On Deepwater Perforating, Deconvolution, and Cleanup Modeling Aurelien Guichard Well Testing Network – Multi Topic Meeting MTM #8 Houston, Texas – May 1 st , 2009
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
On Deepwater Perforating,
Deconvolution, and Cleanup Modeling
Aurelien Guichard
Well Testing Network – Multi Topic Meeting MTM #8
Houston, Texas – May 1st, 2009
WTN – Multi Topic Meeting #8 – Houston, Texas
Agenda
Topic #1: Deepwater perforating
– Characteristics, challenges, and proposed solutions
– Wellbore pressure and downhole forces modeling
– Perforating results and sensitivity analysis
Topic #2: Deconvolution
– How to benefit from deconvolution
– Results from synthetic and actual datasets
Topic #3: Cleanup modeling
– Case study in the Gulf of Mexico
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #1: Deepwater Perforating
Characteristics of deepwater perforating
• Water depth up to 7,000 ft
• Well depth up to 30,000 ft
• Reservoir pressure up to 25,000 psi
• Unconsolidated rocks; low perm; sand control
• Large bore casing program up to 10 1/8 in
• Completion fluid weight up to 15 ppg
• Perforation interval up to 300 ft
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #1: Deepwater Perforating
Challenges: And solutions:
• Potential gun sanding
– None or limited flowback: 10-20 bbl
– Upon perforating, pull up the gun string before flowing
– Gun drop: controlled-tension release & automatic release systems
• Gun shock
– Stronger TCP packer
– Upgraded shock absorbers
– Optimal toolstring design
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #1: Deepwater Perforating
Wellbore pressure and downhole forces modeling
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #1: Deepwater Perforating
Successful
perforations
& frac/pack
in GoM Wellbore Pressure
shows
Dynamic Underbalance
upon Perforating
(Actual Data)
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #2: Deconvolution
Advantages of deconvolution over conventional well test analysis
• Includes entire production history
• Extends radius of investigation
• Confirms reservoir behavior and boundaries where short, individual pressure buildups cannot
• Enhances reserve estimates
To benefit from deconvolution algorithms, importance of
• Well test objective and subsequent design
• Equipment selection, specifically surface rate measurements
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #2: Deconvolution
Synthetic dataset
Production #1: 75 d
Buildup #1: 4.2 d
Production #2: 44 d
Buildup #2: 6.3 d
Total: 130 d
4200
4600
5000
Tes
t Des
ign
[psi
a]
0
200
400
Pro
du
ctio
n [S
TB
/D]
production #1 build-up #1 production #2 build-up #2
Gro
up
s
0 1000 2000 3000
Pressure [psia], Liquid Rate [STB/D], Not a unit vs Time [hr]
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #2: Deconvolution on Synthetic Dataset Conventional model
identification and
history match (red)
One fault
Log-Log plot: dp and dp' [psi] vs dt [hr]Log-Log plot: dp and dp' [psi] vs dt [hr]
Log-Log deconvolution plot: dp and dp' [psi] vs dt [hr]
Bourdet derivative
Deconvolution (blue)
Two faults
History plot (Pressure [psia], Liquid Rate [STB/D] vs Time [hr])
Conventional Analysis: Poor Match
History plot (Pressure [psia], Liquid Rate [STB/D] vs Time [hr])
Analysis with Deconvolution: Perfect Match
130 days 1 hr 100 hrs
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #2: Deconvolution on Actual Dataset
Cleanup
13 days 14000
16000
18000
BH
P3
[p
sia
]
0
2000
4000
6000
Pro
du
cti
on
[S
TB
/D]
build-up #3 production #4 build-up #4 (155.567 hr) production #5 build-up #5 (110 hr)
Gro
up
s
5/10/2006 5/16/2006 5/22/2006 5/28/2006 6/3/2006
Pressure [psia], Liquid Rate [STB/D], Not a unit vs Time [ToD]
Initial
Buildup
6.5 days
Main
Flow
10 days
Final
Buildup
4.5 days
Multiphase
Rates
at Surface
from Vx*
Downhole
Quartz
Pressure
from HPQG
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #2: Deconvolution on Actual Dataset
Log-Log plot: dp and dp' [psi] vs dt [hr]
History plot (Pressure [psia], Liquid Rate [STB/D] vs Time [hr])History plot (Pressure [psia], Liquid Rate [STB/D] vs Time [hr])
Log-Log plot: dp and dp' [psi] vs dt [hr]
A C
Log-Log plot: dp and dp' [psi] vs dt [hr]
History plot (Pressure [psia], Liquid Rate [STB/D] vs Time [hr])
B
Model B
19,241
1,660
415
2,320
Model A
19,500
1,310
513
913
Model C
19,362
1,270
372
1,070
IARF
Pi (psi)
kh (md.ft)
Xf (ft)
Fc (md.ft)
1 hr 100 hrs
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #2: Deconvolution on Actual Dataset
History plot (Pressure [psia], Liquid Rate [STB/D] vs Time [hr])History plot (Pressure [psia], Liquid Rate [STB/D] vs Time [hr])
Log-Log deconvolution plot: dp and dp' [psi] vs dt [hr]
IARF
19,362
1,270
372
1,070
Model
Pi (psi)
kh (md.ft)
Xf (ft)
Fc (md.ft)
Decon
19,246
2,060
181
553
Log-Log plot: dp and dp' [psi] vs dt [hr]
IARF
Deconvolution sees two boundaries
and improves history match!
31
days 1 hr 100 hrs
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #3: Cleanup Modeling
Cleanup Planner, a proprietary tool for the
planning phase of well cleanups:
• Built inside the Petrel RE environment
• Runs Eclipse simulations
• Accounts for drilling/completion fluids lost to
the formation
• Determines optimal flow profile and
drawdown
• Designs cleanup program, choke sequence,
expected duration of cleanup, flowback
volumes, BS&W trends, drawdown behavior,
and resulting well productivity
WTN – Multi Topic Meeting #8 – Houston, Texas
Topic #3: Cleanup Modeling
Case study in GoM: model and actual water cut during cleanup
Related Documents
