Abstract

A comparative analysis of two mid-sized oxyfuel combustion combined cycles is performed. The two cycles are the semiclosed oxyfuel combustion combined cycle (SCOC-CC) and the Graz cycle. In addition, a reference cycle was established as the basis for the analysis of the oxyfuel combustion cycles. A parametric study was conducted where the pressure ratio and the turbine entry temperature were varied. The layout and the design of the SCOC-CC are considerably simpler than the Graz cycle while it achieves the same net efficiency as the Graz cycle. The fact that the efficiencies for the two cycles are close to identical differs from previously reported work. Earlier studies have reported around a 3% points advantage in efficiency for the Graz cycle, which is attributed to the use of a second bottoming cycle. This additional feature is omitted to make the two cycles more comparable in terms of complexity. The Graz cycle has substantially lower pressure ratio at the optimum efficiency and has much higher power density for the gas turbine than both the reference cycle and the SCOC-CC.

1. Introduction

The evidence that anthropogenically generated greenhouse gases are causing climate change is ever-increasing. The Intergovernmental Panel on Climate Change (IPCC) has stated that [1]:

“It is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other types of anthropogenic forcing together. The best estimate of the human induced contribution to warming is similar to the observed warming over this period.”

One of the largest point source emitters of greenhouse gases is fossil fuel based power plants. One of the options to mitigate these greenhouse gases is to utilize Carbon Capture and Storage (CCS) in power plants.

There are three main processes being considered for CCS: the post combustion capture, the precombustion capture, and the oxyfuel combustion capture [2]. This paper focuses on the oxyfuel combustion combined cycle. The oxyfuel combustion fires fuel with nearly pure O2 instead of air and the resulting combustion products are primarily steam and carbon dioxide. This makes it technically more feasible to implement CO2 capturing solutions.

Two competing oxyfuel combustion combined cycles have shown promising potential [3]. These are the semiclosed oxyfuel combustion combined cycle (SCOC-CC) and the Graz cycle. Numerous studies have been done on both of these cycles, including a number of studies that have compared the performance of the two cycles [337].

1.1. SCOC-CC

The SCOC-CC is essentially a combined cycle that uses nearly pure O2 as an oxidizer. After the heat recovery steam generator (HRSG) there is a condenser that condenses the water from the flue gas. The flue gas leaving the condenser is then primarily composed of CO2. Part of the CO2 is then recycled back to the compressor while the rest is compressed and transported to a storage site.

Bolland and Sæther first introduced the SCOC-CC concept in [4] where they compared new concepts for recovering CO2 from natural gas fired power plants. They compared a standard combined cycle with a cycle using postcombustion both with and without exhaust gas recirculation and also the SCOC-CC, along with a Rankine cycle that incorporates oxyfuel combustion. Ulizar and Pilidis [5] were first to present a paper that focused exclusively on the performance of the SCOC-CC. They started with cycle optimization and also simulated off-design performance. They did more extensive work exploring the selection of an optimal cycle pressure ratio and turbine inlet temperature [6] and on operational aspects of the cycle [7]. Bolland and Mathieu also published studies on the SCOC-CC concept [8] comparing its merits with a postcombustion removal plant. Amann et al. [9] also compared the SCOC-CC with a combined cycle using a postcombustion plant and made a sensitivity analysis regarding the purity of the O2 and the corresponding energy cost of the air separation unit. Tak et al. compared the SCOC-CC with a cycle developed by Clean Energy Systems and concluded that the SCOC-CC seemed to be advantageous [10]. Jordal et al. proceeded to develop improved cooling flow prediction models [11] and Ulfsnes et al. studied transient operation [12] and further explored real gas effects and property modelling [13]. Other researchers have started with a conventional natural gas combined cycle as a starting point in the modelling of the SCOC-CC. Riethmann et al. investigated the SCOC-CC using a natural gas combined cycle as a reference case and concluded that the net efficiency of the SCOC-CC was 8.3% points lower compared to the reference cycle [15]. Corchero et al. did a parametric study with regard to the pressure ratio at a fixed turbine entry temperature of 1327°C [14]. Yang et al. [16] modelled the SOCC-CC along with the ASU and the CO2 compression train at different pressure ratios and two different turbine inlet temperatures, 1200°C and 1600°C. They concluded that the optimal pressure ratio is around 60 and 90 for the turbine inlet temperatures of 1418°C and 1600°C, respectively. With optimal design conditions, the net cycle efficiency is lower than the efficiency of the conventional CC by about 8 percentage points for both of the two turbine inlet temperatures. Dahlquist et al. optimized a mid-sized SCOC-CC [17]. They concluded that although the optimum pressure ratio was 45 with regard to net efficiency, it would be beneficial to choose a lower pressure ratio for the cycle. Choosing a lower pressure ratio would only penalize the efficiency by a small amount but facilitate the design of the compressor by a great deal. Sammak et al. looked at different conceptual designs for the gas turbine for the SCOC-CC [18]. They compared a single and twin-shaft design and concluded that a twin-shaft would be an advantageous design for the SCOC-CC because of the high pressure ratio.

The main results of these papers have been summed together in Table 1. It is clear that there is no consensus in the power requirement of the air separation unit (ASU) as it ranges from 735 to 1440 kW/(kg/s) based on the mass flow of the O2 generated. This can be explained partly by the fact that the purity of the O2 stream varies from 90% to 96% and also that compression of the O2 stream is included in the ASU energy demand in some of the studies. The pressure ratio (PR) varies considerably between studies, where the lowest is 24.5 while the highest is 90. This can partly explain that only some of the studies aimed at optimizing the cycle with respect to pressure ratio, while others only selected a pressure ratio based on, for example, experience. Corchero et al. [14] and Yang et al. [16] looked at different pressure ratios and show similar trends as has been found in the current study. The resulting efficiencies from these studies have a very large spread and ranges from 36.7% to 53.9%. An important fact to take into consideration is that the oldest reference [4] is from 1992 and the state-of-the-art cycle efficiency for a combined cycle at that time was much lower than is today. Bolland and Sæther found that the combined cycle efficiency, without carbon capture, was around 52% and that the SCOC-CC was around 10% lower than the reference cycle. The studies that are presented in Table 1 are not based on the same assumptions, such as condenser cooling methods and turbine cooling which of course influence the net efficiencies reported.

1.2. Graz Cycle

The basic principle of the Graz cycle was developed by Jericha in 1985 [19] and was aimed at solar generated O2-hydrogen fuel. Jericha et al. modified the cycle in 1995 [20] to handle fossil fuels.

The Graz cycle, similar to the SCOC-CC, uses nearly pure O2 as oxidizer. In the Graz cycle a major part of the flue gas is recycled back to the compressor; that is, the water is not condensed out of it. Furthermore, the turbine and the combustion chamber are cooled using steam from the steam cycle. This means that the working fluid is mainly steam in the gas turbine. The other part of the flue gas goes through a condenser where the water is condensed from the flue gas. The flue gas is then in major parts CO2 which can be compressed and transported to a storage site. The full design of the Graz cycle incorporates a second bottoming cycle that uses the heat from the condensation of the flue gas. The second bottoming cycle is a subatmospheric steam cycle, since the condensation returns low quality heat, that is, low temperature.

There has been extensive research at the Graz University of Technology in designing the cycle, but the main focus has been on design of the turbomachinery components [2129]. These studies have not been focused on optimizing the cycle performance with respect to pressure ratio. The main results of papers that study the Graz cycle are shown in Table 2. The design of the Graz cycle has been evolving and the most advanced cycle layout is the S-Graz cycle which was presented in paper [26]. Publications published later all study the S-Graz cycle concept.

1.3. Comparison of Cycles

A number of papers have compared the two different cycles along with other carbon capture technologies for natural gas fired power plants [3036]. The main results of papers that present cycle results are shown in Tables 1 and 2. Kvamsdal et al. compared nine different carbon capture options for natural gas fired power plants [30, 32]. Among them were the SCOC-CC and the Graz cycle; the results for the cycle simulations are shown in Tables 1 and 2. It was concluded that concepts that employed very advanced technologies that have a low technological readiness level and high complexity achieved the highest performance. Franco et al. evaluated the technology feasibility of the components in 18 different novel power cycles with CO2 capture [31]. One of the conclusions was that the SCOC-CC would be one of the cycles that incorporates gas turbines that would require the least effort to turn into a real power plant. Sanz et al. made a qualitative and quantitative comparison of the SCOC-CC and the Graz cycle [33]. Their thermodynamic analysis showed that the high-temperature turbine of the SCOC-CC plant needed a much higher cooling flow supply due to the less favourable properties of the working fluid than the Graz cycle turbine. They, in comparison to Franco et al. [31], concluded that all turbomachines of both cycles showed similar technical challenges and that the compressors and high-temperature turbines relied on new designs. Woollatt and Franco did a preliminary design study for both the compressor and the turbine, in both the SCOC-CC and the Graz cycle [34]. They concluded that the turbomachinery can be designed using conventional levels of Mach number, hub/tip ratio, reaction, and flow and loading coefficients. They furthermore concluded that the efficiencies and the compressor surge margins of the components should be similar to a conventional gas turbine. Thorbergsson et al. examined both the Graz cycle and the SCOC-CC [35]. They conceptually designed the compressor and the turbine for both cycles. They concluded that the Graz cycle, in the original version including the second bottoming cycles, is expected to be able to deliver around 3% points’ net efficiency benefit over the semiclosed oxyfuel combustion combined cycle at the expense of a more complex realization of the cycle.

Comparative work on the two cycles has suffered from not having the same technology level in the design of the two cycles. This results in the fact that it is difficult to draw conclusions from the comparisons. The aim of the current study is to assess the two cycles using the same technology level and in addition have comparable complexity levels. The current study goes into more details regarding the optimal pressure ratio and turbine entry temperatures for the oxyfuel combustion cycles then past publications. This is accomplished by establishing a reference cycle, which has a technology level that could enter service around year 2025. The fuel is assumed to be natural gas for all three cycles. The reference cycle is then used as the starting point for the modelling of the oxyfuel combustion combined cycles. In the current study the pressure ratio has been varied to locate the optimal net efficiency with respect to pressure ratio. Previous work has reported around 3% points’ benefit for the Graz cycle [32] including work carried out by the authors [35]. It was viewed that a majority of these benefits would be attributed to the use of a second bottoming cycle as included in the original implementation. To make a fair comparison of the two alternatives it was decided to exclude this cycle feature from the original Graz cycle. It should be noted that it is quite feasible to introduce such a bottoming concept also for the SCOC-CC if the target would be to achieve maximum efficiency. The two simpler implementations were preferred in order to keep down complexity and make practical implementation more feasible.

2. Methods

The heat and mass balance program IPSEpro is used to simulate the power cycles [38]. The systems of equations, which are established using a graphical interface, are solved using a Newton-Raphson based algorithm. The simulation program was modified to incorporate the thermodynamic and transport properties program REFPROP to calculate the physical properties of fluids [39].

2.1. Cooled Turbine
2.1.1. Cooling Model

The cooling model is very important when studying the performance of gas turbine based cycles. The cooling model used is the model and is based on the work of Halls [40] and Holland and Thake [41]. The model is based on the standard blade assumption, which assumes that the blade has infinite thermal conductivity and a uniform blade temperature. The model used in this study was originally implemented by Jordal [42].

The main parameters for the cooling model are first the cooling efficiencywhere is the temperature of the cooling flow at the inlet, is the temperature of the cooling flow at the exit, and is the uniform blade temperature. The cooling efficiency is set to a moderate limit of .

Second the cooling effectiveness is defined aswhere is the hot gas temperature.

The model is a first-law thermodynamic, nondimensional model. The model is based on the dimensionless mass flow coolingwhere is the cooling mass flow, is the heat capacity of the cooling fluid, is the convective heat transfer coefficient on the hot gas side, and is the area of the blade. The main parameter of interest is the coolant mass flow ratiowhere is the heat capacity of the hot gas, is the average Stanton number of the hot gas, and is the cross-sectional area of the hot gas. The relations between the cooling mass flow and the temperature differences areThe Stanton number is defined aswhere is the density of the hot gas and is the flow velocity.

To estimate the cooling requirements for each cooled turbine blade row, it was assumed that the cooling parameters were constant. The parameters were chosen to represent a cooled turbine that will enter service around 2025.

The uniform blade metal temperature is set to 850°C. This means that the maximum temperature will be around 950°C and the average temperature at the gas side of the blade around 900°C. The uniform blade metal temperature is used as the temperature limit for the cycle simulations.

It is assumed, as has been done in other studies [11, 4346], that the Stanton number is constant, , in regard to both the change in the working fluid and the change in the design parameters. The parameters that are assumed to be constant in the cooling model are shown in Table 3. The geometry parameter, , which is the ratio between the wetted blade and adjacent cooled surface areas over the average gas cross-sectional area, is also held constant between all cases. This parameter is unknown for a thermodynamic analysis where the key dimensions of the turbine have not been designed. El-Masri [44] estimated that this parameter is slightly less than 4.0 for a cascade blade row and around 8.0 for a stage, allowing for a row-to-row spacing. Jordal [47] concluded that when taking into account that rotor disks and the transition piece from the combustion chamber to the first stage nozzles are also subject to cooling, an average value should be around 5.0 for a stage.

The cooling model was used to reproduce the results in [48] and showed good agreement.

2.1.2. Expansion

The expansion in an uncooled turbine is modelled aswhere is the gas constant for the working fluid, , are the pressure and entropy, respectively, is the inlet, and is the outlet of the turbine stage. This model was evaluated against different models such as Mallen and Saville [49], using numerical integration and the model used gave good agreement with the numerical integration.

For the cooled turbine, the mixing of the coolant and the main stream gas flow result in a loss in stagnation pressure. This irreversibility is taken into account by defining a new polytropic efficiency [42, 50], defined aswhere is the stagnation pressure at the inlet of the rotor blade row, in is the inlet to the turbine, and out is the outlet of the turbine. Parameter is specific to each turbine and models the losses. It is typically in the range of for a turbine that has good performance and around for a turbine that has poor performance [51]. The polytropic efficiency is set to . The losses are taken into account by assuming that the factor is for all cases. Dahlquist et al. examined the empirical loss models used to design turbomachinery, which are generated using air as the working fluid, and concluded that the loss models generate similar results for the working fluids in oxyfuel cycles [52]. This indicates that it is possible to achieve a similar technology level for the oxyfuel turbines as for state-of-the-art conventional turbines.

2.2. Compressor

The compression is modelled using polytropic efficiency,where is the gas constant for the working fluid, and are the pressure and entropy, respectively, is the inlet, and is the outlet of the compressor.

It is assumed that the polytropic efficiency is constant for all cycles and all cases and is assumed to be . Similar to the turbine, it is assumed that it is possible to achieve a compressor design for the oxyfuel compressor that is on the same level as the state of the art of compressors in conventional gas turbines.

2.3. Combustor

The combustion is a simple energy model based on the assumption that all of the fuel is completed in the combustion, that is, 100% combustion efficiency.

The amount of excess O2 is calculated aswhere is stoichiometric combustion. For the oxyfuel cycles the combustion is nearly stoichiometric; that is . It is preferred that the combustion takes place as close to stoichiometric conditions as possible to reduce the amount of O2 that the ASU needs to produce. Such a low amount of excess O2 is very different compared to traditional combustion in gas turbines, which have much larger amount of excess O2. It is assumed that it is possible to have the combustion under near stoichiometric conditions while the emissions of , CO and unburned hydrocarbons are within given constraints. Sundkvist et al. found that using excess of 0.5% of O2, , resulted in 400 ppmv of CO at the turbine outlet [53] and increasing the O2 ratio resulted in reduced levels of CO, while increasing the energy penalty from the ASU, as expected.

The pressure drop in the combustion chamber is assumed to be 4%. And a compressor is used to increase the pressure of the fuel above the pressure in the combustion chamber.

2.4. Air Separation Unit

O2 is produced with an air separation unit (ASU). The ASU is assumed to be a cryogenic air separation plant. Modelling of the ASU is not within the scope of this current study. ASU power consumption is highly dependent on the purity of the O2 stream. It is therefore an economic trade-off between purity and cost. Typical state-of-the-art cryogenic ASU can produce O2 with 99.5%-volume purity at a power consumption of 900 kW/(kg/s) [54]. By decreasing the purity, it is possible to reduce the power consumption of the ASU. At a purity level of 95%, the power consumption can be assumed to be around 735 kW/(kg/s) [17, 55]. In this study this has been taken into account and a purity level of 95% for the ASU is used. The corresponding O2 composition is shown in Table 4.

The ASU unit delivers the O2 stream at a pressure of 1.2 bar and with a temperature of 30°C. An intercooled compressor is used to increase the pressure of the stream to the working pressure in the combustor. The compression process has been modelled in the cycle simulation.

2.5. Flue Gas Condenser

The main purpose of the oxyfuel combustion cycles is to produce CO2 along with power generation. Because the flue gas consists mostly of CO2 and steam, the most convenient method is to condense the water from the flue gas to produce the CO2. The flue gas will also contain small amounts of Ar, N2, and O2. There will also be traces of harmful acid gases along with particles such as soot. By using a direct contact condenser, these harmful gases and the particles can be removed from the flue gas when the steam is condensed. The condenser will therefore also act as a scrubber.

The efficiency of the condenser is defined aswhere is the amount of water that is condensed from the flue gas and is the amount of water in the flue gas that enters the condenser. The flue gas condenser efficiency is a simple way to evaluate the performance of the condensers [56]. The parameter does not represent an efficiency in its true sense but is a metric commonly used to describe the performance of condensers [56].

2.6. CO2 Compression

The CO2 stream from the condenser that will be sent to storage needs to be compressed to a higher pressure and the remaining water vapour and noncondensable gases need to be removed. This process, the CO2 recovery and compression process, is not within the system boundaries of the current study. It is instead taken into account by assuming a fixed energy cost, kW/(kg/s) of wet CO2 [25]. This energy cost assumes that the stream is compressed to 100 bar. This value also takes into account the removal of water and other gases that are present in the CO2 stream.

3. Power Cycles

The fuel is assumed to be natural gas and the composition is shown in Table 5. Common assumptions used in the simulations of the cycles are shown in Table 6.

3.1. Reference Cycle

A reference cycle was modelled that is in the mid-size range. The mid-size range is from 30 to 150 MW [42]. Here we have aimed at keeping the gross combined power output from the gas turbine and the steam turbine constant at 100 MW. The reference cycle has been modelled as a gas generator and a separate power turbine, that is, a two-shaft gas turbine. The gas generator turbine consists of two cooled stages. The cooling flow is bled from the compressor. The steam cycle for a power plant in this power range usually employs single or double pressure levels and does not use reheat. Here we have used a dual-pressure steam cycle without reheat. The steam turbine is a single-casing nonreheat. The pressure was set to 140 bar and the maximum temperature to 560°C at the inlet to the steam turbine. If the exhaust temperature from the gas turbine goes below 585°C the steam temperature decreases so that the temperature difference is 25°C. A schematic of the cycle is shown in Figure 1.

3.2. SCOC-CC

A schematic of the SCOC-CC is shown in Figure 2. The SCOC-CC is based on the reference cycle. Now, however, the fuel is combusted with O2 that is produced in the ASU. The fuel is combusted near to stoichiometric ratio, meaning that nearly no excess O2 is produced. This minimizes the power demand of the ASU. The combustion chamber is cooled using the recycled flue gas, after most of the steam is condensed from it, in the condenser. The flue gas leaving the combustion chamber is mainly CO2 and also a small amount of steam. The gas turbine layout is the same as the reference cycle with a gas generator and a power turbine. The turbine in the gas generator has two stages, which are both cooled. The cooling flow is also bled from the compressor, similar to the reference cycle. The layout of the steam cycle is unchanged from the reference cycle. The flue gas goes to the condenser after the heat recovery steam generator, where the major part of the steam is condensed from the exhaust gas. The flue gas is cooled in this process. The CO2 stream that leaves the condenser has near 100% relative humidity. A small part of the CO2 stream is sent to the compression and purification process and is then transferred to the storage site. The major part of the CO2 is recycled back to the compressor. The water in the CO2 stream can possibly condense at the entry to the compressor, which could have a deteriorating effect for the compressor. The CO2 stream is therefore heated before it enters the compressor using the heat from the flue condensation.

3.3. Graz Cycle

The main features are that the gas turbine cooling is implemented with steam and that the flue gas is sent straight to the compressor after the HRSG without condensing the steam from it. Part of the flue gas is sent to a condenser where a major part of the water is condensed from it; after this it is sent to the CO2 compression and purification process. The CO2 is afterwards transferred to the storage site.

The most common layout of the Graz cycle incorporates two bottoming cycles. The first one uses a typical HRSG and a steam turbine, which only expands, however, to the pressure of the combustion chamber. This is because the steam is used for cooling both the combustion chamber and the gas turbine blades. The second bottoming cycle uses the enthalpy of the condensation and assumes that the pressure at the outlet of the condenser is 0.021 bar, which is particularly low.

It is hard to imagine that the first design of the Graz cycle will deviate so greatly from the current layout of the combined cycle. Here we have taken the reference cycle as the basis and implemented the major design features of the Graz cycle. A schematic of the Graz cycle is shown in Figure 3. The cycle incorporates an intercooler to reduce the temperature of the gas at the exit of the compressor as well as steam cooling. This layout, not implementing the second bottoming cycle, is considered more reasonable for the first generation design of the cycle. It also makes the complexity level of the SCOCC-CC and the Graz cycle more comparable, by not including improvements that could be implemented on both cycles.

The cycle illustrated in Figure 3 should therefore be understood as a simplified variant of the Graz cycle.

4. Results

A parametric study of the two oxyfuel cycles was performed by varying the turbine entry temperature (TET) and the pressure ratio (PR) of the gas turbine. The turbine entry temperature is the temperature at the exit of the combustion chamber and is therefore also the temperature at the entry to the first stator in the gas turbine. The temperature has been varied from 1250°C to 1600°C. The pressure ratio was varied freely until the design constraints were attained.

4.1. Reference Cycle

The results for the cycle net efficiency are shown in Figure 4 as a function of pressure ratio. The turbine entry temperature is also shown in Figure 4. The net efficiency takes into account the power needed for the pumps in the cycle. The entry temperature for the power turbine has been set to 850°C, to eliminate the need for cooling in the power turbine. If the temperature goes above 850°C, which is the metal temperature limit for the blades, then the first stage in the power turbine would then need to be cooled.

Figure 5 shows the cooling mass flow ratio for the reference cycle. The ratio is defined as the total cooling mass flow divided by the inlet mass flow to the turbine.

4.2. SCOC-CC

Figure 6 shows the gross efficiency for the SCOC-CC as a function of pressure ratio and turbine entry temperature. The gross efficiency is the total power delivered by the gas turbine and steam turbine generators divided by the energy content of the fuel, based on the lower heating value.

As the pressure ratio decreases, the amount of steam in the low pressure steam is also reduced. The lower limit for the pressure ratio is reached when the mass flow of the low pressure steam approaches zero. The higher pressure ratio limit is reached when the temperature difference for the high pressure steam and the flue gas in the preheater approaches 5°C.

Figure 7 shows the net efficiency for the SCOC-CC as a function of pressure ratio and turbine entry temperature. The net efficiency takes into account the fuel compressor power, the power needed for the pumps, the energy needed for the production of the O2, the O2 compressor power, and the power needed to compress the CO2. The largest decrease in the efficiency comes from the power requirement for the O2 production and compression. As the pressure ratio is increased, the O2 compression power consumption increases very rapidly. This results in there being an optimum pressure ratio.

Figure 8 shows the power entry temperature as a function of pressure ratio and turbine entry temperature for the SCOC-CC. As can be seen in Figure 8 the power turbine entry temperature for all cases is above the blade material temperature limit, 850°C. This means that the first stage in the power turbine needs to be cooled.

Figure 9 shows the cooling mass flow ratio as a function of pressure ratio and turbine entry temperature for the SCOC-CC. The cooling mass flow ratio is higher for the SCOC-CC than the reference cycle, since the heat capacity for the working fluid is lower in the SCOC-CC than in the reference cycle.

4.3. Graz Cycle

The Graz cycle was studied at turbine entry temperatures of 1250°C, 1450°C, and 1600°C. Figure 10 shows the gross efficiency for the Graz cycle as a function of pressure ratio and turbine entry temperature. It can be seen in Figure 10 that there is no global optimum for the gross efficiency.

Figure 11 shows the net efficiency for the Graz cycle as a function of pressure ratio and turbine entry temperature. The net efficiency is calculated as is done for the SCOC-CC. The major reduction in the efficiency comes from the power needed for the O2 production and compression. The relative power consumption of the O2 compression increases as the pressure ratio increases, which results in an optimum in the net efficiency. Figure 12 shows the power entry temperature as a function of pressure ratio and turbine entry temperature for the Graz cycle. When the power turbine entry temperature is over 850°C, the first stage in the power turbine is cooled.

Figure 13 shows the cooling mass flow ratio as a function of pressure ratio and turbine entry temperature for the Graz cycle. Since steam is used as coolant for the turbine blade cooling for the Graz cycle, the cooling mass flow ratio is considerably lower than for the reference cycle and the SCOC-CC. This is the result of the fact that the steam has a substantially lower temperature than the compressor discharge temperature and that the steam has a higher heat capacity than the working fluid.

4.4. Optimum Cycles

The results for the cycles with the optimum performance are shown in Table 7. The optimum reference cycle is determined to be a cycle where there is no need to employ cooling for the power turbine. This means that the power turbine entry temperature is 850°C or lower. The pressure ratio that gives the optimal efficiency is 26.15 and the turbine entry temperature is 1400°C, which results in a net efficiency of 56%. The turbine exhaust temperature is only 526°C, which results in the high pressure steam having a temperature of 501°C since the pinch temperature difference has a minimum value of 25°C in the high pressure (HP) superheater.

The optimum SCOC-CC has a relatively high pressure ratio, or around 57.3, and the turbine entry temperature is 1450°C. Even though the pressure ratio is so high, the compressor outlet temperature is only 474°C, which is below the temperature limit of the blade material. The steam turbine produces more of the power in the SCOC-CC compared to the reference cycle. The exhaust gas is cooled down from 618°C to 65°C in the HRSG. This comes from the fact that the working fluid achieves a better fit to the steam cycle. The main decrease in power comes from the O2 production and the O2 compression. The gross efficiency for the cycle is close to 60% but, taking into account the O2 production and compression, CO2 compression, and also the pumps, this is lowered to 46%. The SCOC-CC cycle produces 10.3 kg/s of CO2, which is about 890 tonnes per day. The SCOC-CC also produces about 170 kg/s of water with a temperature of 46°C in the flue gas condenser.

The optimum Graz cycle has a pressure ratio of about 36.5 and a turbine entry temperature of 1450°C. Even though the pressure ratio is lower in the Graz cycle than in the SCOC-CC and the compression is intercooled, the compressor outlet temperature is much higher or around 605°C. The reason for such a high temperature is mainly the fact that the compressor inlet temperature is 100°C. The working fluid saturation temperature is around 95°C so, to avoid condensation at the inlet of the compressor, the temperature needs to be higher than the saturation temperature. To be able to withstand the high temperature at the outlet of the compressor, the blade material will be more expensive than is normally used in compressors. The gas turbine produces a larger share of the power compared to the reference cycle and the SCOC-CC. This is because the cooling in the gas turbine uses steam from the steam cycle. The amount of steam needed for cooling is around 15.4 kg/s, which is about 50% of the steam produced in the steam cycle. This steam will therefore be expanded in the gas turbine and not in the steam turbine. This will have a negative effect on the efficiency since the steam will be expanded to 1 bar instead of 0.045 bar as it is in the steam turbine. One aspect of the Graz cycle is that the power density is much higher compared to both the reference cycle and the SCOC-CC. The compressor inlet mass flow is only 40% of the reference cycle mass flow and 50% of that of the SCOC-CC. The gross efficiency for the Graz cycle is around 59%. The major deduction in efficiency comes from the O2 production and compression. However, the compression power consumption is lower in the Graz cycle than the SCOC-CC because of the lower pressure ratio. The CO2 compression is similar to that in the SCOC-CC cycle. This results in nearly the exact same net efficiency as for the SCOC-CC or 46%. The Graz cycle produces slightly more CO2, or around 10.8 kg/s, which is about 933 tonnes/day. The Graz cycle produces significantly more water than the SCOC-CC, or about 270 kg/s of water with a temperature of 60°C in the flue gas condenser.

5. Discussion and Conclusion

The study compared three combined cycles, a conventional cycle, the SCOC-CC, and the Graz cycle, at the mid-size level power output. The gross power output for all cycles was set to 100 MW. The conventional cycle was used as the basis for the modelling and as a reference for the oxyfuel combustion cycles. A detailed literature review was conducted for the oxyfuel combustion combined cycles. The literature review showed that there is no consensus on the power requirement for the air separation unit. It also showed that the comparison of the SCOC-CC and the Graz cycle has lacked consistent assumptions and agreement on the technology parameters used to model the cycles.

A parametric study was conducted by varying the pressure ratio and the turbine entry temperature for the cycles. A constraint for the conventional cycle was set on the power turbine entry temperature to eliminate the need for cooling in the power turbine. The resulting optimal conventional cycle achieved a 56% net efficiency at a pressure ratio of 26.2 and a turbine entry temperature of 1400°C. The optimal SCOC-CC achieved only a 46% net cycle efficiency at a pressure ratio of 57.3 and a turbine entry temperature of 1450°C. The optimal Graz cycle also achieved a net cycle efficiency of 46% at a pressure ratio of 36.5 and a turbine entry temperature of 1450°C. The main reduction in efficiency for the oxyfuel cycles comes from the O2 production, which reduced the power output from the cycles by more than 10 MW. An additional reduction of the power output comes from the compression of O2 to operating pressure. This is about 7 MW and 6 MW for the SCOC-CC and Graz cycle, respectively. The difference comes from the higher pressure ratio of the SCOC-CC.

One of the benefits of the Graz cycle is the high power density of the gas turbine. This results in smaller turbomachinery for the gas turbine in the Graz cycle, which lowers the cost of this machinery. One of the main penalties of the Graz cycle is that the large amount of steam, which is generated in the HRSG, is not expanded in the steam turbine but in the gas turbine. The result is that the steam does not expand to the condenser pressure of the steam cycle.

The SCOC-CC is considerably simpler than the Graz cycle as it does not implement steam cooling and does not require an intercooler. The optimal SCOC-CC, however, has a much higher pressure ratio than both the reference cycle and the Graz cycle. The efficiency does not vary greatly with the pressure ratio, however, and it is possible to reduce the pressure ratio without significantly penalizing the net efficiency. This would facilitate the compressor design substantially.

Nomenclature

:Hot gas convective heat transfer coefficient
:Dimensionless mass flow cooling
:Cooling mass flow
:Cooling efficiency
:Cycle net efficiency
:Polytropic efficiency
:Condenser efficiency
:Ratio of oxygen
:Hot gas density
:Stanton number of the hot gas
:Cooling effectiveness
:Coolant mass flow ratio
:Blade area
:Annulus area
:Heat capacity of the cooling fluid
:Heat capacity of the hot gas
:Pressure
PR:Pressure ratio
:Gas constant
:Turbine loss parameter
:Entropy
:Uniform blade temperature
:Cooling flow exit temperature
:Cooling flow inlet temperature
:Hot gas temperature
TET:Turbine entry temperature
TIT:Turbine inlet temperature
ASU:Air separation unit
CCS:Carbon Capture and Storage
GT:Gas turbine
SCOC-CC:Semiclosed oxyfuel combustion combined cycle
ST:Steam turbine.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

This research was funded by the Swedish Energy Agency, Siemens Industrial Turbomachinery AB, GKN Aerospace, and the Royal Institute of Technology through the Swedish research program TURBOPOWER. Their support is gratefully acknowledged. The financial grant from Landsvirkjun’s Energy Research Fund is gratefully acknowledged by the first author.