To Wireline Business Managers

From SKK-Technique

cc WTH

Subject DSI "Best Practices"

Date August 27th 1999

The DSI tool continues to generate solid revenues for Schlumberger and should do so for several
years to come. The upgrade to DSI-2 solved most of the measurement problems of the original DSI
and dramatically improves log quality in all formations.

Also, a new GeoFrame processing software, BestDT, has been released which solves most of the
previous processing issues.

We now have a service that can compete very favourably against the X-MAC of Baker Atlas.

However, in SKK, we receive many reports of bad logs being recorded and being delivered to our
clients. These poor logs are seriously tarnishing the image of the DSI and reducing our
competitiveness. In many cases these bad logs have cost us revenue both in rebates and lost
future jobs. Particularly hit is the BCR market where it seems we almost routinely fail to produce a
good log.

The purpose of this document is to highlight to the Operations Managers what they need to enforce
in their locations to get our DSI service back on track.

A. Acquisition software and operating procedures.

The latest version of OP should always be used, currently OP 9-C0-413 and OP9-KP2.
Using the DSI Job Planner is the best method of ensuring that acquisition and processing
parameters are correct, yielding optimal field and final answers. It can be found as part of the Tool
Planner package at:

http://www-maxis.austin.apc.slb.com/intouch/kituser.htm

The current WRM at:

http://manuals.wireline.slb.com/dynaweb/latest-opwrm/dsst/@Generic__BookView

provides up to date details on running the tool but of special note are the following points:

1. The DSI tool should always be run centralised.

Use CME-Z to centralise only up to 10 degrees deviation. In higher deviation wells use stand-
offs, cut to size. Make every possible effort to centralise the tool.
Use enough flex joints (minimum two) to decouple the DSI from excentered tools. Note that
there is only 3 deg of flex per AH-180, providing just 1" of decoupling for one, and only 4.2" for
two combined AH-180s, without separating them by another tool.
 When the hole size is greater than 17.5 inches, it may not be possible to get slowness
measurements when the tool is centralized. Also in large, very highly deviated holes it may be
impossible to centralise the tool.
In these cases only, some excentralisation can be considered, but only as a last resort.
Use the maximum size stand-off, to minimize the excentalisation, and positively excentralise
the tool with a bow spring. Use the dipole normal to the direction of the bow spring.

2. Whilst logging always monitor “Data Copy Indicator (DCIn)” and “ Waveform Copy Indicator
(WCIn)” to avoid repeated frames, and keep below “Max logging speed” as displayed in the
IO monitor.

3. Monitor Head voltage and stay between 250-275 volts,. The upper range will produce less
voltage fluctuation.

4. The preferred acquisition modes are P&S mode, Stoneley mode (to QC the dipole), and BCR
mode. These acquisition modes should always be recommended to our clients to acoustically
characterise the wellbore and the surrounding formation.

5. For BCR mode, a GPIT is required in the same run and whether the tool is run in open or
cased hole.
Centralisation of the tool is mandatory for BCR mode.

6. The GPIT must always be put between non-magnetic housings, this is especially critical in low
deviation wells.

7. For BCR mode, a four-arm caliper is a must (not necessarily on the same run) for hole
ovalisation evaluation. If no dipmeter run is planned, use an EMS in combination with the DSI.

8. In very unconsolidated formations where the compressional slowness may approach or exceed
the mud slowness (a ballpark guideline is 180 us/ft), then an “Expert” mode needs to be run to
accurately measure compressional DT.
Such jobs should be planned with a Geo-Market ID specialist.

9. The DSI tool was designed to optimise recording of formation slowness in open hole and not
through casing. Therefore it is highly recommended to record the DSI whenever possible in
open hole.
If open hole acquisition is not possible, caution should be taken when attempting acquisition of
DTc through casing where:

 The cement to casing bond and the casing to formation bond is less than perfect
 The formation slowness is between 40 and 70 usec/ft or greater than 165 usec/ft
 The cement thickness is greater than 4 inches
 Where road noise is high (small casing diameter)
 There are multiple strings of casing between the tool and the formation

In these environments, it is unlikely that you will recover useable compressional data unless an
“Expert” mode is used. Such jobs should be planned with a Geo-Market ID specialist.

Dipole recordings are further compromised by casing vibrations. Even in the best of conditions,
these casing vibrations may preclude any means of recording valid formation flexural slowness.
The likelihood of measuring formation shear slowness through casing decreases as the
slowness increases. An upper bound for safely measuring formation shear slowness, in casing,
is 300 us/ft

Stoneley waves in casing will not provide any information about the formation.

Prior to any cased hole job, the customer should be informed of these limitations.

SKK are working on improving and simplifying acquisition in CH using high power, low
frequency monopole to minimise the effects of the casing. Guidelines will be issued later this
year. Meanwhile do not propose cased hole DSI unless you have determined with your Geo-
market ID specialist that you have a good chance of success. Contact SKK if needed.
10. You may use DSI-2 tools with a standard SMDR-BD sleeve up to shear slowness of 500 s/ft.
For higher shear slowness, values up to 1200 s/ft, you must use S-DSI, which uses SMDR-
BE. (SMDR-BE = SMDR-BD + slow formation sleeve).
All jobs with shear slowness above 650 s/ft must be planned with the help of your Geomarket
ID specialist and/or SKK..

11. Use of the NG-1 mandrel in jobs requiring TLC (Drill Pipe conveyance) should be discouraged
due to it’s strength limitations.

The chart below is intended to be a general guide to the use of the DSI, in applications where
Dipole Shear evaluations are required. Individual jobs with specific or special circumstances,
should be planned with the help of your Geomarket ID specialist and/or SKK..
.

A. When the formation Shear slownesses are expected to be less than 700 usec/ft , the DSI-2
without the NG-1 mandrel (DSST-BD) should be considered acceptable, and preferred in
drillpipe conveyed logging (TLC).

B. When the formation Shear slownesses are expected to be in excess of 700 usec/ft, the DSI-2
with the NG-1 mandrel (DSST-BE) should be required, and low frequency drive employed.
B. Hardware

1. DSI–2 upgrade

The upgrade to DSI-2 is mandatory and all tools should be modified by Q1 next year. It is not
just an upgrade for slow formations.

The DSI-2 modification of the SMDR is complicated and should only be performed by those sonde
men trained by SKK. A school was held and each Area has trained personnel . Another school will
be held if necessary. Alternatively, if you wish SKK to do the modification, they are willing to do so
for a nominal fee.

Before the modification of the SMDR is performed a receiver check must be made in casing and
the results sent to SKK for analysis. Do not modify a tool until SKK approves its modification based
on the results of the analysis. Details of how to perform this check are given in the DSST WRM
section at:

http://manuals.wireline.slb.com/dynaweb/latest-opwrm/dssttp/@Generic__BookView

ASCII and PDS results files of Electronics Sum Check, Electronics Diff Check, and Receiver Check
are required and should be sent to Keiki Shibuya in SKK (shibuya@fuchinobe.skk.slb.com).

The table below shows the status of SMDR upgrade kits versus number of tools by area.

ASA ECA MEA NSA TOTAL

Kits planned for 1999 30 22 8 50 110
Kits ordered 30 22 8 35 95
Number of tools 34 40 28 58 160
% Planned kits /tools 88% 55% 29% 86% 69%

Those areas that do not have sufficient kits on order need to act now if they are to complete the
upgrades by Q1 next year.

2. Routine Checks

The most common causes of poor or useless waveforms are:

 Eccentralised tool.
 Mismatched receivers/transmitters.
 Mis-oriented or wrongly polarised configurations.
 Poorly maintained tools, particularly SMDR.

a. Receiver Matching

Opposite receivers signals are SUMMed and DIFFerenced downhole, prior to digitisation and
recorded as the raw data. Any mismatch, in any receiver pair, will affect the data quality. Any
mismatching cannot be “processed out”.

A poor contact or bad receiver may cause this mismatch. The receiver matching test is part of the
test phase of the DSI and should be performed after maintenance and on a monthly basis:

http://manuals.wireline.slb.com/dynaweb/latest-opwrm/dssttp/@Generic__BookView
b. Receiver Orientation Check

There were several jobs acquired with the SMDR receiver array rotated 90 or 180 degrees within
its housing. This was caused by wrong assembly of the SMDR or SSIJ sondes after maintenance.

Such errors are surprisingly common and are severely detrimental to log quality to the point of
giving wrong answers. This could prove very expensive to our clients and potentially to us.

All SMDRs must be checked for rotated and/or wrongly wired receivers ASAP and from then
on every time a sonde is Q-checked or disassembled.

The instructions on how to do a receiver orientation check can be found at:

http://www.fuchinobe.skk.slb.com/intouch/products/dsst/smdrorientationcheck.pdf

c. Transmitter Polarity and Matching Checks

The matching of transmitter power and polarity is of almost the same importance as receiver
matching; mismatched transmitters are likely to result in a useless log. The SMDX Maintenance
Manual details how to check the power and polarity of the transmitters.

d. Routine Maintenance

A recent, detailed failure analysis of the DSI shows that apart from the problems above we have a
large number of failures due to lack of basic routine maintenance. For example, the large number
of crushed receiver cans and bellows, crushed monopole cans/ceramics, and dipole transmitter
units suggests we do not routinely top off the oil, or burp the tool to eliminate trapped gas.

Following the prescribed maintenance schedules will greatly reduce the risk such failures and avoid
the heavy cost of replacing these parts.

C. GeoFrame Processing

SKK has developed and recently commercialised a vastly improved processing chain on
GeoFrame called BestDT-1. Ultimately the final log quality is only as good as the weakest
component in the chain, therefore BestDT is a must.
Please inform your GeoQuest counterpart that it is a requirement to use BestDT to process all DSI
logs.

BestDT replaces STC processing on GeoFrame for DSI, SDT, SSLT, and ISONIC; it is currently a
GeoFrame product only. An OP version is planned for Q1-2000.

SKK-Technique