IPTC 14475

Detection of Cross-flows Behind Casing before Perforations, and

Cement Isolation Diagnosis Based on Temperature Analysis
(TUNU field, Indonesia)
Sébastien Perrier, TOTAL

Copyright 2011, International Petroleum Technology Conference

This paper was prepared for presentation at the International Petroleum Technology Conference held in Bangkok, Thailand, 7–9 February 2012.

This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s).
Contents of the paper, as presented, have not been reviewed by the International Petroleum Technology Conference and a re subject to correction by the
author(s). The material, as presented, does not necessarily reflect any position of the International Petroleum Technology Conference, its officers, or members.
Papers presented at IPTC are subject to publication review by Sponsor Society Committees of IPTC. Electronic reproduction, distribution, or storage of any part of
this paper for commercial purposes without the written consent of the International Petroleum Technology Conference 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 where and
by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435

Abstract
In multilayered depleted gas fields, cement isolation is essential to avoid water or gas cross-flows
behind casing, and possible subsequent detrimental effects after perforation, which usually
require long and expensive remedial actions.
In the Tunu Field, a ma ture gas field operated by Total producing 200.000boepd in Eastern
Borneo (Indonesia), a new method has been developed to detect these cross-flows prior to
perforation.

Although it shares similar interpretation concepts as traditional temperature surveys (analysis of

deviations to geothermal gradient), originality of the method lie s in the fact that it u ses logs
acquired before initial perforations on the well. Unlike production logging methods, temperature
profiles are thus exclusively affected by fluid movements behind casing, allowing optimal
resolution.

This timing of logging, provided that it respe cts a minimum stabilization time after cementation
job, is validated by a survey on more than 300 recent Tunu wells: a/ we could not identify in Tunu
any proven water channeling on depleted reservoirs without anomalies being already present
before perforations. b/ temperature anomaly patterns can be associated to different levels of risks
of water channeling. In particular, “heating” anomalies indicate the highest risk of water
channeling, while risk associated to o ther patterns can vary depending on reservoir stacking
configuration and anomaly patterns.

This timing of logging is ideal for reservoir engineers to adapt perforation strategy and anticipate
on eventual remedial jobs before any damage is done to other reservoirs.

Advantages over conventional bond logs interpretations at reservoir scale are comme nted, as
well as the improvements in water monitoring interpretations that this method can bring.
2 IPTC 14475

A. Introduction

A1. Introduction
Tunu is a multilayer field, with thousands of gas-bearing sand reservoirs interlocked with water
reservoirs, spread over a developed area of 600 km2, and up to 4000m of vertical gross
thickness.

If we exclude the recent development of th e most shallow sands, production on the main thick
Miocene deltaic series of the field “main zone” (between 2500 and 4700 mTVD) started to decline
in 2010, after several years above 1.2 Bcf/d of gas production (200.00boepd), and 8Tcf produced.
After an initial phase of production, more than 450 wells have been drilled in several consecutive
phases of infill drilling in the Tunu Main Zone, with current smaller spacing equal to 560m. These
infill wells target both smaller reservoirs at virgin pressure (not intercepted by previous wells) and
re-access to depleted to very depleted reservoirs for additional recovery.

In such a context, pressure differentials between the different gas layers crossed by the infill wells
(typically vertically distant by 2 to 20 meters) can be in the range of a few hundred psi to 5000 psi.
Should the cement bond not be perfectly sealing locally, this creates a favorable ground to cross-
flows behind casing (fig.1).

xx

xx

4550 psi

xx
10m

Pxxx 1300 psi

Pxx1

xx

25m

4600 psi

xx

Water
Gas

Example of high pressure differential
(around Pxxx)
Fig. 1: Typical reservoir stacking and high pressure differential in Tunu
IPTC 14475 3

A2. Tunu perforation strategy and impact of the presented technique

Infill wells in Tu nu are completed as tubingless wells, with perforations performed regularly as
light well interventions by swamp wireline barges. T his provides flexibility to access all kinds of
reservoirs, no matter their thickness, depletion level, etc.

Sequence of reservoir perforation is decided by reservoir engineers and mon itoring teams,
usually following a “bottom-up” progression, keeping generally open other flowing reservoirs: we
add up perforations on the well. In this context, any perforation of reservoirs facing severe water
channelling, can be detrimental to open reservoirs, damaging them irreversibly: main objective of
the technique developed in this paper is to prevent this risk.

Fig. 2: Typical perforation sequence. LWO# indicates the sequence.

Casing patches can be used to isolate perforations if necessary (here LWO-03)

Indeed, anticipating the hazar d of de trimental perforations due to channelling is a pr eventive

method to protect connected reserves, to ensure adequate production delivery of the field, as well
as to decrease costs associated to detrimental perforations or water shut off.

Other advantages associated to the implementation of the technique will be commente d in

section C.

B. Presentation of the technique, and results

B1. Why do cross-flows behind casing matter?

As Smolen and Smolen state [3]: “The main purpose of cement over the production interval is to
provide isolation between neighboring zones (…) Failure of isolation can cause a myriad of
problems such as water production, depletion of gas drive mechanism, loss of production to
neighboring zones (…) “.

By cross-flow behind casing (also called channeling behind casing or channeling), we refer to a
lack of is olation between zones: some gas or water is by-passing the cement in the ann ulus
between formation and outer casing, and is circulating from highest pressure reservoirs into more
depleted reservoirs in their vicinity.

Depending on what kind of reservoirs the channelling connects, we may have:

• water cross-flows, from water reservoirs into depleted gas reservoirs
• gas to gas cross-flows between gas reservoirs
4 IPTC 14475

• a combination of gas and water flows, if several reservoirs are connected to the channelling.

Cross flows involving water are obviously the most detrimental in gas fields. Their consequences
can include:
‐ before their perforation: damage of reservoirs where water is injected, by water blocking
‐ after perforation: severe damage to o pen reservoirs invaded by water, negative impact on lift
performance due to an increase in Water Gas Ratio, or total loss of er uptivity of th e open
reservoirs.

Consequences of gas-to-gas cross-flows are in general less severe than water-to-gas.. However,
these cross-flows can generate sand production or even sudden formation local collapse, if high
velocities in the “small” channels start to “erode” large volumes of sand or shale (see fig.10).

B2. Origin of the technique

Temperature logs recorded in POOH, at low velocity ( 10m.min-1), were available for most infill
wells drilled since 2004. They were all re corded in tandem with conventional CBL tool string,
using high resolution gauges. These logs were performed between 10 to 160 days after cement
job, before any other intervention on the well (perforation or unload).

A review of a few cases of water channellings proven after perforation led us to observe that all
had in com mon the fact that temperature anomalies were present on these logs around the
affected reservoirs (see fig.3, and fig.5 to fig.8).

As logs used are the very first log performed post cement job, before any intervention on the well,
we can conclude that all the channellings identified after perforation were in fact initiated before
the production of the well starts, and that we could have avoided them by not perforating the
affected reservoirs.

This strong association between anomalies and channellings was later confirmed by a systematic
review of 250 w ells drilled between 2004 and 2010. Detailed results about the association
between temperature anomalies and channellings issues are presented in section B5. Mor e than
60 wells have been drilled since then, on which we identified and in most cases did not perforate
the most risky anomalies.
IPTC 14475 5

117 oC 118 119 120 121 122 123

Interval (m)

100

Fig 3: Example of a temperature anomaly observed on log performed after well completion. Anomalic
pressure and water production after perforation proved the existence of a channelling.

Definition or diagnosis of “ proven” channelling post perforation was based upon one of the two
following criteria.
‐ Anomalic BHSIP criteria
When perforating a depleted gas reservoir A, if we produce water and observe BHSIP
(bottom shut in pressure) significant superior to reservoir pressure from gas reservoir A, we
can prove that the perforation is connected to a water reservoir B through a channel behind
casing.
‐ Production Logging criteria
Once several reservoirs are open with different pressure regimes, BHSIP criterion is almost
impossible to use. In these cases, channelling identification depends on Production Logging
post detrimental perforation.

B3. Temperature analysis key concepts

It is widely known in the industry that “temperature profiles have been used extensively for
problem diagnostics related to downhole leaks and cross-flows behind casing”. [1]

These applications are based on 2 concepts, well known in production logging:

• Joule-Thompson thermodynamical effect
a) For gas: Gas expansion is associated to a significant drop in te mperature
(“cooling”) in adiabatic conditions (conditions typically me t in constant volumes,
with gas flow)
b) For water: if produced between 2 medias with a significant pressure differential,
water “expansion” generates a temperature effect. Typically, at 130oC, with 100b
of pressure differential, water will get warmer by +1oC (“heating”) [2]
• Heat exchange between formation/tubing and fluids.
a) Good heat capacity of w ater allows it to bring calories from warm zones to
“cooler” zones.
b) If moving quickly from a cool zone to a “warmer” zone, water may cool the tubing
on its way.

NB: The amplitude of the Joule-Thompson effects for water observed in Tunu usually match well
6 IPTC 14475

the thermodynamical values [2], when we know the pressure differentials: they ranged typically
from 0.5oC to1oC, with extreme values close to 3oC.

B4. Temperature anomalies classification

We observed a limited number of typical patterns of anomalies, although some combinations of

anomalies are possible.

Anomalies can be grouped into 2 main categories (fig.4):

‐ the heatings: temperature is locally higher than what it should be, if we follow the geothermal
gradient
‐ the coolings: temperature is locally lower that geothermal gradient

This last category must be split into two very distinct sub-categories:
‐ The “true coolings”, associated to the circulation and expansion of gas by JT effect.
‐ The water-cooled down-flow channellings (that we will call “ flat down”), that are associated
to cold water flowing down into a depleted reservoir located deeper, and co oling slightly its
path on its way.

Fig.4: Temperature anomaly classification, and their interpretation

It is essential to distinguish “ flat down” from coolings, as th eir interpretation is associated to

circulation of water, not of gas (this makes them potentially more harmf ul). An inter pretation is
indicated in figure 4.

No need to mention that interpretation of temperature anomaly does not stop with anomaly facies:
reservoir configurations and known pressures must be considered to confirm the consistency of
the interpretation.
IPTC 14475 7

B5. Results: Association between anomalies (before perforation) and channellings issues.

0. No anomaly around a depleted reservoir: no risk

Before commenting on anomalies, it is essential to precise that we could not find any p roven
channelling on depleted reservoirs without any anomaly on temperature log.
Otherwise said, for reservoirs already depleted, meaning that some pressure differential already
existed by definition, the flaw in hydraulic seal proven after perforation was identifiable before
perforation in all cases1.

In the case of virgin r eservoirs, 3 cases of channelling without any clear anomaly before
perforation could be proven using PLT. These cases can be explained by the fact that in the case
of virgin reservoirs, the differential of pressure is created by the perforation.

1. Heatings: High risk

The risk of channelling when perforating heatings is in the range of 90 to 95% in Tunu. We
indeed found mo re than 30 cases of perforated anomalies, and got only 3 without water
channelling.
As mentioned earlier, channelling resulted in water problems ranging from severe water-gas
ratios increase to total loss of gas eruptivity of the well (most cases).

Among the 3 exceptions mentioned above, one was likely associated to data artifact (heating in a
gas reservoir area), which can a lways be considered as a p ossible outcome of logs without
repeat sections, although this case was unique in the large set of data reviewed.

2. Flat down: High risk

Risk associated to “flat-down” is sligh tly more difficult to eva luate, as they are scarcer than
other kind of anomalies in Tunu. However, 5-6 out of 10 cases reviewed led to proven or likely
water channelling, with a few other cas es of no flow not documented eno ugh to be conclusive.
We consider that perforating “flat-down” presents a high risk of leading to massive water
production, in line with interpretation of the anomaly (cool water circulating downwards).

3. Coolings: low risk in general

We found a few more than 30 proven channellings among approximately 300 perforations
associated to cooling anomalies. Risk at individual reservoir scale is then low, between 10
and 15%.

The fact that coolings do not systematically lead to major problems can be explained by two main
reasons:
• A cooling can be limited to a gas-to-gas cross-flow. Fig.10 describes an example where a
significant gas flow (3-4mmscfd2) was proven through channelling, without any water
problem (but with sand problem).
• The link between magnitude of the cement flaw and amplitude of temperature anomaly is
not direct. Amplitude of Joule-Thompson effect depends essentially on pressure

We cannot exclude having missed some other channellings appearing late as a result of production, as
1

diagnosis by PLT is now always easy.

In surface conditions. In downhole conditions, we should consider a volume factor of ~90v/v in this case.
2
8 IPTC 14475

differential, and little on flow rate. Small leaks in a m icro-annulus of a few microns can
generate a cooling, while they can limit water flow to undetectable levels.

The reason why coolings can however lead to major channellings can be summarized in 2 points
also:
• Pressure differential between tubing and formation after perforation is usually superior to
pressure differential between reservoirs. Additional stress induced by perforation on a
“weak” cement area can open the leak.
• The amplitude of coolings, typically equal to several Celsius degrees, can hide the
temperature effect of any anomaly associated to water, especially if the cooling is wide or
complex (fig.5 rig ht). As a re ference, magnitude of heating or flat-downs is seldom
superior to 1-2oC.

4. Heatings + Coolings: High risk

There is a risk th at the magnitude of coolings can hide heatings, but we also fou nd cases of
hybrid coolings+heatings, presented in figure 7: heatings are found on the side of large
coolings. The risk associated to these cases of cooling+heating is very similar to that of
heatings mentioned above.

B6. Anomaly examples

To summarize, figures 5 to 8 sho w real examples of an omalies observed on the field, with
estimated risk based on Tunu Field experience. All these cases resulted in a proven channelling,
except case on right-hand section of fig.5.

The GammaRay log is used for basic identification of sands, while colors denote the position of
gas or water reservoirs. Although geothermal gradient is n ot mentioned, we ensured when
preparing the plots that geothermal temperature trend was close to the trend suggested visually.

These plots (fig.5 to 8, & fig.14) can be used to illustrate the answer to 3 frequently-asked-
questions about stability, validity of the data, and position of anomalies:

• In the case of Tunu, this delay of minimum 10 days ens ures perfect stabilization of
temperature on geothermal gradient. Outside of the sections where we are zoomed in,
sections without anomalies are perfectly regular after 10 days (see fig.14)

• Amplitudes and patterns of coolings anomalies, flat-downs, or heatings with Joule Thompson
effect profiles cannot be explained by tool velocity or cable tension irregularities in POOH3.
Such logging artifacts could in some cases create anomalies similar to heatings without Joule
Thomson effect, so these logging parameters should be reviewed in case of doubt.

• Most anomalies stand in front of the m ost depleted reservoir, the one where the maximal
expansion occurs. This last point is often discussed, but our study of more than 400 cooling
anomalies with pressure data confirms this strong tendency. In par ticular, the “peak” of
cooling is located in 99% of cases in front of the most depleted reservoirs, with a handful of
cases only in front of virgin reservoirs.

As they are recorded with most logs, these logs can be reviewed in quality check.
3
IPTC 14475 9

Fig 5: Two examples of coolings. The well on the left suffered severe channelling from the zone with cooling,
the well on the right did not face water channelling from any of the perforated zones.

Fig.6: two examples of “flat-down”.

10 IPTC 14475

“Heating” type
Risk of water channelling: Very high (90-95%)
Only a couple of heatings were not detrimental

129 oC 130 131 132 133 134 129 oC 130 131

JT
Interval (m)

Interval (m)

120
GR T
150
117 oC 118 119 120 121 122 123 oC 136 137 138

JT
Interval (m)
Interval (m)

80
100
121 oC 122 123 oC 116 117 118
Interval (m)
Interval (m)

70 50

Fig.7: Several examples of heatings, with or without clear Joule Thomson effect
IPTC 14475 11

“Cooling+Heating” anomalies
Risk of water channelling: idem as heating (Very high)
Coolings often hide heatings located nearby. Risk of cooling + heating
should be considered as that of “heatings”
140 oC 141 142 143 144 145
Interval (m)

200

116 oC 117 118 119 120

Interval (m)

100

114 oC 115 116 117 118 119 120 121 122

Interval (m)

JT ?

200

Fig.8: Several examples of cooling + heating. Bottom example shows the risk that coolings makes the
heating not detectable (water channelling was proven on this case)
12 IPTC 14475

C. Results and advantages associated to the technique

C1. Main advantage of the technique: its timing.

Typical completion of Tunu wells is tubingless. All well interventions, including CBL logging and
perforations, are performed by swamp electric-line barges, which can access the well generally
10 days after end of drilling. High-resolution temperature profiles used in this paper have then
been recorded 10 days minimum after cement job, most of them in tandem with CBL tool (POOH,
10m.min-1), before any unload of completion fluid or perforation.

In this context:
‐ completion fluid has not been perturbated since e nd of comp letion job, ensuring a g ood
stabilization of the completion fluid temperature
‐ Logs reflect exclusively events occurring outside the casing, as casing is not perforated yet.

Availability of data before perforation is the key advantage of the technique, as it allows to
include a p reventive diagnosis of risk of channelling in the sequence of perforation of the
well.

Apart from other advantages detailed in next sections, this timing allows also
• To perform complementary logging in perfect conditions (eg: static o r water-flow-log type
Neutron Logging, or a second Cement Bond Log)
• To anticipate on possible remedial cement squeeze jobs. These jobs are usually complex,
and anticipating them can be es sential to ens ure their feasibility, economy, and success. As
remedial jobs usually involve killing the well, anticipating them can also avoid irreversible damage
to open reservoirs.

C2. Advantages of the technique for reservoir monitoring and management

The quality of geothermal profiles obtained before perforation of the well is difficult to obtain from
production logging after production due to increased complexity and multiple disturbances in the
well.

This topic will not be addressed in this paper, but independently of the delay between cement job
and temperature acquisition (between 10 and 160 days in our sample), temperature from logs
before perforation is c oherent in the m aximum range of +/-0,5oF at any given sub sea vertical
depth with that of neighbor wells.
It also follows a clear geothermal trend throughout the structure, and confirms previous
geothermal surveys performed after extra long stabilization.

These geothermal profiles at well scale are more than a by-product of the method, as they can be
used in production logging to detect cross-flows behind casing or water entries with new
generation of quantitative interpretation of temperature profiles (see Petricola & al. [1]).

Apart from diagnosis pre-perforation, identification of anomalies (“heatings” in particular) can

prove to be essential for a good diagnosis of water origin after perforation.

Several examples from Tunu’s experience can illustrate these advantages:

IPTC 14475 13

a. “Heatings” found a po steriori on water perforations can bring an a posteriori explanation to

serious geological inconsistencies in local water-gas-contact definition associated to the
integration of a post-perforation diagnosis of water production.

b. The choice of a ppropriated technique and success of water shut off ( water isolation),
depends much on diagnosis of water origin:
• a chemical water shut-off may fail if water does not come for the reservoir, as the
chemical used is not designed to blo ck channels behind casing but just to degrade
specific ranges of po rosities. A cement squeeze should be preferred if channelling is
diagnosed.
• In case of mechanical water shut off by casing patch on a channelling, we may observe
the water changing its path and “reappearing” in other nearby perforation. An example of
such case is sh own in fig.9. A good diagnosis would have suggested considering
isolating either all perforations or none, especially if isolation creates a restriction to
further isolation below.

117 oC 118 119 120 121 122 123

Interval (m)

100

Fig.9: All reservoirs in red were perforated. Production logging indicated that little gas + massive water
was coming from the bottom perforation. Right after the isolation by casing patch, a second PLT
4
showed that the same amount of water was now produced from top perforation. .

c. Sand production can also be associated to channelling, as illus trated in fig.10. D iagnosing
this possibility can also inf luence the strategy to increase well production, or the c hoice of
most adapted sand control technique, in particular exclude chemical sand consolidation
technique (resin).

A second casing patch could be set in this case with success. As casing patch cannot be set below a
4

casing patch in Tunu, in a situation where water would have reappeared below, we would have been unable
to isolate the water sources…
14 IPTC 14475

Fig.10 : Example of 2 reservoirs connected by a proven gas-to-gas channelling. Sand production

from bottom reservoir was dramatically decreased when decision was taken to perforate top
reservoir.

C3. Advantages of the technique over conventional bond logs

The severity of a cross-flow can depend on sever al characteristics of the poor ce ment sealing,
e.g.:
• Origin of the flaw: cement shrinkage, non homogeneous cement placement, poor
centralization of the casing…
• Geometry (e.g. micro annulus or channel)
• Section size, vertical extension (connecting two or more reservoirs.

Cement Bond Logs (CBL) and similar acoustic bond logs aim at characterizing cement placement
and some of the physical characteristics of the cement bond.

They are usually the only tools used to evaluate cement quality before perforation, by associating
the attenuation and waveform of sonic signal traveling along the casing or reflected by formation
to cement presence and the mechanical quality of the connection cement-casing.

Regarding the detection of leaks, as we will comment further, they are usually inconclusive (or
over-interpreted). As Smolen and Smolen state, CBL sho uld not be considered as an isolation
log: “Acoustic cement bond logs do not measure hydraulic seal” [3].

Our extensive review of channellings in Tunu confirmed this statement: many “poor” CBL have no
channeling problems despite a fa vorable ground in te rm of pressure differentials and distance
between sands. On the other side, several apparently good CBL proved to miss s evere
channellings detectable with temperature anomalies.

Fig.11 to fig.13 illustrates three examples show ing that CBL is unable to evaluate qualit y of
hydraulic seal, and that Temperature logs can advantageously complete cement diagnosis.
• On Fig.11, an “undetectable” leak from CBL, is detected by Temperature log.
• On Fig.12, a locally low CBL amplitude should not be over-interpreted as a local barrier.
IPTC 14475 15

• On Fig.13, we illustrate that when CBL indicates a r isk of “bad cement everywhere”, a
qualitative interpretation of temperature log can bring a better insight than conventional bond
logs on the risk of channelling.

Fig.11: Massive water channelling immediately upon perforation, 12 days after cement job, and 2 days after
CBL and Temperature logging. Responses from CBL and VDL are considered excellent.

Fig.12: “Locally good” CBL amplitudes are often presented as “locally sealing barriers”. Temperature
anomaly indicates a risk difficult to identify from CBL-VDL.
16 IPTC 14475

Fig.13: A more complex example. From CBL-VDL, cement quality seems very poor, especially on top
interval. After a proven channelling on bottom interval, we could have expected from CBL-VDL that second
perforation would also fail as cement seems even worse there. Decision was taken to perforate the reservoir
as the Temp. anomaly was indicating “only” a cooling in an area with several water reservoirs. The
perforation resulted in moderate gas flow, without significant water.

We will not comment on multi-directional cement bond logging tools, as we did not used them at
large scale, but believe they face the same limit ations: leaks may be v ery local and relatively
small. We believe they can fail th e same way a visual inspection frequently misses a le ak in a
bicycle wheel inner tube.

A second major inconvenient of CBL logs is that their resolution is limited to single casing
sections.
Our study shows that in the case of Tun u, amplitude, pattern, frequency of temperature
anomalies; and their statistical association to proven channelling are not impacted by the double
casing.

Several anomalies commented in fig.5 to 8 were in double casing sections. Fig.14 illustrates also
that temperature profiles are perfectly continuous independently of n umber of c asings, and that
anomaly interpretation is possible on both sections.
IPTC 14475 17

5 typical profiles
Plain sections: double casing (3,5" or 4.5" in a 7")
 oC

Dashed sections: single casing (3.5" or 4.5")
150
145
140
135
130
125
120
115 Well 1: many anomalies
Well 2: anomalies only on double casing section
110
Well 3&5: no anomalies, whatever the section
105 Well 4: anomalies in bottom section essentially
100
0 500 1000 1500 2000 2500 3000
Depth, for plot only 
(not suitable for geothermal gradients)
Fig 14: Examples of 5 profiles. Depth are adjusted so that profiles do not overlap. Numerous examples in
fig.5 to 8 were in double casing sections.

To summarize, temperature anomalies can detect flaws in hydraulic seal that CBL cannot,
and can be used in double section of casing where Conventional Bond Logs are not
interpretable.

However, if we consider that many coolings can be associated to “harmless” micro-annulus, the
limitations in distinguishing micro-annulus from more serious flaws in cement bond are a common
drawback for both conventional bond logs and temperature anomaly interpretation technique.

We recommend to run in tandem high resolution temperature sensors and bond logs.

C4. Practical deployment of the technique

A simple extraction of Temperature log is performed systematically in standard comma-separated

format file, including GR, Depth, and Temperature.
A spreadsheet can import the corresponding text file and plot the temperature versus depth, with
respective location of gas reservoirs, and provide zooming options in a few keyboard shortcuts.

Two additional features can be useful, although not essential:

• The derivative of Temperature versus depth, to detect, to filter or to count the anomalies. For
example, the derivative is a good way to distinguish to distinguish “flat downs” from coolings.
• The conversion of measured depths in vertical depth. It is compulsory to define geothermal
gradient, but not essential for anomaly detection.
18 IPTC 14475

fig 15: A view of the spreadsheet used to manage the visualization and detection of anomalies.

Main recommendations regarding interpretation include:

• Reviewing always anomalies at different scales (zoom in - zoom out)
• Adapting the anomaly patterns to ea ch field, as frequency and patterns of ea ch kind of
anomalies can differ from one field to another.

D. Conclusions

Purpose of this paper is to demonstrate that:

‐ In Tunu, the vast majority of channellings on depleted reservoirs is initiated, and thus can b e
detected, before reservoir perforation or any production from the well5.
‐ Risk of channelling after perforation can be interpreted from 4 different patterns.
‐ In practice, temperature anomaly analysis gets closer to the definition of an “isolation log” than
conventional bond loggings (CBL-VDL).

Good reservoir management is about quality of diagnosis. The technique:

1. Is an efficient preventive method of diagnosis of risk of detrimental channellings:

• Association between heatings and problems of channelling is high enough (90%+) to
balance in favor of skipping their perforation or performing remedial jobs.
• Although the technique is less predictive with coolings, it can identify risk w here
conventional bond logs would exclude it.
2. Benefits from a n ideal timing to anticipate on remedial actions strategies, including
additional logging.
3. Provides data for a be tter diagnosis of water origin, with benefits for Water Shut Off
success, for fluid contact monitoring, or interpretation of sand production.
4. Provides high quality geothermal information, useful in production logging [1]
5. Is simple to deploy and interpret
6. Is cheap, especially when high temperature gauges can be combined with other tools.

Bibliography:

5
Channelings appearing long after beginning of production cannot be excluded.
IPTC 14475 19

‐ [1] SPE 25630 Multiwell Application of downhole temperature profiles for c rossflow analysis,
Mario Petricola and Mohamed Watfa
‐ [2] Thermodynamical tables for water, V.F.Ochkov, Moscow Power Engineering Institute
(Technical University) available at http://twt.mpei.ac.ru/ochkov/WSPHB/engindex.html
‐ [3] Cased hole and production log evaluation, by James J. Smolen and Jim J. Smolen,
Pennwell Publishing.