You are on page 1of 58

1 SPE Calgary, Sept 6, 2012

Common Pitfalls of Mini-frac Analysis


Robert Hawkes,
Director of Completion Technologies

Pure Energy Services


2 SPE Calgary, Sept 6, 2012
What They Didnt Tell You About
G-function and log-log plots


3 SPE Calgary, Sept 6, 2012
Bachman
G-function
Reappraisal of the G Time Concept in Mini-Frac Analysis
(SPE 160169, Bachman et al)
Thank you Google Images


4 SPE Calgary, Sept 6, 2012
Pre-Frac Diagnostics.
.are we wasting our time?
Most of our Pitfalls are coming from
conventional thinking.

Closure Pressure
Still needs a holistic approach ........and more work!

Leak-off Behaviour
Pressure diagnostics is interpretive and heavily
influenced by your discipline.

After Closure Analysis
Be careful how you impose a flow regime on the data.






5 SPE Calgary, Sept 6, 2012
Diagnostic Approach
o Step-up or Step-down (neither)
o Low Rate or Treatment Rate?
What's the objective
o Interpretation Pitfalls
ISIP
Closure Identification
Think again
Flow Regime Identification
Not as easy as it seems
Specialized Plots
o Completion Considerations
Hz wellbore (toe), Open Hole, Cemented Liner, Packer Leaks, Fluid Choice.....






6 SPE Calgary, Sept 6, 2012
Source of Maybe Some Common Pitfalls
YES !!
Maybe not
No
Maybe not
Maybe not
sure


7 SPE Calgary, Sept 6, 2012
Fracture Flow Regimes from Cinco-Ley (1978)
(Static)


8 SPE Calgary, Sept 6, 2012
-
0

h
Fluid Loss
Velocity

The idea behind Carter's 1D leakoff
coefficient

if a filter-cake wall is building up it will
allow less fluid to pass through a unit
area in unit time

the reservoir itself can take less and
less fluid if it has been exposed to
inflow

Both of these phenomena can be
roughly approximated as "square-root
time behavior"
Carter Leak-off Model
(Dynamic)


9 SPE Calgary, Sept 6, 2012
Who Has Seen This Shape of G-Function Plot??


10 SPE Calgary, Sept 6, 2012
Who Has Seen This Shape of G-Function Plot??


11 SPE Calgary, Sept 6, 2012
Who Has Seen This Shape of G-Function Plot??


12 SPE Calgary, Sept 6, 2012
Who Has Seen This Shape of G-Function Plot??


13 SPE Calgary, Sept 6, 2012
Who Has Seen This Behaviour on the Log-log Plot??


14 SPE Calgary, Sept 6, 2012
SPE 140136 (2011)
Paper identifies 3/2 slope feature and gives some theory
_______.


15 SPE Calgary, Sept 6, 2012
So What Does the Welltesting Community Have to Offer
to Mini-frac analysis?


16 SPE Calgary, Sept 6, 2012
First Mini-frac Derivative Plot, ~1990, Dr. Ted. Leshchyshyn


17 SPE Calgary, Sept 6, 2012
Well Test Analysis Log-Log Derivative Term
(Agarwal Equivalent Time)
Uses radial equivalent time t
er
for derivative
This is NOT the same as Dt




Versatile even when radial flow not present
This is why it is a universal approach to all well test problems
t t
t t
t Derivative Log
dt
P d
t
p
p
er
er
er
D
D

D
_
) (


18 SPE Calgary, Sept 6, 2012
Nolte G-Function Analysis
Using the Nolte G Function construct the various
plots

Assume no closure and the G Function Solution
runs on forever..

What do things look like?
-
0

h
Fluid
Loss
Velocity



19 SPE Calgary, Sept 6, 2012
Nolte G
Function


20 SPE Calgary, Sept 6, 2012
Classic Nolte
behavior with
no closure
P vs G
GdP/dG vs G
dP/dG vs G
Nolte G
Function


21 SPE Calgary, Sept 6, 2012
Square root-t plot
shows a character
change from early
time to late time
with a constant flow
regime.
Nolte G
Function


22 SPE Calgary, Sept 6, 2012
Nolte G
Function
Early Time Slope = 1

Late Time Slope = 0.5
Derivative is the logarithmic
derivative of t

conventional approach.
Late Time slope = 0.5 is Carter leak-off,
not the conventional linear flow.


23 SPE Calgary, Sept 6, 2012
Early Time Slope = 1

Late Time Slope = 1.5*
*as identified in paper 140136
Derivative is the logarithmic
derivative of Agarwal
equivalent (radial) time or

standard well test derivative.
NO linear flow slope, not
picked up in paper 140136.
m = 1.0
m = 1.5
Nolte G
Function


24 SPE Calgary, Sept 6, 2012
Classic Linear Flow Solution
Infinite acting linear flow
Injection period - classic linear flow
Shut-in period - classic linear flow
Use superposition to compute fall-off response



25 SPE Calgary, Sept 6, 2012
Linear
Flow


26 SPE Calgary, Sept 6, 2012
The G-function plot does
not give meaningful
results as the character
changes from early time
to late time with the same
flow regime.
No closure - infinite acting linear flow

Linear
Flow


27 SPE Calgary, Sept 6, 2012
Square root-t plot also
has a character change
from early time to late
time with the same
flow regime.
No closure - infinite acting linear flow

Linear
Flow


28 SPE Calgary, Sept 6, 2012
Intercept is not closure for
the special case 0.5 to -0.5
slope but is instead an
artifact of the plot.

No closure - infinite acting linear flow

Early Slope = 0.5
Late Slope = -0.5

Linear
Flow
Delta Time


29 SPE Calgary, Sept 6, 2012
Agarwal Equivalent Time
Early Time Slope = 0.5
Late Time Slope = 0.5

Linear
Flow


30 SPE Calgary, Sept 6, 2012
Well Test Analysis Approach
The welltest log-log pressure derivative is the best
flow regime indicator in our arsenal
Theory is very well established in traditional PTA
Cinco-ley and Samaniego, 1981
You need to dust off those old and forgotten welltesters in
your closet and exploit them in the mini-frac world.
Welltest Log-log derivative:
the time function is NOT dt of the flow period
the time function is with respect to Agarwal equivalent time
te = tp*dt/(tp+dt)
The Log-log derivatives in Barrees paper SPE 107877
are all with respect to dt as far as we can tell.



31 SPE Calgary, Sept 6, 2012
Observations and Comments:
1. We need to do flow regime identification before picking closure
pressures
Standard well test log-log derivative plot is key
Flow regime dependent plots
2. Do we need replacements for the Combination G Function plot?
3. We need to calculate the correct derivatives
4. Do we even need ACA plots ?
Why not use standard well test plots to calculate properties ?


32 SPE Calgary, Sept 6, 2012
Classic Barree Case
Jean Marie oil well example


33 SPE Calgary, Sept 6, 2012


34 SPE Calgary, Sept 6, 2012


35 SPE Calgary, Sept 6, 2012
Conventional Approach


36 SPE Calgary, Sept 6, 2012
Barree log-log dt plot
Pfoc = 8300 kPa
Conventional Approach


37 SPE Calgary, Sept 6, 2012
Pfoc = 8400 kPa
PDL ??
Conventional Approach


38 SPE Calgary, Sept 6, 2012
Pfoc = 8400 kPa
Conventional Approach


39 SPE Calgary, Sept 6, 2012
Proposed Approach


40 SPE Calgary, Sept 6, 2012
1. Early Time Linear Flow (0.5)

2. Middle Time Carter Flow (1.5)

3. Late Time Linear Flow (after closure)
Agarwal Equivalent Time
Pfoc = 8400 kPa


41 SPE Calgary, Sept 6, 2012
Carter Equivalent Time
1. Early Time Linear Flow (0.5)

2. Middle Time Carter Flow (1.0)

3. Late Time Linear Flow (0*)
*zero slope SPE 160169




42 SPE Calgary, Sept 6, 2012
Over the time range for
which Carter leak-off
has been identified, the
slope does not need to
go through the origin.

Is a Positive Y-intercept
an indictor of PDL,
..not sure.


43 SPE Calgary, Sept 6, 2012
PPD Curve
1. Late Time Linear Flow (after closure: +0.5)

2. PPD: Late Time Linear Flow (-1.5)


44 SPE Calgary, Sept 6, 2012
Agarwal Equivalent Time Function and their Slopes
Derivative Early Time Carter Late Time Carter
te 1.0 1.5
PPD 0 -0.5
Early Time Linear Late Time Linear
te 0.5 0.5
PPD 0 -1.5
Early Time Radial Late Time Radial
te 0 0
PPD -1 -2
SPE 160169


45 SPE Calgary, Sept 6, 2012
CRDM Example


46 SPE Calgary, Sept 6, 2012
CRDM Example
Establish breakdown and pump 5.0
m3 of fresh water @ 0.45 m3/min
Recorders set @ 1780m KB MD


47 SPE Calgary, Sept 6, 2012
CRDM Example
Establish breakdown and
pump 5.0 m3 of fresh
water @ 0.45 m3/min
See Log-log plot for ISIP
determination.


48 SPE Calgary, Sept 6, 2012
CRDM Example


49 SPE Calgary, Sept 6, 2012
CRDM Example


50 SPE Calgary, Sept 6, 2012
p = EOJ - Pw
CRDM Example
Conventional Delta Time


51 SPE Calgary, Sept 6, 2012
Adjusted Agarwal Time
m =1.5
Welltest Log-log
CRDM Example


52 SPE Calgary, Sept 6, 2012
Adjusted Agarwal Time
m =3/2
Delta-Time
m =1/2
m =-1/2
Overlay Log-log Diagnostic Plot
CRDM Example


53 SPE Calgary, Sept 6, 2012
CRDM Example
PPD
Adjusted Agarwal Time


54 SPE Calgary, Sept 6, 2012
Observations and Conclusions:
1. The starting point for ANY mini-frac analysis should be the standard
welltest Agarwal equivalent time log-log plot.
You must do flow regime identification before picking closure pressure.
The PPD curve has been shown to contain flow regime identification
properties a bonus diagnostic curve to the welltest community.
The industry should discontinue using the delta-time log-log derivative
plot.
2. The G-Function and square root-time plots have been shown to be
poor flow regime (and closure) identification plots?
3. PDL (pressure dependent leakoff) diagnostics using the G-Function
plot needs a holistic approach with the Agarwal equivalent time and
PPD log-log plot.
Needs more study


55 SPE Calgary, Sept 6, 2012
Observations and Conclusions:
4. Traditional welltest after-closure analysis techniques can be used for
reservoir property determination.
5. Mini-frac analysis should no longer be viewed as an independent
discipline.
6. Talk to your software vendor.
7. and finally..say Flow Regime at least once day when in the office.


56 SPE Calgary, Sept 6, 2012
Fracture Closure
T
e
s
t

M
e
t
h
o
d

I
n
t
e
r
p
r
e
t
a
t
i
o
n

M
e
t
h
o
d

Rate
Low High
Shut-in
Surface BHP Both
DFIT/MiniFrac/MDT/XLOT Flowback
G(t) Sqrt(t)
Pressure vs
Returned
Volume
Constant
Rate
Volume
Increments
Stabilized Press.
vs Returned
Volume
Other log(t)
T
i
m
e


F
u
n
c
t
i
o
n
s

Banff SPE Workshop,
April 2012


57 SPE Calgary, Sept 6, 2012
End of Reception
Banff SPE Workshop,
April 2012


58 SPE Calgary, Sept 6, 2012
End of April 24
Banff SPE Workshop,
April 2012

You might also like