¾ Wireline & Testing

P. T. Schlumberger Geophysics Nusantara

CMR Help File

OJW

Adapted from NSR-CMR Procedures of 9 April, 1996, originally written by

Aruna Subramanian
Updated 10 August, 1996
Table of Contents

Introduction

Section 1 - Basic Operations

Safety
Hardware requirements
Software requirements
Before Survey calibrations
Master Calibration
Prejob checks
Downhole calibrations

Section 2 - OBM logging

Section 3 - High porosity/permeability formations

Section 4 - LQC

Section 5 - Appendix
High speed (BFV) logging
Multiwait logging
Hints
List of contacts
Introduction

The CMR is a great new logging tool. It provides information that we could
never provide before, and does it in a very reliable manner. Although this is a
new tool there is already a considerable body of experience around the world.
Unfortunately, there have been some unsatisfactory CMR logs run by people that
didn’t fully understand the way the tool works. This is not their fault. Although
the CMR WRM is a very good basic instruction manual on how to run the tool, it
does not cover some of the difficult aspects of CMR logging. The purpose of this
document is to expand on what is already presented in the WRM, and to address
2 specific problem areas - logging the CMR in oil base mud, and logging in high
porosity/permeability formations. A quick summary of the basics of tool operation
is also provided, however the engineer should refer to the CMR WRM
(MH606602-AA ) for more complete details of proper CMR operation, as what is
presented here is just a brief summary, and the WRM is far more complete.

SECTION 1 - BASIC OPERATIONS

General Comments

It is easy to screw something up if you do not know what your are doing
with this tool. If you are not sure what you are doing, read the WRM, or
telephone someone else that can help you out. The CMR is not that tough to run,
but do not proceed on the assumption that it is just like every other tool you
learned about at training school. It is possible to damage the tool if you are not
careful, or to run a bad log. When in doubt, ask for help.

Safety

1 - As in most open hole logging tools, high voltages on the order of 600V exist in
the sonde. Be careful, lest ye get bit.

2 - High magnetic fields exist near the sonde. Keep away from ALL tools with PM
crystals and tubes. Don’t store the CMR on your truck or tool pallet near your
GR, NGT, GPIT or PGD. Be sure to have several feet of clearance from these
tools. The magnet also interferes with data tapes, floppy disks, pacemakers,
mechanical watches, and especially credit/bank cards. It is good practice to keep
the magnet away from all electronics.
3 - Do not dismantle the magnet unless necessary. Without the proper technique
and tools, the magnets can fly apart or slam together with great force. Without
meaning to you can damage the tool and yourself.

4 - Place the sonde in its metal case and then place both in the wooden case for
transport. Dangerous goods manifest for magnetic material is required. Ensure
the transport/airline company is aware of the presence of magnetic material.

5 - Make sure that the bellows in the sonde are filled before RIH. After every 5
jobs Q check the sonde and clean all metal filings from below the antenna.

6 - The sonde goes into its metal case only one way. Don’t try to force it in the
other way.

7 - DO NOT power up the CMR tool in casing, in its metal shipping box, or with it
up against any metal. You will damage the CMR tool. A good rule of thumb is to
only power up the tool with the faraday shield around the antenna.

Hardware Requirements

1 - The tool components are; CMRS-A, CMRC-A in CMRH-A, Bowspring and

shoes (The shoes are for a 3 5/8” tool, same as an AACT and the bowspring is
the same as for a CNT-G). The tool is a DTC only tool. To have head tension
with the CMR you must use an ACTS.

2 - The CMR ready box which you take to location with you should include: the
master calibration water bottle and thermometer, Faraday cage (green plastic
cylinder about 4’ high by 1.5’ in diameter), mud tester jig, brass Allan keys and
screwdrivers, 2 DAT tapes (One with the latest CMR kit, and one with the CMR
master calibration information), and the CMR WRM.

3 - This tool can be run standalone, or can be combined with some other tools.
Note that although it is DTS, it is not universally combinable. The CMR at this
time will not work in combination with most tools with a mass isolated housing.
This includes AITB, SHDT, FMS, DITE, DLTE, FMI. It has been run with LDT,
CNL, AITH, DSI, EPT,GPIT,NGT,EDAC. For a complete list of combinable tools,
refer to the CMR home page, and down load the CMR combinability report, or
contact the CMR product champion, Greg Gubelin in Houston. (A list of CMR
gurus is presented in the appendix. One of these people could certainly answer
your late night questions if you have any)

4 - To acquire CMR data on the same trip in the hole with tools it can not
normally be run with, a selective switch can be used to separate the tools. This is
a technique that was developed for TLC logging in the North Sea, but has also
been used in other parts of the world. I believe that these switches are used in
Kuala-Lumpur as well. The basic concept is that you take advantage of unused
lines in the CMR tool string. You run the CMR string on top of another tool string,
and provide a switch at the top of the CMR tool that can switch cable conductors
1-7 from the CMR to these alternate lines. At the bottom of the CMR string you
have a crossover that terminates the CMR, and switches the alternate conductors
back to lines 1-7. Using this technique, all tools in the CMR tool string must have
7 unused lines. See the appendix for more documentation on this.

Software Requirements

The current app kit is CMRT-019A/019AX/HPCVA-744/HPCAX-745 - if you want

to use it make sure you check it beforehand with the other appkits you intend to
use.

Setup Phase

The CMR is located in LES in open hole telemetry, the tool string is SGTL-
DTA-CMRT-DTC-ACTS. (Use the adapter ACTS, and not the tool ACTS, or it will
not work.) To ensure good excentralization, some locations run a CNT-H with its
bowspring below the gamma ray. If you do this, don’t use the source, and
change the CNT length to zero to avoid unneccesary memorization.

Before Calibration

A good calibration before logging is critical. If the tool is not tuned properly, the
log will be affected, possibly even worthless. It is not possible to recover from a
bad calibration by recalibrating and doing a playback. If the log is run with a tool
out of tune, the log is JUNK.

A before calibration is actually a misnomer. What you are doing is tuning the tool
to operate at the Larmor frequency. This is the resonant frequency of the
hydrogen nuclei in the formation pores. This should be done before starting your
down log, in a zone in the well that contains a significant amount of free
hydrogen. The best place to calibrate is a clean wet zone. As an alternate, some
engineers have placed the tool in a wash out, such as the one usually found
below the casing shoe, and tuned the tool there. Caution is advised if using this
technique, as the tool will be damaged if calibrated in casing! Ensure that the
magnet is in the washout not the bottom of the toolstring!

Technique

- Place tool with antenna located in a zone with free hydrogen (sand or washout)
- Perform SSD TPD enable
- Power us AC main using normal FTB techniques. The software automatically
reads the master calibration data stored in the tool EPROMs. All of the signals
you need to check proper tool function automatically appear in the IOM. The
important things to look for are;
a - All the status lights should be green.
b - The number of messages should be updating and be incrementing by 1.
c - The Hall probe, which measures magnetic field strength, should read around
50-52 mT.
d - The AC Aux. Isolation Relay = 0. This relay controls where the AC aux. power
goes. Either to the CMR or to caliper tools such as the [Link] default is to the
caliper tools

- The high voltage boxes in the control panel will be red as no power has been
sent via AC Aux. Do not power up AC Aux. yet!

- Observe that the mode in the CMRT control panel is LFST Search Station.
(Default)

- If at surface doing a pre job check, place the water bottle on the antenna(square
bottle filled with green stuff), along with the faraday shield. The shield should be
placed such that the grounding clamp faces uphole.

- Perform Tune Word Search Task, init. Set the search frequency at 2.54e6, then
perform the tune word. This takes about 1 minute. When this finishes you will
observe a PDS file with a nice smooth peak.

- If this looks good, proceed with the LF search task. Do initialize again, and set
the frequency to 2.54e6. Perf LF Search. Maxis will blue box you and ask you to
turn up sonde power. Enable sonde power and, if you are not is casing, or have
metal within 2’ of the antenna, apply power to AC Aux. Watch the HV boxes in
the control panel and slowly turn up variac. Note, the control panel only
updates every 5 seconds. If you get too aggressive you can over power the
tool. The boxes turn green when unregulated HV=290, and regulated HV=250.

- When power is applied to the antenna, the AC Aux. flickers a bit. The rate of
flicker is a function of the wait time selected, and will change for every different
mode you log. Don’t worry about it. In fact, make sure it DOES flicker! If it does
not, stop and restart Acquisition.

- After power is OK, answer yes to the blue boxes at tool power up, and proceed
with the search task. A window will appear, showing the results of the task. Wait
until the standard deviation is less than 2.5, and successive FL values are the
same. Then answer yes to the yellow box. Check the PDS file of the results.
You should see a curve fit peak to the red data points. You want at least 2 red
data points on each side of the peak. If not, record the frequency found, and re
do the initialization and task. Do not leave this until you have the tuning
optimized, which means at least 2 of the red data points on each side of the
tuning peak. If the tuning is not good after several attempts, try moving the tool to
another depth. If the hole is sticky, the tuning can be done while moving the tool
slowly. Just make sure the tool remains in a porous zone.

- When the LFST is done correctly, re do the tune word. This will require you to
power down the antenna. Do the tune word search at the same frequency that
you found in your LFST. It is important that both tune word and LFST are done at
the same frequency.

- Software Bug; Perform close PDS file before clicking on Close on the TW and
LFST Report. Otherwise it might not come up the next time you return to
Calibrate phase. Then the only solution is to stop and restart the configuration.

- If the tuning is done properly, the value of delta B0 from the Hall probe in the
CMRT control panel changes to 0 after this task.

- The tool is now tuned and ready to acquire data. Remember, if at any point
you power the tool down, you must retune it. (TW and LFST). A good practice is
to note the temperature and operating frequency occasionally from the Control
panel so if you do lose power/telemetry you have a pretty good idea of the
Central frequency. Whenever tool is powered down or acquisition is stopped, the
tool loses its operating frequency and needs to be retuned. Noting the Central
frequency and setting it appropriately in the TW and LF init tasks speeds up the
search.

Master Calibrations

The master calibration is very easy to do, and should be done at least
once a month. There is nothing wrong with doing it before every job should you
choose to.

- Set up the CMRT-DTC toolstring with the waterbottle. If possible leave the
water bottle and tool in the same environment so that everything is at the same
temperature before starting.

- Do the tune word and LFST as in a before calibration. Then perform the bottle
calibration tasks. When the software asks for the temperature, physically
measure the temperature of the bottle fluid. The fluid gets heated up a bit by the
antenna on a cold day, so it might read slightly different than air temperature. If
your thermometer inside the calibration bottle is broken (It often happens) Use
the thermometer from the mud checker kit.

- Do the calibration in a low electrical noise environment, and use the faraday
cage! Check that normalised noise <1 and Porosity=1.0. When the calibration is
complete, write the values to the EPROM. (NB : OP8-538 does not write mcal
values to EPROMs. This has apparently been fixed in the new appkit.)

- Go to log phase and do a Sandstone Station log. Porosity should be 1.0, if not
redo the master calibration.

- Software Bug: The Signal to Noise Measurement always comes out as Not
Done during the MSL tolerance check for the current software. If the value looks
reasonable, just ignore the message. Sometimes doing a GET from a master cal
file seems to help. (NB: If ALL the values are zero then something is wrong! Do
not proceed till you have the mcal values)

Log Phase - Prejob Checks

The purpose of these checks is twofold. First we want to be sure that the
tool is working correctly, then we want to measure some of the fluid properties of
the mud and filtrate to help in the post logging processing and interpretation of
the data. To perform fluid measurements with the CMR, it should not have
anything connected below it. This is because tools below the CMR act as radio
antenna, and radio frequency noise can introduce noise onto the measurement.
The best way to check the tool out is to perform fluid measurements with the tool
solo, then hook up your entire tool string, and check that the other tools in the
string work.

- With the calibration bottle on the tool, change the mode to Carbonate mode by
clicking on mode at the top of the CMR control panel.

- Do a couple of station logs. Wait until the standard deviation stops changing
and draws a straight line. (About 3 minutes each) and the S/N ratio  20. The T2
distribution and the T2LM should repeat. The CMR porosity should read very
close to 1 (1.0 +/- 0.03)
If it does not, redo the master calibration.

- At the well site, get samples of mud and mud filtrate about 40 ml each. If this is
an OBM job, this is critical. It can be difficult to get filtrate on an OBM job, so
base oil can be used. There must be no particulate material in fluid when the
filtrate check is made. Be sure the jig is very clean between each fluid sample
test.

- Clean the jig really well, and fill it up with the filtrate/base oil. Change mode to
Expert Station Log. Change the wait time to 8 seconds, and the number of
echoes to 5000. (Wait time shall be referred to as TR henceforth - A Numar
term) Do a station log until signal to noise ratio is about 20. Stop, note T2LM,
and include this PDS file on the log. Do two more stations with TR= 4 and 6 secs
and 1800 echoes. The FF porosity should be the same as the one for the 8 sec
station +/- 1 p.u.
- Clean the jig thoroughly. Put the mud in the [Link] the same stations as
above. Wait until the signal to noise ratio is about 15. Be sure the faraday cage
is on the tool. Record the T2LM of the mud, and include this PDS file on your log
tail as a mud check, and annotate accordingly.

- With the base oil/filtrate still on the tool do a Carbonate Multiwait station log.
Record T2LM and T1/T2 ratio. Include this PDS file on the log as well.

- As a rough guide, T1 (downhole) = 2 * T2LM (surf) * T1/T2 (surf)

- Also, roughly, TR= 2 * (T1 + (TE * echo spacing))

- You are now ready to run in the hole.

Log Phase - Downhole Calibrations

Do Not Power up AC Aux. in Casing! Run in the hole to the depth

where you will do your tune word and LFST. Get an idea from the geologist
where the first available sand will be, and position the CMR skid here. If a down
log is required, and if the first sand is a significant distance below the casing
shoe, consider tuning the tool in the washout below the casing shoe. As long as
there is enough hydrogen present, you can tune the tool. Whether it is a porous
zone, or washout, it doesn’t matter.

- As a guideline, Central Freq. downhole = LF (surf) - 0.805 Khz/Deg C

temperature change. Calculate your new expected central frequency, and start
here when you do your first Tune Word search. The tune word and LFST are
done exactly as you did it at surface.

- If your are logging in medium to low porosity water based mud, proceed as
described below. If you are logging in OBM, or in ultra high permeability or
porosity, refer to the separate sections on logging for these environments.

- Without pulling the CMR into casing with the AC Aux. powered, position your
tool and start your down log. There are 2 options that you can do for down
logging.
1 - You can log at a reasonably high rate of speed, and acquire partial
CMR data; The Bound Fluid Mode. To do this, set mode to BFV in the CMRT
control panel. In this mode you log at 900 fph, and only measure the bound
fluid. The other CMR data is not valid. However, using computed effective
porosity from the LDT/CNL you can calculate free fluid and permeability. This is
a good backup log should you not acquire any other data provided it is a clean
zone.
2 - The second mode requires a bit of foreknowledge. Basically, you
proceed as in step 2, but log down at a high rate of speed. You only slow down
to the required logging speed when the CMR antenna actually travels over the
sand bodies. This significantly improves the logging speed, but has the
disadvantage of requiring the engineer and winchman to keep on top of where
the expected sands are located in the well. To log too fast over the sands is to
acquire meaningless data.

- One note on changing modes in the CMRT control panel. It takes a very long
time. When you switch modes, the CMRT status lights in the IOM go red, then
turn back to green. Wait until they change, and the number of messages start
updating before you do anything else. You have to hit apply each time you
change modes. If you change wait times, count the seconds between
fluctuations in the AC Aux. Don’t change modes or parameters after the Log
command has been issued. The log will be invalid. Also do NOT start the
log unless you are certain the AC Aux pulse spacing is the same as the
wait time.

- After your down log, retune the tool. Get tied in, and position the antenna in a
sand. Remember, the antenna is not at bottom where your odometer depth is,
but is up the hole a bit depending on the tool string. Position the antenna in a
sand. to tune the tool. It is best if this is a clean water bearing sand. This
calibration is similar to what you did at surface, and what you did just below the
casing shoe. Remember, if the tool is not tuned properly, the log is invalid,
and can’t be corrected in playback. If you signal level is low, that means that
your porosity is low. If it is too low, it may be difficult to properly tune the tool.
Reposition the CMR in a different sand if necessary.

- After tuning the tool, set your mode to sandstone depth log, and position the tool
to do your first logging pass. The CMR requires AC Aux., which is in conflict with
the caliper tools which also require it. So, position the tool string, power down the
CMR, open the calipers, power up the CMR, and start the log. As long as you do
not power down AC Main, you do not need to retune the CMR. AC Aux. can be
powered down and repowered.

- Do your first up pass at 600 fph (300 fph in Carbonate mode) or whatever
logging speed is required for the wait time chosen. Try to log your repeat over a
sand body. Be sure that the logging speed does not exceed the limit. The result
on your log from too fast of a speed is incorrect and low porosity.

- If you have LDT/CNL from a previous trip in the hole, it helps to have done a
PERT, and to compare effective porosity from the CMR to what you determined
by doing a PERT.
- After your repeat pass, retune the tool, and observe if it determined a different
Larmour frequency than found before the repeat pass.

- Do your main logging pass,(make sure repeat analysis is active) check for
repeatability, and retune the tool after the main logging pass. If there is a
significantly lower effective porosity than expected, your logging speed may have
been too fast, you might not have had the tool properly tuned. Or, there might
have just been more clay content in the formation than the client expected.
Remember, the CMR effective porosity should be less than the LDT/CNL
crossplot porosity by an amount equal to the clay content of the formation. The
CMR does not see clay, the other tools do.

- After your main logging pass, again retune the tool. Present all of these tuning
PDS files on your log. This is very important from a “cover your ass” point of
view. It is also important for the client to know the data was acquired at the
correct [Link] first thing that Non-revenue Generating Town Pukes do,
when there is a question about a CMR log is ask if it was tuned properly. Having
all of these tune PDS files helps to deflect the questions.

In water based mud systems, with low to medium porosities, the CMR is a
reasonably easy tool to run. It is not idiot proof, but the engineer new to CMR
logging should be able to take it and get reasonable data. However, the tool is
very difficult to run in oil based mud, or in ultra high porosity/permeability. For
the purposes of this paper, ultra high poro/k is defined as porosity over 30 %, and
perm higher than 2 Darcy. If you have occasion to log in these environments,
special techniques are required. Refer to the appropriate section below.

SECTION 2 - LOGGING IN OIL BASED MUD

Logging in oil based mud is significantly different than logging in water

based mud. The default parameters selected in the CMRT control panel for
sandstone or Carbonate depth or station logging are invalid for oil based mud.
To use these modes of operation in this environment will result in a bad log. The
procedures for OBM have been developed in the North Sea, and work quite well.
If this section is not clear enough, refer to the list of CMR guru’s at the end of this
paper for more documentation. Do not head off to blindly run the log if you do not
fully understand the tool or modes of operation. The CMR is a very easy tool to
run a bad log with!
Why does OBM matter? I am glad you asked. You should refer to the
CMR WRM for a complete explanation of why CMR acts the way it does, but
basically it is for this reason. Most reservoirs were deposited in a water wet
environment. The hydrocarbons, if any, invaded the reservoirs at some time after
the original deposition. For this reason, most reservoirs are water wet. The pore
spaces are coated with water, no matter what the actual pore fluid is. If you drill
with OBM, the filtrate which invades the pore space, displaces the original
formation fluid. This results in a flushed zone with the pore spaces filled with
OBM filtrate. The OBM filtrate is held away from contacting the pore surface by
the wetting fluid which is water. What you get is an oil droplet which is
surrounded by water. The effect of this, is that when the hydrogen in the pore
fluid is polarized, and then relaxes during the CMR measurement, it never
crosses the oil/water interface and never contacts the pore boundary. Since
interaction with the pore boundary is the most significant contribution to causing
the hydrogen to relax, the fluid in the pore space ends up relaxing at the bulk rate
for that particular fluid. The bulk rate of relaxation is quite long. The result of this
is that you must use significantly longer wait times than the default. To log with
the default mode of logging is to generate a log which gives a pessimistic value
for free fluid and effective porosity, because the bulk rate of relaxation is longer
that the wait time.
For OBM logging, proceed as indicated in the general guidelines above,
but also do the following:

- Measure the T2LM of filtrate and mud at surface before logging. Include these
PDS files on the field print. Use the values you have measured to get a rough
idea of wait time downhole using the method documented above in the prejob
checks.

- Run in hole, and position the tool in a clean water or oil bearing sand to do your
TW and LFST tasks. Do not do this in a gas zone. Tune the tool as you did at
surface. Compute the expected larmor frequency based on what you found at
surface, and on the downhole temperature.

- The next task is to determine the proper wait time. This is the most critical thing
for OBM logging. The wrong wait time results in bad data. Set the mode to
expert station log. Change TR to 10 seconds. Use 1800 echoes unless T2LM at
surface was greater than 500 msec then use 3000 echoes. Do a station log until
the S/N ratio =~10. Stop the station log, and record the CMR porosity and
POL_C. Change TR to 8 seconds and repeat. CMR porosity should not change
by more than 1 PU, and POL_C <1.15. Keep reducing TR by 2 seconds and
record station logs, until CMR porosity changes by more than 1 pu, or POL_C >
1.15. The previous TR is now the one you should use for logging. For example;

TR POL_C CMR porosity

10 sec 1.01 12.6 pu
8 sec 1.05 12.6 pu
6 sec 1.09 12.5 pu
4 sec 1.14 12.5 pu
2 sec 1.19 12.3 pu

In this example, use TR = 4 sec for logging. As a check, use TR=6 for the repeat
section.
- If the hole conditions are poor, and there is a danger of sticking the tools, set
the mode to Expert Depth Log. Using the LQC format, log 50’ passes until you
find the value for TR where porosity starts to change.

- As a reference, here are the logging speeds for various wait times.

TR LOGGING SPEED SAMPLE INTERVAL

8 sec 164 fph 9”
6 sec 218 fph 9”
4 sec 320 fph 9”
3 sec 411 fph 9”
2 sec 615 fph 9”
8 sec 109 fph 6”
6 sec 142 fph 6”
4 sec 214 fph 6”
2 sec 405 fph 6”
1.3 sec 600 fph 6”

Remember, the vertical resolution depends on the sample interval as well

as the stacking used. Check with the client before you change sample
interval.

- Set T1/T2 ratio = 2.5, use the Coates model for permeability, and log your
repeat section with TR + 2 seconds. Use downhole stacking of 3 if MRT<140
degc, and 5 if MRT > 140 degc. Make a note of the allowable logging speed in
the control panel, and stick with it. Use T2_Distribution and LQC formats. Do
TW and LFST before and after each logging pass, and be sure that there is no
change.

- Record the main log with the same parameters, but use the TR chosen from the
optimization station logs/passes. Observe whether CMR porosity is repeating
with the repeat section. If there is a significant reduction in CMR porosity, your
wait time is not sufficient. Check the AC Aux. pulses. Be sure that the pulses
agree with the wait time you set In the control panel.

- The other significant parameter to determine for proper CMR logging is T1/T2
ratio. Experience in NSR has shown that T1/T2 ratio is usually 2.5 for OBM. This
is a good value to use for your log. However, this can be changed during
playback, and the log corrected if needed. It is recommended that you use T1/T2
of 2.5 for the log, and then do a playback after the log to verify that this is
reasonable. The main reason is to save rig time. To check T1/T2 ratio, play
back the 4 second wait time station log that you used to determine the correct
wait time. Since the 4 second wait time log does not agree with the 10 second
wait time log, the correct value for T1/T2 ratio is that number that makes the 4
second porosity agree with the 10 second porosity. So, just keep playing back
the file with the 4 second wait time in increasing steps until it agrees with the 10
second value. This should be about 2.5 for OBM. After you have found the
correct T1/T2 ratio, playback your log data to get your final product. This can be
done later in the FLIC too as long as you have recorded all the necessary data.

Section 3 - Logging in High Porosity and Permeability

The behavior of the CMR in high porosity and permeability is not well
documented at this time. The conventional wisdom initially was that the default
modes of operation should be satisfactory for logging in these conditions.
However, it appears based on some recent experience that this is not the case.
If the pore size is large enough, the water based mud filtrate within the pores may
have a larger wait time, possibly approaching the bulk rate of relaxation for water.
At surface, the bulk rate of relaxation for water can be of the order of 2 to 3
seconds. Clearly the default sandstone mode wait time of 1.3 seconds will not be
satisfactory.
If you are logging CMR in these conditions, which we will describe as
porosities in excess of 20% and permeability higher than 1000 milli Darcy, you
should be ready to do something different than the default mode. The idea here
is similar to what was done for logging in OBM. Basically, log with a longer wait
time at a slower logging speed.

- Do the measurements of mud and filtrate at surface, as should be SOP. Include

these PDS files on your log print.

- Run in the hole, position the tool to calibrate it in a clean wet sand if possible.

- Determine the Tune Word and LFST.

- While still on station, conduct station logs in expert mode. Use a starting wait
time of 10 seconds, and take stations in 2 second steps down. (i.e. 10, 8, 6, 4, 2)
Also do a 1 second wait time. Record the porosity and POL_C for each of these
station logs. Just as in OBM logging, choose the shortest wait time which gives a
consistent porosity to the longer wait times.

- If the porosity is the same for all station measurements, including the 1 second,
you do not have any problem, and the default sandstone mode of logging will
work fine. (Wait time 1.3 seconds)

- If there is a change in porosity for changing wait times, proceed as follows.

- Use T1/T2 ratio of 2.5.

- Record the repeat section at a wait time 2 seconds longer than you determined
as the expected.

- Record the main logging pass with the expected wait time, and compare to the
repeat section. There should be good repeatability. If the main logging pass
shows a CMR porosity less than the repeat, you must use a longer wait time, and
slower logging speed.

This phenomenon is not yet fully understood. Work on high poro/perm is

ongoing. The information presented here is the most current at the date of the
composition of this document, but there may be more current information
available as you read this. If you have a job in this type of environment, you are
encouraged to contact one of the CMR guru’s.

SECTION 4 - Log Quality

If this is a subsequent run to your LDT/CNL/GR, try to have a FEQL/PERT

done by the time you run the CMR. Compare the effective porosity from the
PERT with what you measure on the CMR. Try to get input from the geologist as
to the clay content of the formations. This will give you an idea if the CMR looks
like reasonable data.

- During logging, the POL_C should be less than 1.15. Otherwise, you might
have to do another pass with a higher TR. Remember, when in doubt, use a
higher wait time, and log slower. POL_C is basically a software gain to
compensate for the fact that the tool is moving too fast to polarize 100% the
hydrogen. It boosts the data by the amount of unpolarized H there is. The bigger
the POL_C, the more uncertainty in your answer. It isn’t the same kind of
correction, but it is similar to delta rho for an LDT. Big is bad.

- Normalized Noise < 1.0

- Standard Deviation <0.02

- Delta B < 0.1. If not, stop the log and recalibrate the tool. Changes in Delta B
are usually due to metal debris collecting on the magnet which changes its
magnetic field strength, and zero reference.

- Maintain a smooth logging speed to avoid stick/slip. Even if your average

logging speed is correct, if the tool is stopping and then jumping, the effect will be
to inadequately polarize the hydrogen, and give you a pessimistic and low
effective porosity. The unit might need to be fitted with a low speed gear box.
- Good pad contact is vital to this log. Be sure to have adequate eccentralization.
Do not expect to see good data in rugose or washed out hole. Check your LDT
caliper. Compare log response to the surface mud check. The log response will
look the same in washouts. See example attached to this letter.

- The Larmour Frequency before and after the main pass should be within +/- 5
Khz if done at the same depth in the hole. Temperature changes will cause this
to change if your borehole temperature is not stable, or if you do this at a different
depth in the well.

SECTION 5 - Alternate Operating Modes

High Speed (Bound Fluid) Logging

The standard logging speeds for the CMR are not very impressive,
especially if it is necessary to use a longer than normal wait time due to OBM.
This can be a problem if the client is concerned about hole conditions, or if the
logging unit can not log slow enough to acquire the data in the standard modes.
Fear not, there is a mode of operation that will allow you to log at a reasonable
speed! This mode basically does not attempt to capture the entire CMR data, but
just acquires the early time data. This is the part of the signal associated with the
clay bound water. If you log in this mode, you do not get accurate CMR total
effective porosity, free fluid or permeability. What you do get is an accurate
measure of bound fluid. If you then compute an effective porosity using the
LDT/CNL crossplot porosity corrected for clay content, you can use bound fluid
volume to calculate free fluid, and then using free fluid, calculate permeability.
Note, that this method relies on an accurate value of effective porosity computed
from LDT/CNL. The only way this will work, is if you have good hole conditions,
and if the matrix type and clay content are well known. You definitely lose the
nice ability of the CMR to make a measurement which is independent of lithology.
This technique will require much messaging of the data at geoquest, so you will
not get a field product. Do not use the field CMR porosity, permeability, or
free fluid if you log in BFV mode! This is not recommended as the primary
was of obtaining CMR information. The best way to use BFV logging is to
acquire data while running in the hole on a down log. If the tool gets stuck during
your good main pass, at least you have some data to salvage a log with.

- For Bound Fluid Measurement, Use the BFV pass format.

Wait time = 0.45
Echoes = 600
Sample interval = 4 in.
Downhole Stacking = 1
Uphole Stacking = 7
Advantage is good vertical resolution of 30 inches. Noise is improved with 7 level
stacking. Faster logging speed reduces stick/slip effect. Make a repeat pass to
confirm repeatability.

- One more note about high speed logging. When you are using the standard
mode of logging the CMR, (Not BFV mode), it makes no sense to log the CMR on
either an up or down log over long distances of shale. The CMR porosity in
Shale is 0. Why waste the time giving the client an extended shale log when the
only data of interest are the sands. It is possible to save quite a bit of time when
you run the CMR in combination with the LDT/CNL. If you have the induction, or
MWD resistivity log, log the tool string at your standard LDT speed of 1800 fph,
and only slow down to your CMR speed when the CMR antenna is crossing the
sands. Using this technique, you can save a considerable amount of logging
time without compromising your data quality
.

Multiwait logging

The problem of determining the correct T1/T2 ratio is supposedly fixed by

doing Multiwait station logs, and multiwait depth logs. The theory behind multi
wait logging, is that by acquiring the CMR data using several different wait times,
and comparing the changes in porosity, the correct T1/T2 ratio should be
determined. This is basically what we do in oil base mud logging, when we
make several different station measurements, and then compare them. It is very
nice to be able to do one multi wait station log, and get the T1/T2 ratio right off of
the top of the insert. This is lots faster than our SOP. Unfortunately, Multiwait
logging does not work very well, and is not recommended at this time. It will not
work at all if you have an ENP tool due to hardware deficiencies. So don’t bother
to even attempt this with an ENP. Even with PS tools, it has given some
answers which are not correct when compared to standard station logs. It is not
recommended to use Multiwait station or depth logs for the acquisition of data at
this time. If you want to experiment in the shop, or if it is for some reason
necessary due to bad hole conditions, here is what you need to know. Use the
following parameters for multi wait logging;

- Number of wait times = 3

- 3 sequences are available, decide on the one relevant to the job beforehand.

Sequence #1
Wait time Echo count Repeat Count
1 1.2 sec 1200 2
2 0.65 sec 1200 3
3 0.38 sec 600 5

This sequence has been successful in both sandstones and carbonates.

Sequence #2
Wait time Echo Count Repeat Count
1 1.2 sec 1200 3
2 0.65 sec 1200 4
3 0.38 sec 600 5

This provides better S/N than sequence #1 and is preferred if the tool
temperature is expected to be above 250 degF

Sequence #3
Wait time Echo Count Repeat Count
1 2.0 sec 1200 2
2 0.65 sec 1200 4
3 0.38 sec 600 5

This sequence provides better S/N ratio than sequence #1 and is preferred if tool
temperature is expected to be above 250 degF. This is preferred to sequence #2
in formations where long T2 times are expected > 1 sec.

Echo spacing = 320 usec

Send echoes - NO
(Multi wait will not work until raw echoes are disabled!)
Auto Frequency = Yes
Auto Cal = Yes
Downhole stacking = 1
Sample interval = 6 in
Regularization = Auto
T1/T2 = 1.0
T2 min = 3 msec
T2 max = 3000 msec
T2 cutoff = as appropriate
T2 mudfiltrate = As appropriate
Number of components = 15
Uphole stacking = 1
H index correction = 1
Permeability model = As appropriate

Note, that this software can be a bit hard to set up for Multiwait logging with the
current software. This is definitely something to try in the shop before you attempt
to run it in a real job.
Hints :

- When running CMR with the nuclear string make sure to have 2 x Ah107s
between the CMR and the CNT to avoid a standoff on the CNT.

- When running the CMR with UBI/GPIT it is advisable to use a SAH-F or

AH120 without its epoxy housing.

- GPIT with CMR is advisable to ensure good pad contact in deviated holes
(TLC jobs)

- If delta B0 in the CMRT control panel turns red and stays that way during
logging, you will have to stop the log and retune the tool. This normally
happens if metallic junk has accumulated on the magnet.

- Make sure the CMR antenna is in front of a porous zone while tuning and
not the bottom of the tool.
This is the help file used in NSR as quick operating instructions for OBM

CMR LOGGING PARAMETERS IN LOW VISCOSITY OBM

T1/T2 DETERMINATION :

Position the antenna in front of a clean, porous zone. Tune the tool - Tune Word and
Larmour Frequency Search Tasks.

Mode = Expert Station Log.

Permeability model = Coates.
Wait time = 10 secs.
Echoes = 3000.
T1/T2 = 2.5.
Perform Station log till S/N > 15. Note FF porosity and T2LM. Ensure POL_C < 1.1.

Number of echoes = 1800 (If T2LM > 500 msec, use 3000 echoes).
Perform stations at 4 , 6, and 8 secs and choose the wait time that gives POL_C<1.15.
This wait time shall be called TR.

If porosities are different, playback all three stations with different values of T1/T2 till the
porosities from all three stations are the same as the porosity from the 10 sec station. Use
this value of T1/T2 for logging.

LOGGING :

These parameters have been found to be suitable for most formations in the North Sea.

Repeat Section :
Mode = Expert Depth Log.
Wait time = 6 secs
Echoes = 1800. (If T2LM of filtrate at surface > 500 msec, use 3000 ).
Permeability model = Coates
T1/T2 = Computed as above
Downhole stacking = 3 (If temperature > 140 °C , use downhole stacking =5)
Uphole stacking = 1
T2 cutoff = 33 for sandstone, 100 for carbonates.
Sample interval can be changed to 9"to increase logging speed.

Main log. :
Wait time = 4 secs
Remaining parameters same as repeat section. Do a real time repeat analysis.
N.B. : If the TR determined from the stations is not 4 secs, use wait time = TR+2 for the
repeat section and wait time = TR for the main [Link], remember to tune the tool before
and after the main log.

CMR LOGGING PARAMETERS FOR WBM IN HIGH POROSITY

FORMATIONS

Mode = Expert Depth Log

Permeability model = SDR
T1/T2 = 1.5
Wait time = 3 secs
Echoes = 1800
Downhole Stacking = 3 (use 5 if BHT > 140 degC)
Uphole stacking = 1
T2cutoff = As appropriate
People to contact :

1. Greg Gubelin (Tool product champion), HPC

2. Chris Morriss , HPC
3. Paul Deutch (FTC), HPC
4. Bob Kleinberg (Physicist)
5. Bob Freedman (Acquisition and data handling)
6. Ning Wang (Software)
7. Bob Davis (IND)
8. Indra Kesuma (AJG)
9. Aruna Subramanian (AJG)