Risk Dsam TR A6983

A guide to
Risk Based
Maintenance and
Reinvestment
Management
TR A6983
A deliverable from the RISK DSAM project

August 2010
TECHNICAL REPORT
SUBJECT/TASK (title)
A guide to Risk Based Maintenance

SINTEF Energi AS
SINTEF Energy Research and Reinvestment Management
Address: P.O.Box 4761 Sluppen
NO-7465 Trondheim CONTRIBUTOR(S)
Reception: Sem Sælands vei 11
Telephone: +47 73 59 72 00 Agnes Nybø Dag Eirik Nordgård
Telefax: +47 73 59 72 50
www.sintef.no/energy CLIENT(S)
Enterprise / VAT No.:

NO 939 350 675 MVA
RISK DSAM project partners
TR NO. DATE CLIENT’S REF. PROJECT NO.
TR A6983 2010-08-03 12X447

EL. FILE CODE REPORT TYPE CONTROLLED BY CLASSIFICATION
DocuLive 2,3,5 Eivind Solvang Open

ISBN N0. APPROVED BY (TITLE, NAME) COPIES PAGES
978-82-594-3442-5 Knut Samdal 60 56

DIVISION LOCATION LOCAL FAX
Energy Systems Sem Sælands vei 11 + 47 73 59 72 50

RESULT (summary)
This report gives guidelines for risk based maintenance and reinvestment management of
electricity distribution assets. The report is one of the deliverables from the RISK DSAM
project, and presents a summary of important results from the project.
The aim of the report is to provide insight in risk based maintenance and reinvestments
management and decision support to distribution network companies, by suggesting answers
to the following questions:
− How to incorporate risk analysis in maintenance and reinvestment management?
− How to develop risk based maintenance strategies?
− When and how to perform reinvestment analysis?
Maintenance and reinvestment management are two closely related processes, and should be
governed by the same philosophies and overall objectives, preferably coordinated by one asset
management group within the electricity distribution companies.
Throughout this report we recommend two main principles for maintenance and reinvestment
management:
− Risk differentiation; permitting the distribution network companies to focus their
efforts where they are needed the most. The differentiation should be based on risk
analysis results for all relevant risk consequence categories, for example economy,
safety, etc.
− Continual improvement; both the maintenance and the reinvestment processes must be
subject to continual improvement, using gathered experience and updated knowledge
to learn and make the processes better.
KEYWORDS
SELECTED BY
Risk Based Asset Management Electricity Distribution
AUTHOR(S)
Maintenance and Reinvestments
and Reinvestment Management
TABLE OF CONTENTS
Setting the scene & Summary of recommendations

I Preface I
II The structure of the report II
III Summary of recommendations III
IV Checklist for distribution system companies VII
V Important definitions VIII
Main Report
1 Introduction to Risk based maintenance and reinvestment management 1
1.1 Maintenance and reinvestments 1
1.2 Deterioration of assets 2
1.3 Risk 5
2 Maintenance Management 8
2.1 Introduction 8
2.2 Maintenance strategies 9
2.3 How to establish maintenance strategies 10
2.4 Summary 19
3 Reinvestment analysis 20
3.1 Introduction 20
3.2 Triggering event 22
3.3 Beware of your alternatives 25
3.4 Evaluation of alternatives 31
3.5 Summary 40
4 Implementation and improvement 41

4.1 From maintenance strategies to plans 41
4.2 The value of condition monitoring 48
4.3 Continual improvement 49
4.4 Summary 54
References and Recommended literature 55
SINTEF Energy Research, 2010

Setting the scene

Summary of recommendations
I Preface
This report gives guidelines for risk based maintenance and reinvestment management
of electricity distribution assets, based on a structured and pragmatic approach. The
report is one of the deliverables from the RISK DSAM project1, and presents a summary
of important results from the project.
The aim of the report is to provide insight in risk based maintenance and reinvestments
management and decision support to distribution network companies, by suggesting
answers to the following questions:
− How to incorporate risk analysis in maintenance and reinvestment management?
− How to develop risk based maintenance strategies?
− When and how to perform reinvestment analysis?
The report also discusses the need for data, and how to provide this through the use of
expert judgment and condition monitoring activities.
The focus of the RISK DSAM project and of this report is preventive maintenance, and
reinvestments triggered by deterioration of assets. Further, the focus is on the
maintenance and reinvestment management of medium voltage2 (MV) distribution
networks. However, the presented methods and recommendations are also relevant for
maintenance and reinvestments in general.
1RISK DSAM (Risk-based distribution system asset management) is a Competence Building

project sponsored by the Norwegian Research Council, and companies within electricity
distribution in Norway, France and Sweden. The project was launched in January 2006 and will
end in 2010. SINTEF Energy Research is the executing partner.
2 1 – 36 kV
SINTEF Energy Research, 2010 Page I
II The structure of the report
This report is divided into three main parts:

− Setting the scene and summary of recommendations
− Main report
− References and recommended literature
Setting the scene and summary of recommendations
Chapter III gives a summary of the main recommendations. These are elaborated in
more detail in the main report.
Chapter IV provides a ‘check-list’ for distribution network companies on how to

implement risk based maintenance and reinvestment management (focusing on aspects
covered by this report), while Chapter V contains a list of important definitions.
Main report
Chapter 1 gives an introduction to risk based maintenance and reinvestment

management and briefly describes maintenance and reinvestment of distribution system
assets, deterioration of assets, and risk and risk differentiation.
Chapter 2 describes an approach for how to use risk assessment to establish

maintenance strategies for asset groups.
Chapter 3 describes a structured approach to reinvestment analysis, including when to

trigger such analyses, which reinvestment alternatives that exist, and how to evaluate
different alternatives.
Chapter 4 describes different aspects regarding how to implement maintenance

strategies and trigger reinvestment analysis. The chapter also address the need for data
and the importance of condition monitoring. Finally, the principle of continual
improvement is described.
Use of examples
Examples are provided throughout the report to illustrate important aspects and
applications of risk based maintenance and reinvestment management. The examples
are constructed for the purpose of this report, but are based on information and cases
from distribution companies and hence correspond to ‘real-life’ applications.
SINTEF Energy Research, 2010 Page II

References and recommended literature
References and recommended literature are provided at the end of the report, including
complementary reading for the interested.
III Summary of recommendations
Overall principles
Maintenance and reinvestment management are two closely related processes, governed
by the same philosophies and objectives, preferably coordinated by one asset
management group within the electricity distribution companies.
Common philosophy and objectives
Risk differentiation
Maintenance Reinvestment
Continual improvement
Figure III.1 Maintenance and reinvestment management as two integrated processes

(Nybø and Nordgård, 2010)
Throughout this report we recommend two main principles for maintenance and
reinvestment management:
− Risk differentiation
o Risk differentiation will permit the distribution network companies to
focus their efforts where they are needed the most. The differentiation
should be based on risk analysis results for all relevant risk consequence
categories, for example economy, safety, etc.
SINTEF Energy Research, 2010 Page III

− Continual improvement
o Both the maintenance and the reinvestment processes must be subject to
continual improvement, using gathered experience and updated
knowledge to learn and make the processes better.
In addition, we stress the need for a structured approach, for documentation and
reporting, and last but not least the importance of good information. Condition
monitoring is hence a key activity in risk based maintenance and reinvestment
management.
This report deals with three main topic areas:

− Maintenance management - how to establish maintenance strategies
− Reinvestment management - how to perform reinvestment analyses for assets
− Implementation and improvement
In the following, key recommendations to each of these areas are listed.
Maintenance management - How to establish maintenance strategies
Chapter 2 describes an approach for how to use risk assessment to establish

maintenance strategies for asset groups. The main objective is to use a structured
approach to identify groups of assets which can be prescribed the same maintenance,
and to use risk analysis in this process. Key recommendations include:
Divide your assets into system units and establish maintenance strategies based
on risk differentiation
 Identify groups of assets which can be prescribed the same maintenance
 Assets associated with high risk should be subjected to more
comprehensive maintenance and be considered for reinvestment earlier
compared to low-risk assets
Maintenance strategies must describe:

 The maintenance actions to be done and how often (or when) they
should be performed
 When to trigger a reinvestment analysis
SINTEF Energy Research, 2010 Page IV

Reinvestment management - How to perform reinvestment analysis

for assets
Chapter 3 describes a structured approach to reinvestment analysis, from triggering of

an analysis to making the decision. Key recommendations include:
Identify the right assets to analyse:

 Define triggering criteria (events etc) which define when to perform
reinvestments analyses, reflecting the risk that the assets represent
Establish company standards for performing and documenting reinvestment

analysis, considering costs, risk and uncertainty
Beware of your alternatives:

 Should we reinvest during the period of analysis or not?
 Maintenance is often an alternative to reinvestment (but not forever)
 Establish standard solutions to ‘known’ problems concerning groups of
assets
SINTEF Energy Research, 2010 Page V

Implementation and improvement – How to set up maintenance and

reinvestment plans, gather relevant data, and ensure continual improvement
Chapter 4 describes different aspects on how to implement risk based maintenance

strategies and trigger reinvestment analysis. The need for data and the importance of
condition monitoring and continual improvement in this context is also emphasised.
Key recommendations include:
Identify information needed in order to implement maintenance strategies

and trigger reinvestment analysis
 What data is needed and what data is missing? How valuable is this
information? How difficult is it to obtain? If necessary; How shall we
obtain it?
 Systematically register and update necessary information
Create an overview of your assets and allocate them to the corresponding

maintenance strategy
 Set up maintenance plans, also addressing need for reinvestments
Use condition monitoring results as input to maintenance and reinvestment

management
 Different needs require different condition monitoring. Cost versus
benefits of condition monitoring must be considered.
Evaluate strategies, processes, and results on a regular basis – and use this
knowledge to improve.
SINTEF Energy Research, 2010 Page VI

IV Checklist for distribution system companies
Based on the findings presented in this report, some checkpoints can be formulated for
the implementation of risk based maintenance and reinvestment management.
Establish a common philosophy and objectives regarding maintenance and

reinvestment
Establish an asset management group responsible to enforce maintenance and

reinvestment management
Develop risk differentiated maintenance strategies for different system units

(overhead lines, substations, cables etc), where more attention is given to the
most risky assets, and less attention is given to the least risky assets
Define criteria which trigger reinvestment analysis
Establish a standardised procedure regarding how to perform reinvestment

analyses and how to document them
Establish a procedure regarding how to get the data required to implement and
improve strategies
Establish dedicated condition monitoring and reporting in order to meet

different needs
− Find and correct deviations
− Reveal need for reinvestment / comprehensive maintenance
− Gather other relevant data
Establish a procedure for continual improvement of maintenance and

reinvestment management, including meeting arenas, key questions to be
addressed and the follow-up of selected indicators
SINTEF Energy Research, 2010 Page VII

V Important definitions
Combination of the probability of an event

and its consequence.
Note 1: The term ‘risk’ is generally used only Source:

when there is at least the possibility of negative ISO/IEC Guide 733
Risk
consequences.
Note 2: In some situations, risk arises from the

possibility of deviation from the expected
outcome or event.
The combination of all technical and

administrative actions, including Source:
Maintenance supervision actions, intended to retain an IEC©60050-191-01-01
item in, or restore it to, a state in which it
can perform a required function.
The maintenance carried out after fault

Source:
Corrective recognition and intended to put an item
IEC©60050-191-01-08
maintenance into a state in which it can perform a
EN 13306:2001
required function.
The maintenance carried out at

predetermined intervals or according to Source:
Preventive
prescribed criteria and intended to reduce IEC©60050-191-01-07
maintenance
the probability of failure or the degradation EN 13306:2001
of the functioning of an item.
Replacement of an existing item with a new Source:

one with the same capacity. The project ‘Value
Reinvestment
adding maintenance in
Note: May include modification power production’
(SINTEF)
3 ISO/IEC Guide 73 Risk Management – Vocabulary – Guidelines for use in standards.

SINTEF Energy Research, 2010 Page VIII
Main report
SINTEF Energy Research, 2010

1 Introduction to risk based maintenance and reinvestment

management
This chapter gives an introduction to maintenance and reinvestments in electricity

distribution systems, and to the role of risk assessment in this context.
Asset management deals with the complex balancing of cost, performance and risk –
making strategies and procedures for balancing preventive and corrective maintenance
actions and reinvestments.
The motivation for using risk methodologies in maintenance and reinvestment

management is for the distribution companies to obtain a better allocation of resources
(money and manpower). This can be achieved by prioritising and directing their efforts
where there is the most to gain; spending fewer resources on low risk assets and more
on high risk assets.
Maintenance and reinvestment management should be governed by a common

philosophy stating principles and objectives which are important in the company’s asset
management. The process of formulating the overall philosophy is not further
elaborated in this report.
1.1 Maintenance and reinvestments
The purpose of maintenance and reinvestment can be summarised as:

− Restore, maintain or improve an asset’s condition
− Provide information about assets and their condition
− Replace and/or modify assets (reinvestments)
Electricity distribution assets typically have long life spans and will for most of their life
be subjected to some sort of maintenance. New assets will in general require little
maintenance. The need for maintenance increases as the asset deteriorates with time.
Reinvestment becomes an increasingly relevant option as the asset deteriorates, or if
maintenance is difficult to perform or disproportionately resource-demanding (e.g.
through lack of spare parts). This is illustrated in Figure 1.1.
Condition monitoring is a vital part of maintenance as it provides information about

assets, which can be used to prescribe maintenance or reinvestment actions. Most assets
are subjected to different kinds of condition monitoring throughout their life, and as
they approach their end of life, the need for information is growing, for example in order
to identify reinvestment needs.
SINTEF Energy Research, 2010 Page 1

Reinvestment
Reinvestment
Maintenance
Maintenance
Maintenance
Maintenance
Investment
Repair
time
Figure 1.1 Example of maintenance and reinvestment actions throughout the life of a given asset
1.2 Deterioration of assets
Network components are designed to withstand a certain level of stress, and this ability
will deteriorate with time. The components’ reliability will be a function of their
condition and the stress they experience, the latter for which the random contribution
can be substantial (e.g. large wind / ice loads). Maintenance and reinvestment actions
influence the components condition, and hence the expected probability of failure.
Figure 1.2 illustrates how the condition (strength) of a given component deteriorates
with time. The normal operating stress and the random stress the component is exposed
to are also illustrated. In this case failure occurs at time T due to normal operating
stress4.
4It is important to keep in mind that both strength and stress are stochastic, and requires a
probabilistic approach.

Stress /
Condition
Condition
Random stress
Expected failure
Normal
operating stress
Expected lifetime
Time
T
Maintenance Reinvestment
Figure 1.2 Strength and stress for a component as a function of time (Heggset et al., 2007b)
As illustrated, the component’s condition deteriorates over time. As previously noted

(and illustrated in Figure 1.2), reinvestments are often considered when the components
(or assets) in question are approaching their end-of-life. Condition monitoring is vital in
order to assess the condition and the associated failure probability and remaining life of
assets.
1.2.1 Condition states and deterioration model
SINTEF has developed a failure model5 where the technical condition of a component
can be characterised on a scale from 1 to 4, and a 5th state which implies fault (Table 1.1).
Further, deterioration over time can be modelled by the transition through the different
condition states.
5The method was originally constructed for hydro power plants, but is currently being adapted
for network components, (Heggset et al., 2007a),(Heggset et al., 2007b).

Table 1.1 Different condition states
State Description
1 No indication of degradation. ‘As good as new’.
2 Some indication of degradation. Condition noticeably worse than ‘as good as new’.
3 Serious degradation. Condition considerably worse than ‘as good as new’.
4 The condition is critical.
5 Fault
Figure 1.3 shows how an asset deteriorates with time, including the different states. The
duration of each state k (Tk) may vary from several years to only a few years or months,
depending on the asset and the stresses the assets are exposed to. As indicated in Figure
1.3, failure is defined to occur in the transition from state 4 to state 5.
Technical condition
State:
Failure
1
5
Time
year
T1 T2 T3 T4
Figure 1.3 Technical condition (state 1-5) and life curve (Heggset et al., 2007b)6
By modelling the transitions through the states 1-5 it is possible to calculate the annual
failure probability based on the technical condition of a component (Heggset et al.,
2007b). Such an approach is based on life-curve models and the use of expert judgement
in assessing the duration of the different condition states. This is a good alternative /
complement to methods based on statistical fault data, which often is scarce (and biased
since the link to component condition often is missing).
6 The figure corresponds to the general deterioration model depicted in Figure 1.2.
1.3 Risk
Risk can be defined as the combination of probability of an event and its consequence,
(ISO/IEC, 2002). The term risk is generally associated with the possibility of negative
outcomes of future events.
The risk for a given asset, process or activity can be addressed by answering a triplet of
questions (Kaplan, 1991):
− What can go wrong?
− How likely is that to happen?
− If it does happen, what are the consequences?
The answers to these questions will give a picture of the risks, where the answer to the
first question describes some kind of undesired event; the answer to the second is a
probability statement, while the answer to the third question is a description of potential
consequences – which typically will be multi-dimensional. For example, if a specified
undesired event occurs, it may have economic, reputational and safety consequences.
Even though the questions may look simple, to provide good answers can be
challenging, both in terms of identifying undesired events and estimating probabilities
and consequences.
1.3.1 Probability estimates
When establishing probability estimates, one try to state how likely it is for an undesired
event to occur in the future. It lies in the nature of the problem that there will be large
uncertainties related to this task.
For the analysis of risks related to distribution system maintenance it will be most
relevant to rely on the opinions of experts, in addition to historical experience as
revealed in e.g. failure and interruption statistics. Experience indicates that there are
limited amounts of relevant information which can be provided by statistical sources
(Nordgård et al., 2007, Nordgård and Samdal, 2010), hence expert judgment is a very
important source of information in this process.
Results from condition monitoring (together with life-curve models illustrated in Figure
1.3) will be an important input to estimate probabilities – at least to identify assets with
relatively high probability of failure compared to the average asset.

1.3.2 Consequence estimates
Distribution companies face risk of different kinds, which can be grouped into the
following categories (Sand et al., 2007):
− Economic risk,
− Safety risk,
− Environmental risk,
− Quality of supply risk,
− Reputational risk,
− Vulnerability risk, and
− Regulatory risk.
Several of the consequence categories can be related and overlapping. For example, both
safety and environmental issues may have a significant impact on company reputation.
The consequences of potential undesired events can be estimated through several

methods; e.g. reliability simulations and safety risk models, but also historical records of
previous experience and the knowledge of experts familiar with the components and
their problems.
1.3.3 Risk matrix
The risk matrix is a useful tool to visualise risk. It shows the consequence of an event on
one axis and the probability on the other. Depending on the combination of its
probability and consequence, an event can be classified as more or less acceptable (as
indicated in Figure 1.4).
Negligible Small Medium Serious Catastrophic
Highly probable
Very probable
Probable
Probability
Less probable
Improbable
Consequence
Figure 1.4 A risk matrix

It is important to bear in mind that the risk changes with time, typically due to
deterioration of assets. Decisions and actions concerning investment, reinvestments,
maintenance and operation, and external factors such as load development and climatic
change will also influence the risk.
Risk matrixes can be used to visualise the expected impact of measures to reduce risk.
With reference to Figure 1.4, measures to reduce the probability of an unwanted event
will shift the associated risk downwards in the matrix, whilst measures to reduce the
consequence will shift the risk leftwards. A combination of the two is also possible. This
is indicated by the arrows in Figure 1.4, which show how risk can change as a result of
different risk reducing measures.
1.3.4 Risk differentiation
A key concept in risk based maintenance and reinvestment management is risk

differentiation. The rationale for risk differentiation can be expressed quite simply: to give
assets with high risk more attention (and spending) compared to assets with lower risk –
i.e. to spend the companies’ resources where it is needed the most.
In practice risk differentiation will imply to give ‘risky’ assets more often and / or more
comprehensive maintenance, and to consider reinvestment earlier compared to other
assets.
Figure 1.5 illustrates how a group of assets are divided into different subcategories (A-
D) depending on their associated risk. The aim is to use risk analysis to identify groups
of assets which represent similar risk and therefore can be prescribed the same
maintenance. This represent a concentration of efforts: Rather than to prescribe the same
maintenance to all MV/LV substations or all MV overhead lines, the maintenance
strategies are different depending on the risk the assets represent.
B C D
’Unsorted’ B C C
Assets B B C
Risk assessment A A B
Figure 1.5 Risk differentiation, an illustration

Maintenance management
Establishing risk based strategies
2 Maintenance management
2.1 Introduction
Maintenance activities are an important part of the management of distribution system

assets, covering a large percentage of operation costs, and triggering a majority of the
reinvestments in the grid. In addition maintenance is vital for the safety of the
distribution company’s own personnel and third parties, and it has a potential
significant impact on the company’s reputation. In such a perspective it is obvious that
maintenance is a key activity in asset management.
Maintenance can be divided into two main categories:

1. Preventive maintenance
- Condition monitoring
- Actions to improve a components condition
2. Corrective maintenance
- Correction of minor deviations
- Restoration after fault
A challenge for distribution system companies is to find a good balance between

preventive maintenance (and reinvestments) and corrective maintenance. The costs of
preventive maintenance must be compared to expected benefits related to e.g.:
− Increased (expected) component lifetime
− Reduced costs of interruptions
− Reduced costs, corrective maintenance
− Improved data concerning components
The focus of this report is on strategies for preventive maintenance, and the following
chapter describes how to develop risk based maintenance strategies.
Implementation and improvement of strategies and processes are discussed in chapter 4,

where we also highlight how important information from condition monitoring is in
order to make good maintenance and reinvestment decisions.

2.2 Maintenance strategies
A maintenance strategy specifies what maintenance actions to do and how often (or
when) they should be performed for a given asset or group of assets.
Due to the vast amount of distribution system assets, there is a need to establish
maintenance strategies which cover groups of assets, instead of a specific strategy for
each asset. In the strategy, when to perform maintenance can be stated as a function of
technical condition, time, operation, or to be done after special events, such as failures,
major storms or heavy snowfall.
Maintenance actions cover a variety of tasks with different degrees of complexity: from
simple visual inspections, via more advanced condition monitoring; to minor routine
maintenance and major revisions.
How often or when different maintenance actions are performed should be decided
based on risk differentiation: Assets associated with high risk should be subject to more
frequent and more comprehensive maintenance actions than low-risk assets.
When to make an asset or a group of assets undergo a reinvestment analysis should also
be subjected to risk differentiation. We recommend that the maintenance strategies
include criteria which trigger a reinvestment analysis. This is further described in
chapter 3.2.
EXAMPLE 1 Excerpts from a maintenance strategy for MV overhead lines
Condition monitoring
- Inspection of MV overhead lines shall be performed annually.
- A thorough inspection of MV overhead lines should be performed every 5
years for particularly important lines, and every 10 for other lines.
Vegetation management
- Vegetation management of MV overhead lines shall be performed every 5
years, or when inspections state that it is necessary.
Reinvestment analysis
- MV overhead lines shall be evaluated with regards to potential reinvestment
after 35 years of operation, or when observations from condition monitoring or
other information indicate such a need.

Work process
The process of establishing maintenance strategies should be carried out in teams,

utilising the knowledge of various experts. The teams can for example consist of:
− The overall responsible for the maintenance strategies
− Workers having experience with operating and maintaining the various assets
− External experts if needed and/or available
− People having knowledge and experience with risk assessment
− Facilitator(s). The facilitator is the person responsible for driving the process
forward. The facilitator can both be internal or external to the company.
2.3 How to establish maintenance strategies
A suggested process to establish maintenance strategies for different system units is

illustrated in Figure 2.1. The different steps are further elaborated in the subsequent sub-
chapters.
Select system units
Identify components
Map existing Perform

maintenance practice risk analysis
Evaluate results
Formulate
Figure 2.1 Process for establishing maintenance strategies. Based on

(Nordgård and Samdal, 2010)

2.3.1 Divide the system into system units
The term System units is used to denote a group of components which naturally belong
together. For MV distribution systems, a natural selection of system units can be:
− MV overhead lines
− MV / LV substations
− MV cables.
The aim the process is to establish a maintenance strategy for selected system unit at the
time. It can be advisable to start with the system unit where the expected gain from
implementing a risk based strategy is expected to be the highest (relatively to the
expected efforts needed).
When all identified system units have been through the process, the company will have
maintenance strategies which cover all assets, based on the same overall principles.
2.3.2 Identify system unit components
Each system unit must further be divided into its components.
EXAMPLE 2 System units and components
For the system units MV overhead lines and MV/LV substations the following
components are identified:
System unit: MV overhead lines System unit: MV/LV substations

Components: Components:
- Wooden poles (incl. traverse, - MV / LV transformer
insulators, etc.) - Switch disconnectors
- Phase conductors - Cable terminations
- Line trace - LV distribution board
- Pole-mounted switches - Substation building
- Cable terminations
- Pole-mounted MV/LV
substations

2.3.3 Map existing practice
The majority of the maintenance actions performed today have their origin in judgments
and experience from decades of operating the grid. It may not at the time of origin have
been called ‘risk analysis’, but the results will in many cases be in accordance with the
principles of such.
An important first step in the work of establishing maintenance strategies is to look

closely into today’s practice and previous experience from maintaining the network. In
addition, it is equally important to clarify if there exist any recommendations, rules or
regulations concerning how and how often various maintenance actions should be
performed.
The mapping should be focused on what is done of maintenance activities on each of the
system unit components or at the system unit as a whole (which typically applies to
condition monitoring). It is important to reveal if maintenance is differentiated in some
way and why.
EXAMPLE 3 Excerpts of mapping of existing practice for MV overhead lines
For the system units MV overhead lines the following existing practice is identified:
− Inspections of MV overhead lines are performed annually. These inspections
are performed using a helicopter.
− Thorough inspections of MV overhead lines are performed every 8 years.
The thorough inspections are performed on the ground, following a specified
checklist.
The mapping of existing practice can also provide information about specific problem
areas in the systems, pinpointing assets which are believed to be more exposed to risk
and how their potential failures influence the system.
One important aspect of mapping existing practice is also to clarify and agree on
terminology. Experience shows that there can be many words describing the same
action, or that the same word is used with different meanings. Hence there is a potential
for misunderstandings and misconceptions, making communication difficult.

2.3.4 Perform risk analysis for components
The next step in the process is to perform risk analyses for different components –
aiming to obtain an informative risk picture as a basis for further evaluation.
The risk analyses should be performed on component level – i.e. on the ‘natural’ level of
resolution. The analyses of the different components will together constitute a risk
analysis for the whole system unit.
We recommend the approach presented in chapter 1.3 (Kaplan, 1991), where risk is
described by answering a triplet of questions:
− What can go wrong?
− How likely is that to happen?
− If it does happen, what are the consequences?
Guided brainstorming sessions is a good way to answer these questions, and thereby
performing the risk analysis. Focus should be on events that affect consequence
categories which have been found relevant in the maintenance and reinvestment
philosophy. It is highly relevant to identify factors which can influence risk; i.e. factors
which can influence the probability of occurrence and/or the consequences of undesired
events. Such factors can be different types of operating conditions, variations in design,
etc.
The risk analysis process can identify a need for a higher resolution of the asset groups.
As an example it may be revealed a need to differentiate between various types of
switch disconnectors. This step thus provides further detail to the grouping of assets that
presumably are exposed to similar risk and hence should receive the same maintenance.
EXAMPLE 4 Differentiation of MV switch disconnectors
Concerning MV/LV substations, MV switch disconnectors should be divided into the

following categories:
− Air-insulated switch disconnectors
− SF6 insulated switch disconnectors
− Epoxy insulated switch disconnectors

What can go wrong?
The answers to the question ‘What can go wrong?’ will provide a list of potential
undesired events.
One major source of information to answer this will be the opinions of experts having
experience from working with the components in question. The expert judgments can be
supported with input from accident statistics, fault and interruption statistics and so
forth, when applicable.
EXAMPLE 5 Identification of undesired events for the component

‘Wooden poles (including traverse, insulators, etc.)’
Though brainstorming and discussions in expert groups the following undesired

events have been identified:
1. Pole breakage
2. Pole askew
3. Fire damage of pole
4. Insulator flashover
5. Conductor falls on traverse / burnt traverse
6. Broken traverse
7. Flashover/discharge of insulator chain
8. Displaced traverse
9. (Partially) defect discharger
10. Person falling down from (and / or with) pole
11. Person touching MV parts (from climbing the pole or nearby trees)
12. Poor earthing connections
13. Insulators destroyed by vandalism
14. Impregnation run-off to water and/or soil
How likely is that to happen?
The answer to this question is some kind of probability statement concerning the
potential occurrence of specific undesired event.
Probability estimates can be formulated in terms of verbal descriptions (improbable, very

probable, etc.) or in terms of semi-quantitative statements, e.g. value intervals.

EXAMPLE 6 Probability scales
P5 – Highly Probable More often than once a year

P4 – Very Probable Once every 1-10 years
P3 – Probable Once every 10-100 years
P2 – Less probable Once every 100-1 000 years
P1 – Improbable Less than once every 1 000 years
For the analysis of risks related to distribution system maintenance, potential sources of
information will be the opinions of experts, in addition to historical experience as
revealed in e.g. failure and interruption statistics. Experience indicate that there are
limited amounts of relevant information which can be provided from statistical sources
(Nordgård et al., 2007, Nordgård and Samdal, 2010), hence expert judgment is a very
important source of information in this process.
However, it is still important to work towards having better statistical foundations to

complement and validate expert judgment – e.g. through maintenance management
systems, the fault and interruption statistics, injury reporting systems etc.
If it does happen, what are the consequences?
The answer to the third question will be an estimate of consequences of the identified
undesired events. The consequences will typically be multi-dimensional, i.e. the
undesired event will have impact on more than one consequence category, e.g. through
having:
− Safety impact
− Environmental impact
− Reputational impact
− Economic impact.
As for the probability estimates, expert judgement will also be an important input to
estimate the consequences for various undesired events. Consequences can be ‘local’
(related with specific assets - typically safety) or at system level (economy, reputation).
In some cases – e.g. with regards to estimation of extent of interruptions – simulation
tools (load flow analyses, reliability analyses, etc.) are recommended to get a better
foundation to estimate the consequences of various undesired events.

EXAMPLE 7 Consequence scales
For the consequence category safety the following intervals may be used:
C5 – Catastrophic One or more deaths – many serious injuries

C4 – Serious More than one person with serious injury
C3 – Medium Medium to serious injuries
C2 – Small Minor injuries
C1 – Negligible No injuries
2.3.5 Presentation and visualisation of results in risk matrices
Results from risk analyses can be plotted in risk matrixes to visualise the risks –
providing a risk picture for each of the components for each of the chosen consequence
categories.
There are no universal rules for how to model the risk matrixes, and different designs
are seen; most often ranging from 3x3 to 5x5 matrixes. For the purpose of analysing risk
in order to establish risk-based maintenance strategies, it is necessary to use matrixes
with sufficient degree of resolution. The motivation for this is to enable differentiation of
risk. Experience indicates that a 5x5 matrix can provide such resolution.

EXAMPLE 8 Using a the risk matrix to visualise risk
Figure 2.2 shows the result of a risk analysis for the undesired events listed in section
2.3.4 plotted for the consequence category safety.
Negligible Small Medium Serious Catastrophic
Highly probable
Very probable
Probability
Probable
Less probable
Improbable 8, 12 1, 6 10, 11
Consequence
Figure 2.2 Example: Risk matrix for safety for the undesired events for wooden poles
(Nordgård et al., 2007)
From Figure 2.2 it can be seen that events 10 and 11 are identified as being the most
critical with regards to safety in this risk mapping:
- Person falling down from (and / or with) pole
- Person climbing in pole and touching live MV parts
Not all of the initial 14 unwanted events (identified in Example 5) are placed in the risk
matrix for safety, because some of them are regarded not to be relevant for this
consequence category.
The results from the risk analysis plotted in Figure 2.2 motivates e.g. for having
maintenance activities related to monitoring the occurrence of rot in wooden poles (to
avoid event 10) and for having sufficient vegetation management activities to avoid
climbing in trees (event 11).
A low level of maintenance can influence the probability of occurrence for various
undesired events, and hence move the probability estimate upwards in the figure. For
undesired events with serious consequences it is therefore relevant to prescribe
maintenance activities which can control the probability of occurrence to a sufficient
low level.

2.3.6 Evaluate results and formulate strategies
Based on risk analysis and mapping of existing practise, improved strategies can be
formulated. The result of risk analysis for different components must be subject to
evaluation and discussions among the experts, concerning whether they give an
intuitively right picture for risks related to the different components; and how this risk
can be controlled through prescribing maintenance actions.
Experience shows that risk matrices are very useful tool for focused discussions within
the expert groups. In the discussions, possible risk mitigating actions must be addressed,
with basis in the companies existing practice. An important part of the evaluation is to
judge whether there are maintenance activities well suited to mitigate the risks, and how
maintenance activities should be differentiated according to identified risks.
It shall be noted that the risk mapping does not prepare for a computation of the “right
maintenance” – but rather serve as a basis for a qualitative evaluation of what is needed
of maintenance activities to control the identified risks.
EXAMPLE 9 Maintenance strategy
Table 2.1 shows an extracts of a maintenance strategy for a group of MV overhead lines,
where the maintenance and trigger criteria for reinvestment are different for lines
representing low CENS (Cost of energy not supplied) compared to lines with high
CENS.
Table 2.1 Maintenance strategy for a group of MV overhead lines (adapted from
(Catrinu et al., 2009))
Action Category A: Category B:

Low economic risk (CENS) High economic risk (CENS)
Preventive maintenance Annual inspections Annual inspections
Thorough inspections every 10 Thorough inspections every 5
years years
Vegetation management every 6 Vegetation management every 3

years or at explicit findings at years or at explicit findings at
inspections inspections
Triggering of reinvestment age > 30 years or if condition age > 20 years or if condition
analysis when monitoring results indicate a need monitoring results indicate a need

2.4 Summary
This chapter has described and exemplified the main principles of a structured approach
to establishing risk based maintenance strategies for distribution system assets.
The principles support the development maintenance strategies which prescribe

maintenance actions which address the identified risks – and allow for differentiation of
maintenance efforts based on risk analysis results.

− Divide your assets into system units and establish maintenance strategies based
on risk differentiation
o Identify groups of assets which can be prescribed the same maintenance
o Assets associated with high risk should be subjected to more often and
more comprehensive maintenance actions and be considered for
reinvestment earlier compared to low-risk assets
Maintenance strategies must describe maintenance actions to be done and how often (or
when) they should be performed. The maintenance strategies should also include
triggering events for when to commence a reinvestment analysis.

Reinvestment analysis
From triggering of an analysis to making the decision
3 Reinvestment analysis
3.1 Introduction
Reinvestments involve replacing or modifying existing system units or components.

Reinvestments are often considered when an asset is approaching its end of life, and one
of the distribution companies’ challenges is that they have a large amount of such assets.
We advocate the need for a structured approach regarding when to perform

reinvestment analysis by defining what we refer to as ‘triggering events.’ Such events
are circumstances which make an asset or a group of assets undergo reinvestment
analysis. It is the analysis which is triggered, and not the reinvestment itself. The
decision of whether to reinvest or not, will be a result of the analysis.
Once a reinvestment analysis has been initiated, the fundamental question is:
Should the asset(s) in question be replaced and /or modified during the period of analysis or not?
We here recommend a method for reinvestment analysis, highlighting the path from
triggering an analysis to decision, see Figure 3.1. In addition, important aspects
regarding implementation and improvement are addressed. Different aspects of Figure
3.1 are explained in the following sections.

Triggering No reinvestment or
Initial evaluation
event standard solution
Establish alternative
solutions Update plans
Evaluation of alternative Reinvestment or no

solutions reinvestment
Trigging Analysis Decision
Figure 3.1 Recommended work process for reinvestment analysis
Some key recommendations, which are highlighted throughout the subsequent chapters
include:
- Identify the right assets to analyse at the right time: Define triggering events
- Beware of your alternatives:
o Should we reinvest in the coming x years or not?
o Maintenance is often an alternative to reinvestment (but not forever and
neither for ‘maintenance free’ assets)
o Establish standard solutions to typical problems concerning group of
assets
- Establish company standards for performing and reporting
o Condition monitoring focusing on revealing reinvestment needs
o Initial evaluation of reinvestment need
o Evaluation of different reinvestment alternatives
- Document the reinvestment analysis even if reinvestment is postponed.

3.2 Triggering event
- Identifying the right assets to analyse
Initial evaluation
Update plans
solutions

Figure 3.2 Triggering events initiates a reinvestment analysis
Network companies should define what we have called ‘triggering events,’ which are
circumstances which make an asset or a group of assets undergo reinvestment analysis.
Examples of such triggering events include:
− Results from condition monitoring
− Expected residual life (age)
− Failures and other history (e.g. overload, voltage deviations)
− Unfortunate / Unwanted design
− Regulatory intervention / requirements
− The risk associated with the asset / asset group has been questioned.
− Opportunity window (Construction activity etc.)
− High maintenance cost, interruption costs, costs of losses etc.
When defining triggering events, it is recommended to include risk differentiation, i.e.

focusing on the assets associated with the highest risk and giving them a ‘shorter way’ to
the reinvestment analysis. A description of triggering events may be included in the
companies’ maintenance strategies (Chapter 2.2).
In many cases a reinvestment analysis may be triggered by observation during

maintenance activities, e.g. if the condition monitoring reveals that replacement may be
necessary due to extensive deterioration, or if the estimated costs or other resources
spent on maintenance are substantial.

EXAMPLE 10 Using input from condition monitoring to trigger reinvestment analysis
Figure 3.3 shows an overview of rot development of wooden poles on the overhead
line. The scale 1-4 is used where 4 indicates substantial rot and 1 indicates that the pole
is as good as new (no indications of rot), see chapter1.2.1.
Poles: Rot
0
1
2
3
4
Figure 3.3 Condition information for an overhead line: rot development, poles
Information such as depicted in Figure 3.3, provides good information on whether

assets should be considered for reinvestments. It is also possible to have dedicated
‘check points’ during condition monitoring addressing the need for reinvestment:
- Is there a need to consider reinvestment?
- Should the asset be prioritised for more thorough condition monitoring in order to reveal
reinvestment needs?

3.2.1 Trigging an analysis of a group of assets
Some groups of assets call for special attention as they represent higher risk compared to
other assets, due to their type and design. Such assets are often identified based on
expert opinion and failure history. In some cases, public regulations might trigger
replacement or modification of groups of assets, usually due to inadequate design
regarding safety.
Some examples of groups of assets that might be eligible for reinvestment analysis are:
− 1st generation XLPE cables Experience: These have a high failure

probability
− Pole mounted substations Regulatory requirement (in Norway) to

operated from platform modify design
− Open low voltage systems in Safety / regulatory concern

MV/LV substations
− MV switch disconnectors with Safety concern

wire fence encapsulation
− OH-lines of types Fe-wire, cobber Experience / observation: Old and brittle

or cobber weld
Such groups should be identified during the process of establishing and / or updating
maintenance strategies (chapter 2.2). Triggering criteria should be defined and included
in the maintenance strategies, for example:
− All first generation XLPE cables should be considered for reinvestment within 10
years
− All substations with non-encapsulated low voltage systems should be considered
for reinvestment within 5 years
If a company has large group of assets that is considered for reinvestment, they should
do reinvestment analysis for these groups and try to identify standard solutions (see
chapter 3.3.3).

3.3 Beware of your alternatives
As mentioned in Chapter 3.1, a reinvestment analysis must provide an answer to the

following question:
Should the asset(s) in question be replaced and /or modified during the period of analysis or not?
Initial evaluation
Update plans
solutions

Figure 3.4 Alternatives include reinvestment (including standard solution) or no reinvestment
For a given asset or a group of assets, the distribution company has several choices that
should be considered:
− No reinvestment
o Postpone the reinvestment. Keep the asset in service some more years
− Reinvest now
o Total / Partial reinvestment. The reinvestment can be a solution
dedicated to the problem at hand or a standard solution applied to
previously analysed problem.
In addition, to remove the asset might be an eligible option. The different alternatives
will represent different risk (and costs) for the company.
Postponing the reinvestment should always be one of the alternatives for the analysis, if
necessary including minor measures to reduce risk such as increased maintenance,
restrictions in operation or minor replacements or modifications. This alternative can be
referred to as the ‘reference alternative’, and will work as a basis of comparison for other
alternatives.
When considering postponing the reinvestment, it is necessary to have the time

perspective in mind; i.e. considering for how long this is a viable option:
For how long is it possible to postpone the reinvestment and does this require some
specific effort (e.g. increased maintenance or operating restrictions)?
3.3.1 Identifying possible reinvestment alternatives

Before starting to evaluate the different reinvestment alternatives one should consider
their overall properties and check whether they are eligible options or not. If it is clear
that certain alternatives will not represent a preferred solution, these can be ruled out
without further analysis. Alternatives which are regarded unacceptable in terms of risk
must be modified to include measures which render them acceptable.
Considering the alternatives listed previously, the answers to the following questions
will provide a good overview of which alternatives should be further evaluated:
− Is reinvestment now a possible and good alternative? Why would we like to reinvest
now?
− Is a standard solution applicable? (see chapter 3.3.3)
− Is partial reinvestment a possible and good alternative? Why?
− Is to ‘remove’ the asset a possible and good alternative? Why?
− Is it possible to postpone the reinvestment? For how long? Are immediate measures
needed?
The answer to the questions above will of course vary depending on the asset(s) at hand,
but some general characteristics can be stated.
The main objective to postpone reinvestment is to exploit the assets lifetime, hence
reducing cost by incurring them later on. This has to be weighted against the positive
effects of reinvestment; mainly associated with the improvement of condition, better
design or more appropriate capacity and /or network configuration. Simply put, the
main drivers for the different alternatives could be summarised as:
Main drivers for reinvestment now (total or partial)

− Improvement of condition, design, capacity, placement
− Reduced costs / use of resources: interruption costs, maintenance, cost of losses
Main drivers for postponement

− Reducing short-term investment cost
Main drivers for removal

− Simplifying the network; reducing costs. Asset no longer ‘necessary.’

EXAMPLE 11 Reinvestment alternatives for a MV/LV substation
For a MV/LV substation the reinvestment alternatives could roughly be described as

follows:
- Reinvest
o Rebuild the entire substation; including replacement of all components
inside – consider placement, capacity and possibility to remove the
substation
o Renovate the building; replace all components inside
- Partial reinvestment
o Renovate the building and / or replace or modify some of the
components inside; switch disconnectors, cable terminations,
transformer, low voltage system
- Remove the substation and transfer the customers to another station
- Postpone the reinvestment. Keep the substation in service

Some relevant measures might be:
o Put restrictions on the operation of switch disconnectors
o Reduce the maintenance intervals; perform maintenance more often and
use more comprehensive maintenance activities
o Consider reinvestment again after 5 years

EXAMPLE 12 First screening of alternatives
For a MV/LV substation the answers to ‘the first screening questions’ presented in
chapter 3.3.1 could be:
Reinvestment now?
Yes, it will improve the condition, resulting in reduced probability of unwanted
events which will improve safety and quality of supply. It will give a long-term
solution.
Partial reinvestment?
Yes, the switch disconnectors and cable terminations are in worse condition
than the rest of the station. It is an alternative to replace only the switch
disconnectors and cable terminations. This is less costly than to reinvest the
entire substation; but will extend the substations lifetime shorter than is the case
for ‘full’ reinvestment.
Removal?
No, this is not considered to be a viable alternative. Nearby substations does not
have the capacity to take over the supply.
Postponement?
Yes, it is an alternative to postpone the reinvestment for 5-10 years. We can ‘save
money’ by reinvesting later. The risk is considered to be acceptable given
condition monitoring each fifth year to reveal further degradation
3.3.2 Reference alternative and time horizon
When comparing reinvestment alternatives, it important to keep in mind:

− What is the reference alternative?
− What is the period of analysis (time horizon)?
This is important both when estimating risk and performing cost-benefit analysis.
Answer to questions such as ‘What is the risk?’ and ‘What are the costs?’ will of course
depend on what time horizon is used. The risk associated with an asset may be
acceptable at the moment, but expected to evolve into an unacceptable level during the
next 10 years.

We here emphasise the need to define a reference alternative (often referred to as

alternative 0), which will work as a basis of comparison of the other alternatives. It is
important that this is a viable alternative for the whole period of analysis. If necessary,
the reference alternative must include minor measures in order to be acceptable in terms
of risk.
We recommend to use ‘Postponing the reinvestment’ as the reference alternative for all
reinvestment analysis.
EXAMPLE 13 Reference alternative and period of analysis
The table below shows two different choices of period of analysis (for the same
reinvestment) and corresponding reference alternatives and their associated cost.
Period of analysis Reference alternative Investment cost
5 years Postpone the reinvestment 0

(keep asset in service).
20 years Postpone the reinvestment. 20 000 Euros in year 10

Expected need to replace most
deteriorated components in year 10.
As we can see from the example, the choice of reference alternative must be consistent
with the period of analysis. In this case, to postpone the reinvestment entirely
(investment cost 0) is not a viable option if the period of analysis is 20 years, thus the
reference alternative must include some reinvestments costs.
In many cases it can be advantageous to use a short period of analysis (5-10 years) for
the reinvestment analysis. The advantages include:
− For a short period of analysis to postpone the reinvestment is a viable alternative
− Assumptions regarding risk, failure probabilities etc. is based on observations
from condition monitoring, and updated information is typically gathered every
5 or 10 years (in some cases even more often)
− The most important question to answer is whether to reinvest in the coming few
years or not. If reinvestment can be postponed even further, only a rough
estimate of when is needed (as a new analysis will be performed when
reinvestment is considered again)

3.3.3 Standard solutions and standard analysis
Due to the vast amount of assets in the distribution system, it is advantageous to define
reinvestment strategies which resolve typical worries for groups of assets. This can be
done by identifying standard reinvestment solutions, for example based on expert
opinion, and estimated years on reinvestment for the company’s different assets.
The question to be asked is:

What are typical reinvestment projects in our area the next 5 years? Further ahead?
The answer to the first question may for example be:

− OH-lines with many rotten poles
− Pole mounted substations operated from platform
− Substations with switch disconnectors of a certain type and condition
− Underground cables assumingly approaching their end of life (based on their age)
Once typical reinvestment projects are identified, one should seek to find standard
solutions.
− What are typical worries with the existing solution?
− Can simple measures resolve these worries?
− What are typical reinvestment alternatives and their associated effects?
The answer to these questions is a basis to define preferred solutions to identified

problems, i.e. define reinvestment strategy for groups of asset.
It will also be profitable to use such a ‘standard reinvestment’ analysis as a basis to do an

analysis for a specific asset or group of assets, even if a standard solution doesn’t apply.

3.4 Evaluation of alternatives
Evaluation of reinvestment alternatives comprises three main steps: Initial evaluation,

establish alternative solutions, and evaluation of alternatives.
Initial evaluation
Update plans
solutions

Figure 3.5 Initial evaluation, establish solutions and evaluation of solutions
The initial evaluation is a coarse analysis of the asset or assets at hand (in order to decide
if further analysis is necessary), while evaluation of different alternatives requires a
more detailed approach.
Proposed methods include risk analysis and cost-benefit analysis; the latter can be given
less emphasis in the initial evaluation. Uncertainty is an important aspect when
evaluating both the existing solution and different reinvestment alternatives, and we
advocate that this must be addressed both in the risk analysis and the cost-benefit
analysis.
In some cases there is a need for technical analyses, e.g. of power flow and voltage to
ensure that the alternatives represent technically good solutions, but this is not further
discussed here.
3.4.1 Initial evaluation
The purpose of the initial evaluation of the asset(s) is to determine whether reinvestment
can be postponed, or if there is a need for further analysis. Risk analysis, based on
information about the assets’ condition is a vital part of the initial evaluation.
The initial evaluation of the asset or group of assets at hand will lead to one of two main
conclusions, as indicated in Figure 3.5.

1) No reinvestment or standard solution

a. Reinvestment can be postponed. May include simple measures7
b. A standard solution is applicable
2) There is a need for further analysis

- Alternative solutions must be established and evaluated
In some cases, especially if the reinvestment analysis is triggered due to age, one can
relatively easily determine whether reinvestment can be postponed or further analysis
are required. This presupposes good and updated information about the object
considered for reinvestment.
EXAMPLE 14 Desk-top evaluation
Consider a MV/LV substation, where the building has reached the age of 60 years,
triggering a reinvestment analysis. Reports from condition monitoring and other
relevant information show no indication of circumstances that call for action. The
reinvestment can be postponed without further analysis, but a thorough control of the
building by a construction engineer should be performed as a part of the next
scheduled maintenance (in 3 years).
In addition to information about condition, design etc, other relevant circumstances

witch affect the need for reinvestment may include:
− Complaints from the general public (placement, aesthetics, quality of supply)
− Negative publicity in the press
− Registered / expected deviations regarded quality of supply (interruptions,
voltage problems etc)
− Removal or replacement of asset should be considered
o Obstruction of traffic / new or planned constructions
o Redundant asset
o Other circumstances
− Need for increased capacity / reinforcements.
− Window of opportunity (possibility to coordinate reinvestments with other
projects in the area)
We suggest a list of questions that should be addressed in order to reach one of these
two conclusions, see chapter 3.4.2. The importance of good and updated information
7Simple measures may for example include minor modifications, replacement of non-costly
components and / or increased preventive maintenance.
about the asset(s) condition is emphasised, and it might be necessary to do an ‘in field’
evaluation in order to acquire this information.
3.4.2 Suggested process for initial evaluation
We here recommend some key questions that should be addressed during the initial
evaluation.
The answers to these questions will reveal if there is a need for further analysis, and
identify possible reinvestment alternatives.
1) Trigging What triggered the reinvestment analysis?
2) Existing plans Do we have existing plans concerning the asset?
3) Standard solution Is a standard solution applicable?
4) Mapping of the asset condition

What are the condition of the asset and its main components?
Summary of relevant condition information
5) Economy and other relevant reasons to consider reinvestments

Are there any economic reasons to consider reinvestment?
High maintenance costs? High CENS? High losses?
Are there other circumstances which influence the need for
reinvestments?
6) Summary of risk
What are the main identified risks and what are possible measures?
7) Is there a need for further analysis?
No No need for reinvestment

Simple measures are sufficient
If no: What measures are needed?

When should reinvestment be reconsidered?
Yes
If yes: Which alternatives should be further analysed?

EXAMPLE 15 Initial evaluation
The following example shows an initial evaluation of a MV/LV substation. The

numbering corresponds to the suggested process presented in chapter 3.4.2.
1) What triggered the reinvestment analysis?

− Results from condition monitoring: Oil leakage on cable termination
− Expected year of reinvestment: Based on age and type of the switch
disconnectors
− Unfortunate / unwanted design: Old switch disconnectors of type RB NEBB.
Operation of such switch disconnectors represent a certain safety risk
2) Do we have existing plans concerning the asset?
3) Is a standard solution applicable?
None existing plans or standard solutions
4) What is the condition of the asset and its main components?
Condition indicator8 Comment

Building 2: some indication of Some graffiti and minor ‘wear and tear’.
degradation
Switch disconnectors 2: some indication of Switch disconnectors not tested.
degradation
Cable terminations 4: critical Substantial leakage / perspiration one of the oil-
filled cable terminations
Low voltage system 1: ‘as good as new’
Transformer 1: ‘as good as new’
Summary: The condition of the cable terminations is not acceptable (minimum one cell
must be replaced). The rest of the substation is in good condition, except from the
roofing paper which must be redone.
5) Are there any economic reasons to consider reinvestment?

High maintenance costs? High CENS? High losses?
Are there other circumstances which influence the need for reinvestments?
No particular circumstances.
8The numbers 1-4 correspond to the condition index developed by SINTEF Energy Research.
The index also includes a 5th state which implies fault (see chapter 1.2.1).

6) What are the main identified risks and what are possible measures?
Risk Comment Measures

Safety x Condition, cable terminations
Replace one or all cable
Quality of Condition, cable terminations: terminations
x
supply Increased failure probability (low CENS)
Old switch disconnectors without New switch disconnectors
Safety x encapsulation. Open bus bars. Uncertainty Functional test / revision
concerning operation of switch disconnectors Restrictions on operation
Safety x Open low voltage system Shielding of LV system
Reputation x Graffiti on door, signs, building Remove graffiti
7) Is there a need for further analysis?
Yes: Wish to look further into the following alternatives:
− A0: Postpone reinvestment

Necessary measures include replacement of cable termination
with leakage and increased maintenance of the switch
disconnectors. Restrictions on operation must be considered.
− A1: New switch disconnectors and cable terminations
For both alternatives: Shielding of LV system and removal of graffiti.
3.4.3 Evaluation of alternative solutions
Once alternative reinvestment solutions are identified (based on the initial evaluation;
see also chapter 3.3.1), the expected effects of these must be estimated and compared to
the reference alternative ‘postpone reinvestment’.
The evaluation of alternative solutions shall follow the same approach and use the same
methods and criteria as the initial evaluation. Keep in mind what triggered the analysis
in the first place, and the results of the initial evaluation.
Risk analysis
The risk analysis performed in the initial evaluation, is the basis to evaluate how
alternative solutions affect the risk.

For each alterative one must determine whether and how it affects the risk compared to
today’s solution:
− What unwanted events are affected?
− Does the probability change?
− Does the consequence change? What type of consequence?
It can be useful to use illustrations such as in the following example to visualise the
effect of different alternatives.
EXAMPLE 16 Illustrating changes in risk for different reinvestment alternatives
This example shows how changes in risk can be illustrated graphically. The analysis
corresponds to the initial evaluation of the MV/LV substation described in example 15,
and considers the following alternatives:
− A0: Postpone reinvestment

− A1: New switch disconnectors and cable terminations
A0 includes minor measures to reduce the risk, but as we can see, the risk reduction is
less than for A1.
Risk Now A0 A1 Comment
Safety
A0: Replacement of cable termination with leakage
Quality of
supply A1: New cable terminations
Safety A0: Functional test of switch disconnectors. Revision/restrictions if needed
Safety A0 / A1: Shielding of LV system
Reputation A0 / A1: Removal of graffiti
Cost-benefit analysis
In a cost benefit analysis, different alternatives are compared in economic terms. The
following elements must be considered:
− Investment cost and remaining value
− Maintenance costs
− Cost of losses
− Interruption costs; Cost of energy not supplied.

Reinvestment decisions will in general consist of a trade-off between immediate cost

(investment) and potential future benefits (reduction in risk, future maintenance costs,
future interruption costs etc). To adjust for the fact that the cost elements occur at
different points of time, the concept of present value is used (Brealey and Myers, 2003)9.
Some general recommendations regarding cost-benefit analysis for reinvestments are
listed in (Nybø et al., 2009) and include:
− Evaluate economical effects of alternative reinvestment solutions by performing
cost-benefit analysis.
o Use the principle of present value
o Compare different alternatives with the reference alternative (‘postpone
the reinvestment’)
− Establish a company standard for cost-benefit calculations including:
o General parameters such as time horizon, discount rate
o How to calculate and document cost elements
o How to perform sensitivity analysis
o How comprehensive the analysis should be; this can differ depending on
the project
o Whether and how to include effects such as regulation and taxes
− Beware of the fact that not all effects of reinvestment can be expressed in
economical terms, and that a cost-benefit analysis must be supplemented by an
analysis of unwanted events (risk analysis)
− Incorporate simple sensitivity analysis10 in the cost-benefit analysis: Highlight the
parameters which are the most uncertain and will influence the net present
values of reinvestment alternatives the most. In many cases this will be the
failure rate.
Key questions to be asked include:

− How much can we gain by postponing the reinvestment costs?
− Given postponement: How much higher are the interruption costs, maintenance
costs and cost of losses compared to reinvestment now?
− Given postponement: How much higher is the risk compared to reinvestment now?
9Norwegian readers can read about present value in ‘Planleggingsbok for kraftnett,’
www.planbok.no
10 Norwegian readers can read about sensitivity analysis in ‘Planleggingsbok for kraftnett,’
www.planbok.no.
EXAMPLE 17 Presenting results from a cost-benefit analysis
Figure 3.6 shows how main results from a cost benefit analysis can be illustrated. A0
(postponement) requires some minor measures, but the investment cost is much less
than for A1 (reinvestment now). On the other hand, A0 will give higher maintenance
costs and interruption costs. In this example, the costs of losses are not affected by the
reinvestment.
120.000
100.000
80.000
60.000
40.000 Interruption
Losses
20.000
Maint.
I-R
0
A0 A1
Figure 3.6 Present value of cost the next 5 years for the reinvestment alternatives
A0 (postponement) and A1 (reinvestment now).
I-R is the investment cost corrected for remaining value.
If the decision maker’s only criterion is net present value, A0 will be preferred. But
risks and uncertainties not included in this calculation must also be considered. In
general, postponement of reinvestment will imply higher risk and uncertainty (about
the assets actual condition, failure rate etc) than reinvestment now, and these are
factors which might lead the decision maker to prefer A1 in this example.
Uncertainty
There will always be uncertainty associated with the estimated effect of different
solutions, and it is important to describe this in the analysis.
Examples include:
− Uncertainty concerning the estimation of asset’s condition and the estimation of
effects on safety, reputation, quality of supply etc.
− Uncertainty concerning the estimation of future interruption costs, maintenance
costs, cost of losses and investment costs.
− Uncertainty concerning load development or other factors which are relevant for
the reinvestment decision

Such uncertainties should be described, highlighting factors with large impact on the
estimated risks and costs. Uncertainty concerning the present value of alternative
solutions can be quantified and visualised through sensitivity analysis.
In some cases it may be favourable to take measures to reduce the uncertainty, e.g. to
perform (a new) thorough condition monitoring if there are doubts concerning assets
condition.
3.4.4 Decision making
As outlined in this chapter, reinvestments typically involve different decision criteria,

and how the different alternatives scores according to these criteria (as shown in
example 16) must be considered.
A summary of the existing solution and associated risk, and changes in risk as a result of
different reinvestment alternatives (see example 16, page 36) and a summary of
associated cost (see example 17, page 38), will be a good basis for making decisions. The
different decision criteria must be considered together and the best overall alternative
selected
There exist multi criteria decision making methods which can be used to aid the process
of merging different criteria into comparable utility values for the alternatives. Such
methods are not further elaborated in this report, but the interested reader is referred to
e.g. (Catrinu et al., 2007).

3.5 Summary
This chapter has described and exemplified a structured approach to reinvestment

analysis, including:
− When to trigger a reinvestment analysis
− Which reinvestment alternatives to consider
− How to evaluate different alternatives
− How to document the analysis

− Identify the right assets to analyse
o Define triggering events reflecting risk
− Beware of your alternatives
o Should we reinvest in the coming period of analysis or not?
o Maintenance is often an alternative to reinvestment (but not forever)
o Establish standard solutions to ‘known’ problems concerning groups of
assets
− Establish company standards for performing and documenting reinvestment
analysis

4 Implementation and improvement
This chapter describes some important aspects regarding how to implement risk based
maintenance strategies, trigger reinvestment analysis and establish a culture of continual
improvement. The need for data, in particularly from condition monitoring is given
special attention.
Recommendations include:
− Create an overview over what kind of assets you have and allocate them to their
corresponding maintenance strategy and plan
o Systematically register and update necessary information about your
assets
o Set up maintenance plans, also addressing need for reinvestments
− Use condition monitoring data as input to maintenance and reinvestment
management
− Evaluate and improve strategies and processes
Chapter 4.1 address the first bullet point, while chapter 4.2 promotes the importance of
condition monitoring and chapter 4.3 describes how to continually improve strategies
and processes.
4.1 From maintenance strategies to plans
In order to allocate the different assets to their appropriate maintenance strategy (see
chapter 2), information about selected parameters must be registered. E.g. if age,
condition, type and CENS/h is decisive for the maintenance of MV cables, these
parameters must be found in the assets database in order to allocate the cables to their
appropriate strategy and set up maintenance plans (as illustrated in Figure 4.1).
Maintenance
strategy
Allocation of assets to Maintenance
Asset appropriate strategy plan
database
Figure 4.1 From maintenance strategy to maintenance plan

Once assets are allocated to their appropriate maintenance strategy, maintenance plans
can be derived, both for individual assets and for groups of assets.
As illustrated in Figure 4.2, the planning horizon typically varies between short
(typically one budget year), medium (5 - 10 years), and long (20-30 years). The level of
detail will be quite high for the short / medium term plans, while more rough for the
long term plan.
Budget
year
5 year plan Long term plan (20-30 years)

year
Detailed  Rough
Figure 4.2 Planning horizon
Typically, many of the same parameters used to differentiate maintenance are also used
to trigger reinvestment analysis. In order to implement the concept of triggering criteria
in practice, the network companies have to define a routine for how to compare assets’
properties against triggering criteria.
Maintenance strategy:
Trigger criteria
Reinvestment
analysis
Asset database
Figure 4.3 Trigging a reinvestment analysis

4.1.1 Acquiring necessary information
In many cases, information needed to differentiate maintenance and trigger

reinvestment analysis, is lacking, and in this case, one should start with what is available
and begin to systematically collect the missing data, for example as part of the next
scheduled inspection.
A first step will be to:

− Identify information needed in order to implement maintenance strategies for
different system units11
o What data is needed and what data is missing? How valuable is this
information? How difficult is it to obtain?
− Create a preliminary overview (categorisation) of your assets and their
associated maintenance strategy based on available information
The most important is to identify the ‘high risk’ assets, to ensure that they get
appropriate maintenance and reinvestment attention. In addition, assets which represent
very ‘low risk’ should be identified in order to prevent to over-spend resources here.
The level of detail can be gradually refined as better information is gathered, and
therefore should not the lack of data prevent the implementation of strategies.
4.1.2 Examples
The following examples highlight important aspects regarding the implementation of

maintenance and reinvestment strategies:
− Necessary information in order to differentiate maintenance and get an overview
of your assets
− Trigger reinvestment analysis and addressing future reinvestment needs
− Setting up maintenance plans
11 This should also be kept in mind when defining the maintenance strategies.

EXAMPLE 18 Necessary information in order to differentiate the maintenance

of MV/LV substations
A group of network companies has developed a risk differentiated maintenance

strategy for their MV/LV substations where the intervals between different
maintenance activities depend on the following parameters:
− Switch disconnector type and voltage level
− Environment, switch disconnectors
− Encapsulation, switch disconnectors
− Age, switch disconnectors
Table 4.1 summarises key properties for a MV /LV substation. Based on information
about the switch disconnectors, which in this case is decisive for the maintenance, the
substation can be assigned to its appropriate maintenance programme and time for
next condition monitoring and so forth can be derived. A rough estimate regarding
reinvestment need is also included, which is useful in order to trigger reinvestment
analysis and for the long term planning. The term ‘estimated year of reinvestment’ is
described on page 46.
Table 4.1 Summary of key properties for a substation.
Substation nr 5012
Switch disconnector type NAL - 12 kV

Encapsulation Fully
Environment Clean
Installation year 1980
Maintenance programme A2 – Low risk
Scheduled inspection (last / next) 2009 2010
Scheduled ‘thorough inspection’ (last / next) 2007 2012
Scheduled functional test (last / next) 2000 2020
2030
Estimated year of reinvestment
First estimate (age)
Factors influencing reinvestment needs None known

EXAMPLE 19 Creating an overview of assets
Based on their maintenance strategy (referred to in the previous example) and a

population of their substations with air insulated switch disconnectors, a distribution
network company has performed a mapping of the number of substations within each
category (Figure 4.4).
The overview shows that a large portion of their substations are associated with low
risk and require little maintenance attention (indicated by shades of green), whereas a
small number of substations will be subjected to more comprehensive maintenance
(indicated by shades of orange / red). In other words, the company has a substantial
potential to reduce risk by giving just a small number of stations extra maintenance
attention. In addition, there is a substantial saving potential in terms of reducing
maintenance for low risk substations.
Encapsulation
Environment Age
Fully Semi Open
Exposed > 24 years 11 14 1
Exposed ≤ 24 years 36 1 0
Clean > 24 years 161 252 1
Clean ≤ 24 years 337 15 0
Figure 4.4 Overview of number of substations within each category.

Age is here used as one of the factors for risk differentiation. When good condition
information exists, this will provide better information about risk than age.

EXAMPLE 20 Addressing reinvestment needs: Estimated year of reinvestment
A group of network companies has defined ‘estimated year of reinvestment’ as ‘the

year it’s most likely that a component must be replaced.’ The idea is to use this to
trigger reinvestment analyses.
The estimated years of reinvestment are assigned to different assets. E.g. for a MV/LV
substation the estimated year of reinvestment12 is assigned to the different objects
building, switch disconnectors, low voltage system etc. When one or more objects in a
substation approach their estimated year of reinvestment, a reinvestment analysis of
the station is trigged.
A first estimate of the year of renewal can be established based on age, type / design,
and environment. As time passes, the estimated year of renewal must be updated
based on results from condition monitoring and other relevant information. An
example is shown in Figure 4.5. For assets close to their estimated year of reinvestment,
technical condition and load development are decisive for the estimated year of
reinvestment. Further ahead, it’s ok to use an initial (rougher) estimate based on age.
45
40
35
Number of assets
30
25
20
15
10
0
10 12 14 16 18 20 22 24 26 28 30
Year
Condition Age Load development Other
Figure 4.5 Estimated year of renewal for a number of assets
In addition to triggering reinvestment analysis, the concept of estimated year of

renewal is also useful in order to develop a long term reinvestment plan and to trigger
condition monitoring. The latter can be done by prioritising assets which approaches
their estimated time of renewal for thorough condition monitoring.
12The year of renewal may be different for different objects in the substation, e.g. the switch
disconnectors might have estimated year of renewal 2015, while the estimated year of renewal for
the building is 2025.
EXAMPLE 21 Maintenance plans
Figure 4.6 shows planned maintenance activities (condition monitoring) for two
selected substations. Plans (and some history) regarding inspection, thorough
inspection and reinvestment need assessment is shown. The plans for condition
monitoring may change as a result of the reinvestment need assessment, and is
therefore not illustrated after this point in time.
History Today 5 years 10 years
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Substation 5012 R
Substation 5013 R
Inspection
Thorough inspection
Reinvestment need assessment
R Estimated year of renewal
Figure 4.6 Maintenance plan for two selected substations
Based on maintenance plans for individual assets (as shown above), more aggregated
plans for groups of assets or the whole asset base can be generated. Such plans can be
useful input to resource allocating, budgeting etc.
5 year m aintenance plan

Number o f for 200 substations
statio ns
200
150
100
50
0
10 11 12 13 14
Year
Reinvestment need assessment

Tho ro ugh inspectio n
Inspectio n
Figure 4.7 5 year maintenance plan (condition monitoring) for a population of 200 substations

4.2 The value of condition monitoring
We have previously emphasised how important it is to have good information about the
condition of assets in order to do sound decisions about maintenance and reinvestments.
As an asset deteriorates it represents a higher risk in terms of an increased probability of
failure, and information about assets’ condition is the main reason to perform
maintenance and / or trigger reinvestment analysis.
Condition monitoring represent a large part of the network companies maintenance

activities, and serves several purposes:
− Find and correct deviations
− Assess the condition of components. Reveal need for reinvestment /
comprehensive maintenance.
− Gather other relevant data
Depending on the purpose, there is a need for different kinds of condition monitoring.
In general, there is need for a rather thorough condition monitoring in order to identify
reinvestment needs.
Assets assumingly approaching their end-of-life can be prioritised for thorough

condition monitoring based on available information on condition, failure history, age
etc.
EXAMPLE 22 Condition monitoring triggered due to failure
The underground cable networks in cities are a main concern for many distribution
network companies for several reasons:
− Many of the cables are old
− Missing documentation: Both age, type and condition is in many cases
unknown
− ‘Unavailability’: Digging imply large costs and much administration
After several failures in an urban area, the distribution network company decided to
prioritise several cables in this area for thorough condition monitoring in order to
reveal the need for reinvestment. Such condition monitoring is relatively costly, but on
the other hand, more information about the cables’ condition will help the company
make a better decision and avoid reinvesting too early.

EXAMPLE 23 Detecting rot in poles
Rot development in wooden poles is a challenge, and many poles are replaced because
their strength is reduced due to rot. Several methods exist to detect and measure rot,
from simple screening techniques using a hammer or a special train dog, to more
sophisticated methods to measure the extent of rot.
For lines assumingly approaching reinvestment, sophisticated methods to measure rot

(and also other methods to assess the condition of the line as a whole) are needed. In
other cases, screening techniques may be sufficient. Where the screening indicates rot,
the pole can be more thoroughly controlled
4.3 Continual improvement
The principle of continual improvement is advocated as a fundamental part of asset

management in numerous disciplines – see e.g. (Deming, 2000, ISO, 2005, British
Standards Institution, 2008a).
The goal of continual improvement is to ensure that the organisation learns over time
and uses this new knowledge to improve their asset management practice13. This is done
through accumulating knowledge, learning and adopting to change, and includes e.g.:
− Acquiring and adapting new knowledge about ‘best practice’
− Adapting to new recommendations and demands
− Updating risk assessments based on new knowledge, information or priorities
− Identifying and implementing standard solutions to recurring maintenance and
reinvestment challenges
To make it practically applicable, it is important to start the process of continual
improvement with a relatively simple approach, e.g. to:
− Establish an asset management team which is responsible for implementation
and follow up (continual improvement) of maintenance and reinvestment
management
− Create meeting areas for this group. They should gather at least 1-2 times a year
with the purpose of identifying areas of improvement
o Go through a list of questions that is regarded as important in order to
reach company goals regarding maintenance and reinvestments
o Compliment with simple analysis of selected indicators.
13 This section is heavily based on (Nybø and Nordgård, 2010).

A potential pitfall is to make the follow-up too complex and time demanding. It is hence
important to start with simple questions and analysis, and to link these with the
companies’ objectives and focus areas.
EXAMPLE 24 Follow-up questions related to maintenance and

reinvestment goals
Table 4.2 provides examples of follow-up questions related to specific goals regarding
maintenance and reinvestments (the list is by no means exhaustive). Such follow-up
questions should be asked regularly. Indicators may be used to monitor trends and
goal achievement.
Table 4.2 Goals and follow-up questions
Goal Follow-up questions

Implement maintenance strategies Have we implemented our maintenance strategies?
- If not, why? How can they be implemented?
- If yes, have we identified any improvement potential?
Compliance with rules and Are there any new rules and regulations affecting our
regulations strategies?
Do we know about deviations from current rules and
regulations?
Achieve a high level of safety Have there been any accidents or near accidents? Have we any
for our personnel and 3. party other indications concerning safety?
- Have this revealed unaddressed safety concerns?
Increased security of supply: Have vegetation management been performed according to

Improved vegetation management plan? (if not, why?)
in certain areas
Do we have any indications of reduced failure rate? (e.g. limited
damage from last storm, positive trend in failure statistics)

4.3.1 Indicators
Continual improvement is closely related to the term ‘indicator’. By combining different

indicators one can create a simplified description of the distribution system and assess
past and future performance. Some aspects concerning risk indicators (indicators
providing information about risk) and their use in distribution system asset
management is discussed in (Sand, 2009). This includes results from a survey among
distribution network companies in Norway, Finland and France on which indicators are
used today.
Using indicators to measure the effects of maintenance and reinvestments can be a

challenging task, because the costs incur now, while the benefits incur in the future and
are less visible (often in terms of a failure or accident not happening).
One way to meet this challenge is to distinguish between two groups of indicators
(based on (OECD, 2003)):
− Activities indicators
o Designed to help identify whether actions believed to lower risks are
taken
− Outcome indicators
o Designed to help measure whether such actions are, in fact, leading to
less probability and / or less consequences of unwanted events.
In other words; outcome indicators tell you whether or not you have achieved a desired
result, while activities indicators they you why the result was achieved or why it was
not.

EXAMPLE 25 Indicators ‘measuring’ the effects of reinvestments
Table 4.3 illustrates different indicators related to specific goals and activities. Note that
the activity indicator is easy to observe and connect to goals / activities, while the
outcome indicators will be observed over a longer period of time and will usually be a
result of many different activities (as well as stochastic factors such as weather).
Table 4.3 Activities indicators and outcome indicators
Goal Activity Activities indicator Outcome indicator

Increased security Replace first generation Km replaced Quality of supply index
of supply XLPE cables
Costs Focus on cable failures
and areas where cables
are replaced
Increased security Improved vegetation Areas / km with Quality of supply index
of supply management improved vegetation
management Focus on failures
related to vegetation
Costs and areas where
vegetation
management is
improved
Improved safety for Improve encapsulation for Numbers of improved Safety index
personnel switch disconnectors in encapsulations /
MV/LV substations
Costs
Quality of supply indexes are normally based on failure statistics, while safety indexes
are normally based on accidents statistics.
There exist models trying to describe the relationship between activities and outcomes,
for example level of maintenance and reinvestments, the condition of grid components,
and quality of supply indexes.

EXAMPLE 26 Indicator describing the need for reinvestments
The British regulator OFGEM has defined a health index and uses this to describe the
condition of selected components in the UK power system, including for example
transformers, switch gear and overhead lines. The health index uses a 1-5 scale to
describe assets’ condition, see Table 4.4.14
Table 4.4 Health Index Definitions
ID Description
HI1 New or as new
HI2 Good or serviceable condition
HI3 Deterioration requires assessment and monitoring
HI4 Material deterioration, intervention requires consideration
HI5 End of serviceable life, intervention required
The use of such indexes can be fruitful in order to get an overview of the network and
assets that require maintenance, reinvestment or investments on an aggregated level. It
may also be used to illustrate the expected benefits from reinvestment (represented by
improved health or slower deterioration).
In order to say more about risk, the health index may be combined with other indexes,
for example an index reflecting the components importance for the system15.
Figure 4.8 shows an example of the use of the health index for a group of High Voltage
switchgear. A group of such assets are categorised in three groups according to their
condition:
− Health index 3 or better (HI1-HI3)
− Health index 4 (HI4)
− Health index 5 (HI5)
The pie to the left shows the current situation (Year 0), the pie in the middle shows the
situation in year 5 without intervention, and the pie to the right shows the situation in
year 5 given that certain actions are undertaken. As we can see, with no actions taken,
the health of the population deteriorates with time; a higher percentage is classified
14 This index is closely related to the condition index described in chapter 1.2.1.
15OFGEM has also defined an index which describes the degree of utilisation for selected
components (using a scale ranging from ‘Significant spare capacity’ to ‘Fully utilised, mitigation
required’), see www.ofgem.gov.uk.
HI5 (‘End of serviceable life’). With investments, the health of the population
improves.
Year 5 Health Index Profile ‐ No Year 5 Health Index Profile ‐ With

Year 0 Health Index Profile
Intervention (% of assets) Investment (% of assets)
(% of assets)
4% 1%
6% 14 %
6%
6%
88 %
80 % 95 %
HI1 ‐ HI3 HI4 HI5 HI1 ‐ HI3 HI4 HI5 HI1 ‐ HI3 HI4 HI5
Figure 4.8 Health index distribution for a group of High Voltage switchgear
4.4 Summary
This chapter has described and exemplified different aspects regarding:

− How to implement maintenance strategies and to trigger reinvestment analysis
− The need for data and the importance of condition monitoring.
− Continual improvement and use of indicators.

− Identify information needed in order to implement maintenance strategies and
trigger reinvestment analysis
o What data is needed and what data is missing? How valuable is this
information? How difficult is it to obtain? If necessary; how shall we obtain it?
o Systematically register and update necessary information
− Create an overview of your assets and allocate them to their corresponding
o Set up maintenance plans, also addressing need for reinvestment analyses
− Use condition monitoring data as valuable input to maintenance and
reinvestment management
o Different needs require different condition monitoring results
− Evaluate strategies, processes, and results - and use this knowledge to achieve
continual improvement

References and recommended literature
References
BREALEY, R. A. & MYERS, S. C. (2003) Principles of corporate finance, Boston, Mass.,

McGraw-Hill/Irwin.
CATRINU, M., SAMDAL, K. & NORDGÅRD, D. E. (2009) General framework for

maintenance management. TR A6879, Trondheim, SINTEF Energy Research.
CATRINU, M. D., NORDGÅRD, D. E., SAND, K. & NORHAGEN, J. K. (2007) Multi-

criteria decision support in distribution system asset management. CIRED - 19th
International conference on electricity distribution. Vienna.
DEMING, W. E. (2000) Out of the crisis, Cambridge, Mass., MIT Press.
HEGGSET, J., SOLVANG, E., CHRISTENSEN, J. S., BERTLING, L., BAKKEN, K. R.,
ENGEN, H., HASSELSTROM, J. & PYLVÄNÄINEN, J. (2007a) Risk estimation in
power networks by using failure models for components, Trondheim, SINTEF Energy
Research.
HEGGSET, J., SOLVANG, E., WELTE, T. M., CHRISTENSEN, J. S. & BAKKEN, K. R.

(2007b) Assessment of remaining lifetime and failure probability for network
components - A practical approach. CIRED 2007. 19th International Conference on
Electricity Distribution. Vienna.
ISO (2005) ISO 9000:2005 Quality management systems

Fundamentals and vocabulary International Organization for Standardization.
ISO/IEC (2002) Guide 73. Risk management - Vocabulary - Guidelines for use in
standards. ISO / IEC.
NORDGÅRD, D. E. & SAMDAL, K. (2010) Establishing risk-based maintenance

strategies for electricity distribution companies. IN BRIS, SOARES, G. &
MARTORELL (Eds.) Reliability, Risk and Safety: Theory and Applications.
Proceedings of the European safety and reliability conference, ESREL 2009, Prague,
Czech Republic, 7-10 September 2009. Leiden, CRC.
NORDGÅRD, D. E., SOLUM, G., HEGGDAL, L., SJURSØ, A. B., FJELLSBØ, J. I. &
OMDAL, E. (2007) Establishing maintenance standards following a risk based
maintenance strategy. CIRED - 19th International conference on electricity
distribution. Vienna.

NYBØ, A., GJERDE, O. & NORDGÅRD, D. E. (2009) General framework for reinvestment
management. TR A6830. Trondheim, SINTEF Energy Research.
NYBØ, A. & NORDGÅRD, D. E. (2010) Principles for maintenance and reinvestments

management of electricity distribution assets. ESREL 2010. Rhodes, Greece.
OECD (2003) Guidance on safety performance indicators. OECD Environment, Health and
Safety Publications. Series on Chemical Accidents. Paris.
SAND, K. (2009) Risk indicators for distribution system asset management. TR A6787,
Trondheim, SINTEF Energy Research.
SAND, K., GJERDE, O. & NORDGÅRD, D. E. (2007) Current risk exposure in the
distribution sector. Initial study. TR A6576, Trondheim, SINTEF Energy Research.
Recommended literature
Introduction to Asset Management of infrastructures
BRITISH STANDARDS INSTITUTION (2008a) PAS 55-1: Asset Management.

Part 1: Specification for the optimized management of physical
infrastructure assets.
BRITISH STANDARDS INSTITUTION (2008b) PAS 55-2: Asset Management.

Part 2: Guidelines for the application of PAS 55-1.
Introduction to Risk Analysis
AVEN, T. (2008) Risk analysis. Assessing uncertainties beyond expected values and
probabilities, Chichester, Wiley.
KAPLAN, S. (1991) Risk assessment and risk management - Basic concepts and
terminology. Risk management - Expanding horizons in nuclear power and other
Industries. Hemisphere Publishing Corporation.

SINTEF Energi AS
SINTEF Energy Research
No-7465 Trondheim
Telephone: + 47 73 59 72 00
[email protected]
www.sintef.no/energy

Risk Dsam TR A6983

Uploaded by

Copyright:

Available Formats

Risk Dsam TR A6983

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Risk Dsam TR A6983

Uploaded by

Copyright:

Available Formats

A guide to

A deliverable from the RISK DSAM project

A guide to Risk Based Maintenance

Enterprise / VAT No.:

TR NO. DATE CLIENT’S REF. PROJECT NO.

TR A6983 2010-08-03 12X447

DocuLive 2,3,5 Eivind Solvang Open

978-82-594-3442-5 Knut Samdal 60 56

Energy Systems Sem Sælands vei 11 + 47 73 59 72 50

Setting the scene & Summary of recommendations

II The structure of the report II

III Summary of recommendations III

IV Checklist for distribution system companies VII

V Important definitions VIII

4 Implementation and improvement 41

References and Recommended literature 55

SINTEF Energy Research, 2010

Setting the scene

1RISK DSAM (Risk-based distribution system asset management) is a Competence Building

II The structure of the report

This report is divided into three main parts:

Setting the scene and summary of recommendations

Chapter IV provides a ‘check-list’ for distribution network companies on how to

Chapter 1 gives an introduction to risk based maintenance and reinvestment

Chapter 2 describes an approach for how to use risk assessment to establish

Chapter 3 describes a structured approach to reinvestment analysis, including when to

Chapter 4 describes different aspects regarding how to implement maintenance

SINTEF Energy Research, 2010 Page II

References and recommended literature

III Summary of recommendations

Common philosophy and objectives

Figure III.1 Maintenance and reinvestment management as two integrated processes

SINTEF Energy Research, 2010 Page III

This report deals with three main topic areas:

In the following, key recommendations to each of these areas are listed.

Maintenance management - How to establish maintenance strategies

Chapter 2 describes an approach for how to use risk assessment to establish

Maintenance strategies must describe:

SINTEF Energy Research, 2010 Page IV

Reinvestment management - How to perform reinvestment analysis

Chapter 3 describes a structured approach to reinvestment analysis, from triggering of

Identify the right assets to analyse:

Establish company standards for performing and documenting reinvestment

Beware of your alternatives:

SINTEF Energy Research, 2010 Page V

Implementation and improvement – How to set up maintenance and

Chapter 4 describes different aspects on how to implement risk based maintenance

Key recommendations include:

Identify information needed in order to implement maintenance strategies

Create an overview of your assets and allocate them to the corresponding

Use condition monitoring results as input to maintenance and reinvestment

SINTEF Energy Research, 2010 Page VI

IV Checklist for distribution system companies

Establish a common philosophy and objectives regarding maintenance and

Establish an asset management group responsible to enforce maintenance and

Develop risk differentiated maintenance strategies for different system units

Define criteria which trigger reinvestment analysis

Establish a standardised procedure regarding how to perform reinvestment

Establish dedicated condition monitoring and reporting in order to meet