Knowledge of Real Fluid Behaviour –the Key to Successful Gas Processing Systems

Lars Henrik Gjertsen, Manager Gas Processing

StatoilHydro Research Centre Trondheim

2

Outline• Definition Gas Processing

• Some Challenges Gas Processing

• Gas-Liquid Separation

• Solubility challenges

– Glycol Solubility in Gas

– Salt Solubility in Glycol

3

Gas Processing - elements

2 phase flowor Multiphaseflow

Sales gasPipeline transportation

Gas conversion

LNG production

•Gas-liquid separation•Dehydration•Gas sweetening (CO2 and H2S removal)•Trace component removal and handling•Glycol reclaiming and regeneration•Heat exchange•Compression/expansion

4

Some ChallengesGas-liquid separation•Compressor breakdown•Upsets in contactors and absorbers•Low efficiency/malfunction of adsorbers and absorbers•Off-spec gas product

Absorption processes•Absorbent capacity and kinetics •Foaming•Emulsions due to additives•Loss of absorbent

Solubility of trace components•Changes with process parameters•Accumulates and deposits

5

DEMISTING CYCLONES –Impact of Fluid and Pressure

40

50

60

70

80

90

100

0 5 0 1 0 0 1 5 0 2 0 0 2 5 0

Liquid load to cyclone

Sepa

ratio

n ef

ficie

ncy

[%]

Constant velocity

Model fluid at 40 bar

Live hydrocarbons at 40 bar

6

40

50

60

70

80

90

100

0 5 0 1 0 0 1 5 0 2 0 0 2 5 0

Liquid load to cyclone

Sepa

ratio

n ef

ficie

ncy

[%]

Live hydrocarbons at 40 bar

Model fluid at 90 bar

Constant velocity

Model fluid at 40 bar

Test pressure effect=> Reduced efficiency

Scaling rule important

Velocity not proper for impact of increased pressure

DEMISTING CYCLONES –Impact of Fluid and Pressure

7

40

50

60

70

80

90

100

0 5 0 1 0 0 1 5 0 2 0 0 2 5 0

Liquid load to cyclone [l/hr]

Sepa

ratio

n ef

ficie

ncy

[%]

Live hydrocarbons at 40 bar

Live hydrocarbons at 94 bar

Model fluid at 90 bar

Test fluid effect=> Reduced efficiency

Test pressure effect=> Reduced efficiency

Constant velocity

Model fluid at 40 bar

DEMISTING CYCLONES –Impact of Fluid and Pressure

8

90

92

94

96

98

100

0 .0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 .7

Liquid load to the cyclone deck

Sepa

ratio

n ef

ficie

ncy

[%] 110 bar

DEMISTING CYCLONES –Impact of Velocity

Acceptable performance possible at high pressure

9

DEMISTING CYCLONES –Impact of Velocity

50

60

70

80

90

100

0 .0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 .7

Liquid load to the cyclone deck

Sepa

ratio

n ef

ficie

ncy

[%]

110 bar110 bar, 25% increased gas flow110 bar, 57% increased gas flow

Capacity deviate largely from low pressure testing

10

Fluid Behaviour – Gas–Liquid Separation

• Fluid impact– Water not representative for hydrocarbon systems

– Large impact of fluid properties such as surface tension

• High pressure separation– Large impact of pressure

• Scrubber elements–Large variation in characteristics

Testing: basis for fundamental understanding and establishment of proper scaling rules

11

Challenge - Gas specification

• High cricondenbar, 127 barg. Contactor determines cricondenbar

• Liquid entrainment from 2nd stage scrubber.

0102030405060708090

100110120130

-20 0 20 40 60 80Temperature [°C]

Pres

sure

[bar

a] Measured dew pointsSimulated based on GC analysis, +100ppm C10+Operating point 2nd stage scrubberOperating point contactor

EOS = SRK

1st stage separator

2nd stage scrubber

1st stage scrubber

Glycol contactor

3rd stage scrubber

Rich gas (RG)

12

After modification of scrubbers

• Simulated dew point curve at higher temperature than scrubber operating temperature indicates liquid carry over from scrubber

• Need for dew point measurements offshore or representative gas samples taken in single phase flow for dew point measurements in laboratory

0102030405060708090

100110

-20 -10 0 10 20 30Temperature [°C]

Pres

sure

[bar

a]

Simulated based on GC analysis of composition

Simulated dew point curve at scrubber operating point

Operating point 2nd stage scrubber

Liquid carry over from scrubber?

EOS = SRK

13

0102030405060708090

100110

-20 -10 0 10 20 30Temperature [°C]

Pres

sure

[bar

a]

Simulated based on GC analysis of composition

Measured dew point curve

Operating point scrubber

EOS = SRK

Problem Solution

• Experimental dew point measurements needed to find correct conclusions

• True dew point curve steeper than simulated as also found for the synthetic gases

Modelling uncertainties. Not a scrubber problem!

14

Phase behaviour

Phase behaviour of natural gas with traces of water (40 ppm(mole)),NG composition (mole): 85 % C1, 10 % C2, 4 % C3, 0.5 % nC4, 0.5 % iC4

0

10

20

30

40

50

60

70

80

90

100

-60 -40 -20 0 20 40

Temperature [ºC]

Pres

sure

[bar

]

Hydrocarbon dew pointHydrocarbon buble pointwater dew pointfrost pointhydrate point

CPA-EoS (Cubic-plus-Association equation of state)

Can calculate dew point, frost point and hydrate pointAlso well applicable for aqueous dew point

15

0

10

20

30

40

50

60

70

80

90

100

-60 -40 -20 0 20 40

Temperature [ºC]

Pres

sure

[bar

]

Hydrocarbon dew pointHydrocarbon buble pointaqueous dew pointfrost pointhydrate pointTEG freezing point

Phase behaviour – glycol solubility

Phase behaviour of natural gas with traces of water (40 ppm(mole)) and TEG (0.5 ppm(mole)),NG composition (mole): 85 % C1, 10 % C2, 4 % C3, 0.5 % nC4, 0.5 % iC4

16

Challenge – Contamination of absorbent

0

1

2

3

4

5

6

7

8

9

10

06 .08 .2 007 1 4 .1 1 .2 007 2 2 .02 .2 008 01 .06 .2 008 09 .09 .2 008 1 8 .1 2 .2 008 2 8 .03 .2 009 06 .07 .2 009 1 4 .1 0.2 009 2 2 .01 .2 01 0

Date

Gly

col i

n A

min

e (%

)

Analysed No effect limit Linear (Analysed)

Glycol dehydration unit in upstream process

Glycol soluble in gas

Accumulates in Amine

Amine capacity affected

17

Scaling in Kollsnes condensate handling

93 bar5°C

25 bar25°C

122°C

7 bar35°C

7bar35°C

7bar 35°C

70°C

122°C Restriction

Valve open to 100%

Reduced condensate rate

18

Questions to task force• Why scale in a condensate system?

• Why now after several years of operation?

• Why NaHCO3 – a salt with high solubility?

• How to remove it without shutting down the production?

– A wash/replace will require 8-12 hours -> loss of 50-70 MSm3 gas

• Which chemicals can we use that will not contaminate the condensate?And please hurry!

The valve is about to get plugged once more!

19

Process analysis• Only one condensate MEG separator

– MEG in condensate: 600 ppm

• Start of second cond-MEG separator

– MEG in condensate: 30-40 ppm

• Improved separation, why problem?

– Less MEG should give less salt and less precipitation?

20

0

10

20

30

40

50

60

70

80

90

100

0 200 400 600 800 1000MEG in condensate (ppm)

MEG

as

liqui

d (%

)MEG evaporation in stripper column• MEG is depressurised to 7 bar and heated to 35°C

– Solubility of MEG and water in gas increases

• What happens when MEG in condensate is reduced from 600 to 30-40 ppm?

600 ppm:50% of water/MEG still as liquid

2-300 ppm:No more liquid water/MEGComplete evaporation

30-40 ppm

21

Scale removal options

• Open/replace valve

– Require shutdown, loss of 50-70 MSm3 gas

• Carbonate salt -> Use an acid (suggested by a service company)

– Require shutdown as acid would contaminate condensate

• Water – NaHCO3 is highly soluble

– Possible, but will increase water content in condensate

– May cause hydrate formation in condensate transfer line

• What about using MEG?

22

Use of MEG to dissolve scale• NaHCO3 is soluble in MEG

• Advantages with MEG

– Already present and will not contaminate the condensate

– MEG is available

– No use of other chemicals

– Spent MEG can be treated in MEG regeneration

0

200

400

600

800

1000

1200

1400

1600

0 20 40 60 80 100MEG concentration (wt%)

NaH

CO

3 so

lubi

lity

(mm

ol/k

g)

35°C

Rich MEG

23

Treatment – MEG injection• MEG injected into condensate upstream stripper

• Injection rate: 40 litre/hour

– Rate adjusted to get accumulation in glycol settler

• Treatment duration: 24 hours

– Total MEG consumption is about 1 m3

• MEG collected in glycol settler and sentto MEG regeneration

24

Summary – key to success• Experimental evaluations

– Evaluations have to be carried out with real fluid systems – model systems will deviate from real systems

– Large impact of pressure; high pressure processing is a challenge

– Establish fundamental data and knowledge of mechanisms

– Developing improved design and solutions

• Modelling

– Experimental data and experiences need to be incorporated into models

– Models to be used in combination with best practices

• Operational experience and problem definition

– Important to identify where data/knowledge is needed

– Combination with experimental experience proven to be successful