IAGT -October 15, 2007 1

Supersonic Ejector for Capturing Dry-Gas Seal Vent Gases

Hasan Imran US EPA Natural Gas STAR 14th Annual Senior Advisor, Climate Change Conference, Houston, Texas October 23, 2007

IAGT -October 15, 2007 2

Outline

1. Motivation and Challenges.

2. Why Supersonic Ejector.

3. Development of the Device.

4. Results of Testing of Two-Stage System.

5. Implementation at Compressor Station.

6. Conclusions.

IAGT -October 15, 2007 3

Process Side

Ambient Side

Clean Buffer Gas

Inert Buffer GasLeakage

Inert Buffer Gas

Typical Primary and Secondary Dry-Seal

IAGT -October 15, 2007 4

Primary Seal Leakage Rate

10-3 Kg/s

1.4

1.1

0.75

0.5

0.25

0 0 30

Shaft (mm)

250

200

150

100

50

60 90 120

Pressure-bar

IAGT -October 15, 2007 5

Ref: Thomas Robinson (TransCanada)

Why Capture Dry Gas Seal Vent?

Seal gas leakage 2.0-8.0 m3/hr ~average 5.0 m3/hr

1.55 MMScf/y # of TransCanada units 348 x 348 @ 50% utilization 269 MMScf/y Tonnes of natural gas 0.0440 lbs/cf

5,369 tonnes conversion to CO2-E 21 CO2-E 113 ktonnes/year

CO2 Emissions trading $15 per tonne of CO2-E Gas value $6 per MCF CO2-E value $1,691,237 Value of Gas $1,614,216 Total $3,305,453 Annual saving per unit: $9,498

IAGT -October 15, 2007 6

Concept of Ejector

Motive Gas

Suction Gas

Discharge Gas

IAGT -October 15, 2007 7

Why Ejector?

� Motive gas is available (fuel gas upstream of regulator).

� Discharge gas from the Ejector can be burned in the GG.

� Auxiliary boilers are not always employed in compressor station, and typically operate intermittently (on/off).

� Some stations do not have boilers (warm climate regions).

� No moving parts.

� Low cost.

IAGT -October 15, 2007 8

Why Supersonic Ejector?

� Low suction pressure 300-550 kPa-a.

� Hence high expansion ratio of motive/suction pressure (~ 14).

� Discharge pressure 2500–3400 kPa-a.

� Hence compression ratio up to 8.

� Low suction flow (5-15 m3/hr).

IAGT -October 15, 2007 9

Development of the SSE

� Two-stage supersonic ejectors:

¾ The first stage is highly supersonic (nozzle exit Mach number ≈ 2.54),

¾ The second stage is moderately supersonic (nozzle exit Mach number ≈ 1.72).

� Extensive testing to arrive at optimum design.

� Final configuration gave

¾ expansion pressure ratio (motive/suction) of the order of 14.0 ¾ compression pressure ratio (discharge/suction) of around 8.1

IAGT -October 15, 2007 10

1st Stage Design

Nozzle Diameter mm 1.6 Nozzle Exit mm 2.8 Half Angle deg 1

Diffuser Inlet Diamter mm 4 Half Angle deg 4.7 Diffuser Throat Diameter mm 3.5 Diffuser Throat Length mm 8 Diffuser Exit Diameter mm 18 Diffuser Exit Half Angle deg 5

IAGT -October 15, 2007 11

1st Stage (CFD Results)

Contours of Mach Number

Nozzle exit Diffuser inlet Diffuser throat Diffuser exit

IAGT -October 15, 2007 12

1st Stage (CFD Results)

Contours of Static Pressure

Nozzle exit Diffuser inlet Diffuser throat Diffuser exit

IAGT -October 15, 2007 13

m P T

P

P

P

T m

R

R R

Bench Testing at Compressor Station in Alberta

Expa

nsio

n R

atio

(P1/

P2)

Effic

ienc

y

1.6x2.8 mm Nozzle @ 20.5 mm position

22 22

20 20

18

16

14

12

10

8

6

Expa

nsio

n R

atio

(P1/

P2) 18

16

14

12

10

8

6

44

2 2

0 0 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Compression Ratio (P3/P2) Ejection Ratio (m2/m1)

0.6 (m& 1 + m& 2 )ψ 3η = 0.5 m& 1ψ 1 0.4

0.3

0.2

0.1

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 14IAGT -October 15, 2007

Compression Ratio (P3/P2)

IAGT -October 15, 2007 15

Implementation

P&ID for Supersonic Gas Ejector Skid

Atmosphere

(7) (8) (9)

(N1)

(5)

(Fisher 310)

(2)

(3)

(4)

M

M

Atmosphere

Liquid & Particles filter

Ejector Skid

Fuel Gas Heater

Fuel Line (suction)

Compressor discharge (outside unit valve)

Primary Seal Leakage

700 kPag 700 kPag

½ T

½ T

1”

1” 2”

(6)

Comp. Suction (yard)

(1) Insulated

Insulated

(N2)

2” 1”

1.5”

1.5”

M

1”¾”

500 kPag 500 kPag

½”½”

Fuel gas to Engine

IAGT -October 15, 2007 16

2nd Stage Ejector Nozzles and Internals

IAGT -October 15, 2007 17

Test bench at Didsbury Compressor Station

IAGT -October 15, 2007 18

Dis

char

ge P

ress

ure

- kPa

(abs

)Test Results of the Two-Stage SSE System

3500

3400

3300

3200P1=4600 kPa

P1=5000 kPa 3100 P1=5500 kPa

3000

2900

0 1 2 3 4 5 6

Suction Flow - kg/hr

IAGT -October 15, 2007 19

P1=4600 kPa p1 =5000 kPa P1=5500 kPa

Test Results of the Two-Stage SSE SystemIn

term

edia

te P

ress

ure

(P3)

- kP

a (a

bs)

1600

1500

1400

1300

1200

1100

1000 0 1 2 3 4 5 6

Suction Flow - kg/hr

IAGT -October 15, 2007 20

Fuel Regulator (Fisher 310)

1200 kPa-a

3400 kPa-a

5000 kPa-a

6000 kPa-a

(P1)

(Pout)

(P3)

(Pin)Fuel Gas Line

P&ID Schematic of the 2-stage Ejector system

500 kPa-a (Leakage from Dry-Gas Seal)

(P2)

IAGT -October 15, 2007 21

Implementation

IAGT -October 15, 2007 22

Implementation

Fuel Gas System to Gas Turbine

Ejector Skid

IAGT -October 15, 2007 23

Project Team After Installation

TransCanada Compressor Station in Alberta

IAGT -October 15, 2007 24

Implementation

Gas-Gas Ejector Skid

First Stage

2nd Stage

Filter

IAGT -October 15, 2007 25

Operating Data

Field Test data collected from Pilot Project

Location # (Refer to Fig. 11) P (kPa-a) T (oC)

Mass Flow (kg/s)

Mass Flow (kg/hr)

1 5090 35.9 0.01770 63.720 DE Seal 506 10.0 0.001607 5.785

NDE Seal 512 10.0 0.001607 5.785 Total Seal Leakage 509 10.0 0.003214 11.570

N1 509 -113.0 0.01770 63.720 2 509 10.0 0.00286 10.285

Vent to Amb 509 10.0 0.00036 1.285 3 1190 12.7 0.021 74.005 4 5990 35.0 0.463 1668.240

N2 1190 -68.0 0.463 1668.240 5 2990 8.2 0.484 1742.245 6 5140 10.0 1.177 4237.500 7 5090 35.9 0.693 2495.255 8 2900 23.0 0.693 2495.255 9 2690 18.4 1.177 4237.500

Collected After Successful Commissioning of the Supersonic Ejector (Data Collected on June 23, 2007).

IAGT -October 15, 2007 26

Operating Data

Continued – Filed Test Data

Location # (Refer to Fig. 11)

P (kPa-a) T (oC) Mass Flow

(kg/s) Mass Flow

(kg/hr)

1 5800 35.9 0.02017 72.612 DE Seal 512 10.0 0.00214 7.704

NDE Seal 517 10.0 0.001696 6.106 Total Seal Leakage 515 10.0 0.003836 13.810

N1 515 -113.0 0.02017 72.612 2 515 10.0 0.00250 9.000

Vent to Amb 515 10.0 0.00134 4.810 3 1400 12.7 0.023 81.612 4 6540 35.0 0.506 1820.880

N2 1400 -68.0 0.506 1820.880 5 3160 8.2 0.528 1902.492 6 5850 10.0 1.177 4237.500 7 5800 35.9 0.649 2335.008 8 3100 23.0 0.649 2335.008 9 2690 18.4 1.177 4237.500

Collected After Successful Commissioning of the Supersonic Ejector (Data Collected on July 20, 2007).

IAGT -October 15, 2007 27

Benefits

Ejector 2nd stage motive flow 0.4634 kg/s Motive gas extra power needed 9,300 W Motive gas turbine extra fuel needed 31,000 W Fuel heating value 39 MJ/m3 Fuel heating value 52,000,000 J/kg Fuel gas burned 0.000596154 kg/s GHG CO2 from above burning 0.00155 kg/s GHG CO2 from above burning 48.8808 tonnes/year

Captured Dry-seal vent gas 9 kg/hr Captured Dry-seal vent gas 0.0025 kg/s Heat Energy Equivalent 130,000 W GHG CO2-E of the captured gas 0.0525 kg/s GHG CO2-E of the captured gas 1,656 tonnes/year

Net Heat Energy Saving 99,000 W Net GHG CO2-E 1,607 tonnes/year

Net Savings

IAGT -October 15, 2007 28

Conclusions

¾ It is possible to achieve a very high expansion ratio (14-18 20 times) with a supersonic nozzle. The challenge is always in the recompression of the expanded gas through a supersonic diffuser.

¾ For the application of a supersonic ejector for the purpose of capturing the vent gas from the primary dry-gas seal, a two-stage ejector was developed:

• Expansion of the motive gas in the first stage ejector nozzle is easy.

• However, recompression to a pressure equal to the fuel gas pressure was the challenge, and that necessitated the second stage to boost the gas pressure to an adequate level.

IAGT -October 15, 2007 29

Conclusions

¾ Implementation design concept and analysis indicated that the above implementation plan, which attended to operational and functional details of both the primary and secondary seals, as well as the adjacent magnetic bearing chamber, is achievable.

¾ The S.S. ejector system was installed and commissioned successfully on an 24 MW GT unit in one of TransCanada’s compressor stations in Alberta, and is currently accumulating hours of operation:

• Reducing 1600 tonnes/year of CO2-E of GHG. • Saving of 99 kW of Heat Energy.

IAGT -October 15, 2007 30

Benefits

9 100% recovery of vented methane gas 9 100% capture of greenhouse gas emissions 9 No moving parts 9 No seal, no shaft, no packing 9 Maintenance free operation 9 Zero operating cost 9 Use is not limited to compressors only 9 Can be used wherever there is low pressure, low flow rate

vent stream and a high pressure stream 9 Small footprint 9 Pay back period 1-2 years

IAGT -October 15, 2007 31

Global Pipeline Honorable Mention Award

Received at Rio Pipeline 2007 Conference,

Rio de Janeiro, Brazil

IAGT -October 15, 2007 32

Acknowledgment

Acknowledgement

TransCanada:Bruce Huynh, Barry Nilsson, Doug Mantai

John Straughan, Les Wagner Liz Sirakowski, Marilyn Carpenter Thomas Robinson, Anthony Tse

Brian McConaghyPaul MacGregor

Consultation with:Dwayne Krause (Starco Eng)

Vladimir Bakalchuk (John Crane)Sorin Necula (Dresser-Rand)

NRTC:Chris Foy

Michel BerghmansRobert McBrien