1

MINIMIZING THE CO2 EMISSION FROM THE LIQUEFACTION PLANT

Yoshitsugi Kikkawa, Dr. of Engineering, Engineering Consultant, Moritaka Nakamura, Fellow of Technology & Engineering Division

Chiyoda Corp., Yokohama, Japan

ABSTRACT

The 1st generation LNG power chain for Japan started with the use of gas supplies from Alaska Kenai LNG, Brunei LNG and ADGAS LNG. The predominant purpose of these 1st generation LNG electric power stations was to reduce the air pollution caused by previous use of higher sulfur content heavy oil power stations. That planned air pollution reduction has been successfully achieved. Use of LNG in electric power generation has contributed to reducing air pollution not only in Japan but also in other countries.

Several ten years after the implementation of those 1st generation LNG power stations, further reduction of CO2 emissions has become the focus for mitigating the adverse effects of the global warming phenomena including abnormal and more extreme weather patterns and their disastrous consequences, such as stronger hurricanes, heavier rainfall, warmer temperatures, flooding, landslides, drought in farming regions, receding glaciers, melting of the permafrost zone in Siberia, melting polar ice caps, rising sea levels, ocean acidification, etc.

After the Fukushima Daiichi Nuclear Power Station accident caused by the March 11, 2011 tsunami, people no longer wish to rely too much on nuclear power, even though it produces very small CO2 emissions.

However, until such time that large scale economically viable renewable energy is successfully developed, people will still rely on the use of fossil fuel for power generation.

Among the various types of fossil fuel, natural gas produces the least amount of CO2 emission per unit of heating value.

Fortunately, the emerging new resource of shale gas will provide increased gas sources to feed LNG liquefaction plants to supplement existing more conventional gas resources, and help reduce overall CO2 emissions.

In the LNG power chain, which includes upstream, liquefaction, LNG tanker, receiving terminal and power generation elements, efforts have been made to reduce the CO2 emissions caused by each element.

Among these elements, the liquefaction plant process has great potential to reduce the CO2 emissions, by way of:

Acid gas removal and carbon capture and storage (CCS)

Optimizing the liquefaction system.

Minimizing the flare load during train start-up and shut down

Optimizing the prime mover system, including e-drive

Carbon capture and storage (CCS) from the flue gas of the plant

2

1. INTRODUCTION

Regarding the reduction of the CO2 emission from the Base load LNG plant, several papers has been presented at the past LNG International Conferences[1][2][3]. To respond to ever–increasing general public social concerns about global warming, this paper contains the results of studies on how to reduce the CO2 emissions from the Base load LNG plant towards zero.

2. STUDY BASIS

The following basis was set for this study paper:

2.1 Location of the Plant

The location was assumed to be somewhere in Oceania, which has great potential for the future installation of a base load LNG plant

2.2 Feed Gas Condition and Ambient Temperature

(a) Feed gas composition was set as per the following Table and is based on typical Australian LNG composition.

Component Mol% CO2 1.0 N2 0.1 C1 86.5 C2 8.2 C3 3.4 C4 0.8 C5 0.0

(b) The feed gas was supplied with the following conditions:

(i) Pressure: 70bar

(ii) Temperature: 27deg.C

(iii) An air cooling system was used for the plant.

2.3 Feed Gas Price

The Base load LNG plant consumes a certain amount of fuel, the cost of which is affected by the price of feed gas at any particular time. This variable price will affect the cost of reducing CO2 emissions. For this study, three (3) different fuel prices were considered - 2US$/mmbtu, 4US$/mmbtu and 6US$/mmbtu.

2.4 Plant Capacity

Plant capacity was taken to be 9-10 MTA (million ton per annum), which equates to two liquefaction trains.

3

2.5 Liquefaction Process

There are several liquefaction processes that have been used on previously constructed Base load LNG plants: (i) C3-MR (ii) DMR (iii) Cascade (iv) SMR (v) AP-X Among these processes, an LNG train capacity of around 5MTA will not make use of AP-X and SMR processes.

The C3-MR (propane – mixed refrigerant) liquefaction process was selected for this study, and the sub cooled process (LNG direct rundown to the storage tank) was applied for the liquefaction due to the small quantity of N2 content. The basic process flow diagram is shown in Fig. 2.1.

Figure 2.1: Typical C3-MR Process Flow Diagram

2.6 Delivery Pressure of CCS

The delivery pressure of acid gas for CCS will be affected by the formation location and depth or by the required user pressure such as EOR. Based on the foregoing variables, the pressure range will generally vary between 120 and 220bar [4]. For the purpose of this study, the delivery pressure was set at 150bar.

4

2.7 CO2 Price for EOR

The CO2 price for this study was set at US$40 per ton of CO2 based on recent NETL data [5].

2.8 Carbon Tax for CO2 Emission

For the evaluation of the CCS cost, carbon tax of US$24 per ton of CO2 was estimated for an Oceania located Base load LNG plant based on reported tax levels[6]. The carbon taxes of several countries are listed in Table 2.1 for comparison [7][8].

Table 2.1: Carbon Tax Example

Country Currency Carbon Tax Currency/ tCO2 Currency/ US$ Carbon Tax

US$/tCO2 Finland euro 20 1.318 26.4 Sweden SEK 1,010 0.153 154.2 Norway NOK 371 0.179 66.3 Denmark DKK 90 0.177 15.9 Australia A$ 23 1.037 23.8

3. STUDY RESULTS

The CO2 emission from Acid Gas Removal (AGR) is derived from the CO2 content of the feed gas, i.e. the emission rate cannot be controlled. The CO2 from the flue gas (Fuel CO2) mainly comes from the driver system of the refrigerant compressor. This emission rate can be reduced by optimizing the driver system. Previous LNG project figures are shown in Fig. 3.1 [9]. Generally speaking, the fuel CO2 is the major part of the total CO2 emission.

Figure 3.1: tCO2 Emission /tLNG from Operating LNG Plant

0.000.050.100.150.200.250.300.350.40

tCO2/tLNG AGR CO2tCO2/tLNG Fuel CO2

5

3.1 Acid Gas Removal (AGR) and Carbon Capture and Storage (CCS)

The CO2 in the feed gas is removed by popular amine process. The CO2 from the Base load LNG plant has been successfully injected into the underground formation on a recent Base load LNG plant such as Snohvit LNG. The injection will require four (4) compressor stages and each stage will have a cooler and the water will be separated from the K.O. drums. The water content of the stream will be removed by dehydrator at the 4th stage inlet. The estimated costs of CCS are shown in Fig.3.2. The estimated costs include: (a) Additional Equipment Costs - compressor, inter/after coolers, separation vessels,

dehydrator, etc (b) Fuel Cost - additional fuel cost for CO2 compression

Figure 3.2: AGR CO2 CCS Cost

The CCS cost for the feed gas CO2 content of 1 mol% indicated above is over US$60 and it is considered as marginal when the Base load LNG plant locates near to EOR operation and can get the exemption of the carbon tax. The CCS cost will be reduced in scale when the feed gas contains up to 5mol%. Fig.3.2 shows the CCS cost will reduce to US$40 level, which is close to the current level of CO2 price for EOR, even if the carbon tax exemption is not taken into account.

3.2 Optimizing the liquefaction system

There are several items that can enhance the thermal efficiency of the liquefaction process: (a) Turbo-Expander Application

J-T valve replaced with expander. There are two (2) types of expander. (i) Liquid Expander

Liquid expander manufactured by Ebara International has been successfully used for Oman LNG and several other LNG projects, and has improved the liquefaction cycle efficiency to some extent.

20.025.030.035.040.045.050.055.060.065.070.0

0.0 2.0 4.0 6.0 8.0

CO

2 U

S$/tC

O2

Fuel Cost US$/mmbtu

1mol %2mol %5mol%

6

(ii) Two Phase Expander Two phase expander has also been successfully used for the NGL recovery plant in Poland by Ebara International [10][11]. The process conditions for the recovery plant are similar to the liquefaction process and similar application on LNG plants will be realized in near future. The cross section of the two phase expander is shown in Fig.3.3.

Figure 3.3: Cross-section of Two-Phase Expander

The expected cycle efficiency improvement is listed in Table 3.1.

Table 3.1: Expected Cycle Efficiency Improvement

Expander Location Liquid Expander Two-Phase Expander LNG 2.5% 3.0% Light MR 0.5% 0.7% Heavy MR 2.2% 2.8%

(b) Wet Surface Air Cooler (WSAC)

This concept has been successfully applied for refrigeration units in the steel, food, chemical and power industries, although it has never been used on a Base load LNG plant. Chevron has studied using a propane refrigeration system for a base load LNG plant [12]. Base load LNG plants are often located in areas with a dry climate where the wet bulb temperature is considerably lower than the dry bulb air temperature. The theoretical refrigeration power is basically determined by the Carnot cycle, i.e. based on following equation:

Carnot Work = Qc (T1/T2-1) where: Qc = Cooling Duty

7

T1 = Heat Rejection Temperature, Kelvin T2 = Cooling Process Temperature, Kelvin

The heat rejection temperature of the wet surface air cooler is the wet bulb temperature while that of the conventional air cooler is the dry bulb temperature. The flow diagram of the WSAC is shown in Fig.3.4. The relationship between wet bulb temperature and air relative humidity at the ambient temperature of 27deg.C dry bulb temperature is shown in Fig. 3.5. The refrigerant compressor power vs. the relative humidity at dry bulb temperature of 27deg.C is shown in Fig.3.6. The base point of 100% is the case for conventional air cooler application. The figure shows that WASC application at low relative humidity of 50% will reduce the required refrigeration compressor power to 93% of that used for the conventional air cooler.

Figure 3.4: WSAC Flow Diagram

Figure 3.5: Wet Bulb Temperature vs. Relative Humidity @ 27 deg.C

10

12

14

16

18

20

22

24

26

28

30

0 20 40 60 80 100

Wet

Bul

b Te

mp.

C

Relative Humidity %

8

Figure 3.6: Ref. Power vs. Relative Humidity of Air for WSAC Application

(c) Gas Turbine Inlet Air Humidification

This item is a similar concept to above idea. The humidification lowers the inlet air and results not only power augmentation but also enhancement of the thermal efficiency. Typical enhancement for GE Frame 7 in term of Heat Rate (HR) is shown in Fig. 3.7.

Figure 3.7: Heat Rate vs. Relative Humidity of Air for GE Frame 7

90%

91%

92%

93%

94%

95%

96%

97%

98%

99%

100%

101%

30% 40% 50% 60% 70% 80% 90% 100%

Ref

. Pow

er

Relative Humidity

98.0%

98.5%

99.0%

99.5%

100.0%

40 50 60 70 80 90 100

Hea

t Rat

e

Relative Humidity, %

HR

9

3.3 Minimizing the Flare Load

(a) Start-up and Scheduled Shut Down To calculate the flaring quantity at the start-up and shutdown is not easy, since it greatly depends on operator skill and the plant configuration. However, the flare load during the start-up on Snohvit LNG was successfully reduced by using dynamic simulation [13]. Therefore carrying out a dynamic simulation for the start-up and shut down will indicate how to reduce the flare load and flare quantities [14].

(b) Flare Load from Relieving Device

The maximum flare load of the LNG plant is usually the flare load in the event of a blocked discharge of the C3 Refrigerant Compressor. Air Products has issued a paper and patent which limits the discharge pressure and minimizes the flare load in the event of a blocked discharge [15]. The idea is to increase the design pressure of the compressor casing and the overload event bogs [slows] down the gas turbine driver and propane hold up is kept under the design pressure.

3.4 Driver Option

The cycle efficiency of the liquefaction process and system is well developed and there is only small room for improvement. However, the driver option has the potential to reduce the CO2 emission to a great extent through the continuous improvement of the thermal efficiency of the gas turbine and the waste heat recovery from the gas turbine exhaust.

The basic performance data of gas turbines for ISO condition is summarized in Table 3.2.

Table 3.2: Performance of Gas Turbine by GE

Name GE Model Type ISO Power (MW) Thermal Efficiency

GT ST LM2500 LM2500+G4 Aero 31 - 40.4% LMS100 LMS100 Aero 100 - 43.7% Frame 6 Frame6B Heavy Duty 42 - 32.1% Frame7 Frame7EA Heavy Duty 86 - 32.7% Frame9 Frame9E Heavy Duty 130 - 33.1% S106B S106B Combined Cycle 38 22 49.0% S106FA S106FA Combined Cycle 67 42 52.9% S109E* S109E Combined Cycle 123 70 53.0%

*Note: The Option 3 configuration is based on this type.

Driver options are considered as follows, Table 3.3, based on the papers presented at the GPA Annual Meeting and AIChE Spring Meeting [16][17].

Table 3.3: Driver Configuration for Driver Options

Case C3 Compressor Driver MR Compressor Driver CCS Option 1 Frame 7 (C3+HP MR) Frame 7 (LP +MP MR) No Option 2 LMS100 (C3+HP MR) LMS100 (LP +MP MR) No Option 3 Steam Turbine Frame 9 No Option 4 Motor Motor No Option 5 Motor Motor Yes

10

Option 1

This is the base case for this study and is considered to be the most popular scheme for 4 to 5 MTA train capacity. The power plant utilizes GE frame type power generators. N+1+1(Stand-by) configuration was applied for the power plant configuration. The required power for MR Compressor is about two times of that of Propane Refrigerant Compressor. To suit the available power of GE Frame 7’s, Split-MR* configuration was applied for this study [18]. Limited heat recovery from the gas turbine exhaust is achieved by heat transfer fluid (HTF) to supply the heat to: Acid Gas Regenerator Reboiler Dehydrator Regeneration Heater *Note: trade name of Air Products. The compressor and driver configuration is shown in Fig.3.8.

Figure 3.8: Option-1 Configuration

Option 2

This case is one of the representative schemes used for a power plant that utilizes GE aero-derivative type gas turbine generators. N+1+1(Stand-by) configuration was applied for the power plant configuration. LM2500+ and LM6000 have already been considered as refrigerant compressor drivers. Shell Global Solutions presented a paper at the LNG16 conference and supported the use of LMS100 as a refrigerant compressor driver [19]. The aero-derivative type gas turbine has better thermal efficiency when compared to Frame Type gas turbines. Limited heat recovery from the gas turbine exhaust is achieved by heat transfer fluid (HTF), similar to Option-1.

11

This will result in less CO2 emissions from the fuel. The compressor and driver configuration is shown in Fig.3.9.

Figure 3.9: Option-2 Configuration

Option 3

This case is aims to fully utilize the waste heat from the gas turbine exhaust. The power plant utilizes GE frame type gas turbine generators. N+1+1(Stand-by) configuration was applied for the power plant configuration. The configuration is so called “co-generation”. However this system is complex and costly when compared with Options 1 and 2, although CO2 emission will be much less than that for Option 1. The GE Frame 9 gas turbine drives the MR compressor and the turbine steam generated from HRSG of the Frame 9 gas turbine drives the propane refrigerant compressor. This waste heat recovery steam generator (HRSG) creates 89bar steam, and part of this steam is utilized for the Dehydrator Regeneration Heater. 4bar steam is also extracted from the steam turbine for the Acid Gas Regenerator Reboiler. Waste heat recovery HRSG is also used for the power generator gas turbines.

The compressor and driver configuration is shown in Fig.3.10.

12

Figure 3.10: Option-3 Configuration

Option-4

This case aims to achieve the highest thermal efficiency in the driver system which will result in the lowest CO2 emission from the Base load LNG plant. This case uses a motor driver (e-drive) system as the refrigerant compressor driver. The electrical power is supplied from high thermal efficiency combined cycle. This case will be costly since the refrigerant compressor driver power is indirectly supplied from gas turbine through electrical power system, although it will be the most reliable system. This option was considered in response to the ever-increasing general public social and environmental demand for the reduction of CO2 emissions, although this option will not be justified from economical viewpoint, given the low cost of natural gas fuel. The compressor and driver configuration is shown in Fig.3.11.

13

Figure 3.11: Option-4 Configuration

Option 5

In order to minimize CO2 emission from the base load LNG plant, this option used CCS (Carbon Capture and Storage) in addition to the Option 4 configuration. This option was raised to estimate the cost of CCS by the comparison with Option 4. The MEA process was applied to absorb CO2 in the flue gas (fuel CO2) [20]. The configuration is shown in Fig. 3.12.

14

Figure 3.12: Process Configuration for Fuel CO2 CCS

The MEA Absorber is provided for each flue gas from the combined cycle unit of GE S106FA. A single MEA Regenerator is provided for the MEA regeneration. The reboiler steam is extracted from the steam turbine of the combined cycle unit. The CO2 gas from the MEA Regenerator is compressed by four (4) stages of centrifugal compressor.

The power plant configuration is summarized in Table 3.4.

Table 3.4: Power Plant Configurations for Driver Options

Case Operation Stand-by Remarks Option-1 Frame 6 x3 Frame 6 x1 Option-2 LM2500+ x 4 LM2500+ x1 Option-3 S106B x2 +Frame6 Frame 6 x1 Option-4 S106FA x4 S106FA x1 Option-5 S106FA x5 S106FA x1

3.5 Comparison of Fuel CO2 Emission

The CO2 emission per unit of LNG production is summarized in Fig. 3.13. Option 1 is the base case and resulted the greatest CO2 emission among Options 1 to 4.

15

Figure 3.13: Fuel CO2 per ton LNG

3.6 CCS Costs Estimation for Fuel CO2

The cost of CCS is estimated by the comparison of Option 4 and Option 5. The results are shown in Fig. 3.14. The cost of CCS is shown to exceed US$150 /tCO2. This cost level is far in excess of the current Oceania CO2 price for EOR operation. However, if the future carbon tax in Oceania exceeds US$150 /tCO2, which would be similar to the current Sweden carbon tax, this option should be considered.

0.000

0.050

0.100

0.150

0.200

0.250

Option-1 Option-2 Option-3 Option-4

tCO

2/tL

NG

Option

tCO2/tLNG

16

Figure.3.14: CO2 CCS Cost for Fuel CO2

4. CONCLUSION AND FUTURE CONSIDERATION

To respond to the ever-increasing general public social and environmental concerns on global warming, this paper has provided wide options to address the reduction of CO2 emissions from the liquefaction plant towards zero.

The AGR CCS will be reasonably justified when EOR operation is located near the LNG plant.

Increasing the thermal efficiency of the driver system will be reasonably justified by reduction of the fuel requirement. However, the CCS of fuel CO2 will be difficult to justify even where EOR can be used at the location.

In future, if the general public social and environmental and/or government request /demand to further reduce the CO2 emission becomes reality, although the additional cost for installing CCS is expensive, the CCS of fuel CO2 will be performed at the LNG plant site since the fuel cost of additional fuel which is required to the CCS facility will still lower than the LNG end user purchase cost, more over LNG plant site often locates near EOR operation, which will contribute the cost reduction of the CCS.

140.0

150.0

160.0

0.0 2.0 4.0 6.0 8.0

CC

S C

ost $

/tCO

2

Fuel Price $/mmbtu

CO2 CCS Cost for Fuel

17

REFERENCES

1 Kikkiwa, Y., Liu, Y.,“Zero CO2 Emission for LNG Power Chain?”, LNG13 International Conference, Seoul Korea, May 14-17 2001.

2 Rabeau, P., et al, “How to Reduce CO2 Emissions in the LNG Chain”,LNG15 International Conference, Barcelona, Spain, Apr.24-27 2007.

3 Coulson et.al. "PO1-14 CARBON CAPTURE OPTIONS FOR LNG LIQUEFACTION" LNG16 International Conference, Oran, Algeria, Apr.18-21 2010

4 http://www.uwyo.edu/owenphillips/papers/energy%20journal%20proof%20022310.pdf

5 DOE/NETL-2011/1504, Improving Domestic Energy Security and Lowering CO2 Emissions with “Next Generation” CO2-Enhanced Oil Recovery (CO2-EOR)

6 http://www.sbs.com.au/news/article/1492651/Factbox-Carbon-taxes-around-the-world

7 http://www.une.edu.au/business-school/working-papers/economics/econwp11-2.pdf

8 http://www.norway.or.jp/news_events/environment/co2ets_seminar/

9 Yost, C., DiNapoli, R.: “Benchmarking Study Compares LNG Plant Costs”, Oil & Gas Journal, April 2003.

10 Katarzyna Cholast et. al.,"Two-Phase LNG Expanders" Gas Processors Association – GTL and LNG in Europe. Amsterdam 24th-25th February 2005

11 Kikkawa et. al."Completing the Liquefaction Train by Using Two-Phase LNG Expanders" AIChE Spring Meeting, Tampa, Florida, USA, Apr.27-30 2009

12 Kuo, J. C. et. al., "49e. New Cooling Application: Total Heat Removal from Base Load LNG Plant", AIChE Spring Meeting, Chicago, IL, Mar. 13-17, 2011

13 Bauer, H., 2012 AIChE Spring Meeting Paper"25a Mixed Fluid Cascade, Experience and Outlook", Houston, TX, Apr.1-5, 2012

14 Masuda, et.al."The Use of Advanced Dynamic Simulation Technology in the Engineering of Natural Gas Liquefaction Plants", GASTECH2012, ExCel London, UK, 8-11, 2012

15 Matthew J., et. al., "26e Reduction of Flare Loading During a Refrigerant Compressor Blocked Discharge in a LNG C3/MR Process", AIChE Spring Meeting, Orlando, FL, Apr.23-27, 2006

16 Kikkawa et. al. "Optimize The Power System of Baseload LNG Plant", 80th Annual GPA Convention, San Antonio, TX, March 12-14, 2001

17 Kikkawa et.al., "Availability of Refrigeration Process of Baseload LNG Plant", AIChE Spring Meeting, New Orleans, LA, Mar.10-14, 2002

18 Liu, Y.N., et. al. “Reducing LNG Costs by Better Capital Utilization” LNG 13 International. Conference, Seoul, Korea. May14-17, 2001

19 Sjarel van de Lisdonk, et.al. "PS3-2 Next Generation On‐Shore LNG Plant Designs" LNG 16 International Conefrence, Oran, Algeria, Apr.18-21 2010

20 Simmonds, et.al. "Amine Based CO2 Capture from Gas Turbines" Second Annual Conference on Carbon Sequestration – Alexandria, Va., USA, May 5-8 2003