Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
1
Gas Lift WorkshopGas Lift WorkshopDoha Doha –– QatarQatar
44--8 February 20078 February 2007
Optimisation of Gas Lift Optimisation of Gas Lift Well CleanWell Clean--up Operationsup Operations
by Juan Carlos Mantecon
2
WELL CLEAN-UPWell Clean-up is the period when drilling debris and fluids are still coming out of the well with produced hydrocarbon
• Minimum Rate and time to clean the well
• Size of testing equipment
• Multi-layer production results.
• Influence on WHT
• Comparison of well completion designs
3
WELL CLEAN-UP• Well Clean-up to MODU or FPSO?
• Well cleanup becomes a concern for the planned unloading scenario– Well stream is not allowed to be brought to the process if
the mud concentration is higher than• 1 vol-% for LSOBM (Low-Solid-Oil-Based-Mud), and• 0.2 vol-% for OB WARP (Oil-Based WARP Mud)
– It can be temporarily routed to the export line, but only for a short period (36 hrs) as the early water break-through can result in off-spec for the export crude
• Minimum Rate required to Clean-up the well and the corresponding time are of extreme importance
4
WELL CLEAN-UP
• Can we unload the well to a production fluid filled line?
• Would artificial lift (GL) be necessary?
• Dewatering only the riser would solve the problem?
• An ESD or a trip when the fluid density at the well or at the riser was at maximum value would create unloading problems?
5
WELL CLEAN-UP – Gas LiftProduction Well Clean-up Schedule – No Drilling Fluids
6
WELL CLEAN-UP – Gas LIftProduction Well Clean-up Schedule – No Gas Lift
7
WELL CLEAN-UP – Gas LIftProduction Well Clean-up – No Gas Lift
8
WELL CLEAN-UP – Gas LiftProduction Well Clean-up Schedule – Gas Lift
When using gas lift (300 scf/min of N2 during the first 4 hours), it took approximately 4.3 hours to clean-up the well.
9
WELL CLEAN-UP – Gas LiftProduction Well Clean-up – Gas Lift
10
7" x Tree ⊕⊕
W/Hd &
SCSSV7
CRA 30"O
20"9-5/8CRA
7-5/8CRA
13-3.8"
9-5/8" CRA
7" CRA
WELL CLEAN-UP – Well Completion Design ComparisonGas/Condensate Big Bore Well
11
7" x Tree ⊕⊕
W/Hd &
SCSSV7
CRA 30"O
20"9-5/8CRA
7-5/8CRA
13-3.8"
9-5/8" CRA
7" CRA
WELL CLEAN-UP – Well Completion Design ComparisonGas/Condensate Big Bore Well – 30 MMscfd
12
7" x Tree ⊕⊕
W/Hd &
SCSSV7
CRA 30"O
20"9-5/8CRA
7-5/8CRA
13-3.8"
9-5/8" CRA
7" CRA
WELL CLEAN-UP – Well Completion Design ComparisonGas/Condensate Big Bore Well – 150 MMscfd
13
WELL CLEAN-UP – Well Completion Design ComparisonGas/Condensate Monobore Well – 60 MMscfd
14
WELL CLEAN-UPDual Well Completion
15
WELL CLEAN-UP – Annular Gas LiftProduction Fluid Filled Flowline Well Completion
16
• Offshore rig rates are at an all time high
• Wells are becoming bigger and longer
• Deeper waters and more complex reservoirs.
• In this environment, well clean-up and flow assurance issues such as slugging and hydrates can significantly extend the planned duration of well-tests.
Well Testing – Transient ModellingWell Testing has become more challenging
17
• The ability to predict and being prepared to deal with such problems can :
• optimise well test equipment design• reduce operational risk
• Potential for saving millions of dollars in :• Rig-time• Wellbore completion equipment damage• Reservoir formation damage
• Minimising safety hazards • Minimising environmental impact.
Well Testing – Transient ModellingWell Testing has become more challenging
18
• Traditional well flow software only model steady-state flow
• New specialised software to model the transient behaviour of wells - unloading of completion fluids from rate=0 until steady state flow conditions are reached.
• Key outputs at any time and at any point in the well include:• slug sizes and frequency• fluid composition and hold-up• pressure-temperature trends
• This information enables optimum equipment design so all parameters are within the allowable operating envelope at any time of the well test operation.
0
100
200
300
400
500
600
700
800
900
1000
0 100 200 300 400 500 600 700 800 900 1000
STANDARD LIQUID RATE [Sm³]
GA
S O
IL R
AT
IO [S
m³/S
m³]
Stable Operating Envelope
Standard Liquid Rate [ Sm³/d]
Gas
Oil
Rat
io [S
m³/
Sm³]
Hydrate Formation Temp. – 18°C
Wax Appearance Temp. – 32°C
Riser Stability – ΔP = 1 bar
Riser Stability – ΔP = 6 bar
Reservoir Pressure – 80 bara Riser Stability – ΔP = 12 bar
Gas Velocity Limit – 12 m/s
Erosional Velocity Limits
0
100
200
300
400
500
600
700
800
900
1000
0 100 200 300 400 500 600 700 800 900 1000
STANDARD LIQUID RATE [Sm³]
GA
S O
IL R
AT
IO [S
m³/S
m³]
Stable Operating EnvelopeStable Operating Envelope
Standard Liquid Rate [ Sm³/d]
Gas
Oil
Rat
io [S
m³/
Sm³]
Hydrate Formation Temp. – 18°CHydrate Formation Temp. – 18°C
Wax Appearance Temp. – 32°CWax Appearance Temp. – 32°C
Riser Stability – ΔP = 1 barRiser Stability – ΔP = 1 bar
Riser Stability – ΔP = 6 barRiser Stability – ΔP = 6 bar
Reservoir Pressure – 80 baraReservoir Pressure – 80 bara Riser Stability – ΔP = 12 barRiser Stability – ΔP = 12 bar
Gas Velocity Limit – 12 m/sGas Velocity Limit – 12 m/s
Erosional Velocity LimitsErosional Velocity Limits
Well Testing – Transient ModellingApplication of New Tools
19
• Different initial conditions (facilities backpressure, underbalanced pressure and drilling fluids types) can be modelled.
• Beanup rates and well-test sequencing can be changed to optimise the operation from well clean-up to steady state conditions, including any emergency shut-down (ESD) and “what if” analysis.
• Ultimately this tool can be the basis of a simulatorused by well-test personnel to “virtually” run through a complete well-test operation.
Well Testing – Transient ModellingApplication of New Tools
20
• The clean-up simulations were started from the initial underbalanced conditions with different choke sizes.
• The well models were allowed to run until the brine, diesel and lost circulation mud (LCM) had been displaced from the well and steady state conditions were obtained.
• LCM was included in the model to assess the effect of backproducing mud lost into the formation during drilling. This provided the worst case conditions for the clean-up of the well and the restart of the well after an ESD.
Well Testing – Transient ModellingOtway Case Study - SPE 101872
21
Initial Conditions
• A diesel cushion length of 1635 m MD was required to obtain a 150 psi underbalanced condition, measured at the bottom reservoir inflow point
• A constant backpressure of 100 psia was applied downstream of the choke in the model – relevant when F < CF
• One inflow point was considered – Single-ForchheimerPressure boundary
Well Testing – Transient ModellingOtway Case Study
22
Boundary Conditions
• The injectivity was assumed to be zero for each zone (this is the worst clean up case scenario), since allowing the injectivity to equal the productivity (default) would permit fluid back into the reservoir during the shut-in simulations.
• No analysis of cross flow between productive units was performed (injectivity=0 eliminates this possibility).
• The average seawater temperature at seabed was assumed to be 13oC with a sea current velocity 0.5m/s
• The average ambient air temperature was assumed to be 28oC, and the average wind velocity 7m/s
Well Testing – Transient ModellingOtway Case Study
23
Choke Size Sensitivity Runs
• Clean-up simulations were started from initial conditions using choke sizes of 0.5”, 0.75”, 1”, 1.5”, 2” and 3”.
• 200 bbl of LCM was included in the model as a mass source from the reservoir to simulate the production and clean up of this extra amount of fluid .
• The inflow of the LCM was constrained by the bottom-hole pressure, so that the LCM flowrate modelled was the maximum that could occur under realistic conditions
Well Testing – Transient ModellingMinimum Clean-up Rate
24
• The clean-up time was defined as the time taken for the original liquid content in the well to be completely unloaded, excluding fluids below the Unit 5 inflow point where no gas was available to lift the liquid (rat hole).
Choke Size
Gas Flowrate Oil Flowrate
Clean-up Time
Water Hold-up Oil Hold-up
[in] [MMscfd] [STB/d) [h:mm] [bbl] [bbl] ½ 13 77 - 93.1 1.5 ¾ 30 59 - 75.7 0.1 1 55 106 15:40 22.5 1.0
1½ 114 196 1:08 22.5 3.7 2 178 348 0:37 22.5 0.3 3 264 519 0:24 22.5 0.3
Well Testing – Transient ModellingMinimum Clean-up Rate – Simulation Results
25
0
10000
20000
30000
40000
50000
60000
70000
0 0.5 1 1.5 2 2.5 3
Time [h]
Liqu
id F
low
[stb
/d]
0
20
40
60
80
100
120
140
160
180
200
Gas
Flo
w [M
Msc
fd]
Oil Flow [stb/d] Water Flow [stb/d] Gas Flow [MMscfd]
0
5000
10000
15000
20000
25000
30000
35000
0 2 4 6 8 10 12 14 16 18 20
Time [h]
Liqu
id F
low
[stb
/d]
0
10
20
30
40
50
60
70
80
90
100G
as F
low
[MM
scfd
]
Oil Flow [stb/d] Water Flow [stb/d] Gas Flow [MMscfd]
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
0 0.5 1 1.5 2 2.5 3
Time [h]
Liqu
id F
low
[stb
/d]
0
10
20
30
40
50
60
70
80
90
100
Gas
Flo
w [M
Msc
fd]
Oil Flow [stb/d] Water Flow [stb/d] Gas Flow [MMscfd]
0
20000
40000
60000
80000
100000
120000
140000
0 0.5 1 1.5 2 2.5 3
Time [h]
Liqu
id F
low
[stb
/d]
0
50
100
150
200
250
300
Gas
Flo
w [M
Msc
fd]
Oil Flow [stb/d] Water Flow [stb/d] Gas Flow [MMscfd]
TM-1 Clean-up - Flowrates 3" Choke TM-1 Clean-up - Flowrates 1.5" Choke
TM-1 Clean-up - Flowrates 1" Choke TM-1 Clean-up - Flowrates ¾ “ Choke
Well Testing – Transient ModellingMinimum Clean-up Rate – Simulation Results
26
TM-1 Clean-up – Total Oil content in well for each choke size TM-1 Clean-up – Temperature at wellhead for all choke sizes
0
50
100
150
200
250
300
350
0 2 4 6 8 10 12 14 16
Time [h]
Tota
l Oil
Con
tent
in T
M-1
[bbl
]
3" 2" 1.5" 1"
0
20
40
60
80
100
120
140
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Time [h]
Tem
pera
ture
[deg
C]
.
0.5" 0.75" 1" 1.5" 2" 3"
Well Testing – Transient ModellingMinimum Clean-up Rate – Simulation Results
27
TM-1 Clean-up - Initial Well Conditions ass choke sizes TM-1 Clean-up - 3" Choke after 120 sec
TM-1 Clean-up - 3" Choke after 220 sec TM-1 Clean-up - 3 “ Choke at SS
Well Testing – Transient ModellingMinimum Clean-up Rate – Simulation Results
28
Well Testing – Transient ModellingMinimum Clean-up Rate – Simulation Results
29
The purpose of this activity was to determine the minimum rate and time required to clean-up the TM-1 well, if an ESD were to occur during the clean-up operation.
Choke Size
Gas Flowrate Oil Flowrate
Clean-up Time
Water Hold-up Oil Hold-up
[in] [MMscfd] [STB/d] [h:mm] [bbl] [bbl] ½ 13 77 - 74.3 13.4 ¾” 30 58 - 74.6 0.1 1 55 106 Table 6 43 0.1
1 ½ 114 196 1:06 22.5 3.6 2 178 348 0:38 22.5 0.3 3 264 519 0:23 22.5 0.3
Choke Size
Clean-up Time
Water Hold-up with Rathole
Water Hold-up above Perf
[in] [h:mm] [bbl] [bbl] 1 1 59 34.5 1 5 47 24.5 1 10 46 23.5 1 24 43 20.5 1 48 41.5 19
Table 6
Well Testing – Transient ModellingMinimum Clean-up Rate – Simulation Results - ESD
30
Well Testing – Transient ModellingMinimum Clean-up Rate – Simulation Results - ESD
31
• The following conclusions are specific to the dynamic simulation on the TM-1 well:
– Choke sizes 1.5” or larger allows rapid clean-up of the brine and LCM from TM-1. The clean-up time for a 1.5” choke was 1:08 hours.
– Flowrates using the 0.5” and 0.75” choke sizes would not provide sufficient gas velocity within the well tubing to completely remove the brine, even though the majority of wellbore fluids were unloaded at moderate or low rates.
• The modelling shows that proper clean-up necessary for accurate well testing could not be achieved with these smaller choke sizes.
Conclusions
32
• The following conclusions are specific to the dynamic simulation on the TM-1 well (cont.):
– Although a 1” choke can cleanup the well, the clean-up time was 15:40 hours. This indicates that the standard welltestpackage with a gas handling limit of ~50 MMscfd may be sufficient to handle the cleanup of TM-1. However it would take almost 14 hours longer to clean the well than if it had been cleaned-up using a 1 ½” choke at 114 MMscfd steady-state gas flowrate, which would require a high rate well-test package.
Conclusions
33
• The following conclusions are specific to the dynamic simulation on the TM-1 well (cont.):
• If an ESD occurred during the course of the cleanup, simulations indicate that:– with choke sizes greater than 1 ½”, all liquids would be
cleaned up from the well. – the 0.5” and 0.75” choke sizes would not provide sufficient
gas rate to lift all the fluids from the wellbore. – it will take a long time to lift all the fluids out of the wellbore
with a 1” choke. Although the bulk of the liquid would be lifted out of the well initially, only 15.5 barrels would be lifted out of the well after 48 hours of cleanup, with 19 barrels remaining inthe well (excluding the volume in the rathole)
Conclusions
34
• General
– The knowledge of the minimum flowrates required to clean up the wells enables appropriate design and sizing of well test equipment.
– In this case, dynamic modelling indicated that a standard welltest package may be adequate for cleaning up this big bore gas well with 9 5/8” production tubing, though the equipment would be operated at or near its limits.
– The simulation indicated that a high rate well-test package costing significantly more may not be required, although it would have allowed a significantly faster well cleanup.
Conclusions
35
• General
– Dynamic modelling allows the investigation of a well’s behaviour during unloading, including the effect of an ESD.
Conclusions
– Analysis of the simulation results pprovides a basis for optimal equipment selection and a better preparation of well completion and testing jobs reducing risk, safety hazards and environmental impact.
– Dynamic well modelling is a useful tool to optimize well-test design.
36
Better Reservoir Analysis from Well Test Data Optimisation
– Better Radius of investigation and reserves calculation -deconvolution technique.
Virtual Downhole Gauge
– Model the well test and validate model during well clean-up– Predict BHP during test from THP
– High resolution data from separator and the time resolution by history matching the rate to the downhole pressures using Dynamic Model.
37
be dynamic
Thank You! Any Questions?