18 Predicting the Onset of Liquid Loading and Dynamic Behaviour of LL Gas Wells

Predicting onset of liquid loading and dynamic behaviour of liquid loaded gas wells 2nd European Conference on Gas Well Deliquification

G. Alberts S. Belfroid E. Biezen W. Schiferli K. Veeken

TNO Science and Industry TNO Science and Industry NAM TNO Science and Industry NAM

Predicting onset of liquid loading

Introduction • Goal : prediction of behaviour of wet gas well • Prediction of the liquid loading point • Simulation of a loading well • Start-up/shut-in scenario’s • Investigation best loading mitigation strategies • Control of loading wells (f.i. Gas injection) • Models • Steady models (f.i. Turner) • Dynamic models (f.i. OLGA, TNO dedicated models) • Natural gas wells, laboratory data (TU Delft Air/Water)

2

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Simulation Well (loading) Well 2 Olga 70000

30

Production field data

Production data 60000

Wellhead pressure

25

Qproduction [m3/d]

20 40000 15 30000

Wellhead pressure

20000

10

Simulation (OLGA)

Wellhead pressure [bara]

50000

5

10000

Shutin and foam job 0 0

100000

200000

300000

400000

500000

600000

700000

800000

900000

0 1000000

t [s] 3

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Reservoir decline simulation Natural gas well 2 140

120

Decrease in reservoir pressure

P(toe) [bar]

100

Olga

80

Pres=85 bara 75 65

60

55 45 40

35

20

0 0

50000

100000

150000

200000

250000

300000

350000

400000

450000

500000

Gas mass flow [Sm3/day] 4

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Prediction liquid loading (Dynamic reservoir depletion (740 days)) 60

Pressure (Bottomhole) [bara]

Total liquid

18

50

16

Droplets

14

40

30

12

TPC decline TPC (ss)

10 8

20

6

Liquid film

4

10

Liquid film mass [Sm3/day]; Liquid ratio (LGR) [e6Sm3/Sm3]

20

2 0 40000

0 50000

60000

70000

80000

90000

100000

110000

120000

Gas flow rate [Sm3/d] 5

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Liquid loading • Critical : Minimum gas flow velocity to lift liquid • Main parameters • Liquid content • Inclination angle • Gravity • Non-uniform film thickness • Condensation/production liquid • Entrainment • Well - reservoir interaction (2006) (small A factors) 6

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Air – Water experiments (TU Delft) E xpe rime nta l da ta TU De lft

Critica l gas ve locity [m/s ]

20

Gravity 15

10

5

0 0

7

Increased film thickness

20 40 60 80 Inclina tio n a ngle (90 = ho rizo nta l) [de gre e s ] Ref: Jos van ‘t Westenende

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Air - Water 25 Experimenta l da ta TU De lft OLGA Turner Fiedle rcorrecte d Turner

20 Critica l ga s velocity [m/s ]

Fiedle r

15

10

5

0 -10

8

0

10

Predicting onset of liquid loading

20 30 40 50 60 Inclina tion angle (90 = ho rizo ntal) [de gre e s ]

70

80

90

TNO Science and Industry, 25 September 2007

Test cases

9

ID [m]

Length [m]

Pwh [bar]

Gas Liquid

Max. Inclinations

Air Water

0.1

12

1

M=28.9

Varying (0 - 90°)

Natural gas Well 1

0.111

3743

14

M=18.6 LGR=22

40 ° Tail 90 °

Natural gas Well 2

0.074

3545

21

M=17.5 LGR=17

40 °

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Overview Liquid loading prediction (ratio prediction vs field data)

Angle

Field/experimental data

Turner Bottomhole @ angle

OLGA

Well 1

40

90000 Sm3/day

1.2

2.1

Well 2

40

45000 Sm3/day

1.0

1.6

Air -Water

0

13.3 m/s

1.1

0.3

30

17.3 m/s

1.1

1.1

60

15.8 m/s

1.1

1.3

± 20 %

± 110 %

Error 10

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Conclusions • The onset point of loading is not predicted correctly • Important parameters in the models: • entrainment • inclination angle • For practical purposes using a shape function suffices • Currently working on • adding inclination angle dependency in models • Validating use of shape function

11

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

We would like to acknowledge the support of Shell and NAM

and their permission to publish the information in this presentation

and the TU Delft for supplying the air –water data

12

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Influence reservoir behavior on onset instability Turner Ratio vs A Factor 4.00

Instability sets in at early rate

3.50 3.00

TR (-)

2.50

y = 1.9313x-0.0953 R2 = 0.1162

2.00 1.50 1.00 0.50 0.00 0

10

20

Courtesy Shell 13

Predicting onset of liquid loading

30

40

50

60

A (bar^2/1000m^3/d)

70

80

90

100

From K. Veeken TNO Science and Industry, 25 September 2007

Influence reservoir behavior on onset instability • Tubing intake curve depends on LGR • In steady state no influence of reservoir IPR • Instability can set in left of minimum TPC • If variations exist in liquid content, the IPR becomes relevant

14

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Influence reservoir behavior on onset instability 2.5

Qmin [kg/s]

2

1.5 WGR=100; D WGR=50 Turner 1

0.5

0 0

1

2

3

4

5

Ares [bar2/(1e3 Sm3/day)]

15

Predicting onset of liquid loading

TNO Science and Industry, 25 September 2007

Air-Water 25 Expe rime ntal data TU De lft Film mode l (s te ady s tate ( TU De lft)) TNO Dyn. S imulator OLGA Turne r angle de pe nde ncy

Critica l ga s ve locity [m/s ]

20

15

10

5

0

16

0

10

20

Predicting onset of liquid loading

30

40 50 60 70 Inclination angle (90 = ve rtical) [de gre e s ]

80

90

100

TNO Science and Industry, 25 September 2007