SPE 141770

Application of Production logging and Noise logging for Well Leak Detection,
Case Study in Persian Wells
Reza Salehi-Moorkani, SPE, NISOC, Maryam Mohamadipour, NISOC
Copyright 2011, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Projects and Facilities Challenges Conference at METS held in Doha, Qatar, 1316 February 2011.

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

There are a variety of methods available to detect the source of downhole leak points include Production logs, down-hole cameras,
thermal decay logs, and noise logs. However, many of these methods are not so much effective specially in case of very small
leaks and can result in gathering data that requires a significant amount of logging assurance to interpret. Continuous logging has
proven to be significantly more efficient in locating anomalies than static logging techniques commonly used in noise logging
operations. Identification of the source of pressure communication between well tubing-casing and casing-casing annuli presents a
huge challenge to petroleum engineers.
This paper presents a brief description of two technologies for leak detection and the findings from the field trial. In this well two
type of logging techniques are used: Production logging and Noise logging. These tools were run in an oil well with a history of
leaking oil 600 meters away at surface. Analytical work prior to the trial had established that the wellhead seals and burning pit
flow line were not the source of leak. The paper describes the interpretation of the results that indicated a possible casing leak.
The results show that in Noise logging the source of leak is clearer than production logging. In some cases two logs have read the
same values.

Introduction

Field and Reservoir Characteristics
Field K is one of the oil reservoirs in Dezful Embayment, north of Persian Gulf which is discovered in 1970 when the presence of
two oil and gas producing formations were established: the Asmari reservoir which consists of undersaturated oil and Bangestan
which is a gas reservoir. Subsurface structure of Asmari reservoir has been characterized with seismic operation in 1972. This field
is surrounded by two thrust faults. In general, Asmari reservoir is divided to three sections: upper, middle and lower.
Upper Asmari section consists of zone 1, zone 2 and zone 3. Middle Asmari section consists of zone 4 and zone 5. Zone 1 and
zone 2 mostly consist of lime stone with fine grains. Zone 3 has dense lime stone formations and shale and gypsum in some layers.
Zone 4 consists of fine grain limestone with very low porosity and this zone is not a production zone. The last zone is zone 5
which mostly consist of marly lime stones with some shale layers. At the end of this zone which is also end of Asmari formation, a
thick gypsum layers exist.[1]
Well history
In J une 2006 the well A was drilled as 8.375 openhole and completed in Nov 2006. In winter of 2007 the well starts producing
with flow rate of 500 bpd. As a result of well head pressure drop, its oil production rate decreased to 100 bpd. Fig.1 shows a
schematic of this well along with its geological formations. In March 2009 a surface leak was detected in the vicinity of this well.
This leak could be seen around 600 meters away from the well location and it was noted to have an effect on the well flowing
pressure. Due to local observations of oil on the rocks and since this well was the closest well to this area, it is considered to be the
2 SPE 141770
source of surface leak. Firstly, all surface facilities from flow line and burning pit line were checked to find the leak source.
However, since the pressure measurements behind casings didnt show any variations, it was revealed that the leak was not from
surface facilities and it comes from subsurface somewhere inside the well. Therefore, to check this possibility, noise logging and
production logging tool was run from surface to bottom hole with objective of downhole leak detection.


Fig.1- Well schematic with geological formations

Noise logging and Production logging were run to achieve three objectives:
identify thief zones and producing intervals
determine flow behind the casing strings to identify source of surface leak
find any casing leak points inside the well
Initially, to diagnose and solve the problem the noise logging tool was run, in combination with Gamma Ray, Temperature and
Casing Collar Locator (CCL) tools.

Noise Logging Results
Noise logging data results indicate that fluid flow is from the Asmari formation behind the 9 5/8 casing shoe. The flow continues
to go up and moves behind the 13 3/8" casing and finally it is lost at depth 234m. Fig.2 while highlighting a separating interface in
cement quality behind the 20 casing around 600m, displays the gradual increase in flow energies observed while approaching the
surface. This trend terminates to a peak energy at 234m (yellow circle in Fig.2), also expressed by a temperature anomaly around
the same depth. Noise logging interpreter believed that the peak shows a possible outflow fracture in the Mishan formation where
the migrating fluids are channeling beyond the well trajectory. A deeper look into the horizontal stress forces around that depth
and especially in the Mishan formation could confirm the presence of this fracture. Since all outer casings display no recorded
pressures, then it can be concluded that most fluids that are migrating behind pipe, have an outlet at this fracture. It should be noted
that at the end of this paper, noise logging and production logging results will be fully evaluated.
SPE 141770 3

Fig.2- flow energy and temperature anomal y at depth 234 m

A casing leak was identified at 2290.5m, however according to noise logging records; this leak is not playing an active role in the
contribution to the flow behind pipe in this well. Generally the results of noise logging and the reason of leak are shown in Fig.3.


Fig.3- Leakage path from bottom to surface according to noise logging results

Production Logging Results
Flowing Survey
Spinner: Spinner bumpiness indicates multiphase flow and spinner smoothness indicates monophasic flow. As it is shown in
Fig.4, in part A and B spinner readings are noisy. Part A is from 1800 m to bottom and part B is from 690m to surface. From
1780m to 690m spinner starts to be smooth because just oil is flowing through wellbore. Flowing passes indicate an inflow of oil
from 3632m. This oil entry is lifting the water column below the entry, up to 1780m however this column of water falls back and
is not able to reach surface. (Fig.4, part A) Since logging was not done to TD there is no information on source of water. At 690m
pressure falls below the bubble point pressure and gas is released. (Fig.4, part B) Gas shows its effect with decrease in density and
temperature. The overall trend of spinner passes is not showing any traces of production except from openhole.


4 SPE 141770

Fig.4- spinner bumpiness indicate multiphase flow and its smoothness indicate single phase flow

Density: Density is showing 1.18 g/cc from top of cement to 3632m, and then it decreases to 1.1 g/cc and remains the same up to
3617 m. As it is shown in part A of Fig.5, density decreases from 1.1 g/cc to unsteady range of 0.9-1 g/cc up to 1780 m. This
shows mixture of oil and water flowing together while heavy column of water is trying to be lifted. Finally Density drops to
0.75g/cc at 1780 m and remains the same with slight decrease from 350m to surface (part B of Fig.5). There is no indication of
water production to surface and water just circulates at the bottom of hole from 1780 m to final depth.


Fig.5- Fluid density changes inside the well
Temperature: In order to see the changes of temperature more clearly a derivative of temperature has been created. Two sharp
changes are observed in temperature derivative; the first one which is shown in part B of Fig.6 happens at 3632m where it shows
the entry from openhole section. The second one happens at 690m showing release of gas due to reaching the bubble point
pressure (part B of Fig.6). Hidden heat of gas vaporization cools the flow below the bubble point pressure.

SPE 141770 5

Fig.6- Temperature deri vati ve changes due to fluid changes

After inspection of the various flowing passes curves, it has been identified all the log responses show normal wellbore behavior
and there is no sign of leak.

Shut-in Survey
Spinner: Spinner passes in shut-in condition show a small decrease at 2290m. Shut-in stations in Fig.7 are showing some spinner
rotation from openhole up to 2290m and above 2290m, station spinners drop to zero value. There is certainly an inflow or outflow
occurring at this depth but there is no sign of this inflow/outflow extending up the annulus to surface. Therefore, this can be a
casing damage. Other step changes in spinner are not indication of any flow or leak.


Fig.7- spinner passes in shut-in condition

Density: Density is showing a slow increase with decrease in depth from bottom to surface. The reason for slow increase in
density is temperature fall due to well shut-in, which results in higher density oil. The increase in density on 500m of top of the log
is due to escape of flowing gas while shut-in, which results in denser oil as well. Fig.8 shows that density has a sharp change at
2290m, which could be possibly a reason for denser fluid falling back and mixture of water-oil entering the wellbore. The density
of oil and water column in shut-in condition is less than flowing condition which was at depth 1780 m.

6 SPE 141770

Fig.8- sharp change in fluid density at depth 2290 m

Temperature: There are so many temperature deflections from bottom to top. Fig.9 shows that at 2290m a temperature change is
correlating with the suspicious inflow/outflow zone. Except in depth 2290 m other temperature variations are probably come from
changes in geothermal gradient according to different layers.


Fig.9- Temperature deri vati ve changes at depth 2290 m

Summary of shut-in passes shows that 2290m is suspicious for an inflow or outflow zone. There is no strong reason to judge either
this zone is producing fluid or taking fluid. With low confidence it is believed that this zone could be an inflow zone due to
behavior of density. But about the amount of fluid being produced at this depth as a leak, there is no information since the amount
of leak, if any, is below the threshold of spinner and so unable to be calculated. Apart from 2290 m there is no other zones detected
as casing damage or leak.

Discussion and Results
The results of production logging and noise logging in accompany with surface leak make oil company to perform work over
operation on this well. The job plan was to plug the suspicious depth in 9 5/8 casing. Prior to perform this task, a pressure test is
done to ensure that there is a leak in the casing.
Both logging methods by low confidence stated that the source of leak maybe from the point 2290 m at 9 5/8 casing. But noise
logging results showed that leak is from 9 5/8 casing shoe. The advantage of noise logging over production logging is that noise
logging can detect flow behind casing but its limitation is that it cannot determine the type of fluid which is flowing behind casing.
To find the real reason of the leak source, a pressure test has been performed. First of all, open hole section is plugged with
temporary cement to ensure that no flow comes from open hole segment. Then the work over fluid with specific gravity of 65 pcf
is pumped inside the wellbore and the pressure has monitored at choke manifold. The pressure at surface reach to 1000 psi and
after one hour no pressure drop observed and the well was in static condition and no flow observed. This test showed that the result
of noise logging was more accurate about this fact that leak is from 9 5/8 casing shoe.
SPE 141770 7
The pressure, usually measured in psi, at the bottom of the hole. This pressure which applied at depth 2290 m can be calculated
using equation 1:

BHP =MW Depth 0.007 (1)

Where BHP is the bottomhole pressure in pounds per square inch, MW is the mud weight in pounds per cubic feet, Depth is the
true vertical depth in feet, and 0.007 is a conversion factor if these units of measure are used.
According to equation 1, during pressure test the bottom hole pressure at depth 2290 m is as below:

BHP =(65 7511 0.007) +1000 =4417 psi

While the bottom hole pressure during drilling of this high pressure formation at the same depth is higher and as follows:

BHP =127 7511 0.007=6677 psi

This means that if there was any casing damage in that point, the pressure will be increased and we need a higher density mud to
control the well. But because there was no communication between formation and borehole the pressure did not change.
But if there was not the leak, this question comes in mind that what was the source of these logging readings. Careful observations
showed that x-y Caliper log has deviations in interval 2278m to 2280m as shown in Fig.10. Geological graphic well log shows a
thick salt layer exists in this interval and maybe as a result of salt creeping this ovality in casing takes place.


Fig.10- caliper deviations

Oil migration to surface: as the results of noise logging stated, the oil is migrating behind casing through a fracture and reach to
surface, but seismic surveys showed that no fracture or fault exist at this depth. However, geological investigations showed that oil
comes to surface because of the well elevation is higher than leak point and also it comes through a high permeable carrier layer.
Daily leak measurement: At first, there was no information on the amount of leak. But after sealing the 9 5/8 casing shoe by
squeezing cement, to ensure that leak is decreasing the daily leak rate has been recorded. The results show a decrease on amount of
leak as shown in Fig.11. This figure shows that leak rate after a peak in the amount of leak it starts to decrease.
8 SPE 141770

Fig.11- dail y leak rate after work over

Conclusions
This case study has proven that noise logging is a useful option in detecting leaks but it is not sufficient by itself
especially in complicated cases.
The advantage of noise logging is detecting fluid flow through different media and behind pipe, but this technology which
is used in this well cannot identify type of fluid flowing behind casing string.
Production logging can help noise logging to have a better understanding about leak points, however to achieve the best
results, logging findings should be confirmed with reservoir conditions and geological features, drilling report and
seismic survey.

Acknowledgement
The authors thank National Iranian South Oil Company (NISOC) for their help and support during all stages of this project.

Nomenclature
bbl =Barrel
ccl =Casing Collar Locator
ft =Feet
g/cc =gram per cubic centimeter
in =Inch
md =Measured Depth
pcf =pound per cubic feet
psi =Pounds per Square Inch


References
[1] Cedric E. Hull, Happy R. Warman, "Asmari Oil Fields of Iran", AAPG Special Volumes, Volume M 14: Geology of Giant
Petroleum Fields (1970)
[2] R. Salehi-Moorkani, M. Mohamadipour, Well Integrity in Iranian South West fields ", National Iranian South Oil Company,
(Oct, 2010)