SPE 142449

“The Seven Pillars of Well Integrity Management”: The Design and

Implementation of a Well Integrity Management System
S.J. Sparke, Tullow Oil; R. Conway, S. Copping, Expro North Sea

Copyright 2011, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE/ICoTA Coiled Tubing and Well Intervention Conference and Exhibition held in The Woodlands, Texas, USA, 5–6 April 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

Today and industry wide, management of well integrity is a common issue, and relates to both old and new
producing fields. It is apparent that many fields were not designed with the concept of ‘well lifecycle management’
and were probably constructed for a life of about 15 - 20 years. A producing well that is 30 years old is quite
common (e.g. North Sea production started in 1975) and for wells in locations such as the Middle East some fields
are more than 40 or 50 years old.

In line with its expansion and transition to becoming a major independent oil company through its recent
discoveries of very large oilfields in Ghana and Uganda, Tullow Oil needed to broaden its existing well integrity
policy and practices. This is to fit with a projected 4 to 5 fold increase in oil production in the next few years and
increase its well count to over 500 wells.

In order to evaluate the processes in-place, the “Seven Pillars of Well Integrity Management” were defined as the
fundamental requirements of an ideal system. These were used to identify gaps in existing practices which were
systematically and thoroughly addressed by Tullow.

Tullow now has one centralized well integrity management process that is described in detail through a company
policy. This incorporates a software system to collect and manage key construction and operational data from all
operated wells. Though monitored and audited from a corporate perspective, the integrity system is run by each
asset to suit local operational demands and requirements.

This paper documents the learning process that was experienced by Tullow in their path to deliver a company wide
system that would suit varying production scenarios in four different operating environments, with a broad range of
cultures and technical challenges and ultimately be expandable to suit the growth of the company.
2 SPE 142449

In line with its expansion and transition to becoming a major independent oil company through its recent
discoveries of very large oilfields in Ghana and Uganda, Tullow Oil needed to broaden its existing well integrity
policy and practices. This is to fit with a projected 4 to 5 fold increase in oil production in the next few years and
increase its well count to over 500 wells.

In order to evaluate the processes in-place, the “Seven Pillars of Well Integrity Management” were defined as the
fundamental requirements of an ideal system. These were used to identify gaps in existing practices which were
systematically and thoroughly addressed by Tullow.

Tullow now has one centralized well integrity management process that is described in detail through a company
policy. This incorporates a software system to collect and manage key construction and operational data from all
operated wells. Though monitored and audited from a corporate perspective, the integrity system is run by each
asset to suit local operational demands and requirements.

This paper documents the learning process that was experienced by Tullow in their path to deliver a company wide
system that would suit varying production scenarios in four different operating environments, with a broad range of
cultures and technical challenges and ultimately be expandable to suit the growth of the company.
SPE 142449 3

Introduction

Tullow Oil plc

Tullow is one of the largest independent oil and gas exploration companies in Europe with a major focus on Africa. It has a
large portfolio of exploration and production assets including interests in over 85 licences in 22 countries. Production is in
eight countries and working interest production for 2009 was 58,300 barrels of oil equivalent per day (boepd). In early 2009, it
began a review of its well integrity management system to determine how the previous arrangements could be enhanced to fit
company growth. This was driven by an increasing number of exploration/appraisal successes that have led to several major
development projects and a need to be class leader in its approach to safety and environmental issues. These new fields include
the Jubilee field in deepwater Ghana with typical production rates of 20,000 BOPD per well, as well as a wide range of land
and lake based developments in Uganda. In total the total well count is likely to exceed 500 wells in coming years. The
anticipated distribution of the well stock is presented in Table 1.

COUNTRY 2010 2015 2020

Bangladesh 6 6 3
Ghana 18 30 40
Uganda 27 100 200
UK 35 25 15
Other 0 10 15
Total 86 171 273
Table 1: Tullow Wellstock Distribution
Like many major oil companies, Tullow’s operating environments are diverse and the range of well types equally broad. Over
the coming years, the wellstock will include onshore, platform and subsea wells, both oil and gas with a wide range of flow
rates. In all, ten different well types are planned in for the different Tullow operated license areas and are shown below in
Table 2.

WELL TYPE WELL LOCATION WELL DESIGN

1 Bangladesh Gas well

2 Ghana Subsea oil

3 Ghana Subsea water injection

4 Ghana Subsea gas injection

5 Uganda Gas well

6 Uganda Oil well – natural flow

7 Uganda Oil well – artificial lift

8 Uganda Water injection

9 UK North Sea Gas well – unmanned platform

10 UK North Sea Gas well – subsea

Table 2: Tullow Well Type Breakdown

Though good engineering practice was an essential element of the Tullow business process, a plan was implemented in the
Wells Engineering Department in February 2009 to review the corporate approach to well integrity and put in place a centrally
4 SPE 142449

managed well integrity system (WIMS). The aim of this still allows each asset a large degree of operational autonomy. Integral
with a revised well integrity policy, each asset will be required to undergo an annual and independent wells audit that will
report back to the corporate office to determine the actual well status in all operated fields and compare this to the goals set out
in the integrity policy. This process will provide management with a retrospective position for the integrity of all wells. This
will in effect ‘raise the bar’ on the operating standards of all assets to a single high level.

Current Legislation

In the areas that Tullow operated, it was only in the United Kingdom Continental Shelf (UKCS) that well integrity formed a
fundamental part of the licensing requirements. Internationally, a number of legislative regimes existed each of which focussed
on well integrity in a variety of ways. Any of these were quite capable of providing a suitable core business and engineering
process that could be applied throughout Tullow. Principal regulatory regimes included;

• UKCS: DCR-96 regulations

• Norway: Norsok D-010 regulations

• Gulf of Mexico: The Mineral and Mining Standards

• Netherlands: Dutch Mining Regulations

With the exception of the DCR-96 regulations which are based on safety case situations, the balance tended to be quite
prescriptive in their design and requirements and did not fit the Tullow business model or management process. Further, the
UKCS regulations had been in use within Tullow for a number of years to complement their various UK assets, were
understood by many and provided a suitable core that could easily be expanded and adapted to all other operating areas. In
many other regions where Tullow operated, regulations tended to only discuss good engineering practice and did not generally
directly discuss well integrity related issues.

A decision was made to embrace well integrity and raise the bar so that Tullow could readily demonstrate a corporate
approach to this issue. As a result, elements of the DCR-96 regulations were adopted by Tullow for worldwide use and in
particular the practise of independent well auditing would become a key component of the integrity management process.

Project Objectives
The design and implementation of WIMS in Tullow had the following objectives:

• Instil a positive and proactive culture of well integrity awareness within the Tullow operational units

• Set a high operating standard based on stringent international legislation to be equalled or bettered by all Tullow
operating units regardless of location

• Establish Tullow Oil as top quartile amongst operating companies with regards effective and auditable well integrity
management
SPE 142449 5

Tullow Strategy

The Seven Pillars Of Well Integrity Management

Early in 2009, Tullow instigated a review of their corporate approach to well integrity. This was achieved by a Gap Analysis
where the existing systems and processes where compared to an ideal. However, in order to facilitate the Gap Analysis the
Tullow/Expro Team based on industry experience, first agreed the requirements of an ideal system, which they have defined as
“The Seven Pillars of Well Integrity Management”:

1. Dedicated Personnel
2. Policies and Procedures
3. Well Design and Operating Envelope
4. Inspection and Verification
5. Failure Management
6. Data Management and Reporting
7. Audit Trail

Taking these into account while reviewing the management methods in place, the Team identified five key components of the
new management system that would need to be addressed in order to achieve the ideal. These were

Wells Register – A single document listing ALL Tullow wells worldwide, regardless of location, function or status.

Well Construction Envelope – Standardised approach to determining operating limits

Well Failure Matrix – Pre-defined dispensation periods for the full range of failure scenarios

Leak Rate Criteria – Identification and adoption of robust leak acceptance standards based on industry best practice

Well Integrity Software – Selection of a software tool suitable for adoption Tullow-wide

These issues are addressed in the following section. As it was realised that the selected software tool could be used in
conjunction with other key components, this is discussed first.

Required Upgrades to WIMS

Well Integrity Software

Industry wide it was found that four separate companies (ATW, Expro North Sea, Exprosoft and Intetech) provided software
that had been written specifically for well integrity management. While these were considered suitable candidates for Tullow,
a review of standalone software from the major oil industry software providers as well as general consumer software
companies had to be undertaken to confirm suitability.

To assist in identifying the most suitable delivery mechanism and data analysis package that would fit the Tullow
requirements, a review of SPE papers, internet and oil industry journals was carried out to understand the breadth of public
domain information and how it might be used to support any decisions made in this project. Additionally discussions were
held with well examiners and industry specialists to gain their view on the various applications available.
6 SPE 142449

Tullow recognised that the software had to be capable of collating, recording and finally reporting in one standardised process
for all its well types and cope with expected growth. The key functions of the system are -:

1. To collect and analyse well production data and compare this with well construction data to confirm that from an
operational stand point, the well was working within its design characteristics.

2. Standardisation of valve acceptance testing criteria and test frequency. This would form part of a well failure matrix
looking at different barrier components and the impact they might each have for well status.

3. Reference documents such as API guidelines, would be stored within the system to be used by each asset as required.

4. Finally, the entire package should provide a simple but rugged paperwork trail to be used by well examiners in the
annual audit process.

These requirements are shown below in Figure 1.

During the software review process it very quickly became apparent that the ‘plug and play’ concept was the most appropriate
route to follow as the four providers mentioned previously understood much of the well integrity concepts, the legislation and
had various models operational around the world. The traditional software packages provided by others could ultimately
provide the package we were looking for, however the costs were indicated to be very high with a long lead time and secondly
the various providers did not have the core understanding of the well integrity issues and it was felt that this would require a
steep learning curve with possibly a number of false starts.

As part of the tendering exercise, the short listed companies were required to give a technical presentation via video
conferencing facility to each Tullow Asset. The purpose behind this was to give a wider company opportunity for product
evaluation as well as ownership for product selection to each Asset.
SPE 142449 7

Expro North Sea was selected as the most technically suited to the Tullow requirements with their package, SafeWells © and
in March 2010 work began on a pilot project to be run on the existing Bangora field in Bangladesh. This site was selected as it
had several years of production history and there existed a variety of well related technical issues that would provide sufficient
software challenges to test the entire system and the WIMS concept.

Note: The component screenshots in Figure 2 are presented in Appendices A to E

Wells Register
Within SafeWells, a matrix was developed that would serve as the core register for all wells operated by Tullow and would
provide the starting point for any well audit/examination scheme. Conveniently, the data stored here was quickly agreed to by
a number of different departments and provided a resource for other software activities in the company. With the current
activity levels and five rigs operating simultaneously, the numbers of wells was quickly growing and needed a control measure
to ensure that all wells were identified and registered. As part of its well planning process, Tullow used a software package
called EIKOS. This was modified so that the software would trigger e-mail notification being sent to the well integrity
department when a well was spudded. This provided the data to allow the creation of a new well on the register.

Some months prior to the start of the WIMS project, the well engineering department began a review of its overall software
requirements. The aim of this was to identify a small number of software tools that would be used by all assets and standardise
the work flow process.

Wellview, was selected as the principal tool for well construction data. This includes casing and tubing and component data
for all wells drilled, and any subsequent modifications that might be made during interventions during the wells life cycle.
8 SPE 142449

SafeWells was set up to electronically communicate directly with Wellview, download data and build new entries as the new
well triggers occurred.

While this automation sounds ideal, it did not replace a manual checking process to act as a double check to ensure that new
entries were appropriately recorded.

Regardless of well type (exploration/appraisal/production) further triggers were set up within the well planning software to
send e-mail notifications to the well integrity department about the wells future; abandon, suspend or produce. Depending on
which of these outcomes was selected, would determine how SafeWells would deal with the well in its recording process, i.e.:
if the well was to be suspended or produced more information would be required to allow full and complete auditing to be
carried out.

Well Construction Envelope

Critically what is important in this element of SafeWells is the recognition of key construction data that allows comparisons to
be made between the well construction envelope and the well operating envelope. Depending on the measurement standards
set by the oil company, generally this issue is approached through one of two processes. This is either Maximum Allowable
Wellhead Operating Pressure (MAWOP) or Maximum Allowable Annulus Surface Pressure (MAASP).

The first, MAWOP is clearly defined in API RP 90 as ‘’a measure of how much pressure can be safely applied to an
annulus….” The MAWOP is measured relative to the ambient pressure at the wellhead for any particular annulus. It
establishes a safety margin in consideration of failure modes for collapse of the inner tubular and burst of the outer tubular’’.
Used in conjunction with API 5C3 to determine the Maximum Internal Yield Pressure (MIYP) the MAWOP is identified as
being the lesser of the following three options -:

• 50% of the MIYP of the pipe body for the casing or production riser string being evaluated; or

• 80% of the MIYP of the pipe body of the next outer casing or production riser string; or

• 75% of the Minimum Collapse Pressure of the inner tubular pipe body.

The principle component missing from the overall calculation is that of any reference to hydrostatic pressure associated with
fluids in any of the annuli. But in the absence of this it does provide a large safety margin due to this unknown and is accepted
in the industry as a conservative approach to allowable annulus pressure.

In the opinion of Tullow, using this approach is considered to be suitable in wells that may have a combination of one or more
of the following issues; unknown fluids in the annuli due to age of the well and lack/missing well records, OR, an older well
that used barite and other components that will have settled out in time providing an unknown annular mix.

In more recent wells where oil based mud systems were used and/or there was more confidence in annular fluid properties,
MAASPs were calculated which allowed a greater range of well conditions (such as remaining pipe strength) to be considered.
SPE 142449 9

Once this element of the construction phase has been completed, SafeWells now has the maximum operating casing pressures
stored within the software in a lookup table. Daily production operating data will be collected from a variety of sources and
compared with the casing allowances. This is seen as a graph and is displayed below in Figure 3

Figure 1: Graph with Trigger Pressure and MAASP

Valve Tests and Annulus Monitoring – A “Well Failure Matrix”

Tullow well integrity policy requires that all that naturally flow to the surface (both onshore and offshore) be completed with a
surface controlled sub-surface safety valve and two actuated Christmas tree valves. These two tree valves would normally be
the upper master valve and the production wing valve. Further the policy also requires that each of these valves be function
tested twice a year and pressure tested once a year. This is inline with common industry practice, is compliant with most
European requirements however, many companies do not require such stringent requirements for their onshore wells.

Rather than looking at tests on a valve by valve basis, Tullow has grouped these together and take an overall view of the well.
This tends to be less onerous or draconian but does place a heavy requirement on each asset to carry out a risk analysis process
in the event that several issues are identified on a well. One of the advantages of the SafeWells system is that it is ‘owned’ by
Well Engineering and as such will have a view of all activities to ensure that good engineering practices are continually
administered. Figure 4 below demonstrates how the greater part of the failure matrix is constructed and how this can be
applied in this case to an onshore gas field. Figure 5 then demonstrates how the summation of scores steers the Asset in the
decision making process which in this case is for an onshore gas well.
10 SPE 142449

Leak Rate Criteria

This has been a most interesting issue to resolve and a considerable amount of time has been spent discussing and forming an
opinion. Broadly, the consensus of opinion is that the presence of any leak does not necessarily mean the risk is unacceptable,
rather it means that any pressure needs to be managed safely on a case by case basis. This is supported from a variety of
documents.

A detailed review of API guidelines was undertaken and this confirmed a number of points. The available documentation can
be divided into several categories; Safety Valves (API 14B and 14H), Valves at Point of Manufacture (API 6A and 598) and
finally Annular Casing Pressure Management (API RP90).

• Both API 14B, (Design, Installation, Repair & Operation of Subsurface Safety Valve Systems) and API 14H,
(Recommended Practice for Installation, Maintenance & Repair of Surface Safety Valves), provides a rugged formula
for leak test criteria when testing these valves. This determines leak rates for safety valves to less than 15 SCF/Min or
400 CC’s/min and provides a suitable formula for the calculation process.

• API 6A, (Specification for Wellhead and Christmas Tree Equipment), and API 598 (Valve Inspection and Testing).
These two specifications provide leak rate criteria for valves at point of manufacture. However they very clearly point
out that the standard does not apply to field use.
SPE 142449 11

• API RP90, (Annular Casing Pressure Management for Offshore Wells). In general for casing pressure, a leak is
deemed acceptable if a) the pressure is less than 100psi, b) the pressure can be bled to zero through a ½” needle valve,
or c) the pressure does not that under certain circumstances a degree of leaking is acceptable. Though this guideline
was written for offshore wells it seems quite reasonable to use it for onshore wells.

API 14B and API 14D are without doubt rigorous guidelines that are referred to in many widely available public domain
documents for the testing of subsurface safety valves and surface safety valves. Clearly, like most operating companies,
adopting these is simple and demonstrates good engineering practice. However, with regards to other tree and annulus related
valves establishing an acceptable leak rate was challenging.

With regards to other evidence of leak rate criteria we have identified the following

• Vetco Gray Paper. It is believed that a paper was written by Messrs McDonald and Cromar and within this a leak
rate criteria was established as acceptable if it was less than 2 cc’s/min/valve inch. Unfortunately an extensive search
has not revealed a copy of the actual document.

• Several international operators had adopted the 2 cc’s/min/valve inch

• From the ‘Completion and Intervention Network’ (COIN) and Industry round tables, over 30 different operators
(major, independent and national), were asked to complete questionnaires. From these, it was identified that more
than 25 operators around the globe all use the leak rates described in API 14 as suitable for all tree valves.

Tullow decided that the general industry practise of applying the leak rates from API 14 to the Christmas Tree valves was not
appropriate for its own operations for several reasons. Firstly to accept a leak as described in API 14 as being applicable to all
valves regardless of diameter was unsuitable. Secondly the rate of leak in API 14 was above acceptable limits particularly
where the valve had been manufactured and tested too much tighter tolerances. Finally, provided preventative maintenance
was carried out as recommended by the manufacturers a tighter leak tolerance should be acceptable. In conclusion, Tullow
adopted a leak rate of 2 cc’s/min/Valve inch.

Implemenation methodology

The Tullow WIMS project was initiated in February 2009 and followed the time-line shown below in Figure 6.
12 SPE 142449

While SafeWells was an “off-the-shelf” working tool already used by other operators, a number of customisations were
required in order to fine-tune to Tullow requirements. These included:

• Electronic ‘input forms’ designed to mimic the paper forms already in use

• Importation of completion and casing data from 3rd party software

• Automated valve leak-rate calculation and acceptance

• An automated, risk-ranked failure matrix

The system was initially rolled out in the onshore Bangora Gas Field in Bangladesh, with training sessions in the field and the
operations office. While facilities are excellent and the Bangora Team highly motivated, this site was selected for the trial due
to occasionally unreliable internet communications between the Tullow offices in Dakha and London.

The trial in Bangladesh was totally successful, attributable in no small way to the enthusiasm of the field team and the full
support from local management.

This success and subsequent reports to senior management precipitated a fast-tracking of the roll-out to other locations,
initially Ghana, UKCS and then Uganda. An ideal mixture of effective change maangement and the buy-in from operational
personnel at all levels has resulted in a smooth, painless and auditable transition to the new system, based solidly on the Seven
Pillars of Well Integrity Management.

Conclusions

• The ‘Seven Pillars of Well Integrity Management’ have been defined and addressed. They now support the Tullow
Well Integrity Management System in all operating areas and management facilities

• The selected software system ‘SafeWells’ manages well data as a single database, therefore all issues formerly
experienced with multiple data-sets have been removed. All personnel have access to the same data, at the same time
with minimal lagtime between data collection and upload

• The move towards an online management system has allowed senior management to maintain a ‘finger on the pulse’
of the Tullow operated assets through ‘live’ display screens and real-time summary reports

• The new WIMS ensures continuity. Tullow is no longer exposed to the risks posed by a transient workforce. Tullow
personnel transferring between assets or operating countries will be using the same system with familiar user
interfaces and processes

• Implementation of the new WIMS is across all Tullow operated assets. It completely replaces all previous systems
and processes while retaining field specific lessons-learned and best practice

• The clear definition of roles and responsibilities has removed all ambiguity with regards data input, approval and
subsequent responsibility for reactive maintenance
SPE 142449 13

• The Tullow WIMS is now fully auditable. The logically indexed ‘electronic well file’ in conjunction with a scrolling
record of all activities plus a hierarchy of approval / acceptance rights (with e-signature) allows the system to
withstand scrutiny from partners, regulatory authorities and independent well examiners.

• The roll-out, training and implementation have instilled a heightened awareness of the need for a robust well integrity
management system across Tullow at every level.

• Tullow has now embraced well integrity as a corporate responsibility. The ‘’bar has been raised’’ to a level above that
perceived as an industry standard and proactively sought to equal or better the accepted ‘best practice’. The project
has received full management support since conception and while cooperation by the operational units is mandatory,
implementation has achieved a positive cultural change and organisation-wide acceptance.
14 SPE 142449

Appendices

A.  SafeWells Home Page 

B.  Valve Leak Test Screen 
SPE 142449 15

C.  Operating Envelope Management 

D.  Database Structure 
16 SPE 142449

E Barrier Schematic