Fuel Processing Technology 92 (2011) 1685–1695

Contents lists available at ScienceDirect

Fuel Processing Technology

j ourna l homepage: www.e lsev ie r.com/ locate / fuproc

Optimal design and integration of an air separation unit (ASU) for an integratedgasification combined cycle (IGCC) power plant with CO2 capture

Dustin Jones a, Debangsu Bhattacharyya a,⁎, Richard Turton a, Stephen E. Zitney b

a Department of Chemical Engineering, West Virginia University, Morgantown, WV 26506, United Statesb U.S. Department of Energy, National Energy Technology Laboratory, Morgantown, WV 26507, United States

⁎ Corresponding author.E-mail address: debangsu.bhattacharyya@mail.wvu.

0378-3820/$ – see front matter © 2011 Elsevier B.V. Aldoi:10.1016/j.fuproc.2011.04.018

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 December 2010Received in revised form 10 March 2011Accepted 9 April 2011Available online 19 May 2011

Keywords:Elevated-pressureASUIGCCPumped liquid oxygen (PLOX)OptimizationIntegration

The air separation unit (ASU) plays a key role in improving the efficiency, availability, and operability of anoxygen-fed integrated gasification combined cycle (IGCC) power plant. An optimal integration between theASU and the balance of the plant, especially the gasifier and the gas turbine (GT), has significant potential forenhancing the overall plant efficiency. Considering the higher operating pressure of the GT, an elevated-pressure air separation unit (EP-ASU) is usually favored instead of the conventional low-pressure airseparation units (LP-ASU). In addition, a pumped liquid oxygen (PLOX) cycle is usually chosen if the operatingpressure of the gasifier is high. A PLOX cycle helps to improve plant safety and availability and to decrease thecapital cost by reducing the size of the oxygen compressor or by eliminating it completely. However, therefrigeration lost in withdrawn liquid oxygenmust be efficiently recovered. This paper considers five differentconfigurations of an ASU with PLOX cycle and compares their power consumptions with an EP-ASU with atraditional gaseous oxygen (GOX) cycle. The study shows that an optimally designed EP-ASU with a PLOXcycle can have similar power consumption to that of an EP-ASU with GOX cycle in the case of 100% nitrogenintegration. In the case of an IGCC with pre-combustion CO2 capture, the lower heating value (LHV) of theshifted syngas, both on a mass and volumetric basis, is in between the LHV of the unshifted syngas from anIGCC plant and the LHV of natural gas, for which the GTs are generally designed. The optimal air integration inthe case of a shifted syngas is found to be much lower than that of an unshifted syngas. This paper concurswith the existing literature that the optimal integration occurs when air extracted from the GT can be replacedwith the nitrogen from the ASU without exceeding mass/volumetric flow limitations of the GT. Consideringnitrogen and air integration between the ASU and the GT, this paper compares the power savings in an LP-ASUwith a PLOX cycle to the power savings in an EP-ASU with GOX cycle and EP-ASU with PLOX cycle. The resultsshow that an LP-ASU with a PLOX cycle has less power consumption if the nitrogen integration levels are lessthan 50–60%. In addition, a study is carried out by varying the concentration of nitrogen and steam in the fueldiluents to the GT while the NOx level was maintained constant. The study shows that when the nitrogeninjection rate exceeds 50%, an EP-ASU with a PLOX cycle is a better option than an LP-ASU with a PLOX cycle.This paper shows that an optimal design and integration of an ASU with the balance of the plant can help toincrease the net power generation from an IGCC plant with CO2 capture.

edu (D. Bhattacharyya).

l rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

Integrated gasification combined cycle (IGCC) is a promisingtechnology for generating clean, affordable, and secure power.However, a coal-fed IGCC power plant has lower net plant efficiencycompared to a conventional natural gas combined cycle (NGCC) powerplant. A recent NETL study shows that the net plant efficiency of a coal-fed IGCC plant with a General Electric Energy (GEE)-type gasifier is38.2% compared to 50.8% efficiency for NGCC plants [1]. The net plantefficiency of the IGCC plant further decreases to 32.5% when the CO2

capture option is considered. In order to make IGCC technology morecompetitive, efforts should be exerted to improve its efficiency.Recently, an optimization study has been carried out for improvingthe efficiency of IGCC plants with CO2 capture [2]. The plant layout isshown in Fig. 1. The study considered a GEE-type entrained flowgasifier mainly because of its high carbon conversion rates andenvironmental advantages [2,3]. In high temperature gasifiers, suchas the GEE-type, the gasifier efficiency increases and the size ofdownstream equipment is reduced when oxygen, instead of air, is usedin the gasifier [4]. Fig. 1 shows that the shifted syngas from the gasifiergoes to a SELEXOLTM solvent unit for acid gas removal. Use of oxygen inthe gasifier increases the partial pressure of the acid gases at the inlet tothe acid gas removal (AGR) unit. This, in turn, decreases the auxiliaryenergy consumption in the AGR unit because of a reduced circulation

Fig. 1. Block Flow Diagram of IGCC Power Plant with Carbon Capture.

1686 D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

rate of the solvent (for a solvent-based AGR technology) and a decreasein the refrigeration duty, if refrigeration is considered. Even though theuse of oxygen has a number of advantages, the NETL study shows thatthe air separation unit (ASU) in an GEE-gasifier-based IGCC plant withCO2 capture, consumes about 64% of the total auxiliary power and about16% of the gross power when the power consumptions in the main aircompressor (MAC), oxygen compressor, and nitrogen compressor areincluded [1].

Three product streams from an ASU are of interest for an IGCC plantas shown in Fig. 1. The oxygen stream is mainly used in the gasifierwhile a small portion of it goes to the Claus unit for combustinghydrogen sulfide. It has been found that an oxygen purity of 95 mol% isoptimal for a cryogenic ASU [4,5]. A low-pressure nitrogen stream,produced from the low-pressure (LP) column of the ASU, is used as afuel diluent for NOx reduction and power augmentation in the gasturbine (GT) as long as the oxygen concentration remains below certainlimits to prevent flashback. The maximum oxygen concentration in thediluent nitrogen is desired to be less than 2 mol% for the unshiftedsyngas [10,15]. In the case of a hydrogen-rich, shifted syngas, thisspecification may differ based on the hydrogen content of the syngas,the type of GT burner (premix-type vs diffusion-type), and the overalldesign of the GT combustor. As this specification for shifted syngascould not be found in the open literature, the maximum oxygenconcentration in the diluent nitrogen is considered to be 2 mol%.Furthermore, a high purity nitrogen (b50 ppm O2) is also producedfrom the top of the high-pressure (HP) column for use in purges, pads,and sootblowing in the fuel transfer system and gasification plant [10].

In the design of an ASU as part of an IGCC plant, the main designdecisions are: operating pressure of the ASU and the phase of theproducts, mainly O2, withdrawn from the columns. Based on theoperating pressure of the ASUs, they can be classified as low-pressure(LP) ASU or elevated-pressure (EP) ASU. The operating pressure of theHP column in an LP ASU is usually 4–7 atm compared to 10–14 atm inan EP-ASU. In addition, O2 can be withdrawn from the bottom of the

LP column in gas phase, known as gaseous oxygen (GOX) cycle or inliquid phase known as liquid oxygen (LOX) cycle.

The ASU is generally operated at elevated pressures for IGCC plantssince both the oxygen and nitrogen are required at high pressures forthe gasifier and the GT, respectively. Even though the powerconsumptions in the oxygen and nitrogen compressors decrease in anelevated-pressure (EP) ASU, the power consumption in the MACincreases to compensate. In addition, the higher operating pressure inan EP-ASU also decreases the relative volatility of oxygen and nitrogen,causing significant separation challenges [8,19,21]. However, anoptimal integration between the ASU and GT can significantly improvethe power production in an IGCC plant as well as its operability. For airintegration between the GT and MAC, a part of the ASU air is extractedfrom the GT. If a considerable portion of the ASU air is extracted fromthe GT, it is generally desirable to have the operating pressure of theASU close to that of the gas turbine in order to maximize the savings inthe oxygen and nitrogen compression [6–8,10,21,28].

For a traditional standalone ASU, a gaseous oxygen (GOX) cycle istypically chosen. This cycle takes only gaseous products from thedistillation columns and is less power intensive when products aredesired at near atmospheric pressures [16]. However, a pumped liquidoxygen (PLOX) cycle, where liquid oxygen is withdrawn from thesump of the LP column, is generally chosen when the gasifier isoperated at high pressures to increase IGCC efficiency and to decreasethe size of equipment [24]. It is generally accepted that the operatingpressure of a wet feed gasifier is limited by the delivery pressure of theoxygen [24]. There are safety related problems in compressing oxygento high pressure [20]. In addition, the efficiency of the oxygencompressors is usually low [20]. When the liquid oxygen is removedfrom the system, one must efficiently recover the refrigerationcontained in the liquid oxygen [12,21]. This can be achieved bycondensing a high-pressure air or nitrogen stream by exchanging heatwith the vaporizing oxygen stream in a cryogenic exchanger [12,21].Pumping the oxygen to an intermediate or the final pressure provides

1687D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

the opportunity to reduce the size of the oxygen compressor or toeliminate it completely. In addition, with GOX cycles, hydrocarbonscan accumulate in the sump of the LP column [23]. To avoid this safetyhazard, a slip stream of liquid oxygen must be removed from thebottom of the LP column in a GOX cycle. A PLOX cycle eliminates thisproblem. A PLOX cycle provides additional degrees of freedom foroptimization by choosing the pressure of the PLOX cycle, consideringvarious sources of the high-pressure streams, and distributing theliquefied high-pressure stream optimally between the HP and the LPcolumn in an ASU.

The discussions above shows that there are degrees of freedomavailable that can be utilized for developing an optimal design of the ASUand for optimizing the heat and power integration between differentsections of an IGCC plant including the ASU. That is the focus of this paper.In particular, the following studies have been presented in this paper:

1. The first study is focused on an optimal design of an ASUwith PLOXcycle. As mentioned previously, the refrigeration lost due to thewithdrawal of liquid oxygen in a PLOX cycle can be recovered withhigh-pressure air and/or nitrogen. If high-pressure air is used forrecovering the refrigeration, the compression ratio required ismuch less compared to nitrogen since a part of the MAC dischargeair is compressed, stream 3 in Fig. 2. In addition, the high-pressureair stream does not cause any additional mass flow to the ASUsystem. However, a larger flow of the liquefied air to the systemresults in a lower flow of the vapor feed to the HP column. Thisresults in a decrease in the reboiler duty of the LP column. A largerflow of the liquefied air directly to the LP column significantlyaffects the final product quality since it bypasses the HP column.Therefore, it becomes difficult to achieve the desired productquality as the amount of liquefied air increases. On the other hand,if a part of the LP product nitrogen is compressed and recycled tothe system as a high-pressure stream, stream 4 in Fig. 2, a penaltyin the operating efficiency occurs because of the higher compres-sion ratio and the additional mass flow [12]. However, a significantportion of this recycle nitrogen, if not all, can bypass the HP columnand can be sent directly to the LP column as top reflux. Thisadditional reflux helps to increase the operating pressure of theASU without sacrificing the product qualities. In this paper, wehave explored the possibility of using an optimal combination ofboth the high-pressure streams in order to further decrease theauxiliary power consumption. In addition, a sidedraw, stream 9 inFig. 2, withdrawn from an intermediate stage of the HP column andrecycled to an intermediate stage in the LP column, has beenconsidered to further optimize the ASU design for a PLOX cycle.

2. The second study focuses on optimal degree of air integration for theshifted syngas as applicable to an IGCC with pre-combustion CO2

capture. Most gas turbines are designed to be operated with naturalgas, with an LHV of 48 MJ kg−1 [7,9,13,17]. When a lower energyfuel is used, such as syngas, with an LHV of 8–9 MJ kg−1, more fuel(approximately 4–5 times the volumetric flow and 5–6 times themass flow as compared to natural gas) [7] must be supplied, whichmay prevent a lot of the compressed air from being sent to theturbine due to volumetric/mass flow limitations. This extracted aircan be used as an air feed to the ASU. Integration of the ASUwith theGT presents the possibility of increasing the efficiency of an IGCCpower plant by extracting air from theGT [1,5–12,19] to use as a feedto the ASU. Several papers have been published on the integration ofthe ASU and GT [5–11,19,21,25–27]. Most of these papers agree thatthe optimal degree of integration occurs when all nitrogen availablefrom the ASU is sent to the GT and air that cannot be sent throughthe GT, due to mass/volumetric flow limitations, is extracted for useas feed to the ASU. This level of air extraction is typically found to bearound 25–50% for a syngas, which has a low heating value (LHV) of8–9 MJ kg−1 [7]. For pre-combustion carbon capture in an IGCCplant, the syngas is shifted and cleaned before being sent to the GT

[7,8]. The shifted hydrogen-rich syngas has a much higher LHV of43 MJ kg−1, close to that of natural gas, which is 48 MJ kg−1

.

3. It has been suggested that in order to maximize the benefit of thehigh-pressure air available through air extraction an EP-ASUshould be used so as to increase the savings from the decreasedload on the MAC [6–11]. No quantitative study has been foundwithin the open literature showing whether equivalent savings arepossible from an LP-ASU using a PLOX cycle at varying degrees ofair integration. Smith et al. [8] suggest that at degrees of integrationbelow approximately 40%, an LP-ASU with a PLOX cycle iscomparable to that of an EP-ASU. In this paper, a study has beencarried out for an IGCC plant with CO2 capture by comparing thenet power production from an LP-ASU with a PLOX cycle with thatof an optimal EP-ASU at varying levels of air integration.

4. Optimal operating pressure for the ASU is thought to depend uponthe operating pressure of the GT, the degree of air integration, andthe fraction of nitrogen injection [5–11,19]. It is also thought thatoperating the ASU at a pressure near the operating pressure of theGT in the case of a high degree of air integration will maximizesystem efficiency, albeit accompanied by a challenge in systemoperability due to the varying pressure of the extracted air [8]. Freyand Zhu [6] found that when syngas humidification is used toreduce NOX formation, in combination with nitrogen injection, anLP-ASU is more efficient than an EP-ASU up until approximately60% of the available nitrogen is injected. A study has been done on asteam injected GT with topping steam turbine [14] that also showsthat at lower degrees of nitrogen injection, an LP-ASU is moreefficient. In this paper, we have studied the effects of the degree ofair integration on the optimal ASU operating pressure in an IGCCplant with CO2 capture when a combination of steam injection andnitrogen injection is performed to reduce NOX formation.

2. Modeling of the ASU

A simplified flowsheet of the ASU discussed in this paper is shown inFig. 2. Fig. 2 shows the six configurations examined in this paper.However depending on the exact configuration, not all streams shownmay be considered. Table 1 shows which streams are relevant for eachconfiguration. Unnumbered streams are considered in all configurations.Table 1 also shows the final operating conditions for each configuration.

The ASU consists of a multi-stream heat exchanger (main heatexchanger) that recovers refrigeration from the product streams andcools the incoming streams. Products leave the main heat exchangerat near ambient temperatures after refrigerating one air stream and, ifapplicable, liquefying another air stream and the nitrogen reflux [21].For Configurations 1, 3, 5, and 6 the refrigerated air stream leaves themain heat exchanger as a saturated vapor. For Configurations 1 and 3,refrigerated air flow represents approximately 70% of total air flowwhereas in Configurations 5 and 6, refrigerated air flow representsapproximately 85% of the total air flow. For all configurations wherenitrogen recycle is considered (Configurations 4, 5, and 6), it leavesthemain heat exchanger as a saturated liquid. The liquefied air streamleaves the main heat exchanger as a saturated liquid. For bothConfigurations 2 and 4, refrigerated air leaving the main heatexchanger is superheated by approximately 10 °C. For Configuration2, the liquefied air stream, approximately 1% of total air flow, leavesthe main heat exchanger as a saturated liquid. The air provided to theASU, for cases when 100% integration with the GT is not chosen,comes from the MAC. The refrigerated air stream is then sent to thebottom of the high-pressure column. The refrigerated air streamprovides the boilup for the HP column and the LP column via thecoupled reboiler/condenser. The liquefied air stream is split with aportion being sent to the nitrogen superheater for subcooling and theremainder being sent to an intermediate stage of the HP column. TheHP column, for studies considered here, operates at a pressure of6.8 atm for the LP-ASU configuration and pressures between 11 and

Fig. 2. Simplified flowsheet of cryogenic air separation unit.

1688 D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

13.5 atm for the EP-ASU configurations. The HP column consists of 50valve trays with a total pressure drop of 41 kPa. The HP column yieldsa crude liquid oxygen product [~33 mol% O2] from the bottom and anultra-pure, high-pressure gaseous nitrogen product [b50 ppm O2]from the top [15].

The crude oxygen from the bottom and liquid nitrogen from thetop of the HP column are subcooled in a multi-stream heat exchanger(nitrogen superheater) against the nitrogen vapor from the top of theLP column. The subcooled streams are sent to the LP column. The LPcolumn, for the studies considered in this paper, operates at a pressureof 1.7 atm for the LP-ASU configurations and a pressure range of 3.4 to4.3 atm for the EP-ASU configurations. The LP column is modeled as apacked columnwith an HETP of 0.35 m [18], 25 equivalent stages, anda total pressure drop of 6 kPa. Packing is generally chosen for thiscolumn because nitrogen product is desired at the highest possiblepressure while still maintaining a 2.8 °C temperature approach in thereboiler/condenser. The low-pressure nitrogen is sent to the nitrogencompressors, consisting of four intercooled compressors with anassumed isentropic efficiency of 89%, where it is discharged at31.6 atm. The oxygen product is sent to the oxygen compressors,consisting of four intercooled compressors with an assumed isentro-pic efficiency of 89%, where it is discharged at 66.7 atm. All productspecifications for the ASU are given in Table 2. It should be noted thatthe low-pressure nitrogen purity requirement would be dependenton the fuel and combustor typed used in the gas turbine. The low-

pressure nitrogen purity requirement shown in Table 2 may bedifferent for the low-carbon, high-hydrogen fuel considered here ascompared to typical syngas. This study considers 100% nitrogenintegration where the entire LP nitrogen stream produced by the ASUis sent to the GT as diluent.

The EP-ASU process model was developed in the steady-statesimulator, Aspen Plus [30], using the Peng-Robinson equation of state.The temperature and composition profiles, and liquid and vaporflowrates in the LP-column of the EP-ASU (Configuration 2) comparedwell with the results of Seliger et al. [18]. Design specifications wereused to control the following process variables:

1) The HP column reflux was manipulated to maintain high-pressurenitrogen purity.

2) The heating duty of the reboiler/condenser was manipulated tomaintain a specified high-pressure nitrogen flow.

3) The liquid air flow split fraction between the HP and LP columnswas manipulated to maintain low-pressure nitrogen purity forConfiguration 3.

4) The nitrogen reflux was manipulated to maintain low-pressurenitrogen purity for Configurations 4 and 5.

5) The refrigerated air temperature was manipulated to maintainlow-pressure nitrogen purity for Configuration 2.

6) The total liquefied air flow was manipulated to control oxygenpurity for Configurations 2, 3, and 5.

Table 1Configuration of Streams for Different ASU Cycles.

Cycle ConfigurationNumber

ActiveStreams

HP ColumnPressure [atm]

LP ColumnPressure [atm]

Nitrogen RecycleFlowrate [kmol h−1]

Liquid Air Flowrate[kmol h−1]

LP with pumped liquid oxygen 1 3,5,6,7 6.80 1.70 – 7150EP with gaseous oxygen 2 3,5,8 12.93 4.15 – 370EP with pumped liquid oxygen and liquid air 3 3,5,6,7 11.23 3.42 – 7150EP with pumped liquid oxygen and nitrogen recycle 4 4,7 13.47 4.32 5750 –

EP with pumped liquid oxygen and liquid air and nitrogen recycle 5 3,4,5,7 12.93 4.15 2320 3750EP with pumped liquid oxygen, liquid air, nitrogen recycle, andsidedraw

6 3,4,5,7,9 12.93 4.15 1570 4870

1689D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

7) The refrigerated air temperature was manipulated to maintainoxygen purity for Configurations 1 and 4.

8) LP column pressure was manipulated to maintain a specifiedtemperature approach within the reboiler/condenser.

Table 3Fuel Properties.

Flowrate [kmol h−1] 17,383

3. Modeling of the HRSG and GT with a topping steam turbine

Fuel produced by this IGCC with CO2 capture system is a shifted,non-humidified, hydrogen-rich syngas. For the studies discussed in thispaper, fuel flow and properties are held constant and their values areshown in Table 3. The shifted syngas is fed to a power production GT,considered, for these case studies, to be similar to a GE 7FB [7–9,15]. Asshown in Fig. 3, the hot exhaust gas is used to generate high-pressure(HP), intermediate-pressure (IP), and low-pressure (LP) steam streamsin a heat recovery steam generation unit (HRSG). The steam producedin the HRSG is then used to drive additional power-generating steamturbines. In addition to the combined cycle configuration, a steaminjected GT with a topping steam turbine configuration has also beeninvestigated [14]. In this configuration, steam generated in the HRSG isfirst used to drive high-pressure steam turbines and then is returned tothe GT as a fuel diluent for inhibiting NOX formation. Fig. 3 shows asimplified diagram of the HRSG and GT with a topping steam turbine.

Increasingly stringent gas turbine emission control regulationsmay require single digit NOX concentrations [29]. To inhibit NOX

formation in the GT, nitrogen and steam dilution is considered. Todetermine NOX concentration in the flue gas from the gas turbine, anempirical relation was developed based on data presented by GE onNOX formation with a hydrogen-rich fuel with steam dilution andsyngas with nitrogen dilution [13,22]. Eq. (1a) and (1b) and Fig. 4 arethe correlations used to predict NOX concentration based on the LHVof the fuel with nitrogen dilution and fit against published GE data.Eq. (2a) and (2b) and Fig. 5 are the correlations used to predict NOX

concentration based on the LHV of the fuel with steam dilution andare fit against published GE data.

For LHV≤6:71MJm3 ;NOX ppmvd½ �¼ 1:918exp 9:032 × 10�4LHV

MJm3

� �� �

ð1aÞ

For LHV N 6:71MJm3 ;NOX ppmvd½ �¼ 10:33exp 5:63 × 10�4LHV

MJm3

� �� �

ð1bÞ

Table 2Product Specifications.

Flowrate [kmol h−1] Purity

O2 5970 95% O2

Low-Pressure N2 22,490 2% O2

High-Pressure N2 680 50 ppm O2

For LHV≤6:71MJm3 ;NOX ppmvd½ �¼ 0:07787 × exp 0:001404 × LHV

MJm3

� �� �

ð2aÞ

For LHV N 6:71MJm3 ;NOX ppmvd½ �¼ 0:6255 × exp 0:001 × LHV

MJm3

� �� �:

ð2bÞ

4. Results and discussion

Configuration 1, described in Table 1, was simulated at an HPcolumn operating pressure of 6.80 atm and a corresponding LPcolumn operating pressure of 1.70 atm. This configuration employsa PLOX cycle. Since, for all studies discussed in this paper, the numberof stages for both the HP and LP columns were left constant and sincethe relative volatility of nitrogen and oxygen is higher for thisconfiguration compared to the other configurations, which are run atelevated pressures, the low-pressure nitrogen purity is higher thanspecification. For this configuration, low-pressure nitrogen purity is99 mol% N2 (1 mol% O2). This configuration was run with a PLOX cyclewhich, without a high-pressure air source, was pumped to a pressureof 255 kPa. Further pumping of the liquid oxygen is possible if abooster compressor is used to compress the air to be liquefied.However, with this configuration, and all configurations with a PLOXcycle, as the pressure of the streams to be liquefied is increased, theflows of these streams must also be increased. As the pressure of thestreams to be liquefied is increased, higher flows are required tomaintain the same amount of heat exchange with the liquid oxygen.For example, when condensing saturated gaseous air, 76% more air at29.6 atm is required for the same enthalpy change as compared to anair stream at 15.4 atm mainly because of an elevated condensationtemperature due to increase in pressure. As shown in Table 4,although this configuration does not require nitrogen recycle, thepower required to compress products to the delivery pressure is large,representing 59% of the total ASU power requirement.

Configuration 2, described in Table 1, was completed at an HPcolumn operating pressure of 12.93 atm and a corresponding LP columnoperating pressure of 4.15 atm. This configuration uses a gaseous

Pressure [atm] 31.3Temperature [°C] 193Composition [mol%]Hydrogen 91.1Carbon Dioxide 4.5Carbon Monoxide 1.9Nitrogen 1.3Argon 1.1Water 0.1

Fig. 3. A simplified schematic of the HRSG and GT with a topping steam turbine.

1690 D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

oxygen cycle. This is a more conventional configuration with little or nointegrationwith theGT and, as shown in Tables 1 and 4, requires neithera booster compressor nor nitrogen recycle. However, this cycle requiresthat all oxygen compression be accomplished through the oxygencompressors, which have safety and efficiency issues associated withthem [20]. In addition, since hydrocarbons can accumulate within theliquid oxygen sumpof the LP column, a liquid draw is often still requiredfor safety reasons [23]. However, withdrawal of only gaseous oxygenfrom the LP column helps in generating more liquid and vapor traffic inboth the HP and LP columns compared to a PLOX cycle.

Configuration 3 was simulated at an HP column operating pressureof 11.23 atm and a corresponding LP column operating pressure of3.42 atm. This configuration utilizes a PLOX cycle. This configurationsplits the liquid air generated in the main heat exchanger between theHP column and LP column after subcooling in the nitrogen superheater.The operating pressure for this configuration is limited due to themedium used to recover the liquid oxygen refrigeration. As mentionedbefore, the increased liquid air flow to either column presentsseparation challenges. To overcome these issues, the operating pressuremust be decreased to increase the relative volatility of nitrogen andoxygen. Due to the decreased operating pressure, the power required tocompress products to delivery pressure increases, representing 53% ofthe total ASU power requirement. This configuration does yield a highpower savings via pumping the liquid oxygen; however, the degree ofpumping is limited, due to the high pressure to which the air to beliquefied must be compressed.

Configuration 4 was completed at an HP column operating pressureof 13.47 atm and a corresponding LP column operating pressure of4.32 atm. This configuration utilizes a PLOX cycle. The refrigeration ofthe liquid oxygen is recovered via a nitrogen streamwhich is then sentto the top of the LP column as reflux. This configuration allows higher

Fig. 4. NOX (ppmvd) vs LHV of the Diluted Syngas (Btu/SCF) when N2 is used as a Diluent[13,22].

ASU operating pressures, which decreases nitrogen compression costs,as shown in Table 4. However, little to no power savings are seen inpumping the liquid oxygen, as shown in Fig. 6, as a nearly equivalentamount of power must be used to increase the pressure of the nitrogenrecycle as that saved in the oxygen compressors.

Configuration 5 was completed at an HP column operating pressureof 12.93 atm and a corresponding LP column operating pressure of4.15 atm. This configuration again uses a PLOX cycle. Both liquid air anda nitrogen recycle are used to recover liquid oxygen refrigeration. Byusing both to recover refrigeration, the operating pressure of the ASUcan be increased via the nitrogen recycle, and it is possible to achieveconsiderable power savings by pumping the liquid oxygen, as shown inFig. 6. A summary of the power requirements is shown in Table 4.

The nitrogen recycle can produce the desired nitrogen and oxygenpurities required for an IGCC power plant even at high operatingpressures; however, it is a very power intensive option [12]. ThereforeConfiguration 6 was considered in an attempt to maintain the highoperating pressure of the ASU and decrease the quantity of nitrogenrecycle required tomaintain product purities. Configuration 6was runat the same HP and LP column operating pressures as Configuration 5,but with the inclusion of the sidedraw, Stream 9 in Fig. 2. Thesidedraw takes liquid nitrogen from an intermediate stage in the HPcolumn and sends it to an intermediate stage in the LP column. Theeffect of this sidedraw, in the region studied, helps to improve thecrude oxygen purity from the HP column, as shown in Fig. 7. This, inturn, helps to recover more refrigeration from the liquid oxygen viathe liquid air, as shown in Fig. 8. As the flow of the liquid air to the HPcolumn increases, the heating duty of the reboiler/condenser isdecreased, as shown in Fig. 9, resulting in less boilup in the LP column.

Fig. 5. NOX (ppmvd) vs LHV of the Diluted Syngas (Btu/SCF) with 50–95% H2 by volumewhen steam is used as a Diluent [13,22].

Fig. 6. ASU power requirements for varying cycles and degree of liquid oxygen pumping.

1691D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

These advantages result in a 1.2 MW decrease in the ASU powerrequirement by decreasing the nitrogen recycle required to maintainproduct purities, as shown in Fig. 10. However, as the sidedraw isincreased further, the separation achieved in the LP column is reducedbecause of a reduced boilup ratio. Therefore, even though the crudeoxygen purity keeps increasing with an increase in the sidedraw flow,the refrigeration recovered through the air needs to be decreased toincrease the boilup ratio in the LP column that is required to achievethe desired separation. This study shows that there exists an optimumflowrate of the sidedraw for minimizing the auxiliary powerconsumption.

The large differences in the power requirements of the differentconfigurations is due to the operating pressure at which eachconfiguration must be run in order to satisfy the product qualityrequirements. For example, as mentioned previously, the maximumoperating pressure in Configuration 3 is limited because of using airfor recovering the refrigeration. Whereas in Configuration 4, theoperating pressure of the ASU can be increased, yielding powersavings in product compression. Additionally, Configurations 3 and 5allow for higher degrees of liquid oxygen pumping and although littleto no savings may be possible, this may still be desired in order toeliminate a stage in the oxygen compressors. However, for Configu-ration 3, increased degrees of liquid oxygen pumping would require adecrease in ASU operating pressure since the increasing pressure ofthe liquid air yields higher vapor fractions when flashed, resulting ininsufficient reflux being provided to maintain product specifications.

This study shows that an ASU with PLOX cycle can have auxiliarypower consumption similar to that of an ASU with GOX cycle, ifoptimally designed. The added advantages of the PLOX cycles areincreased safety, and availability of the plant in addition to savings inthe cost of the oxygen compressor.

5. Integration

5.1. Higher heating value fuels

For the studies discussed in this section, two power producing gasturbines, with a frame similar to a GE 7FB, are considered [13,15]. Forthe studies presented here, a shifted, hydrogen-rich syngas isconsidered with fuel properties and composition as shown in Table 3.The fuel gas has a LHV of 43 MJ kg−1. Because this fuel has a calorificvalue near that of natural gas, mass and volumetric flow limitations aredifferent than when using unshifted syngas. The main combustiblecomponent in the unshifted syngas is CO whereas the shifted syngashas mostly hydrogen. Even though pure hydrogen has lower LHVcompared to CO on volumetric basis, the concentration of combustiblesin an unshifted syngas is 64% compared to 93% in shifted syngas. Thisresults in a higher volumetric LHV of the shifted syngas in comparisonto the unshifted syngas. The assumed mass and volumetric flowlimitations are presented in Table 5 [6]. Results of the study on theeffect of the percent of air integration, defined as the percentage of thetotal ASU air supply requirement being provided by air extraction fromthe GT are presented in Fig. 11 for ASU operating pressures of12.24 atm, 12.93 atm, and 13.95 atm. This study was carried out withConfiguration 5 with the addition of Stream 1 shown on Fig. 2. A

Table 4Power requirements for ASU configurations.

ConfigurationNumber

MAC[MW]

N2

[MW]O2

[MW]Booster[MW]

N2 Recycle[MW]

Total Power[MW]

1 55.10 58.32 19.84 – – 133.262 72.08 41.64 15.94 – – 129.663 66.23 57.81 15.30 1.49 – 140.834 73.20 40.74 14.49 – 6.56 134.995 72.08 41.64 12.83 0.59 3.99 131.136 72.08 41.64 12.83 0.73 2.70 129.98

summary of the individual power requirement of the ASU at theoperating pressure of 12.93 atm and power production in the powerblock are shown in Table 6. Table 6 shows that the power consumptionin the MAC monotonically decreases, as expected, with increase in airintegration. As the nitrogen from the ASU gets completely depleted atabout 9% integration, there is no more increase in the powerconsumption by the nitrogen compressor. As nitrogen gets depleted,the power produced by the GT keeps decreasing asmore air is extractedbeyond this level of integration. The extent of power saving in the MACis less than the loss in the generated power from the GT as the level ofair integration is increased further. This is in qualitative agreementwiththe literature [6–11,19,21] where similar findings have been reportedfor lower energy fuels. Importantly, the ASU operating pressure has anegligible effect on power production. This is due to the cost of oxygencompression for all pressure cases being equivalent to pump the liquidoxygen to the same pressure. However, the pressure that the liquidoxygen can be pumped is determined by the pressure of the streams tobe liquefied, and since the pressure of the extracted air is constant forall cases, the pressure of the oxygen sent to the oxygen compressors isconstant [20]. The advantage of operating the ASU at a high pressure isin the decrease in nitrogen compression costs [6–10,21]. However, thiseffect is relatively small due to the need to reflux more nitrogen tomaintain product purities. The optimal found here is lower than thatfound in other studies [6–11,19,21] since the fuel considered is a higherenergy fuel when compared to syngas. As a result, less fuelmust be sentto the GT for the same power output, thereby making mass andvolumetric flow limitations less of an issue. Since the hydrogen-richfuel has a heating value, on a mass basis, close to that of natural gas,torque limitations due to increased mass flow through the turbine are

Fig. 7. Sidedraw effect on crude oxygen purity.

Fig. 8. Sidedraw effect on liquid air flowrate.Fig. 10. Sidedraw effect on nitrogen recycle power requirement.

1692 D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

less significant. However, as the hydrogen-rich fuel still has a lowerheating value on a volumetric basis when compared to natural gas, sosurge problems may still be present for a hydrogen-rich fuel similar tounshifted syngas. However, the hydrogen-rich fuel considered here hasan LHV of 9.4 MJ m−3, requiring only 3.7 times the volumetric flow offuel as compared to natural gas. Compared to syngas, that has a LHV of7.1–7.8 MJ m−3[7], requiring 5 times the volumetric flow of fuel ascompared to natural gas.

5.2. Optimal ASU configuration at varying degrees of air integration

There are several studies within the open literature that have shownthat optimal air integration occurs at 25–50% with a low energy fuelsuch as syngas [6–11,19,21]. In the study presented above for a shiftedsyngas, the optimal air integration is about 9%. However the coal-fedIGCC plants in the USA, Polk and Wabash River, have 0% air integration[15]. Whereas, the coal-fed IGCC plants in Europe, Buggenum andPuertollano have 100% air integration [7]. When 100% integration ischosen, no MAC is required. In addition, the GT compressors aretypically more efficient than the MAC [7]. However, 100% integrationresults in longer start-up times and presents operability issues [7]. Thestudies in this section are intended to address optimal ASU configu-ration as the degrees of air integration are varied in from 10 to 100%.Although it is unclear whether a GE 7FB is capable of 100% airsideintegration, for the studies in this section, it is assumed that the GTpressure ratio, whatever the frame, is 16.1:1 and there are nolimitations in the GT design for 100% airside integration.

Smith et al. [8] have suggested that at degrees of air integrationlower than 40%, an LP-ASU with a PLOX cycle is comparable to an EP-ASU. Smith and Klosek [11] also suggested that at air integration levelsless than 25%, the operating pressure of the ASU can be setindependently of the operating pressure of the GT. Air extraction, in

Fig. 9. Sidedraw effect on reboiler/condenser duty.

the case of an EP-ASU, yields power savings via the decreased load onthe MAC, which in the case of the EP-ASU can represent as much as61% of the total ASU power requirement [15]. In the case of an LP-ASU,power savings are still seen in the decreased load on the MAC;however, the MAC power represents as little as 42.2% of the totalpower requirement. Since the pressure to which the liquid oxygen canbe pumped is determined by the pressure of the streams to beliquefied, the pressure of the extracted air will set this pressurewhether an EP or LP configuration is chosen [20], assuming thatextracted air flow is sufficiently high to efficiently recover the liquidoxygen refrigeration or that a booster compressor is used to compressa portion of the MAC discharge to the pressure of the extracted air.However, if a configuration like that shown in Fig. 2, Stream 2, ischosen, extracted airflow that is higher than the required liquidairflow, approximately 30% for the LP-ASU and 15% for the EP-ASU,can be sent through a turbine, discharged to the pressure of the MACand used to power a compressor to compress the air stream that willbe liquefied. With the higher-pressure air stream available, pumpingthe oxygen to higher pressure is possible. With the operation of an LP-ASU, a larger pressure drop is attained in the turbine, furtherincreasing the pressure of the air stream to be liquefied. It should benoted that in the studies considered here that the isentropic efficiencyof both the compressor and turbine were assumed to be 72%.

For this study, Configuration 1 is considered for the LP-ASU case,Configuration 5 with an HP column operating pressure of 12.25 atm isconsidered for the EP-ASU with PLOX, and Configuration 2 isconsidered for the EP-ASU with GOX. For all levels of integrationinvestigated in this study, it is assumed all low-pressure nitrogenproduced in the ASU can be sent to the GT without exceeding mass orvolumetric flow limitations. At lower levels of air integration, whereextracted air flow is less than that required to recover liquid oxygenrefrigeration, a booster air compressor, Stream 3 in Fig. 2, is used tofurther compress a portion of theMAC discharge to the pressure of theextracted air [12,20]. At higher degrees of air integration, whereextracted air flow is greater than that required to recover the liquidoxygen refrigeration, approximately 15–30% of the total air flowdepending on the configuration, the extra air is sent to a turbine tofurther compress the flow that will be liquefied and then mixed withthe MAC discharge. Fig. 12 shows the result of this study where powersavings is defined as the difference in the power requirement of eachconfiguration from the base case values at 0% air integration. The totalpower requirement of each configuration at 0%, 30%, 60%, and 100% airintegration is provided in Table 7. This Figure shows that at lower

Table 5Assumed flow limitations for gas turbine.

Maximum mass flowrate [tonne h−1] 1,905Maximum volumetric flowrate [m3 h−1] 244,900

Fig. 11. Power production with varying degrees of integration and ASU operatingpressure for Configuration 5 with air integration. Fig. 12. Power savings in the ASU as a function of increasing air integration with the gas

turbine.

1693D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

levels of air integration, the power savings are as much as 1.2 MWhigher for Configuration 1 than for Configuration 2 and 0.5 MWhigherthan Configuration 5. While at higher degrees of air integration ofapproximately 50–60%, all configurations yield nearly equivalentpower savings. Above this level of integration, Configuration 2 yieldsup to 5.1 MW more power savings than Configuration 1 and up to2.2 MW more power savings than Configuration 5. Although thepressure of the liquid oxygen can always be increased past 20–30%integration for Configurations 1 and 5, as shown in Fig. 13, the powersavings in the oxygen compressors and MAC for Configuration 1 arenot as large as the power savings in the MAC for Configuration 2.Additionally, beyond an air integration level of 40%, the cost ofincreasing the pressure of the nitrogen recycle is higher than thepower savings in the oxygen compressors which causes powersavings to decrease with further liquid oxygen pumping. In addition,little additional increase in the liquid oxygen pressure is possiblebeyond 60% for Configuration 5 due to higher flows of liquid air beingrequired as the pressure of the liquid air is increased, resulting in lessextracted air being sent through the turbine, this can be seen inTable 8. Because of this, little increase in pressure of the liquid air isseen beyond 60% integration. A similar effect occurs in Configuration1; however, since the relative volatility for this case is higher, theeffect is smaller since less reflux is required.

6. Fuel diluent

Other researchers have suggested that the operating pressure ofthe GT, the degree of air integration, and the fraction of nitrogeninjection (in the case where both nitrogen and steam are used as fueldiluents) determine optimal ASU operating pressure [6–8]. As a result,a study is carried out in this paper to determine the effect of the choiceof fuel diluent and degree of nitrogen injection on the choice of ASUoperating pressure. This study is carried out with a constant NOX

concentration of 37 ppmvd in the gas turbine exhaust, using Eqs. (1a),(1b), (2a), and (2b) to predict concentration. The simplifiedconfiguration for the combined cycle unit used in this study isshown in Fig. 3. ASU Configurations 1 and 5 were used for this study.

Table 6Power summary for the 12.9 atm case using Configuration 5 with air integration.

Percent AirIntegration

MAC[MW]

NitrogenCompressors[MW]

OxygenCompressors[MW]

BoosterCompressor[MW]

NitrogenRecycle[MW]

PowerBlock[MW]

3 69.9 38.1 13.4 0.3 3.8 752.36 67.8 39.9 13.4 0.2 3.8 752.79 65.6 41.7 13.4 0.1 3.8 753.012 63.4 41.7 13.4 0.0 3.8 750.614 62.0 41.7 13.4 0.0 3.8 748.6

The region in which these studies are carried out requires no airextraction because approximately half as much steam is required ascompared to nitrogen for the same degree of NOX inhibition [10].Since there is no air extraction required, no change in powerrequirements occurs in any of the equipment except the nitrogencompressor. Nitrogen flow is varied and the diluent steam flow ismanipulated to maintain the NOX concentration. As shown in Fig. 3,any steam not routed to the gas turbine for fuel dilution is sent to theIP and LP steam turbines for power production. As shown in Fig. 14,the degree of nitrogen injection has a large influence on the optimaloperating pressure and at a nitrogen injection rate beyond about 50%,Configuration 5 becomes more advantageous than Configuration 1,which is in good agreement with the findings of Smith et al. [8] andFrey and Zhu [6], the latter study having considered syngashumidification. These findings are driven by the cost of nitrogencompression, which is determined by the ASU operating pressure andgas turbine operating pressure. Frey and Zhu [6] found that atransition occurs at around 50% nitrogen side integration due to theneed to extract air to stay within GT operating limits. In our paper, thistransition would occur at beyond about 80% nitrogen side integrationdue to the higher heating value of the shifted syngas. This transition isnot shown in Fig. 14, as beyond 80% nitrogen side integration, airextraction would be required.

7. Impact of carbon capture on the ASU

In this section, the impact of carbon capture on the optimal designand integration of an ASU for an IGCC power plant is discussed. The firststudy focuses on various ASU configurations. While these configura-tions are not expected to differ with or without carbon capture, thedesign criteria can certainly be different if the carbon capture option isconsidered. The purity specification for the oxygen does not vary basedon the carbon capture option. However, the purity specification of thediluent nitrogen can vary. Because of this, the optimum plantconfiguration may be different if the nitrogen specification is changed.

The second study focuses on the optimal degree of air integrationfor shifted hydrogen-rich syngas. This study is applicable only to IGCCplants with the carbon capture option. As the shifted syngas has a

Table 7Total ASU power requirement for each Configuration using varying degrees of airintegration.

N2 Compressors [MW] 0% 30% 60% 100%

Configuration 1 58.3 133.3 110.4 90.3 66.3Configuration 2 41.6 129.7 108.0 86.4 57.6Configuration 5 44.6 131.1 109.5 88.4 61.8

Fig. 13. Liquid oxygen pressure as a function of percent integration.

1694 D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

higher heating value than the unshifted syngas, optimal degree of airside integration is lower. This is a direct result of carbon capture andwill affect the optimal ASU design.

The third study focuses on the net power production from an LP-ASU with a PLOX cycle with that of an optimal EP-ASU at varyinglevels of air integration. One of the key assumptions of this study isthat the all low pressure nitrogen can be sent back to the GT as fueldiluent without exceeding the mass and volumetric limitations of theGT. This is the most likely scenario for an IGCC plant with carboncapture.

In the fourth study, the effects of the degree of nitrogen integrationon the optimal ASU operating pressure in an IGCC plant with CO2

capture are evaluated when a combination of steam injection andnitrogen injection is performed to reduce NOX formation. This studywas carried out with a constant NOX concentration which is consideredto be a function of the lower heating value (LHV) of the diluted fuel andthe type of diluent (nitrogen or steam). Since the lower heating valueconsidered here is that of the shifted syngas and the correlations,especially the one when steam is used is used as a diluent, are forhydrogen-rich syngas, this study is valid only for plants with carboncapture.

8. Conclusions

An EP-ASU with PLOX cycle can be a feasible alternative to thetraditional GOX cycle for improving the safety and availability of anIGCC plant where the operating pressure of the gasifier is high.However, the refrigeration lost in the liquid oxygenmust be efficientlyrecovered to avoid loss in the ASU efficiency. This paper hasconsidered five different configurations with PLOX cycle for 100%nitrogen integration to improve the overall plant efficiency. Inabsence of a high-pressure air stream in Configuration 1 which is anLP-ASU with PLOX cycle, the PLOX pressure could not be increasedmuch. Therefore this configuration required a very high compression

Table 8Results for Configuration 5 using varying degrees of air integration.

Percent AirIntegration

MAC[MW]

N2 Compressors[MW]

O2 Compressors[MW]

N[M

10 63.6 44.6 13.3 2.20 56.5 44.6 13.1 2.30 49.4 44.6 12.3 3.40 42.4 44.6 11.5 3.50 35.3 44.6 10.9 4.60 28.3 44.6 10.4 5.70 21.2 44.6 10.2 5.80 14.1 44.6 10.1 6.90 7.1 44.6 9.9 6.100 0.0 44.6 9.9 7.

ratio for both LP nitrogen and oxygen, highest among all theconfigurations, resulting in 59% of the total ASU power consumption.Configuration 1 consumes higher power compared to an EP-ASU withGOX cycle which is Configuration 2. If a high-pressure air stream isused alone to recover the refrigeration from liquid oxygen, as inConfiguration 3, the resulting liquid air creates significant separationchallenges. As the operating pressure of the LP column is lowered tosatisfy the product qualities, this configuration is found to consumethe highest power among all the configurations considered in thispaper. If a high-pressure nitrogen stream is used alone to recover therefrigeration, as in Configuration 4, a higher ASU pressure can bemaintained without sacrificing product qualities. However, negligiblepower saving is achieved by pumping the liquid oxygen because of anearly equivalent increase in power consumption in the recyclenitrogen compressor to that saved in the oxygen compressors.Configuration 5, a hybrid of Configurations 3 and 4, is found to havehigher power savings due to savings in oxygen compression and theability to operate at higher pressures, resulting in savings fromnitrogen compression. The power requirement is further reduced byutilizing a sidedraw in Configuration 6. The study shows that there isan optimum flowrate of the sidedraw that results in the minimumpower consumption for an EP-ASU with PLOX cycle due to decreasednitrogen recycle. The power consumption in Configuration 6 is similarto Configuration 2. This shows that an optimally designed EP-ASUwith PLOX cycle can have similar efficiency to that of an EP-ASU withGOX cycle with other added advantages of using a PLOX cycle.

As found by others studies, optimal air integration occurs when airextracted from the GT can be replaced with nitrogen from the ASUwithout exceeding mass/volumetric flow limitations of the GT. In thecase of an unshifted syngas, optimal integration is generally agreed tooccur at approximately 25–50% integration. In the case of a shiftedsyngas where the LHV both on mass and volume basis is higher thanthe unshifted syngas, optimal air integration was found to be 9%. Thisis due to a lower volumetric flow of fuel to the GT, allowing higherflows of diluent nitrogen without exceeding the volumetric flowlimitations of the GT.

High-pressure air extracted from the GT, when used in an EP-ASU,yields power savings due to a decreased load on the MAC. It has beenshown here that an LP-ASUwith a cycle as shown in Fig. 2, stream2, canyield greater power savings than is possible for an EP-ASU with GOXcycle or EP-ASU with PLOX cycle when air integration levels are lessthan 50–60% due to the high power savings possible in oxygencompression. At 100% air side integration, the power requirement of anEP-ASU can expect as much as a 55.6% reduction as compared to an LP-ASU which can expect a 50.3% reduction. This is due to the powersavings from the decreased load on the MAC exceeding the powersavings in oxygen compression.

The pressure to which the liquid oxygen can be pumped isdetermined by the pressure of the streams to be liquefied. With theinclusion of stream 2 in Fig. 2, the pressure of the extracted air streamto be liquefied can be increased when extraction air flows are above

2RecycleW]

Liquid Air Flow[kmol h−1]

Liquid AirPressure [kPa]

Liquid O2Pressure[kPa]

5 5053.9 1586 6076 5079.8 1635 6412 5184.1 1826 7317 5337.4 2016 8274 5448.1 2217 9241 5603.7 2412 10007 5728.0 2612 10412 5892.2 2803 10698 6048.2 2990 10963 6256.4 3167 1110

Fig. 14. The effect of fuel diluent on the net power production for constant NOX

concentration in GT exhaust.

1695D. Jones et al. / Fuel Processing Technology 92 (2011) 1685–1695

the amount required for recovery of the liquid oxygen refrigeration.As the quantity of high pressure air needed to efficiently recover theliquid oxygen refrigeration is approximately 30% of the total air flow,air side integrations greater than 30% allows for greater pressurizationof the liquid oxygen. For that reason, it has been determined that themajor factors affecting the optimal ASU operating pressure are degreeof nitrogen injection and gas turbine operating pressure. Since it ispossible to see nearly equivalent or higher power savings from an LP-ASU with a pumped liquid oxygen cycle with degrees of airintegration below 50–60%, the degree of integration seems to haveonly minor effects on optimal ASU operating pressure below thisrange.

A study was carried out on how the net IGCC power is affected bythe choice of fuel diluents in the GT. Nitrogen and steam wereconsidered as fuel diluents and their concentrationswere variedwhilemaintaining the NOX level constant. The study shows that the degreeof nitrogen injection has a strong influence on the optimal operatingpressure of the ASU. Beyond about 50% nitrogen injection rate,Configuration 5 becomes more advantageous than Configuration 1.

This paper shows that an optimal integration between the ASU andGT and an optimal design of an ASU under various degrees ofintegration can help to increase the net power generation from anIGCC plant with CO2 capture. In future studies, capital cost informationcan be considered to determine optimal techno-economic designs.

References

[1] Cost and performance baseline for fossil energy power plants study, Volume 1:Bituminous coal and natural gas to electricity. National Energy TechnologyLaboratory, http://www.netl.doe.gov August 2007 DOE/NETL-2007/1281.

[2] D. Bhattacharyya, R. Turton, S. Zitney, Steady-state simulation and optimization ofan integrated gasification combined cycle (IGCC) power plant with CO2 capture,Ind. Eng. Chem. Res. 50 (2010) 1674–1690.

[3] D. Bhattacharyya, R. Turton, S. Zitney, Plant-wide Dynamic Simulation of an IGCCPlant with CO2 Capture, Proc. of the 26th Annual International Pittsburgh CoalConference, September 21–24, Pittsburgh, PA, 2009.

[4] B.L. Capehart (Ed.), Encyclopedia of Energy Engineering and technology, 2–3, CRCPress, Boca Raton, FL, 2007.

[5] R.J. Allam, E. Foster, V. Stein, Improving gasification economics through ITMoxygen integration, 5th (IChemE) European Gasification Conference, 8–10 April2002, Noordwijk, The Netherlands.

[6] H.C. Frey, Y. Zhu, Improved system integration for integrated gasificationcombined cycle systems, Environ. Sci. Technol. 40 (2006) 1693–1699.

[7] J.G. Wimer, D. Keairns, E.L. Parsons, J.A. Ruether, Integration of gas turbinesadapted for syngas fuel with cryogenic and membrane-based air separation units:Issues to consider for system studies, Journal of Engineering for Gas Turbines andPower 128 (2006) 271–280.

[8] A.R. Smith, J. Klosek, D.W. Woodward, Next-generation integration concepts forair separation units and gas turbines, Journal of Engineering for Gas Turbines andPower 119 (1997) 298–304.

[9] R.F. Geosits, L.A. Schmoe, IGCC – The challenges of integration, ASME Turbo Expo2005, , June 6–9 2005 Reno-Tahoe, Nevada.

[10] Foster Wheeler, Optimizing IGCC Design, Foster Wheeler Review, Autumn 1999,pp. 17–21.

[11] A.R. Smith, J. Klosek, A review of air separation technologies and their integrationwith energy conversion processes, Fuel Processing Technology 70 (2001)115–134.

[12] A.R. Smith, D.W. Woodward, Operation flexibility considerations in the selectionof a pumped LOX air separation process incorporating a source of high pressurefeed air, Air Products research disclosure, Oct. 1996.

[13] D.M. Todd, Gas turbine improvements enhance IGCC Viability, GasificationTechnologies Conference, Oct. 8–11 2000, San Francisco, California.

[14] M. Jonsson, J. Yan, Humidified gas turbines – a review of proposed andimplemented cycles, Energy 30 (2005) 1013–1078.

[15] M.J. Hornick, J.E. McDaniel, TampaElectric Polk Power Station IntegratedGasification Combined Cycle Project, Final Technical Report, DE-FC-21-91MC27363; prepared for the U. S. Department of Energy, Office of Fossil Energy,National Energy Technology Laboratory (NETL), Morgantown,WV, August 2002.

[16] A.K. Anand, C.S. Cook, J.C. Corman, A.R. Smith, New technology trends forimproved IGCC systems performance, Journal of Engineering for Gas Turbines andPower 118 (1996) 732–736.

[17] F.J. Brooks, GE gas turbine performance characteristics, GE Power Systems, GER-3567H, http://www.gepower.com/prod_serv/products/tech_docs/, RetrievedNov, 2010.

[18] Beate Seliger, Richard Hanke-Rauschenbach, Frank Hannemann, Kai Sundmacher,Modeling and dynamics of an air separation rectification column as part of anIGCC power plant, Separation and Purification Technology 49 (2006) 136–148.

[19] A.R. Smith, E.J. Noga, Air Separation Design for IGCC – Future Directions, EPRIGasification Technologies Conference, Oct 1996, San Francisco, CA.

[20] Bruce K. Dawson, Scott C. Siegmund, Yonggui Zhang, Flowsheet Optimization forMulti-Purpose Air Separation Units, The First Baosteel Annual AcademicConference, May 2004, Shanghai, China.

[21] R.J. Allam, H. Castle-Smith, Air Separation Units, Design and Future Development,Gasification of Coals, Biomass, and Oil EUROTHERM Seminar No 65, July 2000,Enschede, The Netherlands.

[22] T. Raddings, R. Jones, M. Scholz, Gas Turbine Flexibility for Low Turbine Fuels,Presented at IChemE Gasification, 9, March 24 2009, Dusseldorf, Germany.

[23] W. Schmidt, ASU Reboiler/Condenser Safety, Presented at European Industrial GasAssociation, Product Safety Seminar, February 2 2006.

[24] Jacobs Consultancy UK, Impact of CO2 removal on coal gasification based fuelplants, , Feb 2006.

[25] Chan Lee, Seung Jong Lee, Yongseung Yun, Effect of air separation unit integrationon integrated gasification combined cycle performance and NOx emissioncharacteristics, Korean J. Chem. Eng. 24 (2) (2007) 368–373.

[26] Chan Lee, Hyung Taek Kim, Yongseung Yun, Optimal integration conditionbetween the gas turbine air compressor and the air separation unit of IGCC powerplant, Energy Conversion Engineering Conference, 1997, IECEC-97, Proceedings ofthe 32nd Intersociety.

[27] Ola Maursad, An overview of coal based integrated gasification combined cycle(IGCC) technology, http://sequestration.mit.edu/pdf/LFEE_2005-002_WP.pdf2005.

[28] Calin-Cristian Cormos, Evaluation of energy integration aspects for IGCC-basedhydrogen and electricity co-production with carbon capture and storage,International Journal of Hydrogen Energy 35 (2010) 7485–7497.

[29] Marvin M. Schorr, Joel Chalfin. “Gas Turbine NOX Emissions Approaching Zero – Isit Worth the Price?”, GE Power Systems, GER 4172, http://www.gepower.com/prod_serv/products/tech_docs/, Retrieved Nov, 2010.

[30] Aspen Plus, Aspen Technology, Inc., Cambridge, MA; www.aspentech.com.