Você está na página 1de 18

2/17/2018

PCB 4213: ADVANCED WELL TEST ANALYSIS


Interference & Pulse Testing

Credit hour 3
Duration 14 weeks Lecture + Lab/Tutorials

Instructor
Dr. Mohammed Mahbubur Rahman

Outline

 Introduction
 Interference Test
• Type-Curve Matching
• Semilog Analysis
 Pulse Test
• Pulse Test Analysis
• Pulse Test Design

Ref:
Pressure Transient Testing by 3J (SPE Text book series 09)
Well Testing by John Lee (SPE Text book series 01)

1
2/17/2018

Introduction

 Single Well  Multi Well


 PBU  Interference
 PDD  Pulse
 Injection
 Fall Off
 DST
 RFT/WFT etc

Introduction

The multiple-well test requires at least one active


(producing or injecting) well and at least one pressure-
observation well.

2
2/17/2018

Introduction

In multiple-well testing, the flow rate at the active


well is varied while bottom-hole pressure response is
measured at the observation wells.

Introduction • The influence region


(the radius of influence)
for pulse or interference
testing is approximately
as indicated in the
figure.
2rinf

• The major difference is


r that the testing time, t, is
much larger in
interference testing
than in pulse testing.

• Thus, rinf, and the total


2rinf +r influence region are
substantially larger in
kt interference testing
rinf  0.029 than in pulse testing.
  ct

3
2/17/2018

Interference Test: Type-Curve Matching

To analyze an interference test by type-curve matching:


Plot, on a Move the tracing paper
tracing paper, horizontally and
the observation- vertically until the data
well pressure points match the
data as exponential-integral curve
Δp vs t

Estimate the Choose a


permeability & convenient
the porosity- match point then
compressibility read the match
product point values

4
2/17/2018

Permeability is estimated from


141.2 q ( p D ) M
k
h p M
The porosity-compressibility product is estimated from

0.0002637 k tM
 ct 
r2  (t D / rD2 ) M

5
2/17/2018

Example 9.1 (JL): Interference-Test Analysis


by Type-Curve Matching
During an interference test, water was injected into
well A for 48 hours. The pressure response in well B,
119 ft away, was observed for 148 hours. Known
reservoir properties are
depth = 2000 ft βw = 1 RB/STB
q = - 170 B/D μw = 1 cp
h = 45 ft r = 119 ft
pi = 0 psig ct = 9*10-6 psi-1
t1 = 48 hours

Example 9.1: Interference-Test Analysis by


Type-Curve Matching

Observed Pressure Data:

Pressure
build up

Pressure
Fall off

Both PBU and PFO segments can be analyzed

6
2/17/2018

Example 9.1: Interference-Test Analysis by


Type-Curve Matching (analyzing the PBU part)
ΔpM = - 100
(pD)M = 0.96
tM = 10 hours
(tD/rD2)M = 0.94

Example 9.1: Interference-Test Analysis by


Type-Curve Matching

At the match point marked:


ΔpM = - 100 psig @ (pD)M = 0.96
tM = 10 hours @ (tD/rD2)M = 0.94

Results: 141.2  17011 0.96


k  5.1 md
45  100
 ct 
0.0002637 5.1 10  1.01106 psi 1
1192 1 0.94
1.0110 6
  0.11
9 10 6

7
2/17/2018

Example 9.1: Interference-Test Analysis by Type-Curve Matching (analyzing the PFO part)
We can estimate the accuracy of the above analysis by using the data from the
declining pressure part of the test (t > 48 hours).

We extrapolate the solid line in Fig 9.6 by


tracing the curve, and estimate the difference
between Δpext and the observed Δp, ΔpΔt. Table
9.2 shows the computations.

The points (plus symbols) fall on just about the same line as the circles, so we
can be confident of the analysis results.

8
2/17/2018

Pulse Test

Pulse testing is a special form of multiple-well testing.


The technique uses a series of short-rate pulses at the active
well.
Pulses generally are alternating periods of production (or
injection) and shut-in, with the same rate during each production
(injection) period.
The pressure response to the pulses is measured at the
observation well.

Although the flow


time and shut-in
time are equal in
that figure, pulse
testing can be
done with
unequal flow and
shut-in times.
However, all flow
times must be the
same and all shut-
in times must be
the same.

9
2/17/2018

Two characteristics of the pressure response at


observation well are used for pulse-test analysis:
1. The time lag
2. The pressure amplitude

Pulse Test Analysis


The ratio of pulse length to the total cycle length is
defined as t p
F'
tc

The dimensionless time lag is defined analogy to any


dimensionless time,

t L D  0.00026372 k t L
  ct rw

10
2/17/2018

Pulse Test Analysis

Pulse Test Analysis


Using the wellbore radius of the active well, the
dimensionless distance between the active and
observation wells is
r
rD 
rw
The dimensionless-pressure response amplitude is
k h p
 pD 
141.2 q  
where q is the rate at the active well while it is active

11
2/17/2018

Pulse Test Analysis

2
  t L  
141 . 2 q    p D   
  t
 c   Fig .
k 2
 t 
h p  L 
 tc 
0.0002637 k t L
 ct 
 (t ) 
 r2  L2 D 
 rD  Fig .

12
2/17/2018

13
2/17/2018

Pulse Test
Wellbore Storage Effects
Wellbore storage effects at the observation well increase the
time lag (tL) and reduce the response pressure amplitude (Δp)
of the first pulse.
We can approximate that the effect of wellbore storage at the
responding well will result in less than a 5-precent increase in
time lag and a virtually unaffected response amplitude when
the distance between pulsed and responding wells satisfies

0.54
 c 
r  32 
 ct h 

14
2/17/2018

Pulse Test
Example 6.2 (John Lee): Pulse-Test Analysis

Given:
Active well was shut in for 2 hours, then
produced for 2 hours, then shut in for 2 r = 933 ft
hours, etc. μ = 0.8 cp
Amplitude of the 4th pulse, Δp = 0.63 psi h = 26 ft
φ = 0.08
Time lag, tL = 0.4 hours q = 425 STB/D
Bo = 1.26 RB/STB
Analyze the 4th pulse to estimate k, and φct

Pulse Test
Example 6.2 (John Lee): Pulse-Test Analysis

15
2/17/2018

Pulse Test
Pulse Test Design

For best results, pulse tests should be run with a good


combination of pulse length, pulse amplitude, and
pressure-measuring equipment:

Choose value of F` Choose the


For producing maximum Get
wells, try to ΔpD[tL/ΔtC]2 points tL/ΔtC
minimize it from the figures
Calculate
tL Get (tL)D /r2
ΔtC From the
ΔtP figures
Δp

Pulse Test
Example 9.5: Pulse-Test Design

We wish to design a pulse test for reservoir with the


following approximate properties:
q = 100 B/D r = 660 ft
k = 200 md β = 1.1 RB/STB
μw = 3 cp ct = 10*10-6 psi-1
h = 25 ft  = 0.12

16
2/17/2018

Pulse Test
Example 9.5: Pulse-Test Design

1. The pulsing well is a producer, so to minimize shut-


in time we choose a short shut-in pulse, such F` = 0.3
2. For initial design calculation we choose the
maximum ΔpD[tL/ΔtC]2 point from Fig. 9.16 for the
first even pulses:
ΔpD[tL/ΔtC]2 = 0.0042 @ F` = 0.3
tL/ΔtC = 0.33
3. From Fig. 9.20:
For that value of tL/ΔtC  {(tL)D /r2} = 0.122

Pulse Test

Example 9.5: Pulse-Test Design

17
2/17/2018

Pulse Test

Example 9.5: Pulse-Test Design


2
  t L  
141.2 q pD   
  tc   Fig .
 (t L ) D  p 
 ct  r 2  2 
2
 rD  Fig . t 
tL  kh  L 
0.0002637 k  tc 

0.1810 10 6 3660 2 0.122 141.2 1001.130.0042
0.0002637 200 
 5.4 hours
20025 0.332
Cycle time is given by  0.36 psi
tC  t L / tL / tC   5.4 / .33  16.4 hours
The pulse length is
Thus 0.36 psi is the expected
t p  F ' tC  4.9 hours
response amplitude for even
pulse analysis.
We would shut in the well for
5 hours, produce for 11 hours,
and so forth

18

Você também pode gostar