amortization costs affect the distributions of the annualized total costs remarkably. Because the unit investment costs of base plants reduce sequentially from the subcritical plants to the supercritical plants and finally the ultra-supercritical plants, although the unit coal consumptions also reduce sequentially, the ratios of fuel costs increase sequentially, and are 64%, 67% and 68%, respectively. Because the ASU and CPU systems are added in oxy-combustion plants, the ratios of investment costs and O&M costs increase, accordingly, but the ratios of fuel costs reduce 2%-3%. Also, it is worth emphasizing, the ratios of De-SO x and De-NO x costs in oxy-combustion plants decrease greatly, and become almost negligible.

CO 2 avoidance cost
Oxy-combustion technology has been considered to control the CO 2 emission from fossil fuel combustion, and this is the reason why so much attention has been paid to it. The CO 2 avoidance cost (c CAC ) can be used to evaluate the economic property of controlling the CO 2 emission. c CAC is defined as the ratio of the c COE difference to the unit CO 2 emission amounts difference between the CO 2 emission control system (oxy-combustion plant with LIFAC in this paper) and the corresponding CO 2 emission non-control system (conventional plant with De-SO x and De-NO x devices in this paper). It means the additional economic cost of avoiding one ton CO 2 emission, which can be described as 2 in which, e CO 2 is the CO 2 emission amount per unit of power (t/MWh). The c CAC of oxy-combustion plants (with LIFAC) for three different loads are given in Table 3.
Large amounts of CO 2 emission can be reduced in oxy-combustion plants, producing an environmental benefit. Some countries have already begun to tax the CO 2 emission  (19) Figure 4 shows the effect of the unit CO 2 emission tax (T CO 2 ) on the c COE of conventional and oxy-combustion plants and the results show that the oxy-combustion technology could be competitive with the conventional mode if the CO 2 emission is taxed at 140-170 ¥/t. When the T CO 2 equals the c CAC without CO 2 emission taxation, the c COE of the oxy-combustion plant is equivalent to that of the corresponding conventional plant. The c CAC calculation relates to the CO 2 emission reduction (the emission difference between the two plants), and the total tax cost difference of the two plants is also related to the CO 2 emission reduction. This makes the T CO 2 value when the oxy-combustion plant and the corresponding conventional plant have equivalent economic property (named as critical T CO 2 ) is equal to the c CAC without CO 2 emission taxation (see equation (19) and Figure 4).

CO 2 capture cost
Another parameter required to evaluate the economic property of the oxy-combustion technology is the CO 2 capture cost (c CCC ). c CCC is defined as the ratio of the c COE difference to the unit CO 2 capture amounts difference between the CO 2 emission control system and the corresponding CO 2 emission non-control system. It means the additional economic cost of capturing one ton CO 2 , can be described as 2 in which, m CO 2 is the CO 2 capture amount per unit of power (t/(MW h)), r CO 2 is the CO 2 capture efficiency. The c CCC of oxy-combustion plants (with LIFAC) for three different loads are also given in Table 3. The high purity CO 2 captured from oxy-combustion plants can be used in enhancing oil recovery (EOR), carbon Figure 4 Relations between c COE and T CO 2 . fertilizer and beverage production. Therefore, if the CO 2 sale is considered, the c COE of oxy-combustion plants may be further reduced and the CO 2 capture cost will change. The cost of electricity (c″ COE ) and the CO 2 capture cost (c″ CCC ) when considering the CO 2 The CO 2 capture cost is related to the CO 2 capture amount, and the CO 2 sale revenue equals the CO 2 capture amount multiplied by the unit CO 2 sale price (c CO 2 ). From eq. (22), we can see that the critical c CO 2 equals the c CCC without a CO 2 sale. Figure 5 shows the effect of the c CO 2 on the c COE of conventional and oxy-combustion plants. Obviously, the economic characteristics of the oxy-combustion technology will enhance significantly if there are organizations who will purchase the high purity CO 2 product. The critical c CO2 (viz. c CCC ) that makes the c COE of oxy-combustion plants equivalent to those of conventional plants is 110-120 ¥/t.
It is worth noting that the relative CO 2 emission amounts (e CO 2 ,0 −e CO 2 ,1 ) and relative CO 2 capture amounts (m CO 2 ,1 − m CO 2 ,0 ) are not equivalent when the oxy-combustion plants are compared with conventional plants. This is because the thermal efficiencies of the oxy-combustion plant increase, and there is increased CO 2 emitted from oxy-combustion plants. The non-equivalence between the relative CO 2 emission amount and relative CO 2 capture amount (the relative CO 2 emission amount is generally less than the relative CO 2 capture amount) leads to non-equivalence between the critical T CO 2 and the critical c CO 2 , and the critical T CO 2 is generally greater than the critical c CO 2 .

Figure 5
Relations between c COE and c CO 2 .

CO 2 tax and CO 2 sale
The economic characteristics of the oxy-combustion technology were evaluated when the CO 2 tax and the CO 2 sale were considered together. Both the CO 2 tax and the CO 2 sale price significantly affect the c COE , c CAC and c CCC of oxycombustion plants. If in which the critical coefficient   = W net,1 /(W net,0  C,1  e )  (1 C,1 ) / C,1 , is actually the ratio of the critical c CO 2 to the critical T CO 2 . Usually,   <1.
The critical lines where the c COE of oxy-combustion plants equal those of conventional plants for three different loads are shown in Figure 6. Points on a line correspond to critical c CO 2 and critical T CO 2 values for a particular case. Above the line, the economic characteristics of oxy-combustion plants are better, whereas below the line, the economic characteristics of conventional plants are better. For example, for the critical line of the 2 × 300 MW subcritical case, the point A is above the line and it corresponds to 60 ¥/t T CO 2 and 80 ¥/t c CO 2 . In this case, the c COE of the oxycombustion plant is smaller and its economic characteristic is better; on the other hand, the point B is below the line and it corresponds to 80 ¥/t T CO 2 and 60 ¥/t c CO 2 . In this case, the c COE of the oxy-combustion plant is greater and its economic characteristic is worse. This result also reveals the difference between the c CO 2 and T CO 2 .

Effects of parameters
A sensitivity analysis of some important parameters in the oxy-combustion plant, such as coal price, ASU cost, ASU power consumption and CO 2 capture efficiency, was performed under the 2 × 300 MW subcritical plant model, and the results are shown in Figure 7. This shows that c COE is most correlated with c F , and that is because fuel costs contribute 62%-65% of c COE of oxy-combustion plants. The following parameters are  and W ASU , because the net power outputs of oxy-combustion plants decrease significantly because of the ASUs (power consumptions are 16%-18.5% of total loads), and the  directly relates to the oxygen demand and the ASU power consumption. The influences of ASU cost, CPU power consumption, interest rate, loan percentage on the c COE are also obvious, but the influence of CPU cost on the c COE is slight, because its cost amounts to only about 2% of the static investment cost of the oxycombustion plants. For c CAC and c CCC , the nine parameters considered have similar influences on them; and r CO2 influences them most because it directly affects unit CO 2 capture amounts and unit CO 2 emission amounts in oxy-combustion plants. The other important parameters are α and W ASU . The influences of coal price, ASU cost, CPU power consumption, interest rate, loan percentage on them are also obvious. Similarly, the influences of CPU cost on them are slight. In general, the influences of the parameters on these three costs are similar. The results show that the influences of α and W ASU on the c COE of the oxy-combustion plant are less than that of the coal price. But the influences of  and W ASU on the c CAC and c CCC are greater than that of the coal price because ASU consumes much power and the influences of coal price on c COE of conventional plants and oxy-combus-tion plants are similar. In addition, the influences of SO x and NO x emission taxes, S and N contents of coal on the three costs were also analyzed in the paper. The results show that the influences are slight, so they are not shown in Figure 7.

Effects of coal samples
To analyse the influence of different coal samples on the economic characteristics of the oxy-combustion technology, three different coal samples were further chosen to conduct a similar calculation process. The ultimate analysis results and lower heating values of these coal samples are all listed in Table 4.
Considering the 2 × 300 MW subcritical plant for example, the c COE , c CAC and c CCC results corresponding to the four coal samples are listed in Table 5. The results show that the influence of different coal samples on the economic characteristics of the oxy-combustion technology is not obvious, and the results obtained in this paper are universally significant.

Conclusion
In this paper, a techno-economic evaluation of 2 × 300 MW subcritical, 2 × 600 MW supercritical and 2 × 1000 MW ultra-supercritical oxy-combustion coal-fired power plants was performed. The results indicate that the electricity cost of a 2 × 300 MW oxy-combustion plant (with LIFAC desulphurization device) is 500.04 ¥/(MW h) (449.09 ¥/(MW h), 395.93 ¥/(MW h), are the equivalent values for the 2 × 600 MW and 2 × 1000 MW plants), which is 1.39 (similarly 1.38, 1.36) times that of the corresponding conventional plant (equipped with the limestone-gypsum desulfurization system and SCR denitration system); its static investment cost is 1.19 (1.23, 1.25) times that of the corresponding conventional plant; its net power output is 0.73 (0.74, 0.75) times that of the corresponding conventional plant. The increase in the static investment cost is mainly because of the high commercial price of ASU, and the significant decrease of the net power output is mainly because of the high power consumption of the ASU and CPU systems. However, without considering the power consumption of the ASU and the CPU, the annualized costs of oxy-combustion plants increase slightly in comparison to conventional plants. This is because the desulfurization and denitration devices with