You are on page 1of 14

SPE 125028

Improving the Process of Understanding Multiprobe Production Logging


Tools From the Field to Final Answer
G. Frisch, D. Dorffer, and M. Jung, Halliburton Energy Services; A. Zett and M. Webster, BP Exploration and
Production

Copyright 2009, Society of Petroleum Engineers

This paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
The newest generation of production logging tools consists of multiple sensors in multiple locations around the wellbore that
incorporate 12 resistivity and capacitance probes and six spinners. The capacitance array tool (CAT™) determines the water,
oil, and gas holdup in the wellbore. The resistivity array tool (RAT™) determines the holdup of hydrocarbons and water.
Likewise, the spinner array tool (SAT™) consists of six bowspring mounted micro-spinners that enable the measurement of
the velocity profile. These new tools provide a detailed examination of the flowing fluids in all types of wells, including
highly deviated and horizontal wellbores, that is not available with the traditional center sample tools because of the wellbore
conditions, especially with fluid segregation. With these 30 measurements, a system of quality control and processing was
developed to enable both experienced and non-experienced engineers to determine whether or not the data was correct and
valid.
A quick analysis tool was developed to enable the field engineer and company representative to enter raw values from the
two holdup devices and calibration values, and to determine the holdups from the two sensors. Similarly, entering the raw
spinner counts, cable speed, and estimated spinner slopes into the quick analysis tool will provide an estimate of the velocity
profile for the SAT spinners and the other spinners that are run. This quick analysis tool graphically shows the holdups and
velocity in an easy-to-understand presentation for people who are not production logging (PL) experts.
After the raw data in the field is validated, a complete analysis is provided. This analysis includes horizontal, vertical, and
3D images of holdup and velocity profiles; continuous displays of flow profiles; and a complete flow analysis consisting of
the split of oil, gas, and water rates at both downhole and surface conditions. This PL data can be presented in standard log
formats, spreadsheets, and other methods as needed. This process can be modified by either the service company or customer.
Several examples are provided that show the capabilities of the new logging tools and the interpretation method used to
determine the results.

Introduction
Phase segregation occurs in many wells, including those with little deviation from vertical; the lighter phases migrate to the
high side of the wellbore, and the heavier phases migrate to the low side. In highly deviated and horizontal wellbores,
traditional PL sensors, which are center sample tools or have single point measurements, may not provide the most accurate
data as a result of the wellbore and well flowing conditions. These PL tools measure fluid properties, such as velocity,
density, capacitance, temperature, and pressure. Tool position, or more accurately sensor position, may lead to incorrect
interpretations regarding the flow environment of the well.
New PL tools have been developed to help address the issues in deviated or horizontal wells. These new tools include two
types of holdup measurements, capacitance and resistivity, as well as multiple velocity measurements. These new tools will
be referred to as Production Array Logs (PAL) to distinguish them from the standard PL logs. These tools provide a relative
bearing measurement that enables the location of each sensor to be determined. The velocity tool also includes an inclination
measurement to aid in the analysis of the PAL data. The holdup tools have 12 measurement probes, and the velocity tool has
six spinners. These tools, when run in conjunction with the standard tool string, provide multiple measurements around the
entire wellbore. The interpretation of each tool individually is complex and, when combined with the other PAL
measurements, the complexity increases dramatically. A new interpretation process was developed that combines the benefits
of the newer sensors and addresses problems caused by the deviated and horizontal wellbores in the standard PL
interpretation procedures.
2 SPE 125028

The completion of the example well consisted of a 7 5/8-in. casing with three zones of perforations. The deviation of the
wellbore over the perforated zones ranges from 25 to 35 degrees. The logging tools used for this well consisted of the
standard PL string which included a radioactive fluid density, (DENR), fluid capacitance (CWH), pressure (QP), temperature
(TEMP), inline spinner (ILS), and fullbore spinner (CFB). The additional fullbore tools that were run included the three array
tools: CAT, RAT, and SAT.

Holdup Tools
One of the challenges of production logging is the identification of the fluid types and the point of entry of those fluids into
the wellbore. Standard holdup tools normally measure the density and capacitance of the wellbore fluids and are usually
centered sample measurements. The density measurement is based on either a tool with a radioactive source or a gradio
(pressure) arraignment. Both of these tools have advantages and disadvantages that should be considered when selecting the
proper tool. For example, a radioactive source may be considered as an environmental hazard, while a tool based on pressure
is inadequate in horizontal wellbores. The challenge of obtaining accurate holdup measurements is addressed by two tools,
the Capacitance Array Tool (CAT) and the new Resistivity Array Tool (RAT). Both tools have 12 sensors that are located
radially on flexible bowsprings that measure the capacitance and resistivity of the wellbore fluids.

Capacitance Array Tool. The CAT is an upgraded version of the tool originally described by Ryan and Hayes (2001) and
Frisch et al. (2002, 2009). The major difference between the original version and the upgraded version is that the sensors are
now placed on bowsprings, whereas the original tool included sensors located on motorized arms. This modification
improves the reliability of the mechanical portion of the tool and enables logging in both directions, rather than only up logs,
as with the previous tool. The first step in quality control and interpretation is to calibrate the sensor readings so that each
sensor will read the same counts for each type of fluid. This calibration enables a standard interpretation color scheme for
each CAT pass and a quick interpretation of the holdups.
250 RCAP06 S3,U1 [cps] 0 250 CN6N S4,U1 [cps] 0
TEMP ERATURE DENS ITY 250 RCAP07 S3,U1 [cps] 0 250 CN7N S4,U1 [cps] 0
171 177 .1 1.1 250 RC
250 AP08
RAWS3,U1 [cps]
RCAP09 S3,U1 [cps]
0
0
RAW CATMAP 250
250
CN8N S4,U1 [cps] D
CALIBRATE
CN9N S4,U1 [cps]
0
0
CATM AP
P RE SSURE HYDRO 250 CAT CO
RCAP10 UNTS
S3,U1 [cps] 0 HI------LO W--- ---HI 250 CAT
CN10NCOUNTS
S4,U1 [cps] 0 HI-- ----LOW ------HI
2750 2800 0 2 250
250 RCAP11 S3,U1 [cps] 00 50 200 250250 CN11N S4,U1 [cps] 0 50
0 200
ft 250 RCAP12 S3,U1 [cps] 0 50 cps 200 250 CN12N S4,U1 [ cps] 0 50 cps 200

950000
XX

XX5 0

C
Figure. 1–Logs showing the CAT data before and after the calibration routine.

Figure. 1 shows a section form the first log example that illustrates the CAT data before and after the calibration process.
Rather than presenting 12 holdup curves, images are generated from the raw and/or processed data. These images are created
from the input data from each tool, using the relative bearing to determine the position of each sensor. After the relative
position of each sensor is known, the data is interpolated from one reading to the next. The image is presented, from left to
right, as high–low–high. This arrangement will always display the low side of the wellbore in the center of the track. From
the legacy center sampling holdup devices, the fluid density, capacitance, and location of the oil/water (XX86) and gas/oil
SPE 125028 3

(XX60) contacts in this well are evident. The color coding for all the CAT images are based on the normalized data in which
water (blue) has a value of 50, oil (green) has a value of 150, and gas (red) has a value of 200. If the well consists only of two
phases, the colors are modified to show only the phases expected; however, it is always recommended to show water in the
images, even if the production stream is water-free.
The raw and calibrated data are used to create the two images shown in Figure. 1. This routine is performed after
acquiring the data; however, a spreadsheet has been developed to enable the logging engineer to determine the holdups on
location. The spreadsheet will be discussed in a following section. When using this technique, the low side in the middle of
the track should always display the heavier fluid, as shown at A. At this point, the low side of the track shows a pure streak of
oil; immediately above this streak, a mixture of oil and gas trending to almost a pure gas phase displays on the upper section
of the wellbore (left and right sides of the track).
Zone C appears to consist of 100% water, as shown by the hydro and CAT tool, with a density reading of 1.01 gm/cc. The
density tool reads 0.73 gm/cc in zone B, indicating a mixture of oil with a small portion of gas. However, the CAT tool is
detecting a mixture of oil and water on the low side of the wellbore, trending to almost a pure oil, then to a mixture of gas and
oil on the high side.

Resistivity Array Tool. The RAT tool consists of an array of 12 microsensors that detects very small, fast moving bubbles of
conductive water and the non-conductive hydrocarbons. Real-time software provides a mean value and a standard deviation
for the resistance values of the 12 sensors over for the period being summarized.

TEMPERATURE DENSITY
171 177 .1 1.1
PRESSURE HYDRO NEW PUBLISHED YWRATA NEW
2750 2800 0 2 RATMAP RATMAP 1 0
GAMMA AVRATMN RAT MN RAT SD HI------LOW------HI HI------LOW------HI YWRATA
0 100 1 0 .2 1.2 .0 .1 0 1 0 1 1 0

XX00

XX50

C
Figure. 2–Raw and processed RAT data over the same zone as Figure. 1.

Figure. 2 shows the RAT data in the same well over the same zone as the CAT log in Figure. 1. The RAT map uses blue
to indicate water and green to indicate hydrocarbons, with the interpolation between the blue and green as a mixture,
depending upon the percentage of water. If gas/water is present, green is replaced with red to make the visualization of the
images easier. The RAT MN data is the average mean value from the 12 sensors, and the RAT SD is the standard deviation
of the same data. There are two methods that can be used to determine the holdups from the RAT tool. The published method
which is described later uses both the mean and standard deviation data to determine the water holdup. The new method uses
the average mean measurements, a water value, and hydrocarbon value. Both methods show almost the same result when
4 SPE 125028

comparing the Ywrat curves. The published method shows more water than the new method between XXOO and XX85.
Both maps show a slight water bulge in the middle of the track at the hydrocarbon/water contact immediately above C.
Because the RAT tool does not distinguish between oil and gas, the changes in phases at B are not shown on the RAT data.
However, there is a slight water entry on the high side of the wellbore A.

The mean value and standard deviation can be calculated as follows:


∑(R ) ∑ (R − m )
2
m= S= Equations 1 & 2
n n
where: m = Mean value
R = Measured values
N = Number of samples taken
S = Standard deviation
m = (Yw × Rc ) + (1 − Yw) × Ri Equation 3
where: Yw = Water holdup
Rc = Apparent resistance of the conducting fluids (water)
Ri = Apparent resistance of the insulating fluids (hydrocarbons)

The manipulation of these equations leads to a formula for holdup that is based on the measured mean, standard deviation,
and the resistance Ri:

Yw =
(Ri − m )2 Equation 4
( Ri − m) 2 + S 2
This equation was originally used in the post processing of the RAT data, but with additional data sets some
inconsistencies of the equation were discovered. Therefore a new processing technique was developed that uses the mean
value of both the hydrocarbons and water. Both techniques are available in both the post processing software and the
spreadsheets used for quality control.

Fluid Velocity
The other major component of production logging is the determination of the rate of the fluid movement. Historically, the
fluid velocity used in most PL analysis is determined by center sample, fullbore, and in-line or basket flowmeters, with
possible combinations of tools for different flowing environments. Each spinner has its own advantages and disadvantages;
however, each spinner will only enable the determination of one velocity centrally within the wellbore. This velocity
measurement is then used with correlations, lab experiments, or flow matching to determine the velocity of each phase. This
is probably the most challenging aspect of production logging because of the possibility of phase velocities flowing at
different speeds and even different directions, depending upon wellbore conditions.

Spinner Array Tool. The spinner array tool consists of six miniature turbines deployed on bowspring arms, which provide
the fluid velocities and direction around the wellbore. The turbines use low friction jeweled bearings to reduce the
mechanical threshold of the spinner and to improve the sensitivity to the fluid flow. The raw data includes the direction and
speed of spinner rotation, which is used to determine the velocity at each spinner location. With the relative bearing
information, the location of each spinner is known, and therefore the azimuthal velocity profile can be determined.
The SAT spinners are calibrated in the same way as a standard spinner by crossplotting the spinner RPS vs. the cable
speed, using the different logging passes. The slopes and thresholds are determined for each spinner by using both the down
and up passes, as shown in Figure. 3. Each of the six spinners is calibrated the same way, then the user determines the best
slope and threshold to use. Our experience indicates that the thresholds and slopes from each of the six spinners have
approximately the same response for most fluids. The processing software enables a combination of slopes and thresholds to
be used, depending upon the data quality,to independently determine the velocity from each of the six spinners.
SPE 125028 5

6.

Line Speed
0
-80 0 80

-2

-4

Spinner RPS
-6 .

Figure. 3–Spinner calibration plot for spinner 1 using six different passes to determine threshold and slopes.

-15 SPIN1 15 CFB VA -100 VA1 150


-15 SPIN2 15 -100 150 -100 VA2 150
171 TEMP 177 -15 SPIN3 15 ILS VA -100 VA3 150
CFB SPIN
2750 PRES 2880 -50 50 -15 SPIN4 15 -100 150 -100 VA4 150
0 GAMMA 100 ILS SPIN -15 SPIN5 15 AVG.SAT VA -100 VA5 150 SATMAP
-100 LINE SPD 100 -50 50 -15 SPIN6 15 -100 150 -100 VA6 150 -104 152

XX00

XX50

Figure. 4–Log showing both raw data and calculated velocities from inline, fullbore, and SAT spinners.

The SAT, inline, and fullbore spinner data is shown in Figure. 4. The raw data from each of the six SAT spinners are
shown in track 3, with the corresponding calculated velocities shown in tracks 4 and 5. For comparison, the apparent velocity
from fullbore and inline spinners are ploted against the average of the six SAT velocities is in track 4. These three velocities
6 SPE 125028

should track or follow the same general trend, but may not perfectly overlay or match because of the tool position and flow
profiles.
The last track is the image created from the six SAT calculated velocities and the SAT tool relative bearing. The images
from the SAT are divided into two sets of colors, indicating both direction and velocity. White to black indicates downflow,
with black being the highest value and white being 0; likewise, yellow to magenta indicates positive flow up the wellbore,
ranging from 0 to maximum velocity. Between A and B, the pattern indicates fluid flow down the wellbore or possible fluid
fallback. The fluid flowing downhole should be the heavier fluids as a result of gravity segregation and, when compared with
the CAT and RAT logs, it should be possible to determine whether or not this is correct. Therefore, the lightest fluids should
be flowing on the high side of the wellbore with the highest velocity, which is the case for this well.

Raw Data Quality Control


Using the 12 sensors on the CAT and RAT and the six spinners on the SAT, several methods were designed and implemented
for the quality control (QC) of these measurements. The initial QC rests with the field engineer’s decision of whether or not
the sensor readings are accurate and repeatable for a steady and consistent flow conditions. The standard log presentations
provide the basic readings for the 24+ holdup sensors and the 6+ velocity sensors. However, when a reading behaves
differently from the others, is it a result of tool malfunction or possible changes in the flow patterns or wellbore conditions?
Is the data consistent enough to provide a quality PL interpretation, or are more PL runs needed? Is it a tool problem
requiring complete sensor replacement? Or do the multiple sensors provide enough coverage to provide the necessary or
required data to solve the initial problem or needs of the customer? These are some of the questions that must be answered
when evaluating data quality at the well site with these new tools. To help provide a consistent QC platform, a quicklook
spreadsheet, with standard interpretation and 3D images provides a method to help the field personnel decide whether or not
the data is adequate for analysis. An additional benefit from the QC program is that it helps educate both experienced and
non-experienced users of PL data about the complicated fluid profiles found in vertical, deviated, and horizontal wellbores.
There are two ways in which to calculate multi-sensor holdups. The easiest method is to determine the holdup at each
sensor, then divide the number of sensors in the phase by the total number of sensors. The holdups that use the averaging
method in which each sensor has equal weight will be Ywe, Yoe, and Yge for water, oil and gas respectivly. The wellbore
cross-sectional area associated with each sensor must, however, also be considered to make a correct determination of
holdup. The contribution of each sensor to the total holdup measurement will vary as a function of the sensor’s location
relative to the top of the wellbore. The ratio of the total cross-sectional area for the sensors indicating the phase to the pipe
cross sectional area provides another method for calculating for phase holdup. The cross-sectional area indicating water can
readily be computed from elemental geometry. The holdups calculated from this method will be Ywa, Yoa, and Yga for
water, oil and gas respectfully. The post-processing software uses both methods to determine holdups for both the CAT and
RAT logs.

Quicklook Spreadsheets
A spreadsheet was developed (Figure. 5) that uses all of the post-processing algorithms to enable the field engineer or user to
enter the basic logging parameters and logging data into the yellow cells and obtain an estimate of the holdups and velocity
from both the PAL tools and standard PL string. This easy-to-use spreadsheet can be exchanged between the field operations
and interested parties. Various options enable the engineer to select the type of phases in the wellbore and to enter some of
the parameters. The data itself can be manually entered or cut and pasted from an averaging program that was also developed
for the PAL tools.
SPE 125028 7

DENSITY HYDRO
WATER 1.0050 2 ZONE DENR CWH
OIL 0.7600 1 1 1.0049 1.9829
GAS 0.1330 0 2 0.7271 0.9751
CAT MASTER WATER 50 3 0.1645 0.2807
CAT MASTER OIL 150
CAT MASTER GAS 200
PHASES 3 0=water/gas,1=oil/gas,2=water/oil,3=water/oil/gas
HALRAD 0 1=new, 0=published
USERAT 0 0=USE RATHY, 1= RATHY01-12
RATHY 0.9919
RATWAT 0.5059

CAT CALIBRATIONS RCAP01 RCAP02 RCAP03 RCAP04 RCAP05 RCAP06 RCAP07 RCAP08 RCAP09 RCAP10 RCAP11 RCAP12
WATER 55 53 54 50 50 55 53 60 55 57 51 56
OIL 157 139 132 153 131 127 150 145 150 141 141 149
GAS 205 185 181 205 171 171 191 185 191 184 178 185

READINGS FROM LOGS


ZONEDEPTH / DESCRIPTION RCAP01 RCAP02 RCAP03 RCAP04 RCAP05 RCAP06 RCAP07 RCAP08 RCAP09 RCAP10 RCAP11 RCAP12
1 XX88-XX93 55 53 57 49 51 55 54 61 55 60 57 56
2 XX70-XX80 166 150 141 161 119 110 149 144 150 143 152 160
3 XX20-XX30 190 172 158 189 131 163 172 163 171 161 170 182

ZONEDEPTH / DESCRIPTION RATMN01 RATMN02 RATMN03 RATMN04 RATMN05 RATMN06 RATMN07 RATMN08 RATMN09 RATMN10 RATMN11 RATMN12
1 XX88-XX93 0.4931 0.4973 0.5207 0.5074 0.5004 0.4982 0.5104 0.5169 0.4995 0.5053 0.5067 0.5148
2 XX70-XX80 0.9922 0.9923 0.9927 0.9920 0.9923 0.9921 0.9927 0.9926 0.9920 0.9926 0.9923 0.9923
3 XX20-XX30 0.9910 0.9904 0.9910 0.9906 0.9908 0.9917 0.9918 0.9912 0.9913 0.9919 0.9920 0.9910

ZONEDEPTH / DESCRIPTION RATSD01 RATSD02 RATSD03 RATSD04 RATSD05 RATSD06 RATSD07 RATSD08 RATSD09 RATSD10 RATSD11 RATSD12
1 XX88-XX93 0.0078 0.0090 0.0075 0.0081 0.0092 0.0088 0.0090 0.0091 0.0094 0.0090 0.0107 0.0110
2 XX70-XX80 0.0185 0.0186 0.0199 0.0192 0.0195 0.0203 0.0195 0.0197 0.0198 0.0196 0.0189 0.0194
3 XX20-XX30 0.0222 0.0214 0.0206 0.0217 0.0215 0.0200 0.0202 0.0211 0.0206 0.0202 0.0196 0.0225
Figure. 5–The modified spreadsheet enabling a quick look interpretation of the holdup data.

1. 00 1. 00

0. 90 0. 90

0. 80 0. 80

0. 70 0. 70

0. 60 0. 60

0. 50 0. 50

0. 40 0. 40

0. 30 0. 30

0. 20 0. 20

0. 10 0. 10

0. 00 0. 00
XX 88-XX 93 XX 70-X X80 X X20-XX 30 XX88-XX93 X X70-X X80 XX 20-XX 30

CAT holdups RAT holdups


Figure. 6–A quick visualization of the PAL holdups.

The holdups calculated are the averaging method described above because of the complexities involved using the area
calculations. Figure. 6 shows a quick visualization of the holdups calculated from the CAT and RAT sensors. The averaging
method for holdup calculations has some errors that were previously described. Phase holdups from all of the available
sensors are shown in tablular form in Figure. 7.

DENSITY ANALYSIS HYDRO ANALYSIS CAT ANALYSIS RAT ANALYSIS


ZONE DEPTH / DESCRIPTION Yw Yo Yg Yw Yo Yg Ywe Yoe Yge YWRE YHRE
1 XX88-XX93 0.9996 0.0004 0.0000 0.9829 0.0171 0.0000 0.9854 0.0146 0.0000 0.9996 0.0004
2 XX70-XX80 0.0000 0.9475 0.0525 0.0000 0.9751 0.0249 0.0339 0.8481 0.1181 0.0007 0.9993
3 XX20-XX30 0.0000 0.0502 0.9498 0.0000 0.2807 0.7193 0.0000 0.4061 0.5939 0.0015 0.9985
Figure. 7–A numerical presentation of both PAL holdups and conventional PL tools.

A similar spreadsheet is also used to determine the velocities from the SAT spinners along with the inline and fullbore
data. The user can enter data in the yellow cells and the spreadsheet will provide the calculated values in the green cells. The
slopes and thresholds for the tools can be calculated with multiple passes, as previously explained, or use standard defaults.
In this case, there was not a no flow area or shut in passes run, so the analyst used a common value for all sensors.
8 SPE 125028

POSITIVE POSITIVE NEGATIVE NEGATIVE


T HRESH SLOPE THRESH SL OPE
FU LLBORE 5.00 0.567 -5.00 0.522
INLIN E 5.00 0.168 -5.00 0.153
SAT 5.00 0.082 -5.00 0.076

ZONE D EPTH / DESCR IPTION CS ILS C FB SPIN1 SPIN2 SPIN3 SPIN4 SPIN5 SPIN6
1 XX88-XX93 59.83 -8.499 -34.414 -3.983 -2.941 -3.228 -4.419 -0.403 -3.449
2 XX70-XX80 60.00 -10.561 -39.593 -3.960 -0.752 -1.864 -3.449 -5.018 -4.961
3 XX20-XX30 60.91 -0.539 -14.578 -2.581 0.423 -0.854 -4.645 -8.494 -6.270

SAT 1 SAT 2 SAT 3 SAT 4 SAT 5 SAT 6 FULLBORE INLIN E


ZONE D EPTH / DESCR IPTION
1 XX88-XX93 2.4171 16.1342 12.3526 -3.3158 49.5263 9.4421 -11.1012 -0.7266
2 XX70-XX80 2.9028 45.1054 30.4818 9.6199 -11.0196 -10.2801 -20.8475 -14.0227
3 XX20-XX30 21.9586 71.0674 44.6757 -5.2059 -55.8493 -26.5809 27.9863 52.3929

Figure. 8–The modified spreadsheet enabling a quicklook interpretation of the spinner data.

The results of the spreadsheet shown in Figure. 8 for the 6 SAT spinners show both positive and negative flow at the
same depths, which appears to be almost random in the determination of the velocity profile. However, when a simple graph
is generated (Figure. 9), the velocity profile becomes apparent and assists in the evaluation of the data.
15 0 .0

10 0 .0

5 0 .0

0 .0

-5 0 .0

- 10 0 .0

- 15 0 .0
SAT 1 SAT 2 SAT 3 SA T 4 S AT 5 SAT 6 F U L LB O R E IN LIN E

X X 8 8 -X X 9 3 X X 70-X X 8 0 X X 2 0 -X X 3 0
Figure. 9–A graphical representation of the velocity data presented in Figure. 8.

The graph used to evaluate the SAT velocities indicates that spinner 2 has the highest velocity and spinner 5 has the
slowest (actually negative) velocity especially, in the zone xx20-30. This trend suggests that spinner 2 is on the high side, and
that spinner 5 is on the low side of the wellbore. The spreadsheets provide a means to review the data to help determine
whether or not the raw data is suitable for analysis.

Combined Interpretation
Although the PAL data can be used to make interpretations by using each tool individually, the real benefit comes from
combining the multiple tools and sensors into one complete interpretation package. Because the CAT and RAT tools provide
up to 24 holdup measurements and the SAT tool provides six velocity measurements, the PAL logs provide the advantage of
a complete flow analysis.
The basic flow equation is as follows:
Q = A ×V Equation 5
where: Q = Flow rate
A = Pipe area available for flow
V = Velocity
SPE 125028 9

Breaking the equation 5 into phase flow rates yields the following:
Qg = A × Yg × Vg
Qo = A × Yo × Vo Equation 6
Qw = A × Yw × Vw

where: i = Phase component (gas, oil, and water)


Yi = Phase holdup
Vi = Phase velocity

The major unknown in PL interpretation is the use of slip velocities to solve for the phase velocities that are required to
solve for equation 6. Historically, multiple equations and relationships have been developed to determine these slip
velocities, which then are used to match the downhole flow rates to the surface flow rates in most PL software packages.
Because the PAL tools provide both holdup and velocity measurements, the combined results should eliminate the need to
use the slip velocities to determine flow rates for each phase. However, since these tools can be arranged in any position in the
tool string, each sensor may not be in the same radial position in the wellbore as the other tool sensors, as shown in Figure. 10.
Correlating the individual position of each sensor is necessary to solve the flow equations without the use of slip velocities.
Because each PAL tool has an independent relative bearing sensor, the azimuthal or radial position of each sensor can be
determined with respect to the high side of the wellbore. This yields the horizontal and vertical position of each sensor. Also,
assuming a 0 relative bearing, it is possible to determine the area of the wellbore above and below the highest and lowest sensor
position in the well. By examining the data in the vertical plane, the measurements can be separated into segments.

ATOP
1

Y
Y
Y
Y

ABOT
CAT or RAT Vertical slicing of wellbore
SAT
Figure. 10–Possible tool and sensor positions for the SAT, CAT, and RAT, and segmented slicing of the wellbore.

At every depth increment, it is possible to provide the holdup, velocity, and area for each slice and therefore, the flow rate
for each slice. The highest and lowest vertical sensor position for each PAL tool is determined and extrapolated to the slices
labeled ATOP or ABOT. Expanding Equation 6 for each slice will yield Equation 7. By determining the holdup, velocity,
area, and flow rate of each slice, the total flow of each phase can be ascertained without the need to use any slip velocity
correlations. This technique removes several areas of concern related to the typical PL analysis, especially with deviated and
horizontal wells and multiphase stratified flow regimes.
i
Qg = ∑ ( Ai × Ygi × Vgi )
1
i Equation 7
Qo = ∑ ( Ai × Yoi × Voi )
1
i
Qw = ∑ ( Ai × Ywi × Vwi )
1

where: Qg = Gas flow rate


Qo = Oil flow rate
Qw = Water flow rate
Ai = Area of each slice
Yg = Gas holdup of each slice
Yo = Oil holdup of each slice
Yw = Water holdup of each slice
Vi = Slice velocity
10 SPE 125028

PAL Log Presentations


In addition to the horizontal image, a new image enables the visualization of the wellbore in a vertical plane. This image
is presented from the low side to the high side. The high side will always be on the right side of the track and will include ‘X’
in the image presentation name. A slight modification of this image incorporates the TVD data and will include ‘T’ in the
name.

TEMPERATURE
1 71 177
P RESSURE
2750 280 0
HYDRO PUBLISHED PUBLISHED
0 2 CATMAP XCATMAP RATMAP XRATMAP SATMAP XSATMAP
FLUID DENSITY
0.1 1.1
HI--------LOW--------HI LOW--------------------HI HI--------LOW--------HI LOW--------------------HI HI--------LOW--------HI LOW--------------------HI
GAMMA 50 200 50 200 0 1 0 1 -104 152 -104 152
0 100

XX00 A
G
E

XX50

F
D

Figure. 11–PAL horizontal and vertical images showing both phase segregation and fluid downflow.

Figure. 11 provides images from the three PAL tools along with quality control plots. The white dashed lines in the CAT
and RAT logs indicate the low side of the wellbore in the third track, and the magenta line indicates the same position in the
SAT log. The vertical center of the wellbore in the vertical maps is indicated by the same line and shows where the center
sample tools would be positioned when properly centralized. The quality control plots in the CAT and RAT logs are used to
show six of the 12 holdup measurements around the wellbore, with sensor one indicated by the blue line. The length of each
arm indicates the type of fluid, with water having the shortest length and hydrocarbons having the longest length. On the
CAT quality control plot, the gas component is slightly longer than the oil phase. The CAT and RAT quality control plots
indicate the heavier fluids on the low side. The SAT quality control plot shows the six velocity calculations in a cross-
sectional view. The longer arms indicate the highest velocity at that depth, whereas the shorter arms indicate the slowest
velocity, including negative velocity. The blue arm is the location of spinner #1. This log shows that segregated flow is
occurring in the well.
The CAT log indicates a pure gas on the high side of the wellbore at point A. From the center to the right side of the
track, a mixture of gas and oil is found. The almost-pure oil in the lower side of the wellbore, indicated by the CAT log from
A to B, is not sensed by the center sample tools, which indicate almost a 100% gas phase over the same zone. Points B and C
indicate oil on the low side, with a mixture of gas and oil on the higher side of the wellbore with some segregation of these
two phases. The density tool reads approximately 0.7 gm/cc, which indicates a mixture of oil and gas. At point C, the
XCATMAP image indicates a mixture of oil and gas and confirms the center sample tool readings. The CAT and RAT data
indicate water at point D, which correlates with the centralized sensors fluid density (DENR) and capacitance (CWH) logs.
At point F, the interface between the oil and gas is not visible on the RAT log because of the differences between capacitance
and resistivity readings. At point E, the water shown seems out of place, but because this is a perforated interval, it appears
that water is entering the wellbore on the high side of the well. The SAT log shows that segregated flow is occurring in the
well. Point H is characterized by negative velocity on the low side with a higher positive fluid velocity on the high side of the
well. Point G also indicates a segregated velocity profile.
With the myriad of options available to examine the PAL data, the use of the vertical displays seems to enable a better
understanding of the complex flow regimes and fluid segregation in these deviated wells. A definite correlation exists
between the phases sensed by the CAT and the velocity patterns sensed by the SAT. At point A, the fluids are a mixture of oil
and gas on the high side, tending to a pure oil component on the low side. This oil is either stationary or flowing slightly
SPE 125028 11

downhole while the gas/oil mixture is flowing up hole. The high side of the wellbore has a higher percentage of gas and is
flowing at higher velocity. At point B, the CAT log shows an entry of gas into the wellbore on the lower side. Immediately
below this point, there is apparent fluid downflow on the low side with a gas/oil mixture flowing uphole on the high side of
the wellbore.

Flow Calculations
Final processing of the PAL data set was performed using Kappa’s Emeraude software. After the data is ported into
Emeraude, the analyst must determine which of the multitude of options to use to perform the final analysis. With multiple
velocity and holdup measurements, there are multiple variations and options to determine final flow. The standard
interpretation that is normally used consists of the PAL process described earlier using the CAT and SAT data (Figure. 12).

All PASSES All PASSES All PASSES All PASSES DOWNHOLE DOWNHOLE DOWNHOLE
Z PRESSURE
QP TEMPERATURE
TEMP Q GAS
QG B/D Q OIL
QO B/D Q
Q ZONAL
QZT Q INCREMENTAL
QZI
22700
700 psia 2800
2800 1165
65 °F 178-5
178 000 B/D20000
-5000 20000 -5000 B/D 10000
-5000 10000-1000 B/D 15000
-1000 15000 -1
000 B/D 15000
-1000 15000-2-2000
000 B/D 10000
10000

Q ZONE

PERFS

Q ZONE

PERFS

Q ZONE

PERFS

PREVIOUS
ZONE

Figure. 12–Flow analysis using Emeraude over the entire well.

The pressure data indicates that the well is not in a steady state. The first pass in the well is shown in red, and the first up
pass is shown in green. After these two passes are made, the well becomes nearly stable. Likewise, a comparison of the gas
flow rate passes shows that all of the passes are similar, except for the first down pass shown in red (Figure. 12). The other
passes show that the flow of oil and gas is consistent at every depth, even though there are variations in the CAT, SAT, and
RAT data between each pass. Although the individual tools show changes between passes, the total flow rate remains
constant.
The flow rates are the average of the seven steady passes and are at downhole conditions. Emeraude uses a procedure to
match the known inputs (measured) to the calculated inputs (calculated) to determine the final flow with minimal error. The
software enables the user to weight each input and to select multiple PVT and slip options. The more inputs that are selected,
the more possible variations are in the final answer. The use of all of these options is possible with the PAL data; however,
12 SPE 125028

because the downhole flow is already determined, the surface production rate is calculated by using the PVT options to
correct the downhole flow rate of each phase to the surface rate. This calculation removes the problems related to the
standard PL analysis that use slip velocity to break down the calculated velocity into phase velocity.

XX00 XX00

EXPANDED EXPANDED
LOG LOG
DEPTH DEPTH

3-D
3-D LOG
LOG
XY00
Figure. 13–CHIME images including 3D view of CAT-RAT logs.

To further improve QC and interpretation, another software program, CHIME, is used to generate 3D images of the PAL
data. This software is designed so that the user can manipulate the data and derive images over zones of interest. CHIME
expands the zones of interest as shown in Figure. 13. The black squares are 10 ft long and the data enclosed by the square is
expanded to the tracks labeled CATMAP and XCATMAP. The RAT data is also shown and expanded the same way. The last
portion is the 3D image. The gray line in all the tracks highlights the depth that is shown in the cross-section in Figure. 14.
On both the CATMAP and XCATMAP images, a portion of 100% water is present on the lower side of the wellbore, but this
is not displayed on the RATMAP and XRATMAP images.

CAT
RAT
Yg = .013
Ywra = 0
Yo = .893
Yhra = 1
Yw = .094

Figure. 14–CHIME cross-sectional display of logs in Figure. 13. Black dots show the sensor positions; the white dot indicates
sensor 1.

Figure. 14 presents the data in a cross-sectional display for both the CAT and RAT logs. By examining the sensor
position of both tools, it can be determined why the water streak is not detected by the RAT tool. The CAT sensors are
almost in the highest and lowest sections of the wellbore, whereas the RAT sensors are slightly off this vertical plan. The
sensor position is the primary reason that the RAT does not sense the water portion detected by the CAT. Even with multiple
sensors, not all phases may be detected, depending upon sensor position. However, with multiple passes, the possibility of
SPE 125028 13

detecting most phases will be improved because the tool will probably not have the same rotational patterns in all the passes.
The layering pattern on the CAT tool also shows slight differences that are not detected by the RAT, but the components are
gas, oil, and water, and unlike the CAT, the RAT cannot differentiate between the oil and gas.

XX00

3-D
LOG

EXPANDED
LOG
DEPTH

SAT MIN =-101.78


SAT MAX = 60.93
SAT AVG = 60.93

XY00
Figure. 15–SAT CHIME display with 3D images and cross section.

Figure. 16–SAT CHIME display with 3D images.


14 SPE 125028

Figure. 15 and Figure. 16 highlight some of the capabilities of CHIME to present the PAL data in images and cross-
sections that enable easier interpretation of the flowing patterns. In this case, there is definite fluid segregation with some of
the fluids flowing downhole as indicated by the white to black shading while the fluids on the high side are flowing uphole
indicated by the yellow to magenta shading. The images are over the same zone with different presentations.
All of the CHIME data, including the software, can be downloaded to either a USB flash drive or a CD to enable
customer manipulation of the data. This manipulation is limited to only the 3D portion, however, rather than to re-calculation
of the PAL data. In addition to the images shown, CHIME can also present scalar data ranging from the raw data to the final
calculated flowrates.

Conclusions
The availability of the PAL tools provides multiple options to calculate holdups, velocity, and flowrates depending upon well
conditions, flowing phases, and other factors that may not be the most advantageous to conventional production logging.
With all of these sensors, quality control is an issue that is addressed by spreadsheets and analysis programs, including 3D
displays. The spreadsheets help to determine the quality of the raw data and whether more logging passes are required or the
tools need to be replaced. After quality data is recorded, several options are available to maximize the interpretation. Holdups
can be determined by a single PAL tool or by multiple tools, if necessary. There are currently four holdup options that use the
CAT and RAT data. Likewise, the velocity profile can be determined from the SAT, or from the inline or continuous spinners
if the SAT data is inconclusive. The flowrates are then determined by using the desired holdup and velocity data and is
converted to surface rates using Emeraude. The final product can be then visualized and quality controlled by the use of the
CHIME software. CHIME will provide 3D images to illustrate the complexities of the flowing conditions of wellbores,
including deviated and horizontal wells. The same procedure is then used to determine the flow rate of each component
without using any slip velocity correlations. This procedure will help in the analysis of horizontal wells, especially when
crests and troughs play havoc with the flow patterns and when a blocking heavy phase is present.
The use of the PAL tools with new interpretation practices improves the interpretation of production logging in both
deviated and horizontal wells. Although the example shown was a high rate oil/gas well with deviation of approximately 45
degrees, the process has shown definite benefits for horizontal wells. Removing the dependence upon slip velocity
relationships reduces some of the uncertainties in the area of multiphase production log interpretation.

Acknowledgments
The PAL sensors described in this paper were developed in part through cooperation with Sondex. The authors wish to thank
BP for permission to reproduce the log examples used here and for the support in the introduction of the PAL tools. The
authors also gratefully acknowledge the management of Halliburton Energy Services for allowing this paper to be published.
A special thanks to Tegwyn Perkins and Ron Stamm for the development of the 3D software to allow the imaging of the PL
data.

References
Frisch, G., Jung, M., Alldredge, P., Zett, A., and Webster, M. 2009. Providing Accurate PL Interpretation with Multi-probe, Multi-Sensor
Tools in Segregated Flow Environments. Paper X presented at the SPWLA 50th Annual Logging Symposium, The Woodlands, Texas,
USA 21-24.
Frisch, G., Perkins, T., and Quirein, J. 2002. Integrating wellbore flow images with a conventional production log interpretation method.
Paper SPE 77782 presented at SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 29 September-
2 October.
Ryan, N.D. and Hayes, D. 2001. A new multiphase holdup tool for horizontal wells. Paper V presented at the SPWLA 42nd Annual
Logging Symposium, Houston, Texas, USA, 17-20 June.

You might also like