Experimental investigation of the impacts of selective exhaust gas recirculation on a micro gas turbine

International Journal of Greenhouse Gas Control 90 (2019) 102809

Contents lists available at ScienceDirect

International Journal of Greenhouse Gas Control journal homepage: www.elsevier.com/locate/ijggc

Experimental investigation of the impacts of selective exhaust gas recirculation on a micro gas turbine

T

Jean-Michel Bellas, Karen N. Finney , Maria Elena Diego, Derek Ingham, Mohamed Pourkashanian ⁎

Energy 2050, Department of Mechanical Engineering, University of Sheffield, Sheffield, S3 7RD, UK

ARTICLE INFO

ABSTRACT

Keywords: CO2 capture Gas-CCS Selective exhaust gas recirculation CO2 enhancement Gas turbines

Selective exhaust gas recirculation (S-EGR) is an option proposed to augment the CO2 content in the flue gases of gas-fired systems, facilitating integrated post-combustion CO2 capture. However, issues such as flame instabilities and increases in unburned species require careful analysis. The performance of a micro-gas turbine under simulated S-EGR conditions is evaluated here. Maximum flue gas CO2 concentrations of 8.4 and 10.1 vol% were achieved at power outputs of 100 and 60 kWe, respectively; a 4–7 times increase in CO2 content compared to the baseline cases – similar to what can be achieved with S-EGR systems. Impacts on the operational performance of the system were assessed, together with the resulting CO, unburned hydrocarbon and nitrogen oxide (NOx) emission trends. The electrical efficiency reduced slightly under S-EGR conditions and emissions of unburned and partially oxidised species increased, especially at lower loads, where incomplete combustion effects were more prominent. The rotational speed of the engine decreased under S-EGR conditions as a result of the change in the oxidiser properties, with the compressor discharge temperatures also decreasing slightly. Furthermore, NOx emissions were lower in S-EGR scenarios due to the expected lower flame temperatures, which leads to reduced thermal NOx production.

1. Introduction Keeping the increase in global temperature rise to 2°C or below as per the Paris agreement to mitigate the impacts of climate change requires a considerable reduction in global CO2 emissions by 2040 (UNFCCC, 2015). In the EU low carbon economy roadmap, the target is to reduce emissions by 60% over this timescale (European Commission, 2018). Nevertheless, electricity demand is expected to increase up to 70% by that time, with a large fraction of this power still being produced from fossil fuels (IEA, 2017). In this context, natural gas is expected to contribute up to 25% of the energy mix by 2040 (IEA, 2017). Natural gas is considered to have the lowest carbon intensity compared to coal and oil (350–400 kg CO2/MWh in natural gas combined cycle (CCGT) power stations), but associated emissions still require mitigation to meet the proposed targets (IEA, 2017; IEAGHG, 2012). Coupling gas-fired power plants with CO2 capture and storage technologies (gasCCS) can become an attractive option to deliver low-carbon electricity. However, implementing gas-CCS poses several challenges, notably related to the large volumetric flowrates of exhaust gas with low CO2 concentrations (˜3-4 vol%) generated in gas-fired systems, as well as the

⁎

high remaining O2 content. These lead to detrimental impacts associated with the size, energy consumption and economic performance of the downstream CO2 capture plant, as well as issues concerning oxidative solvent degradation if an amine-based CO2 capture plant (ACP) is used (Aboudheir and ElMoudir, 2009; Gouedard et al., 2012; Li et al., 2011a; Supap et al., 2006). To overcome these issues, a number of strategies capable of increasing the CO2 content in flue gases from gas-fired systems have been proposed (see for example, Diego et al. (2017a)). These include supplementary firing, exhaust gas recirculation (EGR), selective exhaust gas recirculation (S-EGR), humidified cycles and oxy-fired gas turbines cycles. This work focuses on SEGR, which has been proposed by Merkel et al. (2013) as a promising option that can facilitate CO2 capture in gas fired plants, whilst minimising some of the limitations of other methods, such as the reduced O2 concentrations at the inlet, as considered in the following sections. The present study evaluates the influence of selective CO2 recirculation conditions on the performance of a Turbec T100 Series 3 micro gas turbine (mGT) at the national Pilot-scale Advanced CO2 Capture Technology (PACT) facilities located in Sheffield, (UK) (PACT, 2018) in order to advance the development of this technology.

Corresponding author. E-mail address: [email protected] (K.N. Finney).

https://doi.org/10.1016/j.ijggc.2019.102809 Received 3 October 2018; Received in revised form 15 July 2019; Accepted 6 August 2019 1750-5836/ © 2019 Elsevier Ltd. All rights reserved.

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

Nomenclature ACP CCGT CCS EGR EINOx FTIR mGT

MW Pe RMC nFG NOx PACT S-EGR TOT UHC

Amine-based capture plant Combined cycle gas turbine Carbon capture and storage Exhaust gas recirculation NOx emission index Fourier transform infrared spectroscopy Micro gas turbine

1.1. Exhaust gas recirculation

Molecular weight Net power output Remote monitor and control Flue gas molar flowrate Nitrogen oxides Pilot-scale Advanced CO2 Capture Technology facilities Selective exhaust gas recirculation Turbine outlet temperature Unburned hydrocarbons

2014). The minimum O2 concentration in the oxidiser necessary for stable combustion under EGR is usually considered to be ˜16 vol% (Bolland and Mathieu, 1998; Ditaranto et al., 2009; ElKady et al., 2009; Li et al., 2011a). Some have demonstrated that stable combustion is possible at reduced O2 concentrations of ˜14 vol% at the combustor inlet and 4 vol% at the exhaust outlet, although this is at the expense of emitting higher levels of CO and UHCs (Ditaranto et al., 2009; ElKady et al., 2009). Maintaining ˜16 vol% O2 at the combustor inlet limits the maximum EGR ratio (i.e. the ratio between the recirculated and total flue gas flowrates) to 40%, which can increase the CO2 concentration in the flue gas of large-scale gas-fired systems from ˜4 to ˜6.5 vol% (Adams and Mac Dowell, 2016; Biliyok and Yeung, 2013; Bolland and Sæther, 1992; Diego et al., 2017b; ElKady et al., 2009; Herraiz et al., 2018; Hu and Ahn, 2017; Li et al., 2011a; Sipöcz and Tobiesen, 2012). Increasing the CO2 concentration beyond this value would require new combustor designs able to operate efficiently at the low inlet O2 concentrations observed with EGR (ElKady et al., 2009; Li et al., 2011a; NETL, 2013). A number of theoretical works have demonstrated the potential of

To date a number of experimental and modelling studies have investigated EGR for gas-CCS (e.g. Akram et al., 2016; Ali et al., 2017; Best et al., 2016; Bolland and Sæther, 1992; Li et al., 2011a). EGR increases the CO2 concentration in the exhaust outlet by recirculating a proportion of the flue gases back to the compressor inlet which mixes with the combustion air (Bolland and Sæther, 1992). This process also reduces the flue gas volume sent to the downstream CO2 capture plant (as a fraction of the inlet air is replaced by the recycled flue gas), leading to decreased energy penalties (Akram et al., 2016; Bolland and Sæther, 1992; Diego et al., 2018; Herraiz et al., 2018; Li et al., 2011a) and reduced capital costs due to the smaller size of the capture system required (IEAGHG, 2012; USDOE/NETL, 2013). However, the recirculated flue gas reduces the amount of O2 in the oxidiser and can cause flame instabilities and combustion efficiency issues, leading to higher CO and unburned hydrocarbon (UHC) emissions (Best et al., 2016; ElKady et al., 2009; Mansouri Majoumerd et al.,

Fig. 1. Schematic of the (a) parallel and (b) series S-EGR configurations in CCGT power plants proposed by Merkel et al. (2013). 2

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

EGR to improve the overall net electrical efficiency by up to 2.7% in comparison to gas-fired plants with amine-based CCS only (Biliyok and Yeung, 2013; Bolland and Sæther, 1992; IEAGHG, 2012; Li et al., 2011b; NETL, 2013). This could lead to reductions of ˜3% in the cost of electricity and ˜10% in the cost of CO2 avoided, compared to an aminebased gas-CCS system without EGR (IEAGHG, 2012; NETL, 2013). Moreover, experimental work investigating the effects of EGR has been carried out on gas turbine combustors (Ditaranto et al., 2009; Elkady et al., 2008, 2009; Evulet et al., 2009; Hasemann et al., 2017), on a mGT (Best et al., 2016; Mansouri Majoumerd et al., 2014; Røkke and Hustad, 2005) and on a downstream ACP at pilot-scale (Akram et al., 2016). Ditaranto et al. (2009) identified that under EGR the levels of UHC and CO increased, whilst emissions of nitrogen oxides (NOx) decreased due to a higher NO re-burning mechanism. Røkke and Hustad (2005) also found that NOx emissions were reduced under EGR due to changes in the flame temperature, affecting both prompt and thermal NOx formation. The decreased O2 concentrations in EGR have also been shown to reduce the combustion reaction rates and lower flame temperatures and speeds, due to the higher heat capacity of the CO2 and the shrinking flammability limits (Elkady et al., 2008, 2009; Evulet et al., 2009; Finney et al., 2018). Hasemann et al. (2017) demonstrated that stable combustion is possible at ˜13% O2 concentrations at the combustor inlet, however they also show that CO emissions increase due to the lower flame temperatures and reduced O2 content. In the experimental work conducted by Best et al. (2016) the CO2 concentration in the flue gas was increased to ˜6.3 vol% under EGR conditions. These authors also indicate that CO and UHC increased due to incomplete combustion, whilst NOx emissions were reduced. They attributed the lower NOx emissions to a higher oxidizer heat capacity and thus a reduction in peak flame temperatures leading to reduced thermal NOx production (Best et al., 2016). Regarding mGT performance, CO2 enhancement leads to a reduction in the turbine inlet temperature, compressor outlet temperature and rotational speed due to changes in the specific heat capacity of the working fluid (Best et al., 2016). Mansouri Majoumerd et al. (2014) identified similar conclusions; however, the influence of EGR on overall turbine performance is negligible. The benefits of applying EGR on a turbine integrated with an ACP include a lower specific reboiler duty, with Akram et al. (2016) demonstrating a reduction of 25% when the flue gas CO2 concentration increased from 5.5 to 9.9 vol%. Furthermore, modelling studies investigating a 40% EGR ratio also demonstrated the specific reboiler duty can be reduced – from 4.00 MJ/kg of CO2 without EGR to 3.72 MJ/kg of CO2 when there is an increased level of 6.6 vol% CO2 at the exhaust (Biliyok and Yeung, 2013).

vacuum or compression stages (Merkel et al., 2013). A key benefit of S-EGR over EGR configurations is that the combustion air is not further diluted with N2 and other components in the recycled gas. This means that higher CO2 concentrations can be achieved in the flue gas stream, whilst keeping high O2 concentrations at the oxidiser inlet (Merkel et al., 2013). It has been calculated that the CO2 concentration in the flue gas of a CCGT with S-EGR can increase up to 18–19 vol%, with O2 levels at the combustor inlet of ˜16 vol% (Diego et al., 2018; Herraiz et al., 2018; Merkel et al., 2013). Higher CO2 concentrations of ˜26 vol% can even be achieved if combustors are able to effectively operate using an oxidiser with ˜14 vol% O2 (Turi et al., 2017). These levels of CO2 enhancement in the flue gas are expected to lead to large benefits in terms of size, energy and cost reductions in the downstream CO2 capture plant, as discussed above towards the end of Section 1.1 (Diego et al., 2017b, ; Herraiz et al., 2018; Merkel et al., 2013). Merkel et al. (2013) initially proposed the parallel and series S-EGR configurations shown in Fig. 1a and b, respectively. In the parallel SEGR process, the flue gas flow is split into two streams: one is sent to the post-combustion CO2 capture plant (e.g. an ACP) and the other enters the CO2-selective membrane (see Fig. 1a). Therefore, both the CO2 capture plant and the membrane unit treat a reduced volumetric flowrate of flue gas with high CO2 concentrations, which leads to a reduction in both the membrane area requirements and the ACP size, which can lower process costs. However, in this configuration the CO2 capture efficiency in both the membrane system and the ACP needs to be high, typically > 95% to maintain an overall CO2 capture efficiency of ˜90% (Diego et al., 2017b, ; Herraiz et al., 2018; Merkel et al., 2013). In the series S-EGR configuration, the total flue gas stream enters the post-combustion CO2 capture unit first (e.g. an ACP), which captures only a proportion of the CO2, and the remaining CO2 is then separated in the membrane system and recirculated back to the compressor inlet, as shown in Fig. 1b. Therefore, the ACP can operate with moderate CO2 capture efficiencies because the membrane system will remove any remaining CO2 in the flue gas to achieve the desired overall capture efficiency (Herraiz et al., 2018; Merkel et al., 2013). Nevertheless, this will lead to larger absorber and membrane areas, as the entire flue gas stream needs to be treated by both systems (Diego et al., 2018). Advanced configurations, even considering combinations of S-EGR and EGR processes or a hybrid of parallel and series designs have been recently analysed (Baker et al., 2017; Diego et al., 2018). Several simulation studies have investigated the advantages of SEGR configurations in CCGTs that use amine-based systems for postcombustion CO2 capture in terms of efficiency gains. Diego et al. (2017b), investigated parallel and hybrid S-EGR configurations, indicating efficiency increases of up to 1 percentage point, compared to a CCGT with an ACP alone. Herraiz et al. (2018) evaluated both parallel and series S-EGR; the parallel configuration showed efficiency improvements of ˜1.7%, whereas series S-EGR improved efficiency by ˜0.6-1.1% compared to a CCGT with just an ACP. The cost of S-EGR systems using membranes or amine systems as the post-combustion CO2 capture technology has also been analysed. The works show the potential economic advantages of S-EGR concepts applied to CCGTs (Baker et al., 2017; Merkel et al., 2013; Turi et al., 2017), with some indicating that improvements to the membranes are still needed to enhance their competitiveness – in terms of CO2 permeance, membrane costs and reductions in their pressure drop (Diego et al., 2017b, ; Turi et al., 2017). From an operating point of view, challenges associated with S-EGR systems include flame stability and combustion efficiency, which lead to augmented CO and UHC emissions under certain conditions that require further investigation (Herraiz et al., 2018). Experimental research on S-EGR configurations and operating conditions are limited. Preliminary S-EGR results on the performance of a 100 kW CO2 selective membrane system have been recently published by Darabkhani et al. (2018); however, this study focuses on membrane performance

1.2. Selective exhaust gas recirculation S-EGR has been proposed to overcome the limitations associated with EGR and achieve higher CO2 enhancement levels in the flue gas without severely compromising the O2 concentration in the oxidiser (Merkel et al., 2013). This process separates CO2 from the flue gas using a CO2-selective system and recirculates this stream back to the inlet of the compressor, as illustrated in Fig. 1. The flue gas is fed to the selective membrane system, for example, which uses an air sweep stream that mixes with the CO2 that permeates through the membrane (Merkel et al., 2013). The resulting air + CO2 stream is fed to the gas turbine system whereas the CO2-depleted flue gas is then released to the atmosphere. Merkel et al. (2013) indicates the CO2-selective separation process could be achieved in the membrane with limited energy consumption, as this system is expected to work under nearly-atmospheric pressure conditions (only a slight pressure increase is needed in the gas inlet streams to circumvent the membrane pressure drop). This is possible thanks to the presence of the air sweep stream (that dilutes the CO2 on the permeate side), which combined with the higher CO2 concentrations in the flue gas/feed stream should offer enough driving force to allow for an effective CO2 separation without the need for 3

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

rather than combustion performance. Marsh et al. (2016, 2017) investigated the combustion and operability limits of S-EGR using a swirl burner with a premixed flame, achieving up to 20 mol% CO2 concentrations in the inlet mixture. Their results show that a near stoichiometric air-fuel ratio is needed at these CO2 levels. These authors also report that increased CO2 concentrations and equivalence ratios result in greater CO emissions but reduced NOx levels because of quenched thermal NOx formation due to the lower flame temperatures (Marsh et al., 2016, 2017). However, the information available on the influence of S-EGR on gas turbine performance is scarce. Some works have experimentally analysed the effect of CO2 enhanced conditions, as discussed in Section 1.1 (e.g. Best et al., 2016; De Santis et al., 2016; Henke et al., 2013; Mansouri Majoumerd et al., 2014). These studies though have been carried out at conditions more relevant to EGR cases, which are significantly different from those expected in S-EGR configurations. Maximum CO2 concentrations in the flue gas of 6.3 vol% (typical of EGR) have been considered (e.g. Best et al., 2016), as opposed to the substantially higher CO2 enrichment levels expected in S-EGR systems, as considered at the start of this section. Therefore, experimental investigations specifically tailored to S-EGR concepts are needed to evaluate the influence that such conditions have on gas turbine performance.

analysed, together with the resulting CO, UHC and NOx emission trends under S-EGR simulated conditions. The novelty of this work includes the appraisal of mGT performance in terms of the compressor, turbine and combustion at part and full load operation under these S-EGR conditions. 2. Experimental methodology 2.1. Turbec T100 micro gas turbine system The UK’s PACT facilities are the national centre for research and development into carbon capture and advanced clean power generation (PACT, 2018). A number of facilities are available, including two Turbec T100 combined heat and power mGTs, both Series 1 and 3 models. The Series 3 model that was used for the baseline and simulated S-EGR experimental campaigns in this work can generate outputs of up to 100 kWe and 165 kWth. The electrical and overall efficiencies are 30% and 80%, respectively, with a recuperator and exhaust gas heat exchanger improving the total efficiency of the system (Turbec, 2009). The mGT scheme illustrated in Fig. 2 shows the electrical generator, centrifugal compressor and radial turbine are mounted on a single shaft, which operates with a maximum turbine speed of 70,000 rpm (Turbec, 2009). The ambient air entering the mGT passes through a coarse pre-filter and internal fine filter before separating into two flows: the primary flow used as the combustion air entering the compressor and the secondary flow for ventilation. The combustion air is compressed to a maximum pressure ratio of ˜4.5:1 at nominal conditions, which subsequently enters the recuperator and is pre-heated with hot exhaust gases prior to entering the combustion chamber. The pre-heated combustion air is mixed with the natural gas which is burnt in the combustion chamber. The flame is swirl-stabilised, with a non-premixed pilot flame used to further enhance flame stability. The fuel-lean pre-mixed combustion chamber ensures low CO, NOx and UHC emissions (Turbec, 2009). The exhaust gases leave the combustion chamber at ˜950°C (the turbine inlet temperature at nominal conditions) and expand through the turbine, driving the compressor and generator (Turbec, 2009). The

1.3. Aims and objectives In this context, the work here analyses the effects of S-EGR on the performance of a micro gas turbine. CO2-enhanced conditions characteristic of S-EGR processes are simulated in a Turbec T100 Series 3 mGT by injecting CO2 into the compressor inlet, mixing with the combustion air. A number of experimental tests have been performed over the 60–100 kWe operating envelope at CO2 injection rates of 100–300 kg/h (up to 9.4 vol% CO2 in the oxidiser stream) to mimic a range of S-EGR scenarios. Flue gas CO2 concentrations up to 10.1 vol% have been tested, which represent more than a 6 times increase in the CO2 content of the flue gas, similar to what is expected in S-EGR systems. Impacts on the operational performance of the mGT have been

Fig. 2. Schematic of the key turbine components and the instrumentation locations for the Turbec T100 mGT. 4

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

mGT control system keeps the turbine outlet temperature (TOT) constant at ˜645°C thus changing the air and fuel flowrates depending on the set power output and ambient conditions in each test. The exhaust gases leaving the turbine are used to pre-heat the combustion air in the recuperator, as shown, which works with an effectiveness close to 90% (Hohloch et al., 2010). The remaining thermal energy contained in the gas stream entering the exhaust gas heat exchanger (at ˜270°C at nominal conditions) is used to heat up the water stream. At PACT, the exhaust gases leave the heat exchanger at ˜70°C (nominal conditions) and are then discharged either to the onsite capture plant or to the atmosphere. The Turbec remote monitor and control (RMC) system monitors various mGT operational parameters, for example, the rotational speed. In addition, the FPServer and FirstOp software supplied by the manufacturer displays and records parameters defined by the operator. To complement the existing parameters measured by the RMC system, the mGT has been fitted with additional instrumentation to measure temperature, pressure and flowrates throughout the cycle, as shown in Fig. 2. A multiple channel National Instruments data acquisition system receives the electrical signals from each device and transfers this information to the LabVIEW software, which records the data measurements every second. The electrical efficiency of the mGT is then calculated using the registered net power output as recorded by the turbine and the measured natural gas consumption during the experiments. The mGT was also modified to incorporate a CO2 injection system into the compressor inlet to simulate the effects of selectively recycling CO2 under S-EGR configurations, as illustrated in Fig. 3. This system was designed to ensure an equal distribution of CO2 around the compressor inlet to aid mixing with the combustion air. CO2 is supplied from an external cryogenic storage tank and flows through an external evaporator and trim heater before entering the gas mixing skid at an equivalent temperature to the air. The skid controls the CO2 to the desired flowrate via two supply lines using pneumatic control valves. Each line is designed to deliver a flowrate of up to 150 kg/h, thus allowing a maximum CO2 flowrate of 300 kg/h into the system. The composition of gaseous emissions in the exhaust are assessed online using a Gasmet Fourier Transform Infrared (FTIR) DX4000 gas analyser and a Servomex ServoFlex mini multi-purpose 5200 gas analyser. The FTIR measures a range of gaseous emissions, including CO2, CO, NOx and a range of UHCs considered in this work. Two Servomex 5200 analysers are used for measuring O2 and CO2 concentrations in the exhaust gas and ventilation air outlet. These allow reporting for any CO2 losses to the ventilation air and checking the CO2 values are consistent with the FTIR measurements. All gaseous emissions in this work are reported on a dry basis. NOx emissions are usually reported corrected to 15% O2. However, ElKady et al. (2009) suggest reporting mass specific NOx emissions instead, based on the net power output under CO2-enhanced conditions, such as S-EGR. These authors suggest that representing NOx emission values corrected to 15% O2 under such conditions may be misleading, as the oxygen content at the combustor outlet is naturally reduced in these cases due to the partial replacement of the combustion air with the recirculated stream. As a result, correcting emissions to 15% O2 could lead to an artificial reduction in the values of NOx (ElKady et al., 2009). Therefore, NOx emissions reported here are presented using the NOx Emissions Index (EINOx, in g/kWh), which is determined from:

EINOx =

(NOx . 10 6. MW . nFG . 3600) Pe

2.2. Experimental campaigns Standard operating procedures for the experimental campaigns were developed to ensure each test followed the same method. Prior to starting any tests, it was ensured that the mGT was at steady state conditions and the TOT was ˜645°C, with the actual power output reaching the user-defined set point, the exhaust gas emission levels constant and the CO2 flowrate stable. Each test outlined below was conducted at least twice for a minimum period of 15 min according to ISO 2314 gas turbine acceptance test. A preliminary reference test campaign without CO2 injection was conducted prior to any mGT modifications (including before the additional instrumentation and the CO2 injection system were added). This was used to confirm that these changes did not affect the normal functioning and performance of the mGT. Subsequently, the performance of the mGT through a 60–100 kWe operating envelope was initially assessed without CO2 injection (baseline conditions) to provide a reference for comparison and to characterise the performance of the mGT with the variation in the load factor and as a function of the ambient temperature conditions. Then, CO2-enhancement experiments were carried out by injecting different quantities of CO2 (up to 300 kg/h) into the compressor inlet, thus replacing some of the combustion air, to simulate a range of S-EGR scenarios at power outputs of 60–100 kWe. Although the mGT can operate down to 40 kWe, tests at power outputs below 60 kWe were not considered as the mGT is far from its design conditions and its steadystate performance was compromised. Moreover, experiments at 90 kWe with a CO2 injection rate of 300 kg/h are omitted in the test matrix in Table 1 due to a malfunctioning of the fuel booster which prevented further tests. The test conditions for the baseline and CO2 enhancement experiments are summarised in Table 1. It is important to note that during the CO2 enhancement tests, the CO2 concentration recorded in the ventilation air outlet stream (see Fig. 1) was low – ranging from 0.2 to 0.4 vol% at the maximum CO2 flowrates tested. This, together with the CO2 concentration readings at the mGT flue gas, confirmed that CO2 losses through the system were negligible and the CO2 supplied through the injection lines in the S-EGR tests was effectively entering the compressor. 3. Results and discussion Table A1 in the Appendix provides a summary of results for streams 1, 2, 3, 4, and 9, as outlined in Fig. 2 – including the CO2 flowrate (stream 9) that was injected in each case. The CO2 concentrations recorded in the flue gas throughout the baseline and S-EGR tests are presented in Fig. 4. For the baseline, the reference flue gas

(1)

where NOx is the reported volumetric FTIR concentration in ppmv on a dry basis, MW is the molecular weight in g/mol, nFG is the molar flue gas flowrate in mol/s and Pe is the net power output in kW.

Fig. 3. Installed CO2 injection system into the compressor inlet.

5

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

The electrical efficiency of the mGT slightly decreased with CO2 injection, which has a higher heat capacity than air. This caused a decrease in the gas temperature at the combustor outlet and thus, at the turbine inlet, leading to a reduction in the power output (Best et al., 2016; Nikpey et al., 2014). To generate the desired electrical power output, the mGT marginally increased the fuel supply flowrate and hence the electrical efficiency decreased as shown (Nikpey et al., 2014). Throughout the tests the power output remained stable and the heat capacity of CO2 leads to the abovementioned effects. This performance is seen for all power outputs, with reductions in the electrical efficiency of between ˜4 and 8% at peak load and 60 kWe, respectively, when compared to the reference case with no CO2 injection. The impacts of S-EGR conditions on the rotational speed of the mGT are also illustrated in Fig. 5. As expected, this variable decreased with reducing electrical power output, as it is a function of the volumetric gas flowrates required at each condition. Focusing on the CO2 injection cases, the turbine rotational speed showed a slightly decreasing trend when the CO2 injection rate increased, with differential values ranging from 650 to 1400 rpm between the baseline and the higher CO2 injection rates. This is the result of a change in the properties of the working fluid due to the addition of CO2, which is denser than the air. The mGT adjusts the amount of fuel and air required to achieve the desired power output under each set of experimental conditions. As discussed above in Section 2.1, a fraction of the air flow is replaced by CO2 in the S-EGR tests, and therefore a similar mass flowrate passing through the turbine is equivalent to a lower volumetric air + CO2 flowrate to deliver the desired power output, thus reducing the turbine rotational speed. This trend was seen for all power outputs, with greater speed reductions observed with increasing CO2 injection rates. Similar trends have been reported by Mansouri Majoumerd et al. (2014) and Best et al. (2016), who both indicate that rotational speed reduces with increasing CO2 content in the oxidiser when testing lower CO2 enhancement levels typical of EGR. Small deviations from this trend are usually due to changes in the air temperature during the experiments, as seen in the 100 kWe case with a CO2 injection rate of 300 kg/h (Fig. 5a). The air density decreases when the temperature rises, and therefore the compressor rotates at higher speeds to maintain a similar mass flow of air through the turbine and thus deliver the same power output. This leads to an increase in the compressor work to generate the required electrical power output, which results in a slightly higher heat input and air demand. This is then associated with the increased rotational speeds in the mGT that offset the expected speed reductions due to the S-EGR conditions in this case. Throughout the simulated S-EGR tests the compressor discharge pressure seldom changed across the operating envelope tested as per the baseline results, with values ranging from 3.4 to 4.3 bara at 60–100 kWe, respectively. Similar unchanged performance is reported by the modelling study of Herraiz et al. (2018), who found that the pressure ratio is not significantly affected under S-EGR operation. The effect of S-EGR conditions on the compressor discharge temperature is shown in Fig. 6 for the different CO2 injection rates. As can be observed, this temperature increased with the power output due to the higher pressure ratios achieved. The addition of CO2 which was mixed with the combustion air changed the properties of the working fluid entering the compressor, which had a lower specific heat capacity due to the presence of CO2. This resulted in a narrower temperature change for compression at a given pressure ratio (Best et al., 2016; Herraiz et al., 2018; Sander et al., 2011). Therefore, it is expected that the compressor outlet temperature decreases with increasing CO2 injection rates for the same power output. However, the expected variation in the heat capacity ratio under S-EGR conditions is small, reducing by around 1–2% at the compressor inlet (Herraiz, 2016; Herraiz et al., 2018). Therefore, large changes in the compressor discharge temperature due to this effect were not expected in the CO2 injection tests, as seen in Fig. 6. Fig. 6 also shows there was a slight decrease in the temperature at

Table 1 Test conditions investigated for the baseline and CO2 enhancement experiments. Set power output (kW) CO2injection rate (kg/h)

60

70

80

90

100

0 (baseline) 100 150 200 250 300

❖ ❖ ❖ ❖ ❖ ❖

❖ ❖ ❖ ❖ ❖ ❖

❖ ❖ ❖ ❖ ❖ ❖

❖ ❖ ❖ ❖ ❖

❖ ❖ ❖ ❖ ❖ ❖

Fig. 4. Flue gas CO2 concentrations for the baseline and S-EGR CO2 injection cases across the 60–100 kWe operating envelope.

concentration of CO2 increased from 1.4 to 1.7 vol% when increasing the power output from 60 to 100 kWe. This is associated with the amount of excess air and the air-fuel ratio, which are higher at part load conditions, as well as the increased fuel usage to generate more power. The simulated S-EGR conditions tested in this work augmented the CO2 concentrations in the flue gas from 1.7 to 8.4 vol% (0–300 kg/h CO2) at 100 kWe and from 1.4 to 10.1 vol% (0–300 kg/h CO2) at 60 kWe. This represents a ˜400% and ˜600% increase at 100 and 60 kWe, respectively, at maximum CO2 addition compared to the baseline conditions; these are in the range of the expected changes in S-EGR systems as discussed in Section 1.2. The key objective of this study was to investigate the effects of these S-EGR conditions on the mGT performance. For this purpose, the impacts on electrical efficiency, turbine rotational speed and compressor discharge temperature are analysed below, as well as the effects on NOx, CO and UHC emissions. 3.1. Influence of S-EGR on mGT performance Fig. 5 illustrates the effects of S-EGR on the mGT electrical efficiency and rotational speed at power outputs of 60 and 100 kWe, showing the temperature at the compressor inlet for each test. Similar trends were also obtained at the intermediate power outputs (70–90 kWe). This is compared to the baseline values calculated at the average air temperature in the S-EGR cases using the baseline results obtained, which characterise the mGT performance at different air inlet temperatures for each power output. Lower values of electrical efficiency were obtained at decreasing power outputs, as expected (Turbec, 2009). 6

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

Fig. 5. Influence of different S-EGR CO2 injection rates on the compressor inlet temperature, electrical efficiency and rotational speed at (a) 100 kWe and (b) 60 kWe.

the outlet of the compressor when increasing the CO2 injection rate at power outputs of 80 and 90 kWe, as expected (7 and 6°C, respectively, between the minimum and maximum CO2 injection rate). However, no clear trend was seen for the other power outputs. This is because the compressor outlet temperature is influenced more by the system air inlet temperature, which varies slightly between experiments, thus concealing the effect of a marginal reduction in the heat capacity ratio.

tested in the mGT across all power outputs. The reduced O2 concentrations in the oxidiser, together with the increased CO2 content, impact the combustion performance. Reductions in the laminar flame speed and changes in the velocity field have been reported under CO2enhanced conditions because CO2 dilutes the combustion mixture and the specific heat capacity changes due to this, thereby decreasing the flame temperature and the burning velocity (De Santis et al., 2016; Hinton and Stone, 2014). As a result, the combustion efficiency was reduced, as illustrated in Fig. 5, where lower O2 availability means the combustion performance is reduced because of the modified working fluid entering the compressor. To alleviate this, reprogramming of the compressor and turbine maps to accommodate enhanced CO2 streams would be required. Furthermore, the emission performance of the mGT was modified, also due to this change in specific heat capacity and thus in the flame speeds and reaction rates, as depicted in Figs. 7–9 for NOx, CO and UHCs. The measured NOx emissions for the baseline and S-EGR cases are shown in Fig. 7, given by the emission index as per Eq. (1) (ElKady et al., 2009). In general, it can be seen that as the net electrical power output decreased, the NOx emissions also decreased for both baseline

3.2. Influence of S-EGR on emissions performance: NOx, CO and UHCs As discussed in the previous section, S-EGR operation changes the properties of the working fluid and reduces the O2 concentration in the oxidiser that enters the combustor, thus affecting combustion efficiency and emissions performance (ElKady et al., 2009; Marsh et al., 2016, 2017). The O2 concentrations at the compressor inlet for the tests included in this work are shown in Table 2 across the 60–100 kWe operating envelope. As can be seen, the O2 concentration of the air + CO2 mixture at the compressor inlet and thus, at the inlet of the combustion chamber, ranged from 19.2 to 19.8 vol% for the maximum CO2 injection rates 7

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

Fig. 6. Influence of different S-EGR CO2 injection rates on the compressor discharge temperature across the 60–100 kWe operating envelope.

This represents a ˜61% reduction from 0.31 g/kWh (at 100 kWe) to 0.12 g/kWh (at 60 kWe) for the baseline tests. Despite some dispersion in Fig. 7, the influence of S-EGR overall on NOx also demonstrates some decreases in emissions with the highest CO2 injection rates compared to the baseline tests across the operating envelope. Similarly to the baseline cases, the reason for NOx emission reductions is due to slower flame speeds and greater radiative heat losses caused by the change in the properties of the working fluid under CO2-enhanced conditions, which leads to a reduction in the flame temperature and therefore lower thermal NOx production (De Santis et al., 2016; Lee et al., 2013; Marsh et al., 2017). It could also be expected that NOx emissions reduce with increasing CO2 injection rates. However, there is no clear trend for the S-EGR cases shown in Fig. 7. A possible reason for this is likely due to the very lean flame conditions under normal operation without enhanced CO2 conditions. If the turbine was reprogrammed to consider enhanced CO2 streams, the NOx emissions would likely follow a clearer trend, because the excess air requirements would change, and hence the flame conditions would be suitable for enhanced CO2 streams, and not just for air-fired combustion.

Table 2 O2 concentration (vol%) at the compressor inlet for the baseline and CO2 enhancement experiments. Power output (kWe)

CO2 injection rate (kg/h)

0 (baseline) 100 150 200 250 300

60

70

80

90

100

21.0 20.4 20.1 19.8 19.5 19.2

21.0 20.4 20.2 19.9 19.6 19.3

21.0 20.5 20.2 20.0 19.7 19.4

21.0 20.5 20.3 20.0 19.8 –

21.0 20.5 20.3 20.1 19.8 19.6

and S-EGR conditions. This can be explained on the basis that the mGT operates under lean premixed combustion conditions where the formation of thermal NOx is associated with the Zeldovich mechanism (Seliger et al., 2015). The mGT NOx emissions are then mainly related to the combustion temperature, which decreases at part load due to the higher air-fuel ratio in these cases and which consequently leads to lower thermal NOx values (De Santis et al., 2016; Seliger et al., 2015).

Fig. 7. Influence of the different S-EGR CO2 injection rates on flue gas NOx concentrations across the 60–100 kWe operating envelope. 8

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

dissociation, due to the reduction in temperatures, mixing and flame speeds considered in the preceding paragraph. At part load operation, higher residence times of the gases inside the combustor would be required to ensure complete combustion and minimise CO emissions, as shown in the study by Zanger et al. (2013) using a Turbec T100 mGT. These effects become more prominent with increasing CO2 injection at each power output, especially at lower load factors, thus leading to the increased emissions shown in Figs. 8 and 9. CO concentrations in the flue gas increased from 2 ppmv (baseline) to 24 ppmv when injecting 300 kg/h CO2 at 100 kWe, but from 96 ppmv (baseline) to 516 ppmv at 60 kWe with a 300 kg/h CO2 injection rate. The reported levels of UHC presented in Fig. 9 also follow a similar trend, thus supporting the incomplete combustion effects just considered. UHC emissions remained at values lower than 2–3 ppmv for all CO2 injection cases when the nominal power output (100 kWe) was set. However, they substantially increased under S-EGR conditions at lower power outputs, rising from 7 ppmv (baseline) to 104 ppmv (for 300 kg/ h CO2 injection rate) at 60 kWe. These results show the emission performance of the mGT varies very little under the S-EGR simulated conditions when it operates at nominal power at which the mGT was designed (100 kWe), with limited changes at nearby power outputs. However, it is notably affected by increases in the CO2 content in the oxidiser at lower load factors. These large increases in emissions, substantially higher than those reported under EGR conditions (Best et al., 2016), indicate that modifications to the design of the combustion chamber might be needed for gas turbines operating under S-EGR scenarios to ensure appropriate full- and part-load emission performance, which will be required for load-following installations. This may have additional consequences on the capture plant; not only may elevated levels of CO and UHC emissions affect the solvent of choice in an ACP, but the plant may also find load-following and rapid changes in the output from the CCGT difficult to accommodate.

Fig. 8. Influence of the different S-EGR CO2 injection rates on flue gas CO concentrations across the 60–100 kWe operating envelope.

4. Conclusions A pilot-scale experimental study investigating the performance of a Turbec T100 Series 3 mGT under simulated S-EGR conditions has been conducted. The mGT has been modified to incorporate a CO2 delivery system to investigate a range of conditions characteristic of S-EGR. The operating conditions tested led to a maximum increase in the flue gas CO2 concentrations of ˜400% and ˜600% at 100 and 60 kWe, respectively, similar to what is expected in S-EGR systems. The experimental results indicated that the electrical efficiency decreased by up to ˜8% under S-EGR conditions (at the highest CO2 flowrate and lowest load factor) compared to the baseline values without CO2 injection, due to the modified heat capacity of the working fluid which leads to reductions in the turbine inlet temperature. The reduction was ˜4% at peak load however. The mGT rotational speed decreased slightly, by 650–1400 rpm (up to ˜2%), under S-EGR conditions compared to the baseline, which is attributed to the higher density of the working fluid. However, variations in compressor discharge pressure were negligible throughout the CO2 enhancement tests. The compressor discharge temperature showed a slightly decreasing trend in the S-EGR scenarios. The inlet O2 concentration decreased marginally with increasing CO2 injection rates, thus impacting combustion and emission performance. The NOx emissions were lower under the most extreme simulated S-EGR conditions (300 kg/h of CO2) compared to the baseline (no CO2 addition), due to the effect that lower associated combustion temperatures have on thermal NOx formation. Further, emissions of CO and UHC increased under S-EGR conditions due to incomplete combustion at lower temperatures, but most notably at lower power outputs; the increases in emissions at higher load factors were limited. The experimental results demonstrate stable mGT operation across

Fig. 9. Influence of the different S-EGR CO2 injection rates on flue gas UHC concentrations across the 60–100 kWe operating envelope.

Measured CO and UHC emissions are illustrated in Figs. 8 and 9 across all power outputs. As can be seen in Fig. 8, CO emissions for the baseline tests increased from 2 ppmv at nominal conditions (100 kWe) to 96 ppmv at lower loads (60 kWe). Similarly, UHC emissions also experienced a sharp increase when the mGT was operated at low power outputs, being close to zero at 100 kWe and rising to 7 ppmv at 60 kWe for baseline conditions (Fig. 9). The increase in the recorded CO and UHC emissions at part load can be associated with incomplete combustion due to poor fuel and air mixing, insufficient flame stability and lower combustion temperatures (Lefebvre et al., 2010; Seliger et al., 2015). The CO + OH → CO2 + H reaction rate is thus reduced which leads to increased CO emissions (Seliger et al., 2015). CO2 thermal dissociation might also be considered as a possible reason for increases in the CO emissions, which typically occurs at flame temperatures above 1500°C (Evulet et al., 2009). De Santis et al. (2016) found that the maximum adiabatic flame temperature in a Turbec T100 combustor under EGR to be 2200°C, which suggests that possible CO2 thermal dissociation may occur under simulated S-EGR. However, Marsh et al. (2017) indicated that increased CO emissions are likely to be associated with incomplete combustion instead of CO2

9

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al.

the operating envelope considered, with very little negative impact in terms of emissions or efficiency at higher loads. However, the emissions performance indicates that modifications to the design of the combustion chamber might be needed for gas turbines operating under S-EGR scenarios to meet emission targets, especially at part loads. Reprogramming of compressor and turbine maps to accommodate CO2 enhanced streams under S-EGR would be beneficial to maintain system performance. Overall, the increased CO2 concentrations reported here through the application of S-EGR to the gas turbine offer the potential for significant economic and energy savings for the downstream CO2 capture plant (Akram et al., 2016).

interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements The financial support of the UK Engineering and Physical Sciences Research Council (EPSRC) to the SELECT project (Selective Exhaust Gas Recirculation for Carbon Capture with Gas Turbines: Integration, Intensification, Scale-up and Optimisation, EP/M001482/1) is greatly acknowledged. The authors acknowledge that the PACT Facilities, funded by the Department for Business, Energy and Industrial Strategy and the EPSRC, have been used for the experimental work reported in this publication.

Declaration of Competing Interest The authors declare that they have no known competing financial Appendix A

Table A1 Results table for streams 1, 2, 3, 4 and 9 shown in Fig. 2. Description

Air Inlet

Compressor Out

Fuel Inlet

Turbine Outlet

CO2 Inlet

Stream Number

1

2

3

4

9

Electrical Output (kWe)

T (°C)

P (bara)

F (kg/s)

T (°C)

P (bara)

P (bara)

F (kg/s)

T (°C)

P (bara)

F (kg/s)

T (°C)

100.0 99.9 100.0 100.0 100.0 100.0 90.0 90.0 90.0 89.9 90.0 80.0 80.0 79.9 80.0 80.0 79.9 70.0 70.0 70.0 70.0 70.0 70.0 60.0 60.0 60.0 60.0 60.0 60.6

17.3 15.2 14.8 14.6 13.9 16.4 20.5 15.3 15.6 15.3 13.8 20.2 15.3 15.2 15.0 13.8 14.9 20.1 15.1 15.4 15.7 13.8 17.9 19.7 14.9 15.2 15.8 15.7 15.5

1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

0.8 0.8 0.8 0.8 0.8 0.8 0.7 0.8 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.6 0.7 0.7 0.7 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6

200.7 197.0 195.9 197.0 190.9 194.9 196.3 187.3 186.8 185.5 181.7 185.0 177.3 175.6 175.6 172.5 170.6 174.9 167.3 166.6 166.5 163.1 168.7 164.2 156.9 156.5 157.2 156.5 155.6

4.2 4.2 4.2 4.2 4.2 4.3 4.1 4.0 4.0 4.0 4.0 3.8 3.8 3.7 3.7 3.8 3.8 3.6 3.5 3.5 3.5 3.5 3.6 3.4 3.3 3.3 3.3 3.3 3.3

6.0 6.0 6.0 6.0 6.0 6.0 3.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0

26.8 26.9 27.1 27.0 26.8 26.7 24.4 24.6 24.4 24.4 24.5 21.9 22.1 21.9 22.1 22.2 21.4 19.6 19.7 19.8 19.9 20.1 20.2 17.5 17.7 17.7 17.8 17.7 17.9

645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 644.7 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 645.0 644.0

1.0 1.0 1.0 1.0 1.0 1.0 1.1 1.0 1.0 1.0 1.0 1.1 1.0 1.0 1.0 1.0 1.1 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

0.0 100.0 150.4 200.0 249.8 300.0 0.0 100.0 149.9 200.0 249.9 0.0 100.0 150.0 200.2 249.8 299.9 0.0 100.0 149.9 199.7 249.6 299.9 0.0 100.0 150.0 200.2 249.9 299.9

14.5 14.6 14.1 14.6 13.8 16.4 17.7 14.5 14.6 14.1 14.0 17.5 14.5 14.3 14.1 13.8 15.1 17.4 14.8 14.5 14.8 14.1 17.4 17.1 14.8 14.5 14.8 16.2 15.7

enhancement on the performance of a micro gas turbine with a pilot-scale CO2 capture plant. Chem. Eng. Res. Des. 117, 11–23. https://doi.org/10.1016/j.cherd. 2016.10.001. Baker, R.W., Freeman, B., Kniep, J., Wei, X., Merkel, T., 2017. CO2 capture from natural gas power plants using selective exhaust gas recycle membrane designs. Int. J. Greenh. Gas Cont. 66, 35–47. https://doi.org/10.1016/j.ijggc.2017.08.016. Best, T., Finney, K.N., Ingham, D.B., Pourkashanian, M., 2016. Impact of CO2-enriched combustion air on micro-gas turbine performance for carbon capture. Energ. 115, 1138–1147. https://doi.org/10.1016/j.energy.2016.09.075. Biliyok, C., Yeung, H., 2013. Evaluation of natural gas combined cycle power plant for post-combustion CO2 capture integration. Int. J. Greenh. Gas Cont. 19, 396–405. https://doi.org/10.1016/j.ijggc.2013.10.003. Bolland, O., Mathieu, P., 1998. Comparison of two CO2 removal options in combined cycle power plants. Energ. Convers. Manag. 39, 1653–1663. https://doi.org/10. 1016/S0196-8904(98)00078-8.

References Aboudheir, A., ElMoudir, W., 2009. Performance of formulated solvent in handling of enriched CO2 flue gas stream. Energ. Pro. 1, 195–204. https://doi.org/10.1016/j. egypro.2009.01.028. Adams, T., Mac Dowell, N., 2016. Off-design point modelling of a 420 MW CCGT power plant integrated with an amine-based post-combustion CO2 capture and compression process. Appl. Energy 178, 681–702. https://doi.org/10.1016/j.apenergy.2016.06. 087. Akram, M., Ali, U., Best, T., Blakey, S., Finney, K.N., Pourkashanian, M., 2016. Performance evaluation of PACT pilot-plant for CO2 capture from gas turbines with exhaust gas recycle. Int. J. Greenh. Gas Cont. 47, 137–150. https://doi.org/10.1016/ j.ijggc.2016.01.047. Ali, U., Hughes, K.J., Ingham, D.B., Ma, L., Pourkashanian, M., 2017. Effect of the CO2

10

International Journal of Greenhouse Gas Control 90 (2019) 102809

J.-M. Bellas, et al. Bolland, O., Sæther, S., 1992. New concepts for natural gas fired power plants which simplify the recovery of carbon dioxide. Energ. Convers. Manag. 33, 467–475. https://doi.org/10.1016/0196-8904(92)90045-X. Darabkhani, H.G., Jurado, N., Prpich, G., Oakey, J.E., Wagland, S.T., Anthony, E.J., 2018. Design, process simulation and construction of a 100 kW pilot-scale CO2 membrane rig: improving in situ CO2 capture using selective exhaust gas recirculation (S-EGR). J. Nat. Gas Sci. Eng. 50, 128–138. https://doi.org/10.1016/J.JNGSE.2017.09.012. De Santis, A., Ingham, D.B., Ma, L., Pourkashanian, M., 2016. CFD analysis of exhaust gas recirculation in a micro gas turbine combustor for CO2 capture. Fuel 173, 146–154. https://doi.org/10.1016/j.fuel.2016.01.063. Diego, M.E., Akram, M., Bellas, J.-M., Finney, K.N., Pourkashanian, M., 2017a. Making gas-CCS a commercial reality: the challenges of scaling up. Greenh. Gases Sci. Technol. 7, 778–801. https://doi.org/10.1002/ghg.1695. Diego, M.E., Bellas, J.-M., Pourkashanian, M., 2017b. Process analysis of selective exhaust gas recirculation for CO2 capture in natural gas combined cycle power plants using amines. J. Eng. Gas Turb. Power 139, 121701. https://doi.org/10.1115/1.4037323. Diego, M.E., Bellas, J.-M., Pourkashanian, M., 2018. Techno-economic analysis of a hybrid CO2 capture system for natural gas combined cycles with selective exhaust gas recirculation. Appl. Energy 215, 778–791. https://doi.org/10.1016/j.apenergy.2018. 02.066. Ditaranto, M., Hals, J., Bjorge, T., 2009. Investigation on the in-flame NO reburning in turbine exhaust gas. Proc. Combust. Inst. 32, 2659–2666. https://doi.org/10.1016/j. proci.2008.07.002. II. Elkady, A.M., Evulet, A., Brand, A., Ursin, T.P., Lynghjem, A., 2008. Exhaust gas recirculation in DLN F-class gas turbines for post-combustion CO2 capture. Pro. Asme Turbo Expo. 3, 847–854. https://doi.org/10.1115/GT2008-51152. ElKady, A.M., Evulet, A., Brand, A., Ursin, T.P., Lynghjem, A., 2009. Application of exhaust gas recirculation in a DLN F-class combustion system for postcombustion carbon capture. J. Eng. Gas Turb. Power 131, 034505. https://doi.org/10.1115/1. 2982158. European Commission, 2018. 2050 Low-Carbon Economy, Climate Action. (accessed 9th April 2019). https://ec.europa.eu/clima/policies/strategies/2050_en. Evulet, A.T., ELKady, A.M., Branda, A.R., Chinn, D., 2009. On the performance and operability of GE’s dry low NOx combustors utilizing exhaust gas recirculation for postcombustion carbon capture. Energ. Pro. 1, 3809–3816. https://doi.org/10.1016/ j.egypro.2009.02.182. Finney, K.N., De Santis, A., Best, T., Clements, A.G., Diego, M.E., Pourkashanian, M., 2018. Exhaust gas recycling for enhanced CO2 capture: experimental and CFD studies on a micro-gas turbine. Ind. Comb., Article number 201610, 1–31. Gouedard, C., Picq, D., Launay, F., Carrette, P.L., 2012. Amine degradation in CO2 capture. I. A review. Int. J. Greenh. Gas Cont. 10, 244–270. https://doi.org/10.1016/j. ijggc.2012.06.015. Hasemann, S., Huber, A., Naumann, C., Aigner, M., 2017. Investigation of a FLOX ® -based combustor for a micro gas turbine with exhaust gas recirculation, in: volume 4B: combustion. Fuels and Emissions. ASME. https://doi.org/10.1115/GT201764396. p. V04BT04A012. Henke, M., Monz, T., Aigner, M., 2013. Inverted Brayton cycle with exhaust gas recirculation – a numerical investigation. J. Eng. Gas Turb. Power 135, 091203. https://doi.org/10.1115/1.4024954. Herraiz, L., 2016. Selective Exhaust Gas Recirculation in Combined Cycle Gas Turbine Power Plants with Post-combustion Carbon Capture. Ph.D. Thesis. The University of Edinburgh. https://www.era.lib.ed.ac.uk/handle/1842/20485. Herraiz, L., Fernández, E.S., Palfi, E., Lucquiaud, M., 2018. Selective exhaust gas recirculation in combined cycle gas turbine power plants with post-combustion CO2 capture. Int. J. Greenh. Gas Cont. 71, 303–321. https://doi.org/10.1016/J.IJGGC. 2018.01.017. Hinton, N., Stone, R., 2014. Laminar burning velocity measurements of methane and carbon dioxide mixtures (biogas) over wide ranging temperatures and pressures. Fuel 116, 743–750. https://doi.org/10.1016/J.FUEL.2013.08.069. Hohloch, M., Zanger, J., Widenhorn, A., Aigner, M., 2010. Experimental characterization of a micro gas turbine test rig. Volume 3: Controls, Diagnostics and Instrumentation. Cycle Innovations; Marine. ASME, pp. 671–681. https://doi.org/10.1115/GT201022799. Hu, Y., Ahn, H., 2017. Process integration of a calcium-looping process with a natural gas combined cycle power plant for CO2 capture and its improvement by exhaust gas recirculation. Appl. Energy 187, 480–488. https://doi.org/10.1016/j.apenergy.2016. 11.014. IEA, 2017. World Energy Outlook 2017. Paris, France. https://doi.org/10.1016/03014215(73)90024-4.

IEAGHG, 2012. CO2 Capture at Gas Fired Power Plants: 2012/8. Cheltenham, UK. . Lee, K., Kim, H., Park, P., Yang, S., Ko, Y., 2013. CO2 radiation heat loss effects on NOx emissions and combustion instabilities in lean premixed flames. Fuel 106, 682–689. https://doi.org/10.1016/J.FUEL.2012.12.048. Lefebvre, A.H., Arthur, H., Ballal, D.R., 2010. Gas Turbine Combustion: Alternative Fuels and Emissions. Taylor & Francis. Li, H., Ditaranto, M., Berstad, D., 2011a. Technologies for increasing CO2 concentration in exhaust gas from natural gas-fired power production with post-combustion, aminebased CO2 capture. Energ. 36, 1124–1133. https://doi.org/10.1016/j.energy.2010. 11.037. Li, H., Haugen, G., Ditaranto, M., Berstad, D., Jordal, K., 2011b. Impacts of exhaust gas recirculation (EGR) on the natural gas combined cycle integrated with chemical absorption CO2 capture technology. Energ. Pro. 4, 1411–1418. https://doi.org/10. 1016/j.egypro.2011.02.006. Mansouri Majoumerd, M., Nikpey Somehsaraei, H., Assadi, M., Breuhaus, P., 2014. Micro gas turbine configurations with carbon capture – performance assessment using a validated thermodynamic model. Appl. Therm. Eng. 73, 172–184. https://doi.org/ 10.1016/j.applthermaleng.2014.07.043. Marsh, R., Giles, A., Runyon, J., Pugh, D., Bowen, P., Morris, S., Valera-Medina, A., Best, T., Finney, K., Pourkashanian, M., 2016. The future of gas turbine technology. 8th International Gas Turbine Conference, Belgium. Paper ID Number (32-IGTC16). Marsh, R., Runyon, J., Giles, A., Morris, S., Pugh, D., Valera-Medina, A., Bowen, P., 2017. Premixed methane oxycombustion in nitrogen and carbon dioxide atmospheres: measurement of operating limits, flame location and emissions. Proc. Combust. Inst. 36, 3949–3958. https://doi.org/10.1016/J.PROCI.2016.06.057. Merkel, T.C., Wei, X., He, Z., White, L.S., Wijmans, J.G., Baker, R.W., 2013. Selective exhaust gas recycle with membranes for CO2 capture from natural gas combined cycle power plants. Ind. Eng. Chem. Res. 52, 1150–1159. https://doi.org/10.1021/ ie302110z. NETL, 2013. Current and Future Technologies for Natural Gas Combined Cycle (NGCC) Power Plants. Pittsburgh, USA. https://doi.org/DOE/NETL-341/061013. . Nikpey, H., Mansouri Majoumerd, M., Assadi, M., Breuhaus, P., 2014. Thermodynamic analysis of innovative micro gas turbine cycles. Volume 3A: Coal, Biomass and Alternative Fuels. Cycle Innovations; Electric Power; Industrial and Cogeneration. ASMEhttps://doi.org/10.1115/GT2014-26917. V03AT07A029. PACT, 2018. PACT – Pilot-Scale Advanced CO2 Capture Technology Facilities. (accessed 23rd August 2018). https://pact.group.shef.ac.uk/. Røkke, P.E., Hustad, J.E., 2005. Exhaust gas recirculation in gas turbines for reduction of CO2 emissions: combustion testing with focus on stability and emissions. Int. J. Thermo. 8, 167–173. https://doi.org/10.5541/ijot.158. Sander, F., Carroni, R., Rofka, S., Benz, E., 2011. Flue gas recirculation in a gas turbine: impact on performance and operational behavior. Volume 4: Cycle Innovations; Fans and Blowers; Industrial and Cogeneration; Manufacturing Materials and Metallurgy; Marine; Oil and Gas Applications. ASME, pp. 123–132. https://doi.org/10.1115/ GT2011-45608. Seliger, H., Huber, A., Aigner, M., 2015. Experimental investigation of a flox®-based combustor for a small-scale gas turbine based CHP system under atmospheric conditions. ASME Turbo Expo 2015: Turbine Technical Conference and Exposition p. V04BT04A004. Sipöcz, N., Tobiesen, F.A., 2012. Natural gas combined cycle power plants with CO2 capture – opportunities to reduce cost. Int. J. Greenh. Gas Cont. 7, 98–106. https:// doi.org/10.1016/j.ijggc.2012.01.003. Supap, T., Idem, R., Tontiwachwuthikul, P., Saiwan, C., 2006. The roles of O2 and SO2 in the degradation of monoethanolamine during CO2 absorption from industrial flue gas streams. 2006 IEEE EIC Climate Change Conference 1–6. https://doi.org/10.1109/ EICCCC.2006.277275. Turbec, 2009. T100 Microturbine System: Technical Description – T100 Natural Gas. D14127–03. . Turi, D.M.M., Ho, M., Ferrari, M.C.C., Chiesa, P., Wiley, D.E.E., Romano, M.C.C., 2017. CO2 capture from natural gas combined cycles by CO2 selective membranes. Int. J. Greenh. Gas Cont. 61, 168–183. https://doi.org/10.1016/j.ijggc.2017.03.022. UNFCCC, 2015. Adoption of the Paris Agreement. United Nations Framework Convention on Climate Change (UNFCCC), Paris. USDOE/NETL, 2013. DOE/NETL Advanced Carbon Dioxide Capture R&D Program: Technology Update. Zanger, J., Thomas, M., Manfre, A., 2013. Experimental investigation of the influence of combustor cooling on the characteristics of a FLOX©-based micro gas turbine combustor. Progress in Gas Turbine Performance. InTech.https://doi.org/10.5772/ 54405.

11