You are on page 1of 10

OTC 10945

Reliability analysis of permanent downhole monitoring systems


S.J.C.H.M. van Gisbergen, Shell International Deepwater Services B.V., A.A.H. Vandeweijer, Shell International
Exploration and Production B.V.

Copyright 1999, Offshore Technology Conference


(i) Permanent downhole flowmeters for liquid-only
This paper was prepared for presentation at the 1999 Offshore Technology Conference held in mixtures,
Houston, Texas, 3–6 May 1999.
(ii) PDG systems in ESP-lifted wells that use the ESP
This paper was selected for presentation by the OTC Program Committee following review of power cable for data transmission,
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Offshore Technology Conference and are subject to (iii) The Fibre Optic Well Monitoring (FOWM) system for
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Offshore Technology Conference or its officers. Electronic reproduction, measuring pressure and temperature.
distribution, or storage of any part of this paper for commercial purposes without the written
consent of the Offshore Technology Conference is prohibited. Permission to reproduce in print
(iv) The Tucatran system for cableless communication.
is restricted to an abstract of not more than 300 words; illustrations may not be copied. The (v) The SCRAMS system.
abstract must contain conspicuous acknowledgment of where and by whom the paper was
presented. The installation of electric gaslift valves, all electric inflow
control devices and reservoir monitoring systems are planned
in the near future. Once all these systems are available, the so-
Abstract
called ‘Intelligent Well’ concept, which can be defined as a
Permanent downhole gauge (PDG) systems have been
modular combination of downhole monitoring and control
installed in many oil and gas wells to date. These systems
systems will become reality.
enable improved reservoir management and production
For all these systems reliability is key: Shell’s targets are a
optimisation. Currently, more complex downhole monitoring
90% probability to survive 5 years for monitoring systems and
and control systems are being installed, in what is often called
a 90% probability to survive 10 years for actuators. In this
‘Intelligent Wells’. For all these systems reliability is a critical
study, we have performed an evaluation of the reliability of
aspect.
installed systems, where we have focused on electric pressure
We have evaluated the performance of 952 pressure and
and temperature measurement systems, since sufficient have
temperature monitoring systems installed since 1987. It was
been installed to allow a statistical evaluation.
found that the 5-year survival probability of these systems
The paper contains a description of a PDG system and its
improved from 40% for systems installed in the period 87/88
components, the method of obtaining field data, the statistical
to 75% for the period 91/92. As from 1992 no further
analysis and the final results. Conclusions are then drawn from
improvements have been observed resulting in a 5-year
the results.
survival probability of 69% for the period 1993-1998.
Failure modes have been evaluated and could mostly be
System description
attributed to the downhole gauge and to the cable. Based upon
A schematic of an electrical PDG system for pressure and
these findings, recommendations have been made to achieve
temperature measurement is depicted in Fig. 1.
reliability improvements.
The sensing element is an electronic gauge that measures
the pressure and temperature in the tubing. The gauge is
Introduction
mounted in a mandrel, which forms part of the tubing string.
Permanent downhole gauge (PDG) systems have been
The topside of the gauge is formed by the cablehead. Here,
installed in several hundreds of oil and gas wells so far. These
the gauge is connected to the electrical metal-sheathed cable
systems form an alternative to wireline-conveyed downhole
that runs along the tubing to the hanger.
surveys. In comparison, PDG systems avoid hazardous
The cable is coaxial and consists of an insulated
operations and offer continuous measurements, which enable
monoconductor that is encapsulated in a metal sheath. This
better reservoir management and production optimisation. An
metal sheath forms the return wire and protects the cable from
increasing number are being installed, mainly in wells where
the medium in the annulus. The sheath is in most cases
wireline surveys are expensive (subsea, offshore, remote) or
encapsulated in a thermoplastic to prevent it from damage
where they result in high deferred production (e.g. horizontal
during the installation.
wells). Electric pressure and temperature (p,T) monitoring
Splices in downhole cables are used in case of insufficient
systems are most common. Other systems that have been
cable length or if the cable breaks during installation. It has
installed are:
2 S.J.C.H.M. VAN GISBERGEN, A.A.H. VANDEWEIJER OTC 10945

always been considered as a weak element, and therefore the Results


average number of splices in an installation has been reduced This section contains the results of the reliability analysis.
significantly over the last years. Continuous cable-lengths of First the number of installations is discussed. Subsequently the
10,000 m can now be made and more care is given to the reliability of the total data set is discussed, and the change in
installation. reliability over time is analysed. Finally the failures are
The cable is supported by protectors at every tubing joint. analysed, subdivided into the failure element.
These hold the cable in place and protect it across the Number of installations
couplings, where the cable is most vulnerable to crushing. We have obtained data from a total of 952 p,T
The wellhead outlet differs for a platform/land and a measurement systems that were installed in the period January
subsea installation. In case of a land or platform system, the 1987 until August 1998. In Fig. 2, the number of installations
cable is normally fed through the tubing hanger with is shown for each year, divided into subsea, platform and land
compression fittings at the top and bottom. The cable can be installations. This figure shows a sharp increase in the number
fed out of the tree through a downhole safety-valve line port or of installations up to 1996 with step increases in 1993 and
a flanged outlet. In order to provide a pressure barrier inside 1996. Before 1993, typically 30 systems were installed yearly,
the tree, a bulkhead splice is often made immediately outside whereas from 1993 till 1995 about 100 installations per year
the tree, which terminates in a connector. An instrument cable were done reaching a peak of 192 systems in 1996. After 1996
is used to couple this connector to the surface acquisition unit. the number of installations shows a decrease. For 1998 this is
In case of a subsea installation, a wet-connector is used partly due to the fact that only installations up to August have
between the hanger and the tree (see figure 1). The male part been taken into account.
of the connector is fitted into the hanger on the rig floor. The Figure 2 also shows that the division into platform, land
female part is fitted into the tree. It is connected to the and subsea installations on average remains fairly constant
wellhead outlet via a metal sheathed cable. The wellhead over the years; 39% of all the installations are subsea
outlet is usually a flanged wet connector, which is coupled to installations, approximately 61% are platform installations
the control pod, to a control line in the umbilical or to an whereas only few land installations have been done.
acoustic telemetry system via a diver mateable connector. Reliability
The data from the PDG is usually fed into an IBM- Over the years, substantial developments of PDG systems
compatible PC via an interface unit. Such a system can handle have taken place. This has resulted in a situation that
a multiple of gauges. It can transmit the data to any location effectively every 2-4 years a new generation of systems have
via any chosen communication system, ranging from satellite come on the market. As PDG systems go through the learning
link to transporting floppy disks. curve, one would expect to see a change in reliability over
time. To study this, the survival probability has been
Field data calculated as a function of operational time for 6 different
The data set of the track record was obtained from the four periods of installation: 1987/88, 1989/90, 1991/92, 1993/94,
major suppliers of these systems. Their data were cross- 1995/96 and 1997/98. The results are shown in Figure 3. In
checked with other sources of information, such as articles in this figure also 90% confidence intervals are displayed that
literature1 and information from Operators. In this manner give a 90% probability that the real value of the reliability is
field data of 952 p,T measurement systems installed from within these intervals. The curves for the period 1995/96 and
1987 till August 1998 were obtained. The reports of all failed 1997/98 are terminated after 4 and 2 years, respectively, since
systems have also been collected in order to determine the that is the longest time systems from these periods have been
failing element. installed.
These curves show that between 1987 and 1992,
Method of analysis substantial progress has been made in the reliability of PDG
The systems under study differ by their inherent properties or systems: the 5-year survival probability has improved from
by the external conditions (pressure, temperature, etc.) under 40% for systems installed in 1987/88 to 75% for the
which they are operating. Also other factors, like the oil installation period 1991/92. However, since 1992 no further
company, the field, the well type, the contractor and the improvements have been observed. This has resulted in a 5-
installation date may have influence on the failure behaviour year survival probability of 69% for the period 1993-1998. If
of the system. A physical model, explaining the influence of we compare this to the target 5-year reliability of 90%, it can
these variables on the reliability, is not available. Hence a non- be concluded that the failure rate is still a factor 3 too high.
parametric method (as described in 2) has been used to analyse Failing elements for different generations
the reliability. The results obtained with this method are only We have divided the failures of PDG systems into their
based on the observed track records of the installed systems, failing elements as follows: failure of (1) the downhole cable,
without the inclusion of any physical model and without (2) fixed connections, such as splices and the gauge-cable
making any assumption about the failure behaviour, as is done connection, (3) mateable connectors, (4) the downhole sensor
by parametric models, e.g. a Poisson (constant failure-rate) or and (5) other failures. We have studied the change in
Weibul life time distribution. reliability for each element over time, which is shown in Fig.
OTC 10945 RELIABILITY ANALYSIS OF PERMANENT DOWNHOLE MONITORING SYSTEMS 3

4. This figure depicts the position 1 year after installation and A decrease in reliability with temperature is expected,
shows the percentage of installations still operating and the mainly because of the expected higher failure rate of
percentage that have failed split by failing element for 6 electronics at higher temperatures: as a rule of thumb, the
different periods of installation. It shows that the failures due failure rate of electronic circuitry doubles for each 8°C
to fixed and mateable connections have dropped over the temperature increase. Also other elements show a decrease of
years: in 1987/88 59% of all system failures after 1 year were reliability with increasing temperature.
due to these 2 elements, whereas in 1995/96 and in 1997/98 The influence of bottomhole temperature on reliability is
this has reduced to less than 7%. Therefore, it can be shown in Fig. 7. This figure indeed shows that there is a
concluded that significant improvements have been made in decreasing trend of the 1-year survival probability of systems
the reliability of fixed and mateable connections. Also the installed in the period 1995 till 1998 as a function of
reliability of the cable has shown an improving trend bottomhole temperature.
throughout the years. The gauges however still cause a large Failure of the cable
number of failures. Throughout the years the failure rate of The downhole cable has traditionally been a weak element of
this element has shown an increasing trend. PDG systems. Until 1990, all cables contained splices, which
were a major source of failures, however, splice-free cables in
Current generation lengths of over 10,000 m are now available. Furthermore,
In the remainder of this paper, we will focus on the failures currently cables with an Incoloy 825 sheath are used instead of
that are encountered in the current generation of PDG systems an AISI 316L sheath. This has contributed to improvements in
to arrive at recommendations to prevent them. We have reliability.
somewhat arbitrarily defined the current generation as those Most cable related failures have occurred during or shortly
installed in the last 4 years, i.e. in 1995-1998. In this period in after the installation. Some of these cable failures occurred
total 557 systems for measuring pressure and temperature have when the cable was crushed by the tubing hanger while
been installed. landing the tree. These failures can be prevented by using a
Fig. 5 shows the split-up of failures of the 557 p,T PDG protective cap/centraliser for the cable in the tubing section
systems installed since 1995. The total number of failures of immediately below the tubing hanger. It should be noted that
these systems was 96. This split-up shows that the dominant most identified cable failures have occurred in platform
failing element is the gauge (40%) followed by the cable installations, but crushing of the cable below the tubing hanger
(17%). has in some cases also occurred in subsea wells, only there the
Failure of the gauge tubing has been pulled and the cable was repaired. This is not
The largest cause of failures is the gauge: 40%. Failures have listed as a failure but results in lost rig-time, which could have
typically been attributed to the gauge, based on evaluation of been saved by using a protective cap.
the electrical behaviour of the downhole system. Another cause of cable failures is when tubing loads crush
Originally the gauges of PDG systems were identical to the the cable, especially in an area with severe doglegs. To
memory gauges used in well-tests. Some service companies prevent these failures it is recommended to use cables with
used identical electronic equipment, assembling methods bumper bars where it is expected that the load on the cable can
and/or testing methods. However, the lifetime requirement for exceed the crush resistance of the cable. Alternatively, one
well-test gauges (maximum several months) is relaxed could consider to use mid joint protectors, although these have
compared to that for PDG systems (5-10 years). Therefore in a tendency to shift along the tubing string.
order to improve the gauge reliability it is recommended to use Finally cable failures also occur when the cable has to be
(i) dedicated electronic circuits with proper QA/QC in place, guided across large components in the completion string.
instead of those used for memory gauges, and (ii) improved During some installations the cable was cut at these
burn-in procedures and vibration testing. Further also components as it provided too little space for the cable to pass
destructive sample testing to verify performance is along. To solve this problem it is recommended to use special
recommended. This should be done at actual conditions, e.g. protectors that better guide the cable along these components.
in the closed housing, rather than electronic boards tested in Another option is to provide a recess in the component for the
open furnaces. cable to pass along.
It should be noted that the failure rate of electronic Failure of fixed connections
circuitry is temperature dependent. Most of the PDG systems Fixed connections include splices and the connection at the
are operating at 100°C or below as can be seen in figure 6 cablehead between the gauge and the cable, but exclude the
where the number of installations is listed as a function of make-break connections, such as wet-connectors.
bottomhole temperature. This figure shows that the Traditionally this was a weak element, but with the improved
distribution is almost gausian. The maximum temperature at cable quality, nowadays fewer splices are used. Also the use of
which electronic gauges are operating is currently 150°C. The (partly) redundant metal-to-metal seals for fixed connections
useful temperature range for electronic gauges is extending has lead to improvements. One should however be careful not
due to the use of dedicated electronics and efforts are ongoing to underestimate the importance of these fixed connections,
to extend this further to 200°C. since it may very well be the case that a significant part of the
4 S.J.C.H.M. VAN GISBERGEN, A.A.H. VANDEWEIJER OTC 10945

unidentified failures is due to fixed connections. It is therefore


recommended to further improve the connections and use fully Conclusions
redundant metal-to-metal seals. This report shows that historically PDG systems for measuring
Failure of mateable connectors pressure and temperature exhibited a low reliability. In the
Most common mateable connectors are the wet-connectors in period 1987-1992, substantial improvements have been made,
subsea completions, but also in platform/land systems, a which has increased the 5-year survival probability to 75% for
connector is used at the wellhead outlet. Failures of these systems installed in the period 1991/92. However, since 1992
elements are often caused by a broken seal at the position no further improvements have been made in reliability. This
where the connector is attached to the cable or by an improper has resulted in a 5-year survival probability of 69% for the
electrical contact between the cable and the connector. installation period 1993 till 1998.
Improvements have been made by using redundancy in the We have identified what improvements need to be made to
metal-to-metal seals. Further improvements can be made by achieve the target 5-year survival probability of 90%:
taking more care during installation as most connector failures technical quality, care during installation and management of
occur at this time. interface. We believe that a system reliability well above the
Other failures 90% reliability over 5 years is achievable if these
The third highest cause of failures is those of other elements. improvements are implemented. Financial incentives could be
Examples are failures of prototypes and of the control pod considered to stimulate this, such as making the correct
interface card. Failures of prototypes are by their nature more functioning of the PDG system part of the well gain-share
likely and can only be prevented by more rigorous shop testing incentive scheme, and/or pay-for-data schemes.
including complete system integration and stack-up testing. Improvements in reliability will be required, not only for
Failures of the control pod interface can be prevented by PDG systems that measure pressure and temperature, but also
testing of the complete system with the control pod and not for more complex downhole systems that are entering the
relying upon system simulators. Functional testing of the market, such as inflow control devices, flow meters, reservoir
complete PDG system after completion should also be monitoring devices, etc. These new systems will form a suite
completed to ensure that the pod interface and downhole of equipment to meet various business needs and will probably
system are performing to specification. Also standardisation of increase the market potential of permanent downhole
the downhole communication protocol will reduce the various installations still further. However, for all these systems
types of control cards that are in use, which will have a reliability will be a key issue in the decision process to install
positive impact on reliability. them. The lessons-learned from pressure and temperature
Summary monitoring systems can be very useful input to achieve the
In Table 1, the failing elements and the recommended actions required high-level of reliability.
are summarised. For many elements, the actual care during the
installation is important. This can be achieved by making the References
functionality of the gauge part of the well gain-share incentive 1.Unneland,T. and Haugland,T.:”Permanent downhole gauges used in
scheme. Further improvements of the system by the service reservoir management of complex North Sea oil fields”, paper SPE
companies can be stimulated by performance based pay, as has 26871 presented at the 1993 SPE Offshore European Conference,
been used e.g. for ESPs. Aberdeen UK.
2.Gertsbakh,I.B.:”Statistical reliability theory”, Marcel Dekker, New
York (1989)
Reliability per service company
The dataset that we have enables us to make a comparison
of reliability between the service companies for the systems
installed since 1997. This is shown in Fig. 8 together with
90% confidence intervals, that give a 90% chance that the real
value of the reliability is within these intervals. It shows that
none of the service companies have likely reached the target of
90% survival probability over 5 years if we extrapolate these
curves. It also shows that differences exist in performance of
the various service companies. It should be noted that these
differences are partly caused by the fact that some service
companies have done more difficult installations than others (a
subsea installation under stormy conditions is different from a
platform or land one under mild conditions). However, Service
Companies should be aware that performance comparisons
may very well impact which company to select, and they
should therefore seriously address their systems performance,
in all aspects.
Table 1 Overview of the failure causes with the number of failures due to each cause as a percentage of all
identified failures and the proposed main action to overcome them, for p,T systems installed in 1995-
1998.
Failure % of identified fail. Main action
Cable 17 More care during installation
Protective cap tubing hanger
Calculate required crush resistance: use bumper bars
Design special protectors for guiding cable across large
completion components
Gauge 40 Use of dedicated electronics
Burn-in and vibration/shock testing
Destructive testing at completely actual conditions
Mateable connection 3 Improved installation (Incentive payment scheme)
Fixed connection 1 Use full redundant metal-to-metal seals
Other elements, e.g. 8 1) No specific action (prototype failures will happen)
1) Prototypes 2) Functional testing of the complete system.
2) Control pod interface
6 S.J.C.H.M. VAN GISBERGEN, A.A.H. VANDEWEIJER OTC 10945

Wellhead outlet

Wet connector

Splice

Downhole cable

Cross-coupling protector

Cablehead

Gauge mandrel

Fig. 1 - Schematic sketch of a subsea PDG system.


OTC 10945 RELIABILITY ANALYSIS OF PERMANENT DOWNHOLE MONITORING SYSTEMS 7

250
number of installations

200

150 land
subsea
100 platform

50

0
87 88 89 90 91 92 93 94 95 96 97 98
year of installation

Fig. 2 - Number of PDG installations as a function of the year of installation (1998 until August).

95
85 87/88
reliability [%]

75 89/90
65 91/92
93/94
55
95/96
45
97/98
35
25
11

16

21

26

31
36

41

46

51

56
61
1
6

operational time [Months]

Fig. 3 - Reliability of PDG systems (p,T) for 6 periods of installation: 1987/88, 1989/90, 1991/92, 1993/94, 1995/96 and 1997/98.
8 S.J.C.H.M. VAN GISBERGEN, A.A.H. VANDEWEIJER OTC 10945

Fig. 4 - Division of the PDG systems (p,T) into those that have survived 1 year, and into the failing element, for those that have failed. This is
done for 6 periods of installation.

Cable 17%

Not Identified 31% . Fixed connections


1%

Mateable connectors
3%

Other 8%

Gauge 40%

Fig. 5 - Division of the failures of the PDG p,T systems, installed in 1995-1998, according to the failing element.
OTC 10945 RELIABILITY ANALYSIS OF PERMANENT DOWNHOLE MONITORING SYSTEMS 9

200
180
Number of installations

160
140
120 subsea
100
80 platform/land
60
40
20
0
0-59 60-79 80-99 100- 120- 150-
119 149 200
Bottom hole tem perature [Degr. C.]

Fig. 6 - Temperature distribution of the PDG systems installed in 1995-1998.

100

90
Re lia b ility [% ]

80

70

60

50

40
0 -5 9 6 0 -7 9 8 0 -9 9 1 0 0 -1 1 9 1 2 0 -1 4 9
Bo tto m h o le te m p e r a tu r e [D e g r . C]

Fig. 7 - The 1-year survival probability of systems installed in 1995-1998 as a function of bottomhole temperature, with 90% confidence
intervals.
10 S.J.C.H.M. VAN GISBERGEN, A.A.H. VANDEWEIJER OTC 10945

survival probability

100
90 Service company A
80 Service company B
[%]

70 Service company C
60 Service company D
50

11
13
15
17
19
21
1
3
5
7
9
operating time [Months]

Fig. 8 - Reliability of PDG systems installed in 1997/98, split-up according to the service company, with 90% confidence intervals.

You might also like