Targeted Review of Asset Health

FINAL REPORT
Targeted Review of Asset Health and

Resilience in the Water Industry
PROC.01.0546
Prepared for
Date: September 2017

Contents
Section Page
Glossary of Terms .............................................................................................................................. v
Abbreviations .................................................................................................................................. vii
Acknowledgements........................................................................................................................... ix
Report authors.................................................................................................................................. ix
Introduction...................................................................................................................................... 5
1.1 Background ......................................................................................................................... 5
1.2 Objectives and approach .................................................................................................... 5
1.2.1 Working definitions of asset health and its context .............................................. 6
1.2.2 Establishing the link between asset health, performance, service and outcomes 6
1.2.1 Measurement of asset health ................................................................................ 6
1.2.2 The influence of asset health on expenditure planning and decision making ...... 7
1.2.3 Communication and the views of the customer ................................................... 7
1.2.4 Asset health assurance .......................................................................................... 7
1.3 Report structure.................................................................................................................. 7
1.4 Accuracy and omissions ...................................................................................................... 8
Understanding of asset health ........................................................................................................... 9
2.1 Context................................................................................................................................ 9
2.2 The questions we asked and their purpose ...................................................................... 10
2.3 What were we expecting to be told?................................................................................ 10
2.4 What were we told? ......................................................................................................... 10
2.4.1 Defining Asset Health .......................................................................................... 10
2.4.2 The relationship between ‘asset health’ and ‘resilience’ .................................... 12
2.4.3 Benefits of the current measures of Asset Health and improvement
opportunities ....................................................................................................... 15
Measuring and monitoring the health of our assets ......................................................................... 17
3.1 Context.............................................................................................................................. 17
3.2 Disclaimer ......................................................................................................................... 17
3.4 What we were expecting to be told ................................................................................. 17
What we were told ........................................................................................................... 18
3.5.1 Overview .............................................................................................................. 18
3.5.2 Asset data ............................................................................................................ 18
3.5.3 Data quality .......................................................................................................... 21
3.5.4 BIM (building information modelling) ................................................................. 21
3.5.5 Analysing and presenting data ............................................................................ 21
3.5.6 How data is used .................................................................................................. 23
3.5.7 Inspection, monitoring, control and analysis ...................................................... 24
3.5.8 Innovation in technology ..................................................................................... 28
Asset health, expenditure planning and decision making.................................................................. 29
4.1 Context.............................................................................................................................. 29
CH2M i
4.4 What we were told ........................................................................................................... 30
4.4.1 Linking asset health to service ............................................................................. 30
4.4.2 The tools and methods used to create linkages between asset health and service
and assess risk...................................................................................................... 31
4.4.3 How are asset health measures used to inform timely and effective
expenditure? ........................................................................................................ 34
4.4.4 The role of criticality ............................................................................................ 38
Understanding the views of the customer ........................................................................................ 41
5.1 Context.............................................................................................................................. 41
5.4 What we were told ........................................................................................................... 42
5.4.1 Customer dialogue and asset health ................................................................... 42
5.4.2 Effective language to convey asset health concepts to customers ..................... 44
5.4.3 The value customers place on asset health ......................................................... 46
Assuring our approach to asset health and maintaining service ........................................................ 48
6.1 Context.............................................................................................................................. 48
6.2 Successful delivery of the MoS ......................................................................................... 49
6.2.1 Delivering on asset health commitments: ........................................................... 49
6.2.2 The main factors for success: .............................................................................. 49
6.3 General assurance (process, data, tools) .......................................................................... 50
6.3.1 What we were told – key observations: .............................................................. 51
6.3.2 Findings and comparison across companies ....................................................... 52
6.4 Emergency response and dealing with asset failure ........................................................ 52
6.5 New assets ........................................................................................................................ 54
Incentives and barriers .................................................................................................................... 56
7.1 Context.............................................................................................................................. 56
7.2 What were we expecting to be told ................................................................................. 56
7.3 What were we told? ......................................................................................................... 56
7.3.1 TOTEX and outcomes ........................................................................................... 56
7.3.2 Innovation barriers .............................................................................................. 57
Lessons from international practice ................................................................................................. 59
8.1 Introduction ...................................................................................................................... 59
8.2 Understanding of asset health and its relationship with service ..................................... 59
8.3 ‘State of the Assets’ reporting .......................................................................................... 60
8.4 Inspection and monitoring technologies .......................................................................... 61
8.5 Big data ............................................................................................................................. 62
8.6 Deterioration and failure forecasting ............................................................................... 63
8.7 Encouraging innovation .................................................................................................... 63
Discussion and conclusions .............................................................................................................. 65
9.1 Understanding asset health .............................................................................................. 65
CH2M ii
9.1.1 What does asset health mean? ........................................................................... 65
9.1.2 Asset health and resilience .................................................................................. 65
9.1.3 Benefits of the current measures ........................................................................ 66
9.1.4 What would a good measure look like? .............................................................. 66
9.2 Measuring and monitoring the health of our assets ........................................................ 67
9.2.1 Data ...................................................................................................................... 67
9.2.2 Technology ........................................................................................................... 68
9.3 Asset health and decision making .................................................................................... 69
9.3.1 Linking asset health and service .......................................................................... 69
9.3.2 The tools for linking asset health, service and risk .............................................. 70
9.3.3 Using asset health to inform expenditure ........................................................... 71
9.3.4 Criticality .............................................................................................................. 71
9.4 The views of the customer................................................................................................ 72
9.4.1 Is there an active dialogue with customers? ....................................................... 72
9.4.2 Has a language for conveying asset health concepts been developed? ............. 72
9.4.3 Do customers value asset health? ....................................................................... 72
9.5 Asset health assurance ..................................................................................................... 73
9.5.1 Delivering asset health commitments ................................................................. 73
9.5.2 Assurance processes ............................................................................................ 73
9.5.3 Asset failures ........................................................................................................ 74
9.5.4 Commissioning new assets .................................................................................. 74
9.6 Incentives and barriers ..................................................................................................... 74
9.6.1 TOTEX and outcomes ........................................................................................... 75
9.6.2 Innovation barriers .............................................................................................. 75
9.7 Lessons from international practice ................................................................................. 76
9.8 Summing up ...................................................................................................................... 76
Appendices
Appendix A: Headline findings from each company interview
Appendix B: Summary of key technologies reported by water companies
Appendix C: Linking asset heath, service and investment
Appendix D: Inspection and monitoring
Table(s)
Table 3-1: Typical asset data
Table 5-1: Customer understanding of Asset Health
Figure(s)
Figure 2-1: Phrases companies used to describe what asset health means to them
Figure 2-2: A conceptual relationship between asset health and the ‘four R’s’ of resilience
Figure 2-3: Understanding and measuring resilience (Severn Trent Water perspective)
Figure 3-2 Tableau report example (Welsh Water example)
Figure 3-3 Data usage: the temporal dimension
Figure 4-1 : The Link between asset health, service and outcomes (Yorkshire Water)
CH2M iii
Figure 8-1: Pure Technologies' SmartBall
Figure 9-1 Nephnet device
Figure 9-2 Condition assessment of ferrous pipes
Figure 9-3 Failed electro-fusion joint
Figure 9-4 Phenolphthalein testing on cement pipes
Figure 9-5 Trunk mains condition assessment
Figure 9-6 Sahara leak detection
Figure 9-7 Magnetic flux leakage device
Figure 9-8 Schematic of ePulse device
Figure 9-9 Hawkeye installation
Figure 9-10 Sonar device
Figure 9-11 Laser scanning
Figure 9-12 SewerBatt™ tool
Figure 9-13 Electroscan schematic signal output
Figure 9-14 P-F curve example
Figure 9-15 Failure curve examples
Figure 9-16 Physical inspection of concrete structure
Figure 9-17 Motor Condition Monitoring
CH2M iv
Glossary of Terms
4 R’s Defined within “Keeping the Country Running: Natural Hazards &
Infrastructure, 2011” as the 4 aspects of resilient infrastructure. (See
resilience, redundancy, response, resistance, reliability)
Asset performance indicator A measure of asset failure events e.g. bursts/km/yr
Critical An asset or component whose failure would result in a loss of service
and a risk event. Often used to describe assets for which the
consequences of failure would also be very severe
Criticality A measure of how critical something (such as an asset) is
Condition indicator Measure of the physical state of the asset e.g. remaining pipe wall
thickness
Consequence The impact of failure. For example, a cost impact, service impact,
environmental impact, damage to health
Context indicator Influencing variable potentially affecting failure probability e.g.
temperature, water aggressivity
Indicator A general term covering measures, metrics and parameters for
measuring risk to service. Can be a qualitative, indicative parameter
Infrastructure A term that refers to assets including water distribution pipes, trunk
mains and sewers that allow the distribution and return of water and
waste. Infrastructure assets are predominantly (but not exclusively)
below ground
Measures A primary and quantitative parameter. Used interchangeably with
the term ‘indicator’
Measures of Success (MoS) A primary measure of service to customers and the environment,
which supports delivery of an outcome/customer promise
Non infrastructure A term that defines water company assets that are not classed as
infrastructure. These may include for example reservoirs, dams,
water and waste water treatment works and pumping stations.
These are predominantly above ground
Outcome Delivery Incentive Reward and penalty based mechanism for incentivizing outcomes:
based on performance against the target for the MoS
Performance Commitment Target level of performance for the MoS
Resilience Ability of assets, networks and systems to anticipate, absorb,
adapt to and / or rapidly recover from a disruptive event
Redundancy Avoiding dependence of a system on single assets or facilities e.g.
stand-by pumps, duplicate water mains
Reliability A measure of the likelihood that an asset or system will function
when it is required to ‘operate’
H2M
v
Resistance Provision of protection for the assets in a system so they are
resistant to known risks
Response/recovery Processes and systems, typically operational, such as early warning
alarms and rapid response plans that can minimise impact and
support the ability of a system to recover quickly so that service loss
is minimised
Serviceability indicator Indicators introduced prior to the current Outcomes based approach,
comprising asset and service indicators aimed at demonstrating the
ability of the assets to deliver service to customers
Service indicator Direct measures of customer and environmental impact
Service measure The monetised parameters used to measure risk and value service
delivery used in many asset management tools.
Sub-threshold indicator A monitored parameter (e.g. water quality) that is exceeding
expected values but not yet exceeded compliance standards or
targets (typically monitored against trigger values)
CH2M vi
Abbreviations
AH Asset Health
AMP Asset Management Plan
BAU Business as usual
CAPEX Capital expenditure
CMPCF Capital Maintenance Planning Common Framework
CMMS (SES) Computerised Maintenance Management System
CSO Combined sewer overflow
DEFRA Department for Environment, Food and Rural Affairs
DWI Drinking Water Inspectorate
EA Environment Agency
EDM Event duration monitor
FOG Fat, oil, and grease
FMEA Failure modes and effects analysis
GIS Geospatial information system
GMEAV Gross Modern Equivalent Asset Value
LoS Level of Service
MEICA Mechanical, electrical, instrumentation, control and automation
MTTF/MTTR Mean time to fail/repair
MoS Measures of Success
OPEX Operating expenditure
ODI Outcome delivery incentive
PR14/19/24 Price review 2014, 2019, 2024
RCA Root cause analysis
RCM Reliability centered maintenance
SDP Sustainable drainage plan
SEMD Security and Emergency Measures Direction
SLA Service level agreement
SM Service measure
TOTEX Total expenditure
UKWIR UK Water Industry Research Ltd
WASC Water and Sewerage Company
WLC Whole life cost
CH2M vii
WOC Water Only Company
WTP Willingness to pay
CH2M viii
Acknowledgements
We would like to thank all of the project participants, all of whom engaged openly and with enthusiasm
for the subject. The review covered all the water companies in England and Wales.
Affinity Water
Anglian Water
Bristol Water
Dee Valley Water
Northumbrian Water
Portsmouth Water
Severn Trent Water
South East Water
South Staffordshire Water
South West Water
Southern Water
Sutton & East Surrey Water
Thames Water
United Utilities
Welsh Water (DCWW)
Wessex Water
Yorkshire Water
Report authors
Paul Conroy, Katie Dixon, Jim Ellis, Elliot Gill, Jackie Leslie, Brett Korteling, Alison Mallows, Peter von
Lany, Alec Yeowell
CH2M ix
Executive Summary
Asset health is a key element in delivering resilient water and wastewater services. It is our belief
that asset health is widely understood to be a vitally important factor in providing resilient services,
both now and in the future.
This targeted review was intended to help Ofwat better understand how water companies in
England and Wales are approaching the measurement and management of asset health, and
importantly, how this contributes to their wider approach to resilience. Our interviews provided
evidence that the water companies are taking their obligations on asset health seriously.
We believe that, whilst there are few immediately serious concerns regarding the impact of asset
health on service, there are some issues, as outlined below, that will require ongoing scrutiny. Many
of these drivers are not new. External challenges to asset health, such as climate change, population
growth and aging assets continue apace, with new pressures such as the regulatory and political
focus mean that it has now become imperative that the industry consider how to move forward.
It is important that the water companies are being proactive in this area and that the role of asset
health is understood in the context of risk to service and investment decision making. Consequently,
the aim of this targeted review was to understand how well companies understand the risks
associated with poor asset health; how they identify and measure those risks and what they are
doing to mitigate them.
In support of these objectives, we undertook structured interviews with senior representatives of all
the water companies in England and Wales across several topics, including:
 Understanding of asset health and the risk to service
 Measurement of asset health
 Asset health in decision making
 Asset health and the customer
 Assurance around the health and impact on service for existing and new assets
 Incentives and barriers to innovation
 International practice
This review has provided a valuable opportunity to gain a systematic understanding of the
approaches and procedures that companies have in place with respect to asset health. Although
limited by the perceptions of those interviewed and the timescales of the project, our purpose was
to gather knowledge, identify general issues and examples of good practice – the headlines below
are evidence that this has been achieved.
It was apparent that, whilst there are areas of best practice, there are areas where companies do
things differently and where further discussion and investigation would be beneficial to raise the bar
across the industry. These issues and opportunities are considered in the main chapters of this
report and are the focus for our final conclusions and recommendations.
In summary, our headline observations are:
Understanding of asset health and risk to service

1) For the majority, asset health is measured and reported as a Measure of Success (MoS) and
is a variant of the former serviceability indicators (not necessarily called asset health). Some
companies believe that this is sufficient, whereas others talked about possible changes being
required so that the measure is more forward looking
H2M
1
2) There is diverse opinion as to what, exactly, asset health is. For example, is it a measure of
asset condition/performance and/or service and fitness for purpose? The belief that asset
health has a number dimensions and is inextricably linked with the concept of resilience was
common. Different facets of asset health may be more relevant for different asset types and
their circumstances
Measuring and monitoring the health of assets
3) All companies use additional data (indicators) to support understanding of risk to service and
to support investment decisions, such as outputs from deterioration models, observations of
repeat asset failures and site-based risk assessment information
4) In general, asset condition was not thought to be a good asset health indicator. Indeed,
some companies do not like talking about asset condition. This may be because early capital
maintenance planning tools were condition based, and:
 A condition based approach is not risk based
 The tools tended to use visual condition grading for assets and the correlation
between visual condition and remaining life or failure probability is often weak
(but not in all cases)
However, some companies recognise that certain types of asset condition data (where the
condition data is a good indicator of failure probability1) can be of value, (e.g. trunk mains,
critical sewers) provided it is used to inform risk. Very low failure rate assets could benefit
from this understanding (see point 6 below). It is noted that this issue has been subject to
previous research by UKWIR (Deterioration of long life assets WM13)
Asset health, expenditure planning and decision making

5) All companies consider the role of asset health in the context of a risk-based planning
framework (the Common Framework/Expenditure Planning Framework) and use a variety of
tools and models to link health (measured through asset performance and service) to service
risk and expenditure need. Some companies stated an intent to develop forward looking
asset health indicators based on better understanding of asset aging/deterioration. We
believe that all companies should consider measures which inform risks to service to current
and future consumers
6) For critical, high consequence infrastructure assets (especially where failure probability is
very low and consequently difficult to predict using statistical extrapolation) there is a risk of
unexpected failure and extensive local damage, including risk to life if the asset is near to a
railway line, road etc. Understanding the state/condition of such assets can inform
probability and risk mitigation; we believe that there is a need to take stock of the costs and
benefits of developing this awareness for critical, high consequence assets, such as trunk
mains. It is also imperative that companies have a good knowledge of where their critical
assets are and the potential consequences of failure in order to support this analysis
7) Asset health (assuming this is a measure of asset condition and asset performance) is not
always an accurate or reliable indicator of service (and vice versa). Service failures can be
managed through operational resilience mitigations and having redundancy and standby.
There is likely to be an ‘economic’ optimum balance in terms of resistance, reliability,
redundancy, response and recovery mitigations, but it will be situation specific. More work is
required by companies to understand this balance
1 Remaining life is another difficult to define concept. It could be used to indicate 1) time to failure 2) time to reach a critical limit state 3)
the point at which it will be economic to replace the asset, etc.
CH2M 2
8) Some companies raised the risk of an asset health bow wave arising because capital
maintenance had not kept pace with asset deterioration, resulting in the risk of a ‘cliff edge’
when many assets start failing. We are not aware of any strong evidence that this is likely to
be an issue in the short to medium term - the very wide statistical distribution of the rates of
aging of some of the assets mentioned as being potential problems (e.g. sewers) would tend
to result in a gradual increase in failure rates. Nevertheless, we believe that companies
should act now to make use of a window of opportunity to improve their understanding of
the relationship between asset health, service impact and underlying levels of expenditure
on capital maintenance, and avoid any cliff edges emerging in future
9) For non-infrastructure production assets, there is typically a lot of inherent resilience to
service loss due to the in-built redundancy (stand-by and duplication). Consequently, the
significance of the health of individual pieces of equipment is less clear. However, it is
important that the complexities of the role played by non-infrastructure assets are
understood, and that effective maintenance regimes are in place. It is likely that there are
major opportunities for optimising maintenance strategies and improving efficiency
through a better understanding of asset health and resilience
Understanding the views of the customer

10) Most water companies do not appear to have consulted directly on asset health in terms of
the state of the assets. Consultation has been around service and outcomes. Companies
need to give careful consideration as to how best to engage with their customers over
asset health
Assuring our approach to asset health and maintaining service

11) All companies have assurance processes in place relating to data, tools and systems as well
as emergency response and commissioning of new assets. The processes stated to be in
place give confidence that they are sufficient, but this has not been tested in detail and
merits some further consideration. We note that companies with ISO 55000 certification will
have some additional mechanisms for checking, audit and continuous improvement. It is
important that companies can demonstrate that their assurance processes are effective and
commensurate with need
12) Some water companies rely heavily on the supply chain (i.e. delivery partners, contractors
and suppliers). There is not always evidence of an independent view and assurance when
new assets (in particular, complex ones) are put into service. We believe this may merit
further scrutiny
Incentives and barriers

13) Financial barriers to invest in innovation were raised, but on balance the UK provides a
mature and stable market place for technology providers to invest and there is great
opportunity for knowledge sharing and learning
14) There have been some assertions that the regulatory regime may inadvertently discourage
technological innovation and lead to some short-term behaviours. Our thoughts on
innovation are that the reported barriers can be surmounted and that the water
companies of England and Wales and the UK are in a good position to be world class
innovators for the sector on the measurement and management of asset health
Lessons from international practice

15) We have seen that the water companies in England and Wales have been active in trialling
and testing technology for measuring and managing asset health; however, adoption to date
is limited and has not yet led to a clear step-change in asset performance or understanding
CH2M 3
of risk to service. Nonetheless, we believe more can be done in this area and we encourage
ongoing efforts to identify cost-beneficial innovations and technologies that improve
understanding of asset health. We note that investment in innovation in certain other
sectors is considerably greater, affording opportunity for benchmarking and learning and
these should be explored more aggressively. So, whilst we note that the UK water sector
appears to be relatively mature in terms of trial and use of technologies compared to some
international water companies, there is great value in maintaining awareness of the work
going on in more innovative sectors elsewhere and the ‘data revolution’ (smart systems,
analytics etc.) is likely to provide significant future opportunities for asset management
CH2M 4
Introduction
1.1 Background
Poor or sub-optimal asset health can lead to deterioration in service, particularly with respect to
water quality and quantity issues, which are of fundamental concern to customers and the
environment. This is not a new concept. For many years water companies in England and Wales have
been aware of the importance of the state of their assets in terms of the ability to deliver service to
customers. Indeed, even before the 1989 privatisation of the water and sewerage companies the
issue of poor asset health and the impact on service had been recognised and was the focus for
development of guidance documents, such as WRc’s Water Mains Rehabilitation Manual (1986). This
highlighted the problems of deterioration of ferrous water mains and the adverse impact on water
quality, pressure, supply continuity and operating costs. The solution was to develop and implement
a structured and integrated planning approach and prioritise mains rehabilitation according to cost
and benefits, thereby improving asset health.
The specific issue of the deterioration of ferrous water mains has been bought under control
through many years of targeted investment and today, water companies are more aware of risks
and have better knowledge of their systems, so they can respond more effectively when incidents
occur. Nonetheless, asset health and the effect of the state of the assets on service to customers and
the environment remains topical and it is important that we remain vigilant and responsive to
current and future risks, remaining appreciative of asset deterioration and potential failure modes to
enable timely and optimal intervention.
We note also that the Water Act 2014 introduced a duty for Ofwat to further the resilience objective
and asset health is a key aspect of the sector’s ability to deliver reliable and resilient water and
wastewater services to current and future customers. Furthermore, asset health is highlighted as a
key issue within Ofwat’s consultation on the outcomes framework for PR19, in which draft
expectations were put forward for how companies should address asset health; how to improve
reporting requirements to provide clarity on companies’ asset health measures, and whether to
increase standardisation in this area.
1.2 Objectives and approach

This targeted review was commissioned by Ofwat to better understand how the water companies in
England and Wales are approaching the measurement and management of asset health. The
purpose was to gather information about current approaches to asset health across the industry and
use this to help determine how well companies understand the risks associated with resilience and
asset health and whether they are taking appropriate measures to mitigate these. Ofwat want to
ensure that the current regulatory approach can support companies in being proactive in dealing
with asset health and help address any (real or perceived) conflicts. Ofwat is seeking reassurance
that the water companies are being forward-looking in this area and that the role of asset health is
understood in the context of risk to service and investment decision making. The targeted review
considers the following key questions posed by Ofwat:
1) How does asset health contribute to the companies’ wider approach to resilience?
2) How well do companies understand the risks associated with poor asset health?
3) How do companies identify and measure those risks, and what are they doing to mitigate
them?
4) What assurance processes do the companies have in place, and how do they respond to
failures when these occur?
To answer these questions, the review compares current approaches across the sector, drawing out
examples of best practice and areas with the greatest potential for improvement. To gather the
CH2M 5
necessary information, we developed a structured questionnaire that we asked the companies to
complete. We then carried out interviews with relevant staff within each of the companies to discuss
their responses and question them in more detail around specific areas of interest or concern. We
asked the companies a wide range of questions, grouped under several core topic areas, each of
which is discussed in more detail in the following sections.
The information gathered from the questionnaire responses and subsequent interviews has been
collated and analysed to help answer the key questions posed by Ofwat at the outset of this
targeted review. This analysis, and the resultant responses to the questions, are described in the
subsequent sections of this report.
Note: we have summarised some of the key findings for each of the companies and this is captured
in Appendix A – Headline findings from each company interview
1.2.1 Working definitions of asset health and its context

The questions in this category were designed to help us to understand how companies view asset
health. In addition, through comparing responses across all the companies, we wanted to
understand the diversity that exists in the working definitions used between companies and how
differences in working definitions are influencing company specific approaches to evaluating asset
health and its implications on risk (of service failure) and the resilience of water and waste-water
systems.
We asked about how companies currently define asset health; whether the definitions vary by asset
group and how companies determine the effect of asset health on the risk (of service failure) and
the resilience of their services.
Company responses to these questions have provided insight into the broader issue of how their
understanding and assessment of asset health fits into their wider plans for managing the resilience
and the performance of their assets and service systems.
1.2.2 Establishing the link between asset health, performance, service and
outcomes
Understanding how asset health will affect service and outcomes is essential for understanding the
costs and benefits of maintaining or improving asset health, and for justifying an asset health
focused business case. Understanding these complex linkages necessitates an investment in data
and information and application of tools and processes that help to elicit and quantify these
relationships. For some assets, there is considerable uncertainty in developing these relationships
and it is important that the risks associated with these uncertainties are understood. We developed
the questions in this category to draw out evidence that the companies have formally explored these
relationships and have compiled an evidence base to help support their position on asset health. We
asked how companies are linking asset health to performance, service and outcomes; what tools
they use to model and predict asset health; how probability and potential consequences of asset
failure are identified and validated and what companies are doing to plan for and mitigate the
impacts of, asset failure.
1.2.1 Measurement of asset health

We asked how asset health is measured, to better understand the metrics used by the companies
for measuring asset health and also how they are identifying and using new technology to quantify
and predict asset health. We wanted to find out what is going on ‘under the surface’ of the headline
indicators, to get an idea of what is measured at the asset level and how this ultimately is used. We
also wanted to discover what technology innovations are being implemented and how these are
benefiting the companies and their customers. We asked what indicators and measures of service
are currently used; how effective these are and how they could be improved; about the quality of
CH2M 6
asset data and about the techniques and technologies companies are using to monitor the health of
their assets.
1.2.2 The influence of asset health on expenditure planning and decision

making
Ultimately, the purpose of asset health information is to support targeted, prioritised and cost
effective/beneficial investment decision making. Customers tend to value service and outcomes as
opposed to asset health per se, so it is important to show how asset health affects service and in
turn to demonstrate how expenditure will affect asset health. These questions were aimed at
understanding how the benefit of investment on asset health has been determined and how asset
health is being used by the companies in their decision-making processes. We also wanted to
understand if/how Ofwat’s current regulatory approach along with wider external factors are
potentially affecting asset health. We asked how asset health data is used to inform investment
decisions; the relationship between asset health to service and investment; how asset criticality
influences companies’ approach to monitoring and investment; how TOTEX and Outcomes may be
affecting asset health and how the health and resilience of new assets should be valued in
investment choices.
1.2.3 Communication and the views of the customer

The aim of this question was to find out how or whether companies are engaging with customers on
asset health, including how they have introduced and explained the concept of asset health to their
customers and what level of value customers place on it.
1.2.4 Asset health assurance

The purpose of the questions in this final category was to understand how companies are ensuring
that their asset health assumptions, metrics and forecasts are robust and transparent. We asked
how companies have been performing against their asset health commitments and what has
influenced this; about the assurance processes being used (for existing and new assets) and about
what the companies perceive as the main barriers to innovation in asset health management and
what can be done in future to help overcome these.
1.3 Report structure

Section 1 provides a general introduction.
Section 2 – 7 summarise the key findings of the company consultations carried out as part of this
targeted review.
Section 8 provides a short assessment of the UK position against international practice.
Section 9 draws out key conclusions.
Appendix A, B C provide company comparisons in tabular format:
Appendix D summarises inspection and monitoring technology reviewed in Section 3.
CH2M 7
1.4 Accuracy and omissions
We have endeavoured to achieve a fair and balanced view of asset health measurement and
management by water companies in England and Wales. To help ensure consistency, we used a set
of structured questions and held regular internal briefings to share knowledge and experience.
We decided that this report would be of greatest interest and benefit if we identified some of the
activities and approach being used by specific companies and similarly provided examples of
technologies in use. We recognise that this can be disappointing for companies if omissions or
inaccuracies are reported and we accept the potential for this, but we (CH2M) cannot guarantee
that we have a perfect analysis of current practice because:
 This targeted review has been undertaken over a short period of time and it has covered a
wide range of topics. To deliver the project programme, several different teams undertook
the water company interviews. Each of the interviewers had expertise in one or more
technical disciplines and in water company assurance. This would have potential to create
some bias and perhaps encourage a differing focus for discussion between the teams. We
note that an extended programme would have allowed a single interview team to undertake
all the interviews and allow further validation of response. This could have had several
benefits:
o Consistency and replicability of interview technique and focus
o A broader understanding of the companies’ answers across the key questions and
across the water company
 In preparing this review CH2M have relied on information provided by each of the
companies during an informal interview and any follow up information they chose to
provide. This information has not been independently verified and we have assumed it to be
accurate and reliable. While we have drawn out examples of good practice and quotes from
the companies interviewed throughout our review, we have been wholly reliant on the
information that was provided. As such it is possible that companies not identified
specifically are undertaking similar work, but simply their responses focused on different
examples. Therefore, while we provide examples, readers should not infer that those not
mentioned are underperforming or not meeting expectations in any way. We also note that
the review is not exhaustive and issues may exist that we have not been able to identify
 The work was conducted in a relatively short period of time and this report has not been
circulated for comment to the companies who are subject to the review:
o Some of the companies provided a significant quantity of follow up materials,
though the time constraints and volume of this have meant that there will be
inconsistencies in terms of reflecting this in our report
o Companies will have chosen to focus on specific topics and the extent of written
materials provided has varied in scope and depth
o We have not documented all of the technologies that were identified and
acknowledge that this is a major technical area and worthy of review in its own
right; the inclusion of a specific technology in this report does not reflect an
endorsement for that type or brand of product
CH2M 8
Understanding of asset health
2.1 Context
The companies interviewed do not have an agreed working definition of what they mean by ‘asset
health’. But, there is broad agreement across companies that ‘asset health’ is a richer concept than
‘asset condition’ in that ‘asset health’ needs to consider not only the physical state of the asset but
also the role and importance of the asset in ensuring that service performance targets and customer
expectations can be met.
We (CH2M) offer the follow definitions to help inform the discussion, though it must be noted that
these are our views and not necessarily agreed by others:
Asset condition: this is a measure of the state of the asset.
Condition is a measure of the ability of the asset to resist load and is a potential surrogate for
remaining life and failure probability. However, there are many ways of assessing asset condition
and some do not give a very precise measure of the remaining life/failure probability. For example,
visual inspection will not show what ‘lies below the surface’. Asset deterioration is also an important
consideration when trying to estimate remaining life.
Fit for purpose: this is a measure of how well the asset can fulfil its function.
The asset can be in great condition, with low probability of breaking, but may be under capacity, or
of the wrong type and cannot deliver sufficient product. We have excluded this from the asset
health review.
Remaining life: this is an estimate of how long before the asset will need to be repaired or replaced.
There are subtle interpretations as to what ‘remaining life’ means. Economic service life is the point
where an analysis of whole life costs would indicate that there is greatest net benefit of
intervention. Some assets may continue to operate well beyond their nominal remaining life,
perhaps in an uneconomic state or simply because their operating environment is relatively benign
and the asset has not yet been exposed to a stress that will cause failure. When exploring remaining
life, it is important to recognise that for a population of similar assets, there will be statistical
distribution around the expected remaining life mean. Local information is essential for determining
where in the distribution the asset is likely to sit.
Failure probability: this is a measure of the likelihood that the asset will fail in some way.
It is important to appreciate that assets can fail in different ways (different failure modes) and this
can have different impacts on service (leak vrs burst for example). Failure probability (like remaining
life) is influenced by the nature of the asset and the operating environment. Some assets exhibit age
related deterioration and the failure probabilities can increase with time. Some assets are
maintainable and maintenance regime can influence failure probability. When an asset fails, service
can be disrupted. However, the speed with which the company responds; the availability of another
asset that can compensate; the presence of storage or alternative routing may mean that asset
failure does not impact significantly on service.
Resilience: this is a measure of the ability to withstand/respond/recover from a serious event; the
current UK water sector definition also incorporates consideration of future change and maintaining
service to customers and the environment:
“The ability to cope with, and recover from, disruption, and anticipate trends and variability in order
to maintain services for people and protect the natural environment now and in the future”
Asset health: this is a facet of resilience, based on the ability of the asset to resist stress and
consideration of reliability of the asset (failure probability). Many consider it is subtler than the
CH2M 9
concept of asset condition and includes consideration of aging of
the asset and how asset failure directly contributes to service While there are
failure, now and in the future. some benefits,
This chapter reports on what companies mean by ‘asset health’ and companies are
how their view the health of their assets helps them to manage the concerned about
resilience of assets and ensure an adequate level of performance of weaknesses in the
the water and waste-water systems within which they operate. current approach
In addition, this part of the review considers what companies
perceive to be the benefits of their current measures of ‘asset
health’ as well as their shortcomings. Companies were asked to
indicate where there is room for improvement in their conceptual understanding of ‘asset health’
and in their ability to assess what is the actual level of ‘health’ of their assets.
2.2 The questions we asked and their purpose

We were interested in eliciting the following information from companies:
 What does asset health mean to them?
 How do they see the relationship between asset health and resilience?
 What are the benefits of the measures of asset health that they currently use?
 Where is there room for improving the concept of asset health and the ways in which it can
be measured?
We hoped that the responses to these questions would provide insight into current concepts of
asset health and help us to understand how companies are working to develop their measurement
and management of asset health. We were particularly interested in how companies measure the
health of assets with a strong influence on water and waste-water systems resilience and the
capacity of companies to manage their customers’ exposure to risks of service failure.
2.3 What were we expecting to be told?

We were expecting to be told that:
 Most companies infer the health of their assets from a number of serviceability and
reliability indicators, in many cases aligned with Ofwat serviceability measures
 Although most companies were using these as a basis for reporting on asset health in PR14,
many companies are looking at developing ‘forward-looking’ indicators that are more
focussed on the performance of assets in the future
2.4 What were we told?

2.4.1 Defining Asset Health
Although there is no consistent working definition of asset health across all companies, most
companies associate good asset health with achieving good levels of service and performance. The
general view held by companies is that good, stable, measured performance is a strong indicator of
good asset health and conversely, that inadequate or declining performance may indicate poor asset
health.
There is widespread use of the original (and the variants of) Ofwat serviceability indicators as a
surrogate or proxy indicator of asset health. Many companies are satisfied with the use of these
serviceability measures as broad indicators of asset health, although some companies have refined
these measures to, for example, improve consistency and eliminate double counting. We note that
CH2M 10
United Utilities have a company specific measure of asset health that is based on the asset
maintenance forecast derived from their Pioneer deterioration models.
Most companies consider that the basket of serviceability indicators and derived measures of
success (MoS) provide a ‘top-down’ view of asset health. If indicators of a MoS lie in target
performance thresholds, especially over successive periods of time, this is taken as a signal that
asset health within the systems is adequate. Conversely, sudden or trend changes in measured
indicators are seen as signals of a possible deterioration in asset health. Some of the common
phases used by companies when describing asset health and what it meant to them are shown in
Figure 2.1 (from questionnaire responses):
Figure 2-1: Phrases companies used to describe what asset health means to them
Some companies are concerned that using serviceability derived measures has conflated the issues
of asset health and the ability of their systems to provide a good service to customers. Although
asset health and systems performance can be strongly related, under some views of asset health, it
is possible, within limits, to provide an adequate level of service to customers even if individual
assets are in poor health, since redundancy and spare capacity within networks and systems can
compensate for an inadequate health of individual assets.
In response to this concern, some companies are developing and testing new definitions for asset
health. These developments are currently internal to these companies and are intended to shadow
their serviceability and reliability measures during PR19 (see further 2.4.3).
An emerging, but far from universal, concept of asset
health seems to include the physical condition of the There is widespread use of
asset taking into account the (dynamic) environmental Ofwat serviceability indicators
and weather conditions to which the asset is exposed and as a proxy indicator of asset
the operating conditions within which the asset is health
expected to function. This concept has an explicit focus on
individual assets or sets of assets and could therefore be described as a ‘bottom-up’ view of asset
health. In this view, information on asset condition is combined with information on its criticality and
the resilience of the network or system, that the asset is part of, in order to create an index of asset
health. Welsh Water outlined plans for developing and testing forms of an asset-health index.
CH2M 11
Developing asset-focussed indicators is generally seen to be more challenging in the cases of
underground assets and wastewater systems than for above-ground assets that are used in water
supply and distribution systems.
The use of asset specific information, on its own or embedded within an index, offers an opportunity
for companies to develop ‘forward-looking’ projections of changes in asset health based on ‘leading’
indicators for these changes. Accurate projections of changes in asset health would strengthen the
ability of companies to make proactive asset management decisions. It is interesting to note that
Thames Water include levels of water network rehabilitation as one of their indicators of asset
health, which impacts on average asset age and introduces an asset remaining life perspective.
United Utilities spoke about the concept of remaining expected economic life, weighted by GMEAV
(gross modern equivalent asset value).
Current working definitions of asset health and the principles on which they are based are
considered by most companies not to vary across asset groups. But many companies also recognise
that the information that would need to be collected to support a more asset-focused assessment of
asset health would need to vary between asset groups.
Specific feedback included:
 Most companies have working definitions of asset health and some used alternative terms
e.g. ‘asset stewardship’ (Northumbrian Water); ‘asset service and reliability’ (Yorkshire
Water). Asset health is an attribute that is difficult to measure specifically and therefore
needs to be inferred from relevant indicators
 Several companies are changing or considering changing the sub-measures included in the
serviceability measures to make them more asset focussed. For example, South
Staffordshire, is applying weightings to the indicators within composite asset health
outcome delivery incentives (ODI’s) that recognise that some sub-indicators carry more
weight than others in contributing to the monitoring of asset health, and are not
represented in other ODI’s. Those measures that are represented in other ODI’s are given a
lower weighting
 Some companies, such as Welsh Water, are exploring the development of measures of
service for asset health that are: asset focussed; forward-looking; and, more easily
understood by customers
 United Utilities is using an internal working definition of asset health that is based on a
combined measure of condition, performance and failure-mode data which are used to
describe the overall ‘health’ of the asset as either ‘Healthy Asset’, ‘Minor Health Issues’ or
‘Poor Health’
 Several companies, including Anglian Water, see asset health and ‘asset operation’ as the
key influences on ‘asset performance’ and service, noting that asset health and ‘asset
operation’ are more difficult to measure than asset performance
 Several companies raised the issue of assets are that are critical, but have a low failure
probability and have potentially severe consequences, noting that such assets require a
special measure of asset health that recognises their criticality and high potential level of risk
should they fail
2.4.2 The relationship between ‘asset health’ and ‘resilience’

Asset health and resilience are linked; asset health is a dimension of resilience. We asked the
companies to tell us how they interpret this interaction and how asset heath contributes to
resilience.
CH2M 12
All companies recognise that asset health contributes to asset and systems resilience. But, they also
recognise that other factors also contribute to resilience - these include network interconnections,
spare capacity, redundancy (e.g. availability of stand-by assets) and emergency responses.
All companies agree that asset health is a building block of asset resilience and as such, it contributes
to the resilience of the systems within which the asset functions.
There is general awareness amongst many companies
There is agreement that asset that more insight into resilience could be gained by
health is a building block of analysing the asset specific data. The granularity of this
asset resilience and therefore data in some areas needs to be increased to make it
system resilience more useful.
Several companies gave examples of how they are
developing insight into the rate and degree of asset deterioration through statistical models that
produce deterioration curves from which the probability or likelihood of failure can be inferred. This
information can be used as inputs to models that simulate the impact of failure of assets on service
and performance.
Several company responses indicate a need for greater clarity in understanding exactly how asset
health influences asset resilience and systems resilience. Many companies are looking at how the
‘four R’s’ of resilience as proposed by The Cabinet Office (2011) that is: Reliability, Resistance,
Redundancy and Response and Recovery2 apply to different asset groups. However, opinions seem
to be divided over whether asset health relates more strongly to the reliability element of resilience
or to the resistance element of resilience, or to both. Figure 2-2 shows a possible model for asset
health within the wider resilience concept.
Some companies are exploring ‘systems’ views of their networks and asset systems to
assess the capacity for resilience within their systems to external ‘shocks’ (e.g. from
extreme weather events, flooding, droughts) as well as from internal ‘shocks’ such as the
failure of critical assets. This systems view is shown in Figure 2.2 and is helping companies
to consider issues such as asset criticality and redundancies within systems alongside asset
condition and asset health when assessing asset and systems resilience
Figure 2-2: A conceptual relationship between asset health and the ‘four R’s’ of resilience
2 Cabinet Office (2011). Keeping the Country Running, Natural Hazards and Infrastructure. HMSO
CH2M 13
Several companies are looking at new ways in which they can better understand, assess and manage
the resilience of their assets and systems. South Staffordshire Water mentioned the development of
a ‘resilience maturity matrix’ and a ‘resilience lens tool’ (a scorecard in which two quadrants refer to
the asset-health related factors of reliability and quality of supply).
Some companies refer to the challenge of determining where ‘asset health’ ends and ‘resilience’
begins - this issue is illustrated by Severn Trent in Figure 2.3 in noting that to fully understand
resilience, a range of metrics should be considered across different time horizons.
Figure 2-3: Understanding and measuring resilience (Severn Trent Water perspective)
We note that this concept diagram reflects Severn Trent Water’s current thinking and is not
necessarily an industry view. Health and resilience overlap and both can have a short and long term
dimension.
Another challenge, referred to by some companies, relates to the issue of measuring changes in the
condition of assets that fail infrequently (e.g. dams and tunnels) – such assets are often ‘critical’ and
when they do fail, this can create impacts and generate consequences that are potentially severe for
the company, their customers and the environment.
Although several of the responses to our question on resilience were not direct answers, key points
made on the relationship between asset health and resilience included:
 Several companies (e.g. Welsh Water and Southern Water) recognise that more data
granularity is required combined with a sharper focus on interrogating and interpreting
the data to get at the root causes of issues in order to determine which issues are asset
health related
 Some companies (e.g. South West Water, Southern Water, Yorkshire Water) point out
that resilience is a function of the capacity of the asset (a form of resistance) as well as
its condition e.g. a structurally sound dam with a half empty reservoir may be an asset in
good health, but not resilient to drought
 One company, Anglian Water, has modelled the impact of failure of all connected assets
on service in order to allow an internally consistent valuation of risks and resilience;
another, South East Water maintains asset risk registers linked to asset failure data
stored on a GIS, with outputs accessed via an ‘asset dashboard’ to help visualize
information on asset related risks
 Southern Water is determining the impact of asset health on system resilience by
applying approaches including: Inter-temporal modelling for water resource
management planning; drainage area / surface water management planning; specific
surveys and investigations of strategic crossings, standalone reservoirs and single
sources of supply and strategic network fault tree analysis
CH2M 14
 Thames Water is evaluating the resilience of their water supply systems by modelling
the propagation and effects of asset failure on the performance of their water supply
networks
 Portsmouth Water is carrying out a ‘critical links analysis’ of water mains as part of a
process of risk-based selection of mains due for renewal by linking asset condition
surveys with criticality information from MISER modelling
 United Utilities Water uses a measure of asset criticality that is informed by several
measures in terms of consequence of failure (failure mode analysis) but also a number
of other factors including the risk reduction that an intervention would provide on a
specific asset linked to strategic drivers such as Pollution or Health and Safety
2.4.3 Benefits of the current measures of Asset Health and improvement

opportunities
All companies recognise the broad benefits of using serviceability performance measures and asset
reliability measures that have been applied in a relatively consistent manner across companies over
many years. The indicators within these measures of service yield historic information on
performance; these can be compared against threshold values over many years from which trends
can be detected and analysed.
Most companies combine serviceability performance indicators relevant to each asset group,
without weighting individual indicators, to derive an overall asset health ‘score’ for each asset group.
One company (South Staffordshire Water) applies weights to their indicators before combining
them.
The following quotes, from Severn Trent Water and Southern Water respectively, illustrate
perceived benefits of using serviceability performance measures and asset reliability measures as a
signal of asset health:
 “they provide a basis for predicting and avoiding future service failures, targeting
investment, improving the robustness and efficiency of longer term planning and
demonstrating to stakeholders that assets and service are being maintained at a broadly
stable level of risk”
 “the greatest benefit of the current measures of asset health is that they have been
established for some time, providing a useful dataset for trend analysis of asset health and
rates of deterioration”
However, most companies expressed some concerns over weaknesses in the prevailing approach,
such as:
 A number of serviceability indicators are sensitive to weather events which can mask
underlying asset deterioration effects
 Systems can perform well despite the presence of deteriorating assets – their influence can
be compensated for by network interconnections, spare capacity, redundancy (e.g.
availability of stand-by assets) and emergency responses, all of which all contribute to the
overall resilience of water and waste-water systems
 Asset health is particularly difficult to infer from performance indicators for non-
infrastructure assets
 On their own, historic indicators are insufficient to support ‘forward-looking’ proactive asset
management decision-making
CH2M 15
 Assessing asset health requires a combination of ‘top-down’ information on systems
performance with ‘bottom-up’ data on asset condition, criticality and the level of resilience
within water and waste-water systems
In response to these weaknesses, several companies are developing analytic methods within their
asset management frameworks. The aim of these methods is to help to understand better the health
of their assets through ‘deep dives’ into data and information they collect, often in real-time, on:
 Asset condition
 Asset behaviour
 Specific aspects of systems or asset performance (e.g. the generation of pressure transients
in water supply and distribution systems; sewer flooding and pollution incidents) that could
be linked to a deterioration in asset health
This information can then be combined observed trends in performance to develop a
complementary mix of ‘leading’ and ‘lagging’ measures of asset health. The aim is to then use these
measures within an overarching asset management framework that is informed by good data
analytics, visualisation of outputs (a number of companies are developing information ‘dash-boards’
to help in this) and the setting of internal ‘targets’ for measuring and managing asset health.
Our key findings relating to improving measures of asset health include:
 Anglian Water is moving away from Serviceability MoS’s by disaggregating asset health from
asset performance by recognising that asset health and asset operation both contribute to
asset performance, whilst noting that:
o Asset Health - is not easily measured, especially for underground assets; it relates to
the deterioration of the asset and is time based
o Asset Operation – the nature of asset operation and its impacts on asset performance,
are not easily measured
o Asset Performance – is more easily measured than asset operation and health (e.g.
mains bursts) and is likely to affect service performance
 Welsh Water is looking at a potential step change to the existing serviceability based
measure by developing a new asset health MoS to be implemented at the next AMP and
included in PR24 that will be:
o Forward looking (drawing on deterioration models and trend analysis)
o Asset focussed (rather than service impact focussed)
o Include sub-threshold ‘at risk’ measures (particularly for non-infrastructure)
o Reflects the new price controls (e.g. headroom and energy generation for bio-solids)
o Use ‘customer friendly’ descriptions of the indicators that comprise the MoS
 Several companies mentioned their participation in water industry research projects into
areas of resilience and asset health – these include, for example, the Discover Water project
which is developing new asset health metrics and the Safe and Sure project on water
systems resilience.
Some companies, for example Northumbrian Water, mentioned that they were looking beyond the
UK water industry at the International Water Association and the Institute of Asset Management for
insight into emerging best practice in asset health assessment and management.
CH2M 16
Measuring and monitoring the health of
our assets
3.1 Context
This chapter looks at the data companies capture store and analyse to determine the probability of
asset failure and impact on service and to inform decisions. It looks at the technologies used to
capture information on asset health and how the companies are developing innovations in this area
and in the intelligent monitoring and control of their assets and systems.
3.2 Disclaimer
It should be noted that there likely to be alternative devices, tools and technologies available that
support asset health analysis and we are unable to make any kind of comparative assessment within
the scope of this study. Where we mention technologies specifically it is because these were
identified during consultation and our intention is purely to illustrate what types of technologies
exist.

We asked companies what indicators they use to measure asset health and the underlying data used
to enable targeted decision making; we asked about how well companies understand their network
and any data deficiencies and we asked about use of monitoring and inspection equipment and how
new technologies are developed.
The purpose was to determine whether the water companies are making full use of available
technology and if there is a mechanism for continuous improvement and innovation in this area.
3.4 What we were expecting to be told

We expected to hear that the water companies collect a lot of asset performance data, such as
bursts, blockages and water quality information, and that this is used to inform risk-based asset
management. We expected to be told that monitoring of physical parameters such as flow, pressure
and water quality is business as usual and used for operational decision making.
We also expected to hear that, periodically, discrete inspection and monitoring of the assets and
their performance is carried out to support maintenance planning and investigate failures or
problems.
We expected to hear that companies had a good, but not necessarily complete record of the physical
assets in their area. The records would include information on the location, age, dimension,
materials and environment. We expected that all companies would have a GIS system for holding
infrastructure data, and that this was being actively used (and updated) to manage infrastructure
assets. Similarly, for non-infrastructure assets we expected to hear that companies have a
comprehensive, hierarchical asset register that holds asset data and links to a maintenance
database.
We also expected to hear that novel, innovative techniques for assessing asset health are piloted by
companies and that they share information and participate in collaborative research projects
through UKWIR to keep the industry informed. However, the cost of some of the technology and the
practical constraints are expected to be a barrier with respect to deployment of some of the
advanced pipeline inspection techniques that have been developed in the oil and gas sectors.
CH2M 17
What we were told
3.5.1 Overview
We have summarised the key findings relating to the use of inspection and monitoring technology
and the tools used for analysing and presenting data. This is captured in Appendix B – Summary of
key technologies reported by water companies.
This appendix tabulates inspection, monitoring and control and data analysis tools used by each
company for each main asset group. This is illustrative and it should not be inferred that if a
specific technique is not shown against a company, that they do not or have not used it.
All companies stated that the quality of data they hold and manage is generally sufficient for the
purposes of both managing their risk and delivering required outcomes. All companies consider that
they are collecting good asset health data as reflected in the serviceability based MoS and the
quality data that is required for annual reporting.
Some of the typical data collected is shown in Table 3.1. It is debatable as to how many of these
parameters should be classed as asset health measures, but they are all used to inform risk to
service:
Table 3-1: Typical asset data
Water Wastewater
Infrastructure Non infrastructure Infrastructure Non infrastructure
 Discoloured water  Reactive maintenance  Repeat sewer flooding  Reactive maintenance

complaints (network (network capacity)
 Water quality  Consent compliance
management)  Sewer collapses (physical
 Ammonia  Effluent quality monitoring
 Poor water pressure condition)
(network capacity)  Nitrate 
 Sewer blockage
 Mains bursts (physical  Membrane integrity
 Pollution events (network
condition)  UV dose management)
 Drinking water  Unit cost of water (fixed and  Rising mains bursts
compliance marginal)
(distribution)
 Tonnes of chemical per Ml of
water

The review found that companies have tested and are now using a variety of technologies while
some have plans to trial other new tools. However, the companies are conscious of the relatively
high costs of some of the advanced pipeline inspection devices and cite barriers relating to
incentives to take risks. To address this, it is important the companies work together and share
knowledge effectively. We believe that UKWIR have an important role in this regard.
3.5.2 Asset data

A variety of asset data and information is collected to help understand asset health and to
determine expenditure needs. Determining what data is important has been based on many years of
experience and learning. In addition, when risk-based planning was introduced formally as part of
the Common Framework for PR04, formal elicitation techniques such as fault tree and event tree
analysis were introduced which encouraged consideration of cause and effect pathways and
determination of failure modes. These formal techniques provided a structured way of identifying
what data was important (the risk factors) and also linking of environment, asset data, performance
data, service risk and expenditure.
CH2M 18
As an example, Figure 3.1 maps out some of the relationships between water pipe risk factors and
thus identifies some of the key data that may be useful in understanding asset health and service
risk; the better these relationships are understood, the more efficiently service can be managed:
Figure 3-1: Map of water pipe risk factors
Expenditure Failure Asset Network

probability performance resilience
factors measures factors
Design,
construction
Asset
and installation condition/ Leakage
Network configuration
Operational response
state failure
Capital Pipe
Operating
maintenance break
conditions
failure
Quality, External Quality
Enhancement, environment failure
Growth
Probability of asset failure Asset failure Probability of

failure leading to
service impact
3.5.2.1 Infrastructure asset data

Typically, for measuring infrastructure asset performance, data such as burst rates, leakage, water
quality, collapses, blockages are collected. This is stored in a variety of databases and can typically be
resolved to specific areas, assets and overlaid with operational data, including customer contacts.
GIS tools are used to observe these relationships.
In addition, companies are collecting information on flows and pressure and events such as surge to
help understand the dynamic nature of the water networks and potential risks. Sewage effluent level
information and asset condition are used to help understand the risks associated with wastewater
networks.
3.5.2.2 Non-infrastructure asset data

In the case of non-infrastructure assets, the focus, as reported by companies such as Welsh Water, is
output quality. However, maintenance activity (particularly reactive and repeat events), is used to
help understand asset risks. Monitoring of asset health using, for example, vibration, acoustic and
CH2M 19
thermographic sensors appears to be relatively well established, albeit with some mixed feelings
regarding effectiveness.
3.5.2.3 Condition data

We can define condition data as information regarding the physical state of the asset, such as the
residual strength, or extent of deterioration or propensity to leach corrosion products into the
supply. For example:
 For a steel water pipe operating under high pressure, the remaining pipe wall thickness is a
fairly direct measure of propensity to leak or burst
 For a non-infrastructure asset, the visual appearance is not a particularly good indicator of
failure propensity or performance
Asset condition is a measure of the physical state of the assets. In an ideal situation, this can be used
to estimate failure probability or remaining useful life. For mechanical and electrical assets, asset
condition monitoring is used to assess wear and deterioration and to inform preventative
maintenance; this is as much about minimising costs as about managing risk.
We also note that many UK water companies do not like talking about asset condition in the context
of expenditure planning. This is because early capital maintenance tools were condition based, and:
 A condition based approach is not risk based
 The tools tended to use visual condition grading for non-infra assets and the correlation
between visual condition and asset remaining life or failure probability is often weak
In former years, visual condition assessment was used extensively to make a case for non-
infrastructure investment. This was not risk based and the condition gradings were often not a very
good indicator of asset failure probability or remaining life and hence it was important to get the
industry to change behaviour and develop a more sophisticated approach. As a result, a
performance and service based approach to risk analysis was encouraged and use of condition, per
se, discouraged.
However, some companies recognise that certain types of asset condition data can be of value,
provided it is used to inform risk. There are some assets where measures of condition can correlate
well to failure probability and hence be used to estimate risk. In these instances, condition data may
be the only viable data that we have to support estimation of failure probability. We conclude that
there is a place for condition data, but it must be used to inform the analysis of risk. Typically, this
may be the case for low probability, high consequence assets (such as trunk mains). It is noted that
this has been subject to previous research by UKWIR (Deterioration of long life assets WM13).
Very few companies systematically and proactively collect condition data for water infrastructure
assets and those that do focus on assets defined as critical or that have a significant consequence of
failure. In the case of CCTV inspections for sewers, whilst companies proactively plan inspections of
their critical/at risk infrastructure, wider inspection do not seem evident.
Most companies have been trialling non-destructive techniques for infrastructure condition
assessment. There were mixed opinions as the effectiveness of non-destructive methods of
wastewater infra pipe condition assessment, with some companies having had successful pilots, and
others having less successful pilots. None of these technologies have crossed over into business as
usual. For water infrastructure, non-destructive technologies are used quite extensively, and surge /
transient monitoring is becoming prevalent, and many companies use a ‘calm’ network as a lead
indicator for water infrastructure asset health.
Most companies have a condition grade for their civil, mechanical and electrical non-infrastructure
assets, and the use of a computerized maintenance management system to compare planned
maintenance against reactive maintenance is common. This reflects the comment made by a
number of companies that asset health is particularly difficult to infer from performance indicators
CH2M 20
for non-infrastructure assets, where redundancy or operational interventions mask the impact of
asset health on performance. For some non-infrastructure assets (e.g. wastewater treatment
works), input variables (such as weather, influent quality) are reported to have a greater impact on
service or performance than asset health.
3.5.3 Data quality

All companies consider that the asset health performance data they are collecting is generally
satisfactory, especially in relation to infrastructure health data.
Several companies acknowledge that their asset register records are not complete and none of the
companies, in our opinion, can say with absolute certainty that they have 100% confidence in the
precise location of all their buried assets. Thames Water, for example, acknowledged that there is
uncertainty in the location of some buried assets. However, Thames Water did not believe that there
were any major data deficiencies that need to be addressed in relation to the current measures of
asset health. Most companies report that they predominantly know where all their assets are that
are critical to service delivery and the risk of these critical assets causing a service failure.
All the companies have similar mechanisms in place for updating data and it is typical practice to
update and improve records when and where errors are found. South Staffordshire Water, Bristol
Water and Northumbrian Water mentioned the exercises they had been through historically to
update GIS and the allocation of confidence grades; we believe most if not all of the companies will
have undertaken a similar exercise.
Based on what we have been told, and our many years of working with the companies, we would
agree that on balance there is generally adequate awareness of assets and asset details, and that
where there are issues, these are acknowledged. We do believe that there is room to improve
understanding of risk for critical assets, especially where it is difficult to predict failure probability.
3.5.4 BIM (building information modelling)

Some companies, for example South East Water, reported that they were exploring BIM standards
for managing their lifecycle data. BIM is being promoted by HM Government and the following
definition is taken from the 2012 report: Building Information Modelling (available on the
www.gov.uk web site):
“Building Information Modelling (BIM) is a collaborative way of working, underpinned by the
digital technologies which unlock more efficient methods of designing, creating and
maintaining our assets. BIM embeds key product and asset data and a 3 dimensional
computer model that can be used for effective management of information throughout a
project lifecycle – from earliest concept through to operation. It has been described as a
game-changing ICT and cultural process for the construction sector.”
This is only indirectly related to asset health, but we expect more companies to adopt these
standards in future and there may well be benefits in terms of promotion of consistent data and
information and efficient storage and usage. Whilst peripheral to the issue of asset health, it has
been suggested by some companies that their data is an asset and hence is relevant.
3.5.5 Analysing and presenting data

All companies now hold their network asset data in a GIS system which can be used for identifying
failures at the asset level, supporting operational responses and targeting investigations and
rehabilitation activity.
Asset registers and maintenance databases enable non-infrastructure maintenance activity and
failure information to be resolved at asset and component levels and these tools are generating
information that will improve future maintenance strategies.
CH2M 21
All are using dashboards to some degree, to give an understanding of service risk, and although no
water companies have a smart, fully real time control system covering all of their assets, some
appear to be moving in this direction and it is their aspiration to introduce a live system across their
assets.
Some companies were able to show us live dashboards of asset and failure data compiled in near
real time. This provided a live demonstration of the water assets on GIS, which resolves the asset
health and performance data down to the individual asset level. This helps the company to identify
root causes of service failure at an asset scale, and is used both operationally and for long term
planning.
All companies have deterioration models that link asset performance to service for infrastructure
assets. Many deterioration models predict measures of asset performance e.g. bursts, blockages,
collapses, which can be considered to be direct indicators of how ‘healthy’ the asset is. Some predict
rates of corrosion/degradation of the asset, which then needs to be linked to performance, so these
are more indirect indicators of asset health. These tools are well established and have been subject
to research by UKWIR and independent scrutiny by third parties and assurance providers.
We have no concerns as to the general suitability of these models, provided that they are subject to
validation and used to inform decisions alongside other information and data.
The use of GIS, dashboards and deterioration models is covered further in section 4.4.2 & 4.4.3.
There have also been some interesting developments in visualization and tools such as Tableau have
been adopted to support data mining and analysis. Figure 3.2 is an extract from a Tableau report.
This specific report presents water quality complaints by zone and also enables daily data to be
examined. In this case, the high priority zone based on total complaints was determined to be of
much lower priority as problems were primarily all recorded on a single day and determined to be a
transient issue.
CH2M 22
Figure 3-2 Tableau report example (Welsh Water example)
3.5.6 How data is used

In our discussions, it became clear that asset health data can be used in different ways, from
supporting operational responses relating to unexpected failures or performance through to
informing long term decision making. The box-diagram at Figure 3.3 maps out some of the data
types being collected by companies and the planning activities that the data supports. These
activities are important to ensure a healthy system. It is necessary to monitor and react when trends
and patterns are revealed within an AMP period and make effective asset health decisions
accordingly.
CH2M 23
Figure 3-3 Data usage: the temporal dimension
3.5.7 Inspection, monitoring, control and analysis

3.5.7.1 Introduction
Inspection and monitoring of assets is not a particularly new concept. In the past, visual inspection
on non-infrastructure assets was commonplace, as was the use of pipe sampling for assessing the
condition of water pipes and CCTV inspection for assessing the condition of sewers.
It is interesting to note that only Affinity Water, Portsmouth Water, South Staffordshire Water,
Southern Water and Thames Water mentioned undertaking trunk mains condition assessment,
some of which was stated as being ‘historical’. They noted that this activity was focused on the more
‘critical’ assets. We do know from other sources that other companies also undertake this analysis to
some degree.
We have adopted a different approach with our review of inspection and monitoring technology and
undertaken additional research, based on public domain information, to provide information on the
technologies that are available. We considered this to be important to help provide a summary of
the technologies in use or under review. Note: it is important that all technology and innovation,
whether ‘hard’ or ‘soft’ needs to be evaluated and appraised in terms of the costs and benefits that it
provides and that we do not advocate any specific solutions.
See Appendix B for examples of inspection, monitoring and control and data analysis tools used by
each company for each main asset group.
Appendix D provides further commentary on the types of tools and techniques being used and
trailed for inspection and monitoring.
CH2M 24
3.5.7.2 Water pipes
General
In the case of water pipes, and particularly for distribution pipes, the key information collected is
burst rates, leakage, water quality data and pressure.
 For water pipes, GIS tools are used by all companies to map failures and target investigations
and rehabilitation as appropriate. Deterioration models are used to forecast expected rates
of asset failure
 As far as we can establish, the companies use deterioration models that use burst data to
predict future rates of failure and service to support investment planning and targeting
 Pressure, flow and water quality are monitored as a matter of routine and pressure transient
monitoring is becoming more widespread. We cannot say how effective transient
monitoring is for detecting and avoiding bursts but with an increasing uptake in the
technology, data will become available to enable the cost benefits of the technology to be
assessed
On balance, there is a good range of options available for monitoring and assessing water
infrastructure assets and the use of statistical models and deterioration analysis has become
relatively routine. There are limitations in terms of the apparent availability and cost of advanced
condition monitoring for trunk mains, but investigation into options continues and is encouraged.
Distribution pipes
With regard to inspection of the physical asset, there is little condition sampling currently being
carried out on distribution pipes (Affinity Water and Southern Water noted as an exception). This is
because the other data available is considered better for identifying and targeting rehabilitation.
We were also told about the use of technologies. Ice pigging, for example, is quite widespread for
cleaning water mains. It has been used with mixed success by companies including Dee Valley,
Bristol Water, South Staffordshire Water, Welsh Water and Yorkshire Water.
From a monitoring perspective, there was some reference to water quality monitoring in distribution
as well as pressure monitoring. Surge analysis and leak-noise analysis were also mentioned by a
significant number of the companies, both in terms of becoming established and as a basis for
potential trials.
Flow and pressure monitoring and use of portable pressure loggers also seem well established for
supporting leak detection and management activities.
For managing distribution pipes, trending burst data and use of statistical models can be considered
best practice; these pipes tend to be of ‘lower criticality’ (low consequence of failure in terms of
service impact) and so a reactive approach to maintenance, based on developing failure models and
targeting the worst performers using GIS analysis should be an effective basis for managing asset
health and service.
Trunk water mains

For trunk water mains, the use of failure data to infer residual life/need for rehabilitation is more
problematic. These tend to be low failure rate assets and although models do exist (UKWIR Long Life
Asset project WM13) that can help estimate likely condition, these are often unreliable at the
individual asset level. Consequently, more direct condition survey may be considered for these
assets.
CH2M 25
Some of the options available for understanding the condition of trunk mains include discrete
sample and destructive testing of cut-outs as well as leakage monitoring tools such as Sahara and
SmartBall. Interestingly, there was little mention of these techniques (Welsh Water referred to using
Sahara), though we know that others have used these techniques with success in the past (e.g.
Thames Water) and United Utilities are looking at SmartBall technology to assess rising main
condition.
Inspecting trunk mains can be expensive and hence it is important to be able to prioritise where to
invest in inspection and monitoring. A priority is understanding the health of the most critical and
high consequence of failure trunk mains because the damage caused by failure can be extreme.
We note that few companies stated that they are currently active in terms of undertaking cut-outs,
samples and doing non-destructive testing on trunk mains (Affinity Water, South Staffordshire
Water, Southern Water, Thames Water did mention this approach). South East Water were the only
company to mention the use of CCTV for internal inspection of trunk mains.
Tools such as Syranix Pipeminder have become more established and this is being explored by an
increasing number of companies for monitoring pressure transients and inferring failure potential
(see previous section).
Anglian Water and Severn Trent Water told us that they will be trialling Echologics ePulse acoustic
condition assessment tool and Anglian water also told us that they are looking at thermal imaging
for leak detection.
Over the last 20 years, the water companies have explored some of the in-line inspection devices
used in other sectors (intelligent pigging) either directly through trials or as part of industry research
projects. These are expensive options and there remain issues regarding their practicality when
being considered for use in water mains e.g. the need to drain down and clean, to remove
obstructions, and questions about the usefulness of data relative to the cost. Trialling these
technologies has tended to be the domain of the larger water companies, that have (or had)
significant research and development departments and budgets.
Regarding monitoring and control in trunk mains, we have already mentioned the increasing use of
pressure transient monitoring, though we cannot comment on the cost benefits of these
technologies as they are still relatively new.
Northumbrian Water described the use of flow cytometry to monitor bacteria in drinking water and
this is an interesting, though perhaps indirect, tool for looking at asset health.
Finally, regarding remote/real time control of trunk mains networks and pumps, there is evidence of
pressure and flow management, based on data from loggers in the network (Sutton and East Surry
Water, Southern Water, South West Water, Yorkshire Water).
We expect that most companies are exploring control of their networks, but currently most of this is
for operational management of flows and pressure and water quality, as opposed to managing the
health of the trunk mains in the system. The surge and transient analysis being explored offers
potential for some interesting developments in this area.
3.5.7.3 Sewers
As with water infrastructure, all companies (the WASCs) use GIS to map blockages and collapses and
help to target investigations. Sewers have relatively low collapse rates so targeted CCTV is used to
help understand issues. As with water mains, deterioration models are used to predict failures, but
the low rates of failure mean that this is more problematic. Outside this review, we are aware that
CH2M 26
alternative models such as Markov chain models are/have been used by some to forecast failure
rates, but these use mathematical assumptions and are not suitable for predicting failure of specific
sewers as the models are representative of the asset population.
The primary method determining the condition of sewers is CCTV. This is well established and ideally
all critical, high risk sewers will be scheduled for regular survey. The CCTV identifies defects such as
collapses, occlusions and FOG (fats, oils and greases) and informs rehabilitation and cleaning needs.
All WASCs use CCTV as the primary tool for assessing the condition (and health) of their critical
sewers (the most important ones e.g. highest risk). We would also like to see that repeat CCTV data
and survey of a wider sample is conducted to help develop robust deterioration models, but we are
not aware of which companies undertake this additional analysis.
There are some examples of the use of inspection technologies such as Sewer Batt for early blockage
detection and Electroscan for infiltration. The effectiveness of these techniques is being explored.
Wessex Water also mentioned the use of chemical testing for concrete sewers, and laser and sonar
scanning for large sewers. These are more established techniques and have niche application.
United Utilities referred to the use of SmartBall for live assessment of rising mains issues, which
appears to be a relatively novel application of the technology in the UK.
Although not mentioned generally, we know that companies also have drainage area planning and
sustainable drainage planning processes that implement the risk based methodology defined in
WRc’s SRM5 (sewerage risk management manual).
We believe that sewer condition data is valuable, as are repeat inspections to inform understanding
of deterioration. We also agree that collection of this data this should be prioritised based on risk,
but there is also value in gathering additional information on lower risk sewers to inform the
modelling.
In terms of monitoring, EDM monitoring at CSOs has become established and helps to identify
pollution events, but we do not necessarily consider this to be an asset health monitoring technique.
Yorkshire Water and Northumbrian Water described extensive sewer level monitoring programmes,
but this is perhaps more indicative of hydraulic capacity issues than health (though there is an
infiltration consideration that is worth noting).
There was no clear example of sewer network real time control, though there are likely to be a
number of interesting control systems relating to pumped storage.
3.5.7.4 Treatment works and pumping stations

When considering complex assets, such as a treatment works, there is a common view that the
inherent redundancy, stand by and duplication means that asset failure due to poor health does not
typically result in a service failure. Therefore, asset failure is considered more from a cost and
maintenance perspective.
We agree that where there is redundancy and standby, that a ‘run to fail’ maintenance policy may
be acceptable, but only if the failed asset is repaired so as to preserve the level of redundancy in the
asset. Ideally, the company will be making these decisions having examined issues such as the
criticality of specific components; considered the potential costs of repair and MTTR (mean time to
repair) and linked the policy to the spared strategy. The company needs to be considering
component availability in the context of the asset system.
Based on our discussions, it appears that companies are making some use of techniques such as
vibration monitoring to inform predictive maintenance, but it is interesting to note that there are
mixed views on the effectiveness of these tools. We are not able to comment on this further and it
merits investigation.
CH2M 27
Some companies mentioned maintenance planning approaches such as RCM (reliability centered
maintenance), FMEA (failure modes and effects analysis), reliability block analysis and RCA (root
cause analysis). The discipline of maintenance management and optimisation is significant and
sophisticated and merits further investigation.
These approaches have been developed in sectors other than the water industry and there is
probably a significant opportunity to learn and apply at least some of this knowledge and the
associated tools.
As a final point, we have considered if there is an apparent technology gap opening up between
companies and we believe, at the present time, if there is a gap, it is one of choice and there are
examples of being innovative and utilising technology for monitoring asset health and resolving asset
health issues by the WASCs and WOCs.
3.5.8 Innovation in technology

All companies told us they have an innovation strategy and all believe that they are using innovation
well and investing the right amount in innovation to improve their data capture and analytics. The
majority are actively trialling new innovative techniques for assessing condition or other asset health
measures. All the companies participate in UKWIR projects and a number, such as Anglian Water and
Bristol Water provide defined opportunities for the supply chain to identify and trial innovative
technologies.
While there are a number of advanced monitoring systems that can be used to measure and predict
asset and service failures, a number of larger companies told us that they faced barriers to
implementation. These included challenges in justifying investment in the technology internally and
technological barriers. Companies noted that the need to get a return within the AMP period can be
a barrier to innovation. We consider these issues further in section 7.
There is plenty of evidence that our companies want to be innovative, including examples such as
Anglian Water’s ‘Shop Window’.
CH2M 28
Asset health, expenditure planning and
decision making
4.1 Context
This part of the review was aimed at exploring asset health forecasting; asset health impact on
service and asset health investment decision making.

To help understand how companies make expenditure decisions from an asset health perspective,
using knowledge about the state and the failure likelihood of their assets, we asked:
 How they established the linkage, or relationship, between asset health and the service that
is delivered to customers and the environment?
 If they used tools and methods to elicit and quantify these relationships and how risks
relating to asset health are identified and validated?
 How are asset health measures used to inform timely and cost effective interventions?
 How does criticality influence the approach to asset health measurement?
The purpose behind these questions was to make sure that companies understand and have
assessed how asset health affects risk, the service delivered to customers and the environment and
the role of asset health in informing investment decision making.

We expected all the companies in England and Wales to have effective risk-based asset management
planning systems in place, which enable quantification of the probability and consequence of failure
in a systematic and transparent way.
Part of this process should involve being able to assess risk to service based on an assessment of
asset failure probability, understanding the likelihood that asset failure causes a service failure, and
being able to scale the consequences in terms of severity, duration and type of event. This can be
summarised as:
Level of Risk = (the likelihood that a service event

could occur, given that an asset might fail) x (the
degree of severity, extent and duration of the impacts
of the service event)
In the case of specific asset health risks, the failure event is the asset failure, for example a burst, a
breakdown, a collapse, internal corrosion and the release of corrosion products into supply. We
expected companies to be aware of the options for managing risk; the 4Rs of resilience and a need
to get the right balance.
We expected to find that companies view asset health as being part of the wider resilience picture,
though asset health is more about the intrinsic asset ‘reliability’ and perhaps ‘resistance’ aspects,
rather than the ‘response’ and ‘recovery’ aspects of the 4Rs which are more closely aligned with the
mitigation of consequences.
CH2M 29
We know that since the introduction of the CMPCF (Capital Maintenance Planning Common
Framework) for PR04 there have been significant efforts by companies to understand risk, and part
of that has been the deployment of tools such as FMEA (failure modes and effects analysis) to
understand how and why assets fail and how that failure may affect performance and service. Such
tools have been used to help understand how failure of a component in a complex system could
affect overall system performance. We also know that companies are using asset specific
deterioration models supported by statistical analysis techniques to help predict and anticipate
performance of their assets and quantify the effects on service. In recent years, the UK water
companies, through UKWIR, have been updating the CMPCF and a new expenditure planning
framework has been developed. Service forecasting tools have been reviewed as part of this and
advice given as to their potential application (Framework for Expenditure Decision Making:
development of Service Forecasting Approaches 15/RG/05/43).
We know that water companies are using asset specific deterioration models covering many of the
asset groups and we expected all the companies to be able to explain how they use these to help
forecast asset health and service and be able to validate their use. We know that many companies
have participated in the 2015 UKWIR study on Asset Performance Indicators (Leading Indicators
Linking to Performance Commitments, 15/RG/05/42) and some are using this to help develop their
approach to identifying suitable lead indicators.
In summary, endeavouring to link asset health in terms of failure probability to the service event is
not a new concept and so we were expecting that companies would have tools and methods in place
and had evidence that the assumptions used in the models are valid.
4.4 What we were told

The key points made by each of the companies have been compiled and are summarised in
Appendix C – Linking asset health, service and investment.
4.4.1 Linking asset health to service

All companies could articulate the concept of a hierarchy of indicators, measures and outcomes and
explain where asset health belonged in this. Figure 4.1 (from Yorkshire Water) illustrates this:
Figure 4-1 : The Link between asset health, service and outcomes (Yorkshire Water)
CH2M 30
In this example, asset lead indicators are essentially early warning predictors of asset failure. These
could be measures of asset condition, energy usage, environment, deterioration model outputs etc.
These are all failure probability risk factors.
All companies also use asset management decision support tools that link asset health, service and
investment. Many of the companies’ deterioration and failure analysis tools are embedded in
investment planning tools, and all the WASCs and some of the WOCs have some form of integrated
toolkit. These toolkits are quite sophisticated and comprise several components, including a variety
of specialised mathematical analysis techniques for prioritising potential schemes and optimising the
planned investment programme.
However, a key and common component is that these tools are based on a set of service measures
or KPIs that can be forecast and can be monetised so that the costs and benefits of expenditure can
be quantified. Many are planning to use their PR14 tools for PR19. Others such as Bristol Water and
Yorkshire Water are making more radical changes to their decision support processes and tools,
which are used to estimate the investment required to achieve service targets. These tools rely on a
body of information that reflects the different asset hierarchies and structure for infrastructure and
non-infrastructure assets. Decision making data can be aggregated and dis-aggregated to
understand performance and service at different levels of granularity and using different timeframes
for analysis.
All companies emphasise they (and their planning tools and processes) are using a risk-based
approach, consistent (as far as we can tell) with the current revision of the CMPCF (UKWIR’s
expenditure planning framework). We are told that the tools enable forecasting of service, which
may be directly related to an asset performance measure for example collapses or bursts and enable
the modelling of the effects of investment on service.
It will be important to be able to demonstrate that the assumed linkages and relationships between
asset health and service are appropriate and are validated accordingly.
4.4.2 The tools and methods used to create linkages between asset health
and service and assess risk
Dashboards, scorecards and GIS
We found that almost all companies used a range of dashboards to communicate and manage
measures of asset health at various levels in the business. The best examples of these dashboards
were served with data from a single and assured point in the business. We were often told that
companies were working to consolidate, verify and improve data sources. Companies were intent on
eliminating data duplication and the need for people to hold or locally manage their own data. We
often heard phrases like ‘one version of the truth’.
Most companies included indicators that were very similar or identical to Serviceability measures in
their dashboards. Some companies such as Yorkshire Water had added or are intending to add
leading measures to their dashboards. Severn Trent Water have utilised the ‘driver tree’ process
(Leading Indicators Linking to Performance Commitments, 15/RG/05/42) to identify a series of
measures that could be summarised under an asset health banner, but prefer to monitor the trends
in these indicators at the specific indicator level.
It was very apparent that dashboards which include measures of asset health are being used to
communicate and manage service, performance and outcomes in the near, medium and long term:
 For near-term management, asset health data is being used to target reactive maintenance
and responses to failures. For this purpose, the best examples of dashboards we saw were
being used to create a frequent line of sight between operational activities and asset
CH2M 31
performance. We saw examples of visualisations of asset health data, such as hotspot maps,
being used at this level
 For short-term monitoring and management, the dashboards were being used to identify
emerging trends in leading measures, asset performance and the delivery of planned
maintenance activities. At this level the dashboards were used to create a line-of-sight to
outcomes and supported the identification of the need for slightly longer-term risk
management activities, such as asset specific proactive management programmes. In this
role, the dashboards were often used in internal company reporting and routinely circulated
to senior management positions. We saw examples of visualisations of asset health data and
data analytics being used to support this function
 Dashboards were also used to report asset health concepts during board meetings and asset
health data, trends and information were shown to be used to plan and prioritise investment
medium-term within the AMP cycle. Examples of this included analysis of asset health data
being used to support changes in procurement practices, planned preventative
maintenance, ‘Drainage Area Plans’, zonal planning, specifications for new assets and
improving emergency response provision
 To support long-term investment planning, trends in asset health data were being used by
all companies to support the strategic modelling activities that inform AMP to AMP planning
typically based on regression/trend analysis to update and verify deterioration models
GIS tools are used primarily for recording information on infrastructure assets. The GIS is effectively
the infrastructure asset register, storing information on asset type, material, diameter etc. as well as
being used for viewing and visualising event data, including bursts, collapses, customer contacts etc.
The GIS is typically used to support analysis of risks and target investment, based on frequency of
failure and the recorded impacts on customers and the environment. Some GIS have in-built
analytical capabilities, but it is also typical to export the data into specialist analytical and statistical
tools that enable more sophisticated analysis, such as deterioration modelling and service
forecasting, and presentation of the data.
Statistical models and elicitation for understanding the probability of failure
To help predict the failure of assets and service, companies are typically using statistical and data
analysis tools and techniques to examine the failure history and build models that forecast expected
or remaining asset life, understand the probability of failure and how the probability of failure
changes with time (deterioration analysis). Most companies, if not all, have relatively mature water
infrastructure deterioration models.
For assets that fail less often companies acknowledge that statistical approaches may be less
reliable. For civil assets we saw some examples of inspection and monitoring techniques being used
to understand the health of key assets. However, for strategic forecasting purposes, the companies
tended to fall back on some form of statically derived model or expert judgement and accept or
manage the uncertainties associated with the related forecasts.
Tools for estimating the consequence of failure
To help predict the consequence of asset failure, companies employ a range of tools and techniques
which included:
 Using detailed hydraulic modelling to identify impacts
 Statistical and data analysis to quantify consequences, such as analysis of events and impacts
 GIS-based tracing and proximity analysis and workshops to develop risk maps
 Failure modes and effects analysis (FMEA
 Reliability and Maintainability assessments (RAM)
CH2M 32
At least one company is actively engaged in a programme of work to quantify system resilience using
models that can simulate conveyance, asset reliability, system control and resilience hazards
underpinned by a 4Rs type assessment. This approach draws on a range of asset health and attribute
data along with the configuration and operation of the system to understand and quantify strategic
system risks.
Infrastructure assets
We were told that water distribution infrastructure can generally be managed reactively. When
these pipes fail, the consequences are relatively minor and can be addressed most expediently with
an effective response/recovery. Replacing a distribution pipe line to avoid a single burst would not
typically be cost beneficial. However, if a distribution pipe is in poor health and fails repeatedly the
economics of the decision may change and analysis of whole life costs and the risk may merit a
replacement decision. The type of analysis used to understand and predict service risk (WLC and
deterioration modelling) is now routine.
Large, high consequence of failure trunk mains present a different problem. Failure probability is
very low and often the asset in question has never failed (so asset performance is not useful as the
health measure), difficult to predict and consequences of failure can be severe. It is difficult to link
asset health to service for these assets and gathering asset health data, such as condition data, is
expensive. A criticality based, pro-active assessment and monitoring approach may be justified. The
issue of this type of asset is discussed in section 4.4.4.
Similar considerations apply to sewers, though condition based inspection using CCTV is well
established and used to inform potential service risks and expenditure needs.
Non-infrastructure assets
We were told that production assets are complex and have a lot of built in redundancy so that they
do not ‘fail’ when individual components fail and a variety of operational responses can be deployed
to address issues. We also acknowledge the message that the works output quality is used primarily
to monitor performance.
Taking this at face value, the operational resilience and redundancy can compensate to a greater or
lesser extent for poor asset health resulting in asset failure. However, asset failure has a financial
cost and if several assets fail simultaneously of failed assets are not repaired in a timely fashion then
the implications could be serious.
In theory, the contribution of each asset component could be quantified probabilistically by using
tools such as reliability block analysis, but generally it appears that water companies use more
pragmatic approaches for understanding the reliability of the components that comprise their
infrastructure assets.
We feel that more work can be done to help better understand complex non-infrastructure assets.
We believe that companies are moving in the right direction and are capturing equipment failure
data, trialling condition monitoring tools and applying maintenance approaches such as RCM, but
this does feel a little ad hoc and a more concerted approach is advocated.
The Expenditure Planning Framework (the new Common Framework)
The water companies in England and Wales have been acknowledged internationally for the
implementation of the Common Framework to support risk-based asset planning at PR04.
The Common Framework was introduced for PR04 to raise the standards of asset management
planning in the sector and to help ensure that risk-based planning was at the heart of this approach.
Whilst this is primarily a guiding framework, the documentation provides guidance on implementing
the various stages of the framework and is generally perceived to represent a good practice
approach. The Common framework was recently refreshed by UKWIR (the Expenditure Planning
Framework), but it is still in essence a risk based approach that encourages companies to quantify
CH2M 33
probability and consequences of service failures; value the impacts and model the costs and benefits
of interventions in order to maximise benefit.
4.4.3 How are asset health measures used to inform timely and effective
expenditure?
Strategic forecasting of risk and the benefit of interventions
All the companies utilised some form of software to quantify or qualify future risk and identify future
risk mitigation activities. The approach considers the probability and consequences of failure and
looks at the costs and benefits of expenditure. The value of the benefit is informed by customer
valuation of service measures in the service measure framework. To support this, many of the
companies (all of the WASCs) use integrated asset planning tools (such as SEAMS’ WILCO,
Tynemarch’s PIONEER, ICS’s Investment manager, Yorkshire Water’s LEADA) to support risk
assessment, deterioration modelling, service forecasting and expenditure optimisation. The
elements of the risk based approach can be seen in Figure 4.2.
Figure 4-2: Elements of a risk-based approach to asset management
4.Consequence
2.Probability of 3.Likelihood of 5.Value of
1.Asset Health of service
failure service impact investment
impact
As we heard (and is reported elsewhere in this document), the failure probability can be derived
from analysis of failure and performance data, modelling and extrapolation, inferred by expert
panels or indicated by pertinent asset condition measures. In some cases e.g. where failure data is
plentiful or degradation mechanisms are well understood, deterioration can be used to estimate the
effects of age related deterioration on assets. Context information such as temperature and rainfall
data are typically included as causal factors in the statistical models used to predict rates of failure.
We note that statistical models are inherently uncertain and this needs to be factored into their use
as decision making tools. Companies need to be assured (and be able to assure others) that the data
is accurate.
Companies told us that they are working to understand how best to manage their assets, or, rather,
the value that they deliver, and this requires a broader understanding of resilience and the inherent
level of resilience of their asset systems.
Companies do not make investments purely based on analysis from models and theoretical linkages
between asset health, service and expenditure. Experts are consulted and actual failure and asset
CH2M 34
data is examined before committing investment. Areas and tools subject to ongoing development
are:
 Asset performance scorecards and use of asset health data
 GIS
 Understanding resilience
 Optimising health and resilience
Identifying risks and targeting investment
Several companies acknowledge that their asset data records are not complete and that in some
cases they do not know where all of their infrastructure assets are. But they believe that this is not
an issue in that it does not stop them delivering their outcomes. By and large, they believe that they
know where all of the assets critical to service delivery are and the risk of these critical assets
causing a service failure.
Performance data is mapped against assets using GIS and scorecard tools to identify the assets that
are failing in reality (predictive models tend to give general results and contain uncertainties) and
enable targeting and expenditure according to observed risk.
All companies now hold their network asset data in a GIS system. As mentioned in section [4.4.2 it
was also apparent that companies are continuing to develop scorecards/dashboards of asset health,
performance, service and outcome data to manage service delivery. These dashboards were usually
linked to corporate data systems and created important lines of sight through the business. Nearly
all of the companies we interviewed were able to demonstrate dashboards that drew on asset
health data to manage and communicate performance, service and outcomes on a regular basis.
The makeup of those dashboards varies: some are live, some are a snapshot, and some track trends
over time.
In most cases these dashboards and GIS provide a line of sight between asset health measures and
performance, service, outcomes and expenditure. Some companies could show us live dashboards of
asset and failure data operating in real time. This represented a live demonstration of the water
assets on GIS, which resolves the asset health and performance data down to the individual asset
level. This allows the company to identify root causes of service failure at an asset scale, and is used
both operationally and for long term planning of expenditure. The best examples of these
dashboards were built on validated and assured data from a single source.
Asset failure data combined with service impact data is typically used to target where assets will be
replaced or rehabilitated. This is typically done in the case of network assets where the failures and
events can be mapped and analysed and the root cause of failure identified.
This works well where failure data is plentiful and because the assets can be managed reactively, it is
OK to use hot spot/cluster analysis (using lag indicators) to target capital maintenance effectively.
This targeting approach is used in combination with asset failure rate deterioration analysis to help
estimate the envelope of investment. For these lower consequence and relatively high failure
frequency assets, this type of approach is probably sufficient for managing service, risk and asset
health. Indeed, the current basket of serviceability based asset health indicators bear this out as
service is largely stable across the industry.
Asset health is not always an accurate or reliable indicator of service (and vice versa). For example,
for non-infra assets (and if we assume that asset health is a measure of the propensity of the asset
to fail due to break-down) then asset health and the consideration of maintenance optimisation are
potentially more about efficiency and reducing WLC (whole life costs) of running the plant, rather
than optimising the output performance of the works.
For lower importance assets, targeting expenditure to performance and monitoring the average
rates of deterioration for the asset cohorts is probably fine for ensuring that they do not create an
CH2M 35
asset health ‘bow wave’. For critical assets, especially where failure probability is low, there is a
danger that it becomes too late to intervene in a timely manner. Understanding the state/condition
of critical trunk mains is advantageous, if expensive in terms of the inspection. For major water
production assets, the exact role of asset health is less clear and a wider understanding of risk and
resilience is probably important.
Companies recognise the need to develop more leading indicators and that improvement is possible
but also point out that service is being managed effectively using existing information.
This is a slightly mixed message, but it reflects the fact that a) service levels are generally satisfactory
and b) despite this, there is probably more that can be done in terms of developing lead indictors of
asset health in order that improved forecasting of future service can be achieved.
We consider that there is work to be done in understanding how best to balance asset health
investment against operational and system resilience measures and that there is opportunity to
improve non-infrastructure maintenance planning by further deploying maintenance optimisation
approaches.
Clearly, asset health measures, on their own, do not fully inform expenditure decisions. We have
already discussed that asset health is part of the wider resilience picture and so resilience of
operational responses, system configuration and asset redundancy also need to be considered. Our
(CH2M) thoughts are developed as follows:
Understanding resilience
Likelihood of service impact can be influenced strongly by other resilience factors such as the ability
to respond operationally, or the presence of a back-up supply. A simple risk-based model and
statistical extrapolation of asset failure and service impact works particularly well where there is a
direct and easily quantifiable link between asset failure and a consequence of service impact.
Companies recognise that this is difficult where there was redundancy and standby built into the
configuration of complex asset systems and efforts are now being made to better understand and be
able to model these decision-making parameters.
Optimising health and resilience
Several companies are exploring the relationship One company is actively building
between asset health and resilience in more detail. system models that are able to
At least one company is considering whether it is simulate the conveyance, asset
possible to identify an economic level of resilience. reliability, system control and
As discussed in section 2.4.2, companies recognised resilience hazards aspects of their
the complex interplay between asset health (in terms water strategic supply systems.
of low failure probability, resistant and reliable The models provide a way of
assets) and operational resilience (system drawing together asset health
redundancy, effective response and recovery) and
and resilience concepts
the need to better understand these relationships.
Figure 4.3 illustrates the possible relationship
between asset health (resistance, reliability dimensions of resilience) and operational resilience
(redundancy, response & recovery) investments to manage service.
In this figure, there is an assumption that service is relatively stable across the whole range.
However, at the extremes, service is likely to start to fail as it becomes impractical to adopt a wholly
asset heath (using reliability and resistance) based, or wholly resilient system (using operational
response and redundancy) based, policy. Asset health is an integrated part of the 4Rs of resilience
and this example, where they are forced apart to illustrate a concept, is somewhat artificial; they are
integral and not diametrically opposed.
CH2M 36
However, the point we want to make is that there is likely to be an optimum mix (delivering greatest
value over whole life) of the options i.e. the 4Rs. This is likely to be situation specific and may change
with time. These are important issues and merit further investigation.
Figure 4-3: Illustrative relationship between asset health and resilience
It is important that no element is neglected or overlooked when considering risk factors and
exploring the best balance of options.
Companies need to decide if replacing/maintaining to create a more ‘reliable’ healthy asset base is
better than building in redundancy/connectivity, response and recovery or other operational
approaches provides the best investment efficiency.
To achieve an economic level of resilience, now and in the future, the water companies will need to:
 Improve understanding of Asset Health and better define the probability of failure and
develop strategic repair, renewal and procurement plans
 Better define customer impact of asset failures based on detailed hydraulic understanding
and improved GIS representation of networks. This improved level of information also
enables better response and recovery operations
 Complete strategic evaluations of network distributions to identify isolated communities
that rely on a single asset
 Clarify the best value intervention in terms of operational management, repair or renewal. A
more detailed understanding of asset health makes it possible to assess whether to invest in
short term actions or long term capital projects
 Further clarify the best form of intervention in terms of operational, repair or renewal and
focus long term investment on enhancing resilience
Prioritising investment based on service measures has resulted in significant improvements to the
asset base and generally stable service and these improvements give companies the opportunity to
evaluate other factors that influence the long-term provision of reliable service. When the network
system is considered, it becomes apparent that risk may be mitigated by considering system
configuration changes, but it is recognised that this is difficult to quantify where there is redundancy
and standby built into the configuration of complex asset systems. Some companies are therefore
building system models that are able to simulate the conveyance, asset reliability, system control
CH2M 37
and resilience hazards aspects of their water and wastewater strategic supply systems. The models
provide a way of drawing together asset health and resilience concepts. There several emerging
simulation tools (simple system model example is shown in Figure 4-4) which may help to identify
risks and identify the best forms of interventions in line with a 4Rs classification methodology.
We believe that these present an innovative and promising method to integrate and balance asset
health and operational resilience investments.
Figure 4-4: Reliability and resilience simulation
Deteriorating raw water
Bore hole pumps WTW and Storage Pumps to distribution
main
4.4.4 The role of criticality

The concept – of ‘criticality’
The was
Reliability explored with all companies. From discussions it was evident that a
Block
range of viewssimulating
and processesthe reliability
were used ofto
a try and qualify criticality. All companies described
raw water main the system.
criticality (and consequence) concepts being used in decision making and how criticality was used to
define appropriate risk mitigation regimes for assets such as inspection regimes and development of
strategies such as such as ‘fix on fail’ or planned preventative maintenance.
Most WASCs have maintained the use and definition of ‘critical’ sewers and have been using this
definition to inform CCTV survey programmes.
Company responses reflected the use of different measures of criticality being applied across
different assets in the same company.
Some companies have used tools and models to generate criticality ‘scores’ that are stored in
corporate systems at an appropriate level. An example of this approach is the use of water hydraulic
models to understand the consequences of asset failure by systematically failing each pipe in the
network. The scores generated by this approach can be assigned to individual network assets and
used across a range of decision support functions.
Non-infrastructure assets have also been studied and the criticality assessed. Tools and methods
including reliability block diagrams and reliability concepts have been used and the findings captured
at an appropriate level in the asset hierarchy.
Companies articulated the statutory obligation that are placed on them to inspect some of their
‘critical’ assets at specific intervals such as large raised reservoirs.
Several companies articulated formal definitions of criticality for water networks or zones of water
networks. These definitions considered the sections of the network linked to sole sources of supply
serving more than a specific number of customers.
In general, companies articulated a view of criticality that was either:
 Criticality is about the consequences of failure, or
 Criticality is about risk
When we spoke to companies about asset health and ‘critical assets’ we asked specifically about
how best to manage ‘high consequence, low probability’ assets.
CH2M 38
We believe that for many water company assets it is perfectly acceptable to use asset performance
and failure data and statistical analysis to determine failure probability and deterioration and
thereby plan timing and extent of expenditure. Typically, these are assets whose failure would have
a small or negligible impact on service and where a reactive maintenance approach can be effective.
However, if the asset is one where the consequence of failure will be very high and the failure of
probability is very low, then it may not be possible to extrapolate historical failure data to help
estimate failure risk. It is inevitable that there will be a number of critical, high consequence of
failure assets, such as major trunk mains, where companies are not fully aware of the state of these
assets and hence the risk of failure. It was not so evident from responses as to the level of
confidence that companies have that failure probability of the critical assets is fully understood;
some companies have told us that they are prioritising monitoring and inspection based on
consideration of the criticality and consequence of failure of these assets.
In these cases, it is important to know where these assets are, what the consequences of failure will
be and to consider the value of gathering additional asset health data, such as physical condition, to
help inform the asset management process and plans put in place for dealing with potential failures.
We were told by some that assets with a failure history are the priority, but this will be of little
consolation if the ‘first burst’ results in devastating consequences and the company was not
prepared.
In the case of these important assets, we (CH2M) believe that a key element of an asset’s health is
the ability to resist load. Many of our assets are exposed to aggressive operating environments and
as they age, the resistance (R) decreases and the asset will become more vulnerable to failure under
load (S). A typical example would be a steel trunk main under high operating pressure that has a
damaged external coating and is being corroded by an aggressive soil. There is uncertainty around
both the resistance, the stress and indeed the rate of aging. However, as the asset deteriorates, the
distributions of stress and resistance start to converge and asset failure probability increases. The
capacity to resist load is a potential measure of asset health.
Figure 4-5: Asset deterioration and failure probability
CH2M 39
If the asset is still relatively young, it may be many years before the two distributions (R & S) start to
overlap and failures start to occur3.
There is a widely held belief that at some point investment in asset renewal (or rehabilitation) will
need to increase to maintain customer service. This was also a conclusion of a soon to publish
UKWIR project on long-term investment in infrastructure. This is because companies’ investment has
not been keeping up with asset deterioration. The effects of this on customer service have, to date,
been masked by other elements of resilience (operational response and recovery, and redundancy).
But eventually, the lack of investment will impact on service, as other resilience measures are
overwhelmed. If many of the assets reach this point at the same time, there may be a “cliff edge” of
unmanageable failures.
We would anticipate that, at present, rates will increase slowly due to natural variation in
deterioration rates (the assets won’t all reach the end of their life at the same time). Hence, whilst it
is important to understand aging, monitor assets and be sustainable in the long term we (CH2M) do
not consider a ‘bow wave’ of poor asset health and a ‘cliff edge’ of failure to be imminent. However,
it would be prudent to continue to research these issues; develop a sound understanding of the
asset health and resilience trade-offs and have a mechanism in place to detect change in a timely
fashion so that efficient measures can be taken if and when necessary.
The more immediate concern, therefore, is the unexpected failure of individual critical assets, the
state of which may not be known with any accuracy. If this is a critical asset (such as a trunk main)
and one where the impact of failure is significant, then there is great uncertainty regarding the risk it
presents. There is value in trying to understand the current state of the asset, specific deterioration
processes and the stresses in order that we can determine when the failure probability is becoming
a significant factor in terms of the overall risk. We also need to understand consequence of failure
and criticality as influenced by wider consideration of resilience, so that risk can be managed
optimally. The asset manager will also need to be able to determine at which point they will need to
focus on asset health as opposed to reliance on system resilience and operational response.
It is tempting to recommend that such assets are prime candidates for undertaking condition
assessment to help understand the failure probability and we note that companies do undertake
such appraisals. However, we acknowledge that assessment approaches such as discrete sampling
may not be practical or accurate enough; that in-line inspection may be expensive and extrapolation
from cohort models too imprecise which may limit the extent of pro-active inspection, but this is an
area where the industry should come together to share insights and research solutions.
3 For a population of similar type assets it may take some time before service deteriorates noticeably despite maintenance or replacement
of assets is not being balanced against deterioration. A population of assets will also tend to exhibit a wide range of rates of deterioration,
which explains why ‘cliff edges’ are not common for populations of long life assets. However, where assets age at a more similar rate and
are of a similar starting condition, a more pronounced effect could be observed. This would be more likely for shorter life assets operating
in aggressive environments.
CH2M 40
Understanding the views of the customer
5.1 Context
As part of our review we wanted to understand the extent that individual companies have
developed an understanding of asset health and subsequently to what extent an asset health
dialogue has been undertaken with customers. It could be considered that the two concepts are
interrelated, hence if the company has a specific definition of asset health it is more likely to
undertake a dialogue with customers explicitly about asset health.
In addition, if a company has undertaken a dialogue with its customers, has it has developed an
effective language to convey the key components of asset health to its customer base? Key issues
we wanted to consider here were whether, based on the dialogue with the company, customers
were likely to understand asset health.
A final focus for this section was whether once an effective and discreet definition of asset health
has been conveyed to customers, customers have been able to assign a value to asset health.
Although customer engagement on asset health and more especially resilience has been mentioned
in the past, it is only in more recent communication with WOCs and WASCs that it is being discussed
in issues related to the business plan submissions for PR194. For example, in Ofwat’s November 2016
consultation on the outcomes framework for PR19 three of Ofwat’s six draft asset health
expectations for PR19 related to customer engagement including: “2. Companies should engage with
their customers and CCGs on how their asset health measures protect current, future customers and
the environment.” (Appendix 3, page 24).

We asked the following questions in order to form a view of the extent that asset health has played
in customer consultation and eliciting customer views and values:
 Have you had a dialogue with your customers regarding asset health? The focus of this
question was to understand if companies have had an explicit and direct discussion with
customers in relation to asset health
 Have you developed an effective ‘language’ for conveying asset health concepts to
customers? The focus of this question was to understand if companies had attempted a
dialogue with customers, whether that was effective in conveying issues related to asset
health, or whether, despite the attempt it was too difficult to convey asset health principles
to customers
 What value do your customers place on asset health and how is this evidenced? The focus
of this question was to understand whether given an effective description of asset health
customers are able to assign a value to asset health more generally
4 E.g. A consultation on the outcomes framework for PR19 Appendix 3 – Better reflecting resilience in outcomes, Ofwat 11/2016 ;
Outcomes – Water 2020 stakeholder workshop 21/2/17; Ofwat’s customer engagement policy statement and expectations for PR19
25/5/2016; Ofwat workshop on outcomes 14/6/16.
H2M
41
We expected a varied response to these questions from companies due both to the absence of an
accepted definition of asset health and that companies are at different stages of their engagement
with customers for PR19.
We expected some companies to have explicitly considered asset health issues in relation to
customer dialogue and engagement. Our expectation was that this would be where there has been
active engagement for PR14 and/or an ongoing focus on asset health and/or resilience since then,
partly because of PR19 expectations and partly because asset health has actually become embedded
into the organisation as an operational driver.
Where asset health has not been embedded or played a key role in PR14, we expected to find less
focus on asset health in customer engagement to date. This expectation was driven partly by a lack
of explicit direction for engagement on asset health from key regulators; we note that Ofwat did
emphasise the importance for consultation on performance commitments, which includes those
currently described as asset health measures. In such instances our expectation was that companies
would continue to assess broader outcomes and focus on maintaining acceptable service levels,
broadly in line with the common framework approach. This does not reflect a lack of interest by
companies in asset health, only that it takes time to reprioritise, plan and incorporate new issues
along with existing components into their plans.
Where companies have explicitly considered asset health, then we would expect to hear mention of
plans and issues around asset health and how it had or could be incorporated in deliberation and
quantitative work, the role of valuation as well as how it would influence decision making processes.
5.4 What we were told

5.4.1 Customer dialogue and asset health
We asked companies “Have you had a dialogue with your customers regarding asset health?”
Companies told us that for PR14 they engaged with customers to understand service priorities, views
on future Outcomes and Measures of Success and the overall acceptability of their plans. Plans for
PR19 were at different stages in different companies, though few shared these in significant detail.
Most expected to make enhancements to their approach at PR14. Enhancements included:
 Deliberative assessment using amongst other methods, focus groups, deep dive interviews,
and off- or on-line customer research panels
 Quantitative work involving surveys including stated preference (Willingness to Pay (WTP))
and in some cases revealed preference studies. Some companies had two phases: Phase 1
informed, and prioritised outcomes and the value of changes in service attributes and was
often used as an input to optimization modelling; Phase 2 assessed the acceptability of the
business plan service choices and their cost, identified through household bills. In some
cases, revealed preference studies (e.g. cost of a trip to the river) were used to value
attributes and provide a comparator and/or an element of triangulation
 Specific topic studies including for example vulnerable groups, investment for the future,
drinking water priorities, resilience and acceptability
All companies engage with their customer groups through a variety of communication mechanisms
and all are undertaking or planning customer research to value service delivery. However, asset
health does not feature as an explicit topic in most cases.
WOCs and WASCs provided a clear response on the extent of asset health dialogue that has been
undertaken with customers. In the majority of cases WOCs and WASCs advised that they did not
undertake a direct discussion with customers in relation to asset health. However, in many of these
CH2M 42
cases WOCs and WASCs feel that other questions/queries with customers can be considered as
appropriate proxies for asset health or, rather, the service delivered by assets that are healthy. For
example, all companies had performance commitments related to asset health at PR14 and
companies engaged with their customers and CCGs on their performance commitments.
Figure 5.1 sets out the proportion of companies that had an explicit dialogue with their customers
and those companies where an explicit dialogue in relation to asset health was not undertaken (note
that this is a composite view based on PR14, business as usual and preparation for PR19 dialogues)
Figure 5-1: Levels of asset health dialogue (including at PR14)
29 71
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Direct Dialogue No or Little Dialogue
Based on the findings of our review we understand that almost three quarters of companies had not
undertaken an explicit dialogue with their customers in relation to asset health at this time. The key
reasons companies had not engaged with customers on asset health were cited as:
 Companies believe that it is difficult to engage with customers on asset health as customers
do not necessarily understand the concept and have difficulty relating to it
 Asset health is considered a complex issue which customers can struggle to understand
 Asset health is considered to be part of ‘business as usual’ operations and customers expect
this to be managed on their behalf by the company
 Customers are more interested in the more immediate issues such as resolving repeated
supply interruption
 Asset health and its ‘symptoms’ are implicit in discussion of direct issues such as quality of
supply, leakage and interruptions
Our review found that a number of companies did discuss asset health issues with customers –
though we are not clear how often the term ‘asset health’ is used directly.
Wider evidence suggests that asset health issues can be important to customers. For example the
deteriorating health of the Butterley reservoir spillway provoked a community controversy and press
and social media discussion. A member of public launched a crowd funding appeal to fight or change
ongoing plans. This illustrates the importance and provides an opportunity to understand customer
priorities and trade-offs in relation to the asset health.
The numbers who contributed and joined the social media discussions reveal the importance of
participation by certain customer groups for some types of asset.
CH2M 43
Examples of asset health discussions with customers
 Yorkshire Water have an Online Community of customers as part of ongoing customer
engagement activity for PR19. The community is made up of a representative sample of
customers from across the region. The community is involved in discussion across many topics
regarding service and communications that customers receive from Yorkshire Water. Asset
health is one of a number of topics currently being discussed with the online community.
 Other approaches mentioned include wider stakeholder contact including bi annual feedback
from MPs, the use of company magazines to inform customers on company news including
new and replacement assets. Several companies said customers had identified the need for
better understanding and education in relation to asset health
 South Staffordshire Water has a customer panel who have been engaged on asset specific
schemes which impacted on asset health at PR14. This has been supplemented with recent
CCW/customer panel engagement on current asset health decisions.
 Direct customer surveys have been undertaken by Southern Water to understand whether
customers value affordable, stable asset health and service resilience including for future
generations.
Some companies said they would undertake further work on asset health but were not specific in
what form this would take. Others, mainly water only companies, said they expected to continue to
develop their existing approaches using valuation of assets through service and performance
outcomes.
Northumbrian Water told us they recognised the need for a clearer steer on customer involvement
in asset health. In preparation for PR19, they plan to talk more to customers about what they do and
establish what customers want to have a say in and what they are happy to allow the company to
plan and decide on their behalf without a further steer.
5.4.2 Effective language to convey asset health concepts to customers

It is clear that only a small number of companies have had an explicit asset health dialogue with their
customers and companies can consider the concept too difficult for customers to consider directly.
Four themes on language and asset health concepts emerged from company responses. These
themes were around:
 Ease of customer comprehension and the complexity of asset health
 Perception of risk and how it was framed
 Ability to target different groups using appropriate communication channels and content
 Use of scenarios
5.4.2.1 Asset health complexity and comprehension

Companies told us customers do not generally understand what is meant by asset health and they
are more interested in asset service and performance, which are concepts they can relate to.
Companies tend to engage with customers through these elements and from this infer the
importance of asset health. In other words, it is not always necessary to have a dialogue with
customers on asset health in order to assess asset health. Three responses highlighting the issues
are provide in Table 5.1 below.
CH2M 44
Table 5-1: Customer understanding of Asset Health
Company Message
Thames "Customers understand that maintaining the current level of service is impacted by investment in Asset
Health both now and in the future. They can engage with this concept and recognise that spending in this
area may need to rise to deal with the effect of ageing assets in order to maintain the same level of service
they currently enjoy.”
Anglian Evidence from Anglian Water’s Domestic Customer Survey [1] suggests investment in current and future
maintenance is an important customer priority (after tackling leaks). However this was also the area that
participants in qualitative research and engagement activities found the most difficult to comment on.
Severn “It is not straight forward to engage with customers on this critical issue. As such, when we engage with
Trent customers about asset health we typically do this using terms and concepts which they are more likely to
relate to”. This includes:
 understanding their willingness to pay for service improvements that can be a manifestation of changing
asset health, for example supply interruptions and sewer flooding;
 understanding attitudes towards risk and resilience, and a willingness to pay for a reduction in risk, and;
 more generally views on maintenance, infrastructure and protecting our services for future generations.”
5.4.2.2 Risk and risk perception

Several companies told us that customer perception of risk is affected by the way in which scenarios,
choices or questions are framed and ordered. Severn Trent, Thames and Anglian each plan to
implement research (qualitative and quantitative (WTP)) with people previously impacted by service
failures due to asset health or related to resilience. Companies did not say how they plan to use any
emerging findings. However, they could provide case studies and insight to improve situational
reality and understanding for others, they may form a comparator group for valuation studies or be
used for triangulation.
Severn Trent, Thames, Anglian and Northumbrian Water are all planning to use gamification or
simulation in developing an ‘immersive’ approach to explore how better to help customers
understand and appreciate the potential impact of a service failure on their daily lives. This work
may have some significance with respect to asset health.
5.4.2.3 Targeting and appropriate communication

Before developing their digital strategy in 2015 Yorkshire Water told us they engaged with
customers and used segmentation to understand how different types of customers would like to be
communicated with in different situations. The work undertaken also led to guidance being
developed on tone of voice and what was appropriate for Yorkshire Water to use on digital channels
when communicating with their customers.
Companies such as South West Water, Thames and Anglian have undertaken analysis using
demographic and locational factors to improve their understanding of customer segments.
Companies also segment customers in terms of billing arrangements. Companies appear to use
conventional approaches based around geographic or socio-economic factors. No companies
mentioned the use of more holistic market research approaches for groups of customers with similar
concerns or drivers and interests using methods such as principal components.
5.4.2.4 The use of scenarios

Anglian, Severn Trent and Thames provided examples of research to understand customer
understanding and valuation in qualitative and monetary terms of asset resilience and asset health.
For PR14 Severn Trent researched a range of issues around ‘resilience’ including the Elan Valley
Aqueduct and the resilience of supplies to Birmingham. The core focus was the then current supply
situation for Birmingham and four possible future options. Customers expressed a clear willingness
to pay for greater resilience in order to avoid extended interruptions.
CH2M 45
Thames has undertaken phased surveys to investigate the influence of demographics and geographic
location on awareness and attitudes to the Thames Tideway tunnel and its affordability.
5.4.2.5 Summary
In summary, different ways to discuss asset health and resilience are being used. The use of
scenarios in stated preference surveys has had some success. Research to understand targeted
digital communications has supported digital and on-line discussion on asset health and other
subjects.
5.4.3 The value customers place on asset health

We asked companies what value do your customers place on asset health and how is this
evidenced?
As already discussed, companies have told us that most customers find it difficult to value asset
health per se. They understand service and performance outcomes and, in general, are able to
prioritise and value changes in these. Valuations are undertaken at individual service level impact
and in many cases follow a similar approach to PR14.
Key elements of the valuations of individual outputs or service levels are:
 Companies have used contingent analysis surveys to assess willingness to pay for a change in
a particular situation
 Situations can be hypothetical or specific questions can be linked more directly to alternative
options. These can give customers the opportunity to more clearly indicate if one solution is
preferred or valued in monetary terms over another
 Some companies have used alternative revealed preference methods to elicit valuation of
outputs
 Some firms such as ICS and Accent have used academic review of the studies as well as CCG
groups
 Companies have undertaken deliberative in-depth interviews, focus groups, customer panel
surveys, WTP studies and revealed preference work5 which can be used together to
triangulate and support (or not) valuation work. Company interviews and responses did not
explore the strengths and weaknesses of valuation or different valuation methods
 Anglian Water explained some companies used stated preference approaches such as
conjoint analysis, which can value individual service or performance attributes. Optimization
decision support programmes use these in investment prioritisation. The value of asset
health is inferred using this approach but could be explored further if assessed explicitly as
an attribute
 Southern Water reported,6 “While WTP research in its current form provides a strong
evidential basis for capital investment planning, there have been challenges, not least from a
number of CCGs, on the extent to which customers really understood the stated preference
exercises or whether they were too detached from real-world decisions”
In terms of the asset health service values drawn from PR14 Stated Preference survey reports the
amount of directly evidenced support for asset health is limited. We do know that:
5 which assess the value consumers place on goods and services from their purchasing habits or user activity
6 Water 2020 Customer engagement: lessons and opportunities, Southern Water
CH2M 46
 Anglian Water said service attribute responses and the values assigned provide a level of
customer demand. This can be mapped to, and compared with different supply side service
solutions. The solution can include operational as well as asset health and capital solutions
 Business plan acceptability work for Southern Water suggested that dual-service household
customers were willing to pay up to 8.2% more (£38) for an intermediate package of
improvements, and 16.1% (£75) for a maximum improvement package. Some of the
improvements within the packages were asset health related but it may have been difficult
for customers to identify this
 As part of the Thames PR14 customer acceptability testing, they explained asset health as
‘maintaining a safe and reliable water and wastewater system’
 For certain services, and with some approaches7 some companies are using failures and
levels of performance as a proxy to infer values of embedded asset improvement in a
package of services. However, business needs and priorities do change and funding for a
given outcome or service could be spent flexibly within the boundary of the outcome or
service and may include asset health solutions or not
 For Anglian Water current and future maintenance appeared in their list of priorities, with
56% of customers indicating that they would support an increase in maintenance investment
from a baseline level
 Most companies told us that their customers have indicated they want a reliable and good
quality service and that they trust the companies to plan and deliver this accordingly
7 conjoint analysis to assess service attributes
CH2M 47
Assuring our approach to asset health and
maintaining service
6.1 Context
The companies were asked four questions around the subject of assurance. We wanted to
understand how they apply themselves to assuring that the knowledge of their asset health is
robust, and therefore that their ability to provide the service to their promised standard is soundly
reported and based on good information.
We asked the following questions, in the order they are discussed in the report:
 How successful have you been in delivering your asset health commitments and what have
been the main factors?
 What assurance processes do you have in place to ensure that your asset heath related
obligations (the general duty to maintain your system) are being met? Do they address:
o Asset data sufficiency and forecasting of service and timing of investment
o IT systems
o Understanding of risks to service
o Effectiveness of models
o Resolution and targeting
 Response to asset failure:
o How do you identify and assess the implications of potential asset failures?
o In cases where failures cannot be prevented, or are unexpected, what are your rapid
response emergency measures and longer term care plans and how do these
mitigate the impacts of failure?
o How do you capture lessons learnt from the above and incorporate them into the
way you manage asset health, especially when considering proactive and innovative
interventions?
 What assurance and testing is done on new assets to help ensure that they will achieve their
intended design life?
The responses and discussion around these questions will provide an understanding of how
companies view their current performance in maintaining asset health and the key factors that
support that performance.
Following through the responses will illustrate how company assurance processes are supporting the
stated view of performance through successful scrutiny of their data, models and systems, and
whether or not these allow a comprehensive understanding of the risks of asset failure and how to
respond to such failures. Responses will also illustrate how, when failure incidents do occur, a
company then feeds the lessons and information learned into a continuous improvement regime.
Finally, the responses will show what assurance is in place when developing new assets to provide
confidence that design lives will be met and how is this influenced by experience with existing
assets, the present assurance processes and recent lessons from asset failures.
CH2M 48
6.2 Successful delivery of the MoS
We posed the question to all companies, “How successful have you been in delivering your asset
health commitments and what have been the main factors?”
The responses received allowed a qualitative assessment of their relative success in meeting their
individual performance expectations. Responses also provided elaboration on the main factors
affecting performance against expectation and the context in which these factors should be viewed.
The question is searching for information and insight that would allow a view to be taken of the
appropriateness of the Measure of Success metric used by a company; the associated factors
contributing to success or otherwise of a company in performing against the measures would
hopefully provide demonstrable evidence of that appropriateness.
6.2.1 Delivering on asset health commitments:

Universally, companies state that they consider they are meeting their target measures for asset
health. The performance is generally substantiated with reference to meeting serviceability and
customer service target levels of performance over a period of time, commonly AMP 5 to the
present. Some companies noted longer periods back to when specific monitoring data recording was
begun in earnest, for example Portsmouth Water’s reference to Burst Rates since 1997.
Several companies caveat their responses in some areas where singular failures of an asset or
compliance requirement have occurred and detracted from otherwise stable trends in reported
performance. The place in the system, and thus the criticality of the asset that failed, was generally
not included in company caveats and this is considered to be an indication that company awareness
of the health of some of their critical assets in terms of failure probability could be improved, both in
terms of coverage and quality.
Adverse weather conditions were a common factor in such caveats, but as reported by some,
weather conditions can also be benign and allow on occasion significant outperformance against
targets. Care is required with such cases as establishing what the target value is that accurately
reflects that asset health is being maintained in the long term is not free of problems as record
periods can be relatively short.
6.2.2 The main factors for success:

It is evident that there is no single factor and hence remedy that can be applied to companies to
bring them all up to a common standard of performance; the responses provided confirm that
company circumstances and local conditions preclude this and they develop and manage their
approach accordingly. The individual approaches adopted would appear to suit and be appropriate
for the circumstances found, as in the main, the companies are of the opinion that they are meeting
their asset health commitments using their chosen measure and none have indicated that the
measure(s) could be inappropriate with respect to ensuring and maintaining asset health.
There is large variation in the level of detail provided by companies on what they considered to be
the main factors for the success being reported. Overall, from analysis of the responses there are a
number of common themes or elements evident. The success factors are manifold and the
descriptions provided show some commonality at WOC and WASC level where at a high level:
 WOC’s point to specific details applied to a more focused group of assets
 WASCs highlight systems, monitoring and evaluation, consistent data sets and approaches,
and good timely and consistent communication
In the case of Thames Water, it was a case of both being provided with detailed description and
itemisation of specific tools and approaches provided by system and asset category.
We have drawn out the most frequently mentioned factors and factors of particular note as:
CH2M 49
 Targeted investment, ensuring investment is well focused and appropriate and investment
plans are flexible
 Risk models and forward looking risk models
 Tools and systems
 QA, QMS to PAS 55 / ISO 55000 approach
 Monitoring and evaluation, routine and monthly, with review by executive leadership teams
or specialist communication cells to promote uniform, consistent and timely view of
performance and effectiveness. On occasion, this factor arose under the broader heading of
Governance
 Mature alliances, mentioned in the context of delivery partners and similar therefore in
nature to continuity of approach and people
 Common data sets across business processes
 Well-defined measures, in one case expanded further to note that a sensible balance
between mature data and new measures for innovation was a factor
 New and renewed focus, mentioned in the context of reporting and monitoring and more
transparent decision making
 Fresh minds and contribution from outside the water and sewerage sector (petro-chemical
industry and highways both mentioned as sources)
 The success factors have evident linkages and dependencies between them, reinforcing the
higher level factors such as ISO 55000 systems, monitoring and evaluation and
communication and common understanding of performance. There is therefore the
recognition of the need for relevant performance data to be available to ensure forward
looking models are appropriate. This requires that the measures and metrics used are well
defined and this was recognised in a number of company responses
 A theme of consistency over time is evident; some companies referring to mature alliances
and approaches and mature data
 Larger companies, predominantly WASCs, refer to systems and processes, and good
consistent communication within the organisation as key contributors to success. There is no
indication that the lack of mention of this among smaller companies is accompanied by a
lower level of performance and assurance is frequently a tacit element of their approach; it
is more the case that systems and processes are of a form appropriate for the size and
complexity of the organisation. It is not evident that smaller companies pay less attention to
or apply less effort to assurance of their asset health knowledge and processes. It is
reasonable to expect company approaches to assurance to be risk based and from the
qualitative responses provided, once the size of the company is taken into account, the level
of assurance effort does not appear to vary greatly
 Several company responses to questions on the subject of innovation, incentives and
barriers referred to the lack of incentive and reward for the greater risks incurred in
adopting innovative approaches. Greater collaboration between companies was cited as a
possible positive contribution to overcoming this barrier; this would also be a positive
contribution to the assurance supporting the quality of the innovation and mitigate any
influence that company size may have
6.3 General assurance (process, data, tools)

The assurance framework under which companies are operating underwent significant change
during AMP5 with Ofwat requiring companies take responsibility for the assurance that they provide
CH2M 50
across the whole business. Concurrently and in the area of current concern, asset management and
therefore asset health, the traditional role of the Ofwat appointed independent third party Reporter
was removed.
We therefore posed the following question to illicit from companies how they had responded and
structured their assurance to manage the risks in order to fulfil the new obligation placed upon
them: “What assurance processes do you have in place to ensure that your asset heath related
obligations (the general duty to maintain your system) are being met? Do they address: asset data
sufficiency and forecasting of service and timing of investment, IT systems, understanding of risks to
service, effectiveness of models, and resolution and targeting?“
We would expect to see that companies had implemented rigorous assurance at three levels, at the
operational management level and built into the culture as a matter of business as usual, in the
internal risk management and compliance function of the company, and with the internal audit
function. We would also expect significant third party involvement in these assurance activities in
order to bring independent and specialist expertise where companies recognize models and
techniques are new, often complex and still under development. Given the very varied size,
circumstances, and risks present across the companies it would be reasonable to expect a very wide
range of areas and approaches to assurance provision but always built around good asset and quality
management principles and practice.
6.3.1 What we were told – key observations:

The level of assurance that companies describe covers a wide spectrum – from at one end the
rigorous implementation of the three lines of defence to the other extreme of description of
individual pieces of assurance carried out on specific elements of company activities.
The three lines of defence was specifically used to describe the assurance approach in only two
companies, Severn Trent and United Utilities.
The deployment of an independent assurer in a Reporter role
for assurance of annual reporting was universal across all
companies with the scope of that assurance usually based on There is a noticeable
assessed risks. trend, with larger
The asset management processes were stated to be in line companies having
with ISO 55000 principles by seven companies, of which four clearly identifiable
currently have accreditation for their systems and by assurance processes
implication regular independent assurance checks carried out and teams in various
on them.
areas of the company
Of the four companies accredited to ISO 55000 Northumbrian
reported that their wider processes for asset management
were covered by an integrated quality management system
accredited to ISO 9001. Quality management systems to ISO
9000 were reported in four companies in total of which three have sought and achieved
accreditation.
Specific examples of independent assurance were provided by all companies over a range of subject
areas, in the majority of cases there was reference to the assurance of infrastructure and non-
infrastructure deterioration and service impact models. As responses generally referred to specific
examples, it was not evident that assurance was comprehensive and covering all models used, nor
that the assurance was targeted at those considered to have the most risk attached to them and to
be difficult to assure.
CH2M 51
6.3.2 Findings and comparison across companies
There is a noticeable trend in the responses with the larger organisations having more structured
responses to the question and able to point to clearly identifiable assurance processes and teams in
various areas of the company.
We note that it was beyond the scope of this project to look in any detail at the assurance processes
and are unable to say categorically whether they are fully effective.
Some smaller companies would appear to be less likely to have all the same formal assurance
elements and structure in place. Their tacit understanding of a smaller stock of assets, and
theoretically greater ease with which they can understand their assets and their performance, may
drive a perception that there is lower risk to their assets and operations, therefore some of the
assurance processes required of a larger company are perhaps not such a priority for a smaller
company.
The cyclic nature of business planning and price reviews may also be an influence in this regard; the
majority of companies are implementing revisions to systems or models on some scale in
preparation for PR19 and either are or will be seeking further assurance on these developments.
This is recognizable from the manner in which these will be procured from recognised and
accredited providers, the application of internal company procedures and governance and additional
independent review and assurance on specific areas of concern identified by companies.
6.4 Emergency response and dealing with asset failure

Integral to delivering an approach to managing asset health and a satisfactory service to customers is
understanding the consequences to the service of unexpected failures and circumstances.
For the extreme cases and as directed by DEFRA, a company is required to have SEMD plans in place
and assured every year at a lower level that is focused at an asset level. In these plans companies
have a free hand to develop appropriate systems and although there may be an element of overlap
between the two, it is specifically the lower asset level incidents which are of interest.
The questions asked focused on identifying consequence of and response to failures, and the
subsequent inclusion of the lessons learnt in the future management of the assets, in three parts as
follows:
 How do you identify and assess the implications of potential asset failures?
 In cases where failures cannot be prevented, or are unexpected, what are your rapid
response emergency measures and longer term care plans and how do these mitigate the
impacts of failure?
 How do you capture lessons learnt from the above and incorporate them into the way you
manage asset health, especially when considering proactive and innovative interventions?
Our expectation of companies is that their systems can or have identified in a structured manner the
element of the service they provide that is dependent on an asset. To ensure such a task presents a
manageable workload this would be targeted at critical / high risk assets, that response
management plans would cover a range of failures and that the process would be the subject to a
comprehensive feedback process to improve identification of consequence of failure, management
of asset health and improved targeting of investment.
CH2M 52
Responses show all companies have processes in place to identify the consequences of failure. Risk
based approaches have been employed, criticality of assets
reviewed and networks modelled to understand the parts
of the networks that would leave the company vulnerable
were they to fail. Companies refer to consequences in
terms of the accepted measures of interruption to supply, Lessons learned are
water quality, sewer flooding, pollution events etc., no
frequently used in
innovative measures of consequence were drawn to our
attention, however there is variability in the extent and investment planning,
manner to which consequence is valued. but assurance of
lessons learned is less
Pending improvement in understanding of failure
clear
probability in the critical areas there is a risk that response
measures are not comprehensive, potentially lacking in the
assessment of the scope and extent of potential failures.
Further enquiry into the detail of company response
processes across the wide variety of asset types that they
operate would be required to gauge whether they are not robust and in need of revision sooner and
to a greater extent than the normal review and feedback improvement process that companies are
operating will provide.

There is evidence of continued work being done to
improve understanding of failure consequences,
through, for example, critical links analysis and Risk based approaches have been
modelling and coverage of across the approaches are employed, criticality of assets reviewed
being extended to increase coverage of networks. and networks modelled to understand
There is further work identified by companies as the parts of the networks that would
required for them to have comprehensive view of leave the company vulnerable were
their complete network in a uniform manner and in they to fail
some instances the quality of the assessment is
potentially limited by the incomplete knowledge of
the assets.
In terms of how companies respond to failures, the use of Emergency Management Plans was a
universal response and were acknowledged by some as being an area they would like to improve.
Anglian Water mentioned the strong partnership between operational units and alliance (delivery)
partners; it is not clear how a consistent level of response is achieved through this approach, nor
how it caters for potential shortfalls in asset knowledge that hinder both understanding of
consequence and responses to failures.
The feedback of lessons learned, post incident reviews and the like, are a key element to asset
management and the improvement of long term asset performance; it is formal requirement for ISO
55000 certification. All companies responded that they have lessons learned processes in place, in
the case of Northumbrian it was noted as being improved under the continued certification process
for ISO 55000, reinforcing the benefit of such system implementation programmes.
CH2M 53
A common theme from company responses
Companies are approaching assurance of is the use of lessons learned in the
new assets from three directions: preparation of subsequent asset
investment plans, not so evident is whether
i) Design and technical standards of the they point to needs for updated modelling
assets to identify consequences of failure and
ii) Comprehensive commissioning plans and associated data for these such as
consequence values, or for improved
processes
understanding of the probability of failure
iii) Post project reviews of assets. Such recommendations might
arise from an assurance exercise carried out
on systems were one to be carried out; however, there was no evidence and narrative from
companies on assurance of this aspect.
6.5 New assets

To understand and have confidence in the future health of new assets and confidence in the
assumptions of asset life used in the investment decision for those new assets is needed, we asked
the question:
”What assurance testing is done on new assets to help ensure that they will achieve their intended
design life?”
Our expectation was that measures would be presented in the areas of acceptance and
commissioning of new assets with respect to the technical standards, on site and at point of
manufacture. We expected narrative around how such measures are accommodated in the varied
procurement arrangements to be provided.

Companies are approaching assurance of new assets from three directions: design and technical
standards of the assets, comprehensive commissioning plans and processes, and post project
reviews. Company responses varied greatly in the details and where their emphasis lies but in
general elements of all three areas were evident in all company approaches.
There was clear differentiation in how the processes are applied to infrastructure and non-
infrastructure assets and details provided were of the systems and approaches and their
implementation rather than specific testing methods for the various asset types. Few if any mention
the use of technology to assist in the assurance process, either at asset creation or later; for example
in the use of designed and built in diagnostic facilities.
A concern of several companies was the reliance on contractors and the supply chain to carry out
some of the quality control and assurance, particularly on infrastructure assets. In this regard a
number of companies presented early steps on their delivery partner selection and procurement
processes to ensure quality management processes were aligned, and expectations of acceptable
quality standards understood from the outset.

Significant effort in the assurance of asset performance is in place, the extent and manner in which
this has been described is influenced by the size of the company and the extent to which they
therefore rely on detailed and well documented systems.
Responses have covered not only new assets but also the processes used during subsequent periods
of an assets life; focus has not necessarily been on what can be done differently when creating an
asset to improve the chances of economically attaining its design life.
CH2M 54
There has been no mention of approaches used when rapidly changing technologies are employed
or short life assets are installed as part of a solution. These areas will however be the subject of
lessons learned as the numerous approaches described for follow up understanding of asset
performance are implemented.
Significant reliance is placed on the delivery partner / contractor / supplier in the partnership
approach to delivery and there is not always evidence that there is an independent view and
assurance when new assets (in particular, complex ones) are put into service. This will place greater
emphasis on the terms and conditions supporting the delivery of the specific scheme or service and
the risks pertaining to these will be greatly influenced by the maturity of the relationship and the
experience of it with the asset type in question. It is likely to be more of an issue with complex non-
infrastructure projects and a common theme was the dynamic and iterative process required for the
creation and delivery of new assets if the desired performance and quality is to be achieved.
Responses on the general assurance processes
and the success factors for meeting asset health
goals also referred to the benefits of mature and
stable supply chain relationships.
There was evidence of good practice in many of
Significant reliance is placed on
the company responses. By way of example,
the delivery partner /
Thames Water refers to procedures in place with
their Eight2O Alliance that provide details and contractor / supplier in the
assurance of every joint completed and tracking partnership approach to
of quarantined failures. There will be a cost delivery and there is not always
associated with implementing such procedures evident that there is an
and this will be more easily born by the larger independent view and
projects, contracts and alliances. Smaller assurance when new assets (in
companies with smaller programmes may be at a particular, complex ones) are
disadvantage in this regard, a view supported in
put into service.
to an extent by the observations by some smaller
companies that they are reliant to an extent on
the control and assurance processes of the
contractor.
CH2M 55
Incentives and barriers
7.1 Context
This section of the report focused on examining two issues:
 Whether the twin drivers of ‘TOTEX’ and ‘outcomes’ are impacting on underlying issues of
asset health. We hypothesized that an over-focus on the customer experience (as measured
through the outcome delivery incentive, ODI) and a more balanced take on operational and
capital solutions (encouraged by the TOTEX approach) can lead to a situation where short
term service is maintained at the same time as underlying asset health is deteriorating. Such
a situation would be unsustainable in the long term, as ultimately asset failures would
overwhelm the ability of the utility to respond in a timely manner and maintain service. This
raises issues of intergenerational equity and whether utilities are properly balancing the
needs of current and future customers
 Whether innovation to support good asset health is hampered by structural or behavioural
issues. It is hypothesised that regulation and poor return on investment disincentives the
uptake of technologies and good practices which would support stable and improving asset
health
And we asked how the TOTEX and outcome based approach affected underlying assets health, and
how the health and resilience of new assets is valued when making investment choices. This allowed
us to examine whether current management practices support or undermine long term asset health.
We also specifically asked about barriers to innovative approaches, and what Ofwat could do to
overcome these barriers.
7.2 What were we expecting to be told

We expected to be told that a focus on customer service, the right choices around the balance
between operational and capital interventions and risk-based investment planning do adequately
manage asset health. We expected to hear underlying concerns that current approaches are not
encouraging sufficient infrastructure renewal to prevent the ultimate presence of 1000 year old
sewers8. We expected to hear that eventually an increase in investment levels would be necessary
to address an ageing infrastructure asset base.
On innovation, we expected to told about the piloting of various technologies, but hear concerns
over the cost-benefit of making them business-as-usual approaches.
7.3 What were we told?

7.3.1 TOTEX and outcomes
Our findings in relation to TOTEX and outcomes can be summarised as follows:
 Feedback suggests a general confidence that the TOTEX and outcomes management
approach does not impact adversely on underlying asset health. Companies say they know
this because they see and predict stable performance as experienced by customers
 However, and simultaneously, it was acknowledged that there is a focus on providing service
to current customers and their stated priorities and preferences are for the short term. It
8 Generally, serviceability (the de facto current asset health measures) are stable at current levels of investment. However, the renewal
rates are cited as being indicative of asset lives of many hundreds of years which seems incongruous. The exact reason for this is not
known, but it is argued that the impact is depletion of asset health.
CH2M 56
was recognised that this could lead to a deteriorating asset base in the longer term and
more volatile asset capability. There was no declared evidence that this was currently the
case and it was accepted that a properly designed planning approach should incorporate
long term aspects
 Some companies highlighted issues around achieving performance commitments year-by-
year or at the end of an AMP being a disincentive for long term investment planning. They
cited a focus on near term rather than long term performance and the tendency for benefit-
cost methods to promote ‘patching’ repairs rather than wider renewal. For example, South
West Water is concerned that the focus on service outcomes with customers tends to
incentivise ‘just-in-time’ reactive responses rather than encouraging proactive investment in
the asset base; they and others (e.g. Southern Water) are concerned that under investment
in asset health could result which would increase the financial burden on future generations
as assets deteriorate with time. This contention may merit further debate – the outcomes
approach is intended to encourage a long term and sustainable approach
 TOTEX and outcomes management based on whole life cost analysis has improved the
uptake of preventative maintenance programs and public campaigns to address sewer
misuse and fats, oils and grease (FOG)
 Good, long duration trending data on asset health is an important asset in itself and
essential to monitor the impact of TOTEX / outcome management approaches
 A minority of companies recognised the theoretical risk of a TOTEX driven program
becoming overly OPEX biased with ultimate consequence for the statutory accounts and a
concern about ratings agencies’ response affecting their access to capital
 We did not get elaborate answers regarding how asset health and resilience are valued
other than that they are not given an intrinsic value. They are valued in terms of their costs
and benefits with respect to service
7.3.2 Innovation barriers

Our findings around innovation barriers can be summarised as:
 Large WASCs generally saw few fundamental barriers towards implementing innovative
methods to manage asset health. Many were piloting new technologies but conversion of
trials to ‘business-as-usual’ was mixed
 Some companies cited that return on investment for innovation was poor and that
incentives were lacking, especially when the benefits of efficiencies were returned to
customers rather than shareholders. Reference was made to innovation incentive schemes
in energy sector as something that Ofwat could investigate
 Longer business planning cycles could give greater certainty over investing for the future
through more innovative methods. In period ODIs were said to discourage innovation
because of risk of penalties. They were said to encourage ‘fast followers’ rather than
genuine innovators because of risk aversion. End of AMP ODIs were thought to give more
time for innovations to bed in or for an alternative strategy to be applied. Some considered
that five-year regulatory business cycles were not sufficiently encouraging of innovation
 Some small companies cited an absence of in-house innovation resources (on affordability
grounds) and decried a recent drop-off in industry knowledge sharing and collaboration to
spread good practice
 Many companies indicated their preference for a common asset health / resilience measure
(or measures) to focus attention on this issue and act as a driver for innovation. However,
they disliked proposals that make absolute comparisons between companies disregarding
different legacies and circumstances
CH2M 57
 The maturity of asset data systems and analytics varies across the companies and those
lagging behind saw the imperative of enhancing their approach as a precursor to being
innovative in this area
 Balancing short term and long term performance is a challenge with the former being
prioritised at the moment. Companies have reported that they would welcome support from
Ofwat to guide how they balance the needs of current and future customers, and that they
would like to see greater engagement from Ofwat in industry initiatives like Water UK’s ‘21st
Century Drainage’ and UKWIR’s ‘long-term investment in infrastructure’ would be welcomed
 Drinking water quality regulations were a perceived barrier to innovative invasive inspection
methods, even those proven internationally. However, we (CH2M) note that these
regulations safeguard the public and have been effective in preventing potentially unsuitable
products being used in England and Wales
We can see that there is a possible inconsistency of view as to how a TOTEX and outcomes approach
would impact asset health and innovation. We believe that Ofwat’s objective is to encourage both
long term focus and short-term outperformance, incentivised by appropriate ODIs. However, some
have asserted that these instruments could, in effect, create a more short-term outlook.
We have no detailed evidence of this (though the significance of inferred asset lives of certain long
life infrastructure assets is intriguing) and were not able to explore these issues in more detail. If
these concerns are material, it is important that the long-term health and resilience of our
infrastructure is not eroded by short-term imperatives and this merits ongoing discussion and
consideration.
CH2M 58
Lessons from international practice
8.1 Introduction
We undertook a horizon scanning exercise to find out what water and sewerage utilities in other
countries are doing differently that the UK water industry might be able to learn from.
To do this, we contacted our asset management, conveyance and maintenance experts who
collectively enabled us to get a view of the approach to ‘managing asset health’ from Australia and
New Zealand, Singapore, Canada and the US. We asked our colleagues to consider the following
questions:
 Is asset health a concept that is used by your water clients and if so how is it measured and
used? If not, what is used to ensure assets are in suitable condition and will perform
adequately?
 What tools and technologies are typically used for measuring and managing asset health as a
matter of routine? (inspection, monitoring and control)
 What state of the art/emerging tools and technologies are you aware of that are being
evaluated by water companies and where?
.
8.2 Understanding of asset health and its relationship with

service
Although there were variations in the regulatory and institutional arrangements between the
different countries we looked at, we found the water industry’s understanding of asset health and
its linkages to service to be broadly similar, with similar areas of uncertainty. There is no evidence
that water and sewerage companies in other countries are significantly different to the UK in how
they understand and use asset health as a factor in investment planning. Other water utility
communities do not appear to be using the phrase ‘asset health’ explicitly as a driver of investment.
For example, in the context of infrastructure assets, whilst the UK companies talk about asset health
and service, in the US some companies talk about the consequence of failure (COF), and likelihood of
failure (LOF), which are combined to quantify risk. Evaluation of the LOF for a pipeline includes
consideration of performance, condition, maintenance status, and resiliency. Evaluation of COF for a
pipeline, includes consideration of service impact, in addition to environmental consequences,
financial costs, and social impacts. The resultant risk rating for a pipe asset that accounts for these
categories is used to identify deficient pipes and plan improvements.
LOF is informed by:

 Inspection data (condition)
 Hydraulic modelling (performance)
 Inspection data and work order history analysis (maintenance status)
 Desktop analysis (resiliency)
COF is informed by:

 Hydraulic modelling (service impact)
 GIS, and ecological penalty records (environmental consequences)
 GIS, and cost estimating for emergency repairs (financial costs)
 GIS with attention to critical facilities, roads, and cultural resources (social impacts)
CH2M 59
This is a risk assessment and is not so different to what we would expect to be being done by UK
companies
In Canada, condition and performance is used to inform the Likelihood of Asset Failure; where
condition typically refers to the physical condition, and performance typically incorporates the
following factors:
 Functional Condition: Are there design related deficiencies? Is the component deficient
against current policies and standards?
 Capacity Condition: Are there sizing related deficiencies?
 Efficiency Performance: Are there efficiency related deficiencies?
 Safety Performance: Are there any safety hazards or other deficiencies?
Our US colleagues provided the following perspective on non-infrastructure assets:
“For vertical assets we use risk. They key parameter we look for in asset health is
condition. We develop asset type (Pump, motor, MCC, Tank, Valve) criteria to determine
condition. Once the condition is determined, we determine useful remaining life (RUL). This
is key for the CIP (capital investment planning) process. Remaining life is done by
subtracting a percentage of Useful Life based on condition. i.e. Condition 1 (the best
condition) we would say 95% of the original useful life is remaining. Condition 2 gets about
80% of original life. 3=50%, 4=25%, 5= 0 to 5%. Asset Health to me is performance and RUL”.
The following condition monitoring tools are used for monitoring MEICA assets, to manage health
and avoid failure (ref: P –F curve, see section 3.5):
 Vibration Analysis
 Oil Analysis
 Ultrasonic Sound Signature
 Voltage and amperage balance
 Insulation resistance
 Motor circuit analysis
 Dissolved gas analysis
We were referred to some state of the art usage of RFID (radio frequency identification devices) that
measure temperature and vibration and report remotely in real time and also the use of these
devices in a way that that measures each machines performance and state in a “train” (upstream
valve, first pump, mid-stream valve, second pump, conveyor, etc..) so that the train is “coordinated”.
Defect elimination as a maintenance practice was also highlighted, but unfortunately we have not
been able to follow up on these techniques and cannot comment further on their potential value.
8.3 ‘State of the Assets’ reporting

One thing which stood out as a potential example for the UK was the “State of the Assets” Report
(SoAR) produced by some of the more mature water companies in Australia and New Zealand. These
come in various forms, but tend to provide a high-level overview of key asset metrics like:
 Average remaining life (% of design life)

 Average condition rating by asset type
 Renewal funding vs demand
 Proportion of PM vs CM
CH2M 60
 A number of reliability metrics (actual vs published MTBF, operational hours vs duty hours,
etc.)
Some report water quality data in the SoARs as well, although the link to asset condition and
performance is more tenuous for these. SoARs are underpinned by regular Asset Condition
Assessments and the mandated regular Asset Re-valuations. SCADA historian data and other stored
remote sensing data are also interrogated to populate the metrics. The take-up of SoARs in one form
or another is increasing across Australia and New Zealand, generally driven by company Boards’
requiring greater understanding of how the operational side of the business is tracking. At the
moment, these documents are generally discrete and tend to provide more detail on the asset-
related areas reported in the Annual Plan. The documents feed into the budget allocation process to
varying degrees: sometimes through the AMP process, where tracked metrics might support funding
need and other times as a reference at budget estimates time.
8.4 Inspection and monitoring technologies

There do not appear to be any major differences in terms of the inspection technologies commonly
in use for either infrastructure or non-infrastructure assets.
A useful summary of technologies and their application potential was provided by our colleagues in
the US. This is a list of techniques used in the US and we note that there are no stand-out examples
that are directly applicable that have not been used or considered in the UK (Figure 8-2 & 8-3).
The most commonly used monitoring tool is CCTV, which could be down to its relatively low unit
cost as much as the quality or value of the data it provides. Across the countries we looked at, there
are a number of examples of innovative emerging technology being piloted to good effect; however,
as in the UK, progress is restricted by individual utilities’ budgets and consideration of the practical
limitations of the technology.
US Environmental Protection Agency report EPA/600/R-10/101 (August 2010) discusses technologies

that can be used to gather more comprehensive data (than CCTV) on below ground infrastructure.
The report notes that the use of laser, sonar, and electrical scanning to evaluate such features is
established in the US, and that emerging technologies such as impact echo, spectral wave analysis,
and ultrasonic testing are also being explored for application to sewer condition assessment.
In Australia and New Zealand, drones are increasingly being used for asset inspection and review
purposes. For example Melbourne Water has trialed the use of drones for inspecting large assets
such as reservoirs and treatment plants. The trials found that drone inspections provided higher
quality, more accurate and usable data, as well improving efficiency and reducing costs. In addition,
all data recorded by drones are spatially tagged so they can be incorporated readily into a utility’s
existing GIS. It is worth noting, however, that Melbourne Water found that clouds and gusty wind
caused problems with the quality of certain types of imaging, which could be an issue in the UK,
where drone usage is gradually on the increase (e.g. Welsh Water commenced a pilot using drones
for asset inspection in 2016).
In the US, the number of non-destructive condition assessment technologies that use
electromagnetic or acoustic fields to measure the level of asset deterioration grows each year. Much
of this technology comes from the oil and gas market and is slowly infiltrating in to the water
business. Alongside the development of devices that generate and read these fields, significant
progress has been made in the development of software algorithms that can provide meaningful
interpretation of this data. Some UK water companies have already made use of this type of
technology: in fact, Thames Water won a Pipeline Industries Guild award in 2013 for its pilot of JD7’s
CH2M 61
internal ultrasonic condition assessment tool PipeScan+ in partnership with JD7 and Morrison
Utilities. UK companies including United Utilities, Anglian Water, Thames Water, Severn Trent Water
and South West Water are already investing in research and development in this area.
In Canada and the US, one of the emerging technologies is Pure Technologies’ SmartBall for leakage
detection, which has been used by utilities including the City of Ottowa. SmartBall is a free-
swimming, in-line, leak detection technology designed to operate in live large diameter water mains.
This highly sensitive acoustic sensor can conduct long surveys on a single deployment, collecting
data about leaks and air pockets, with no disruption to service.
Figure 8-1: Pure Technologies' SmartBall
8.5 Big data

There is increasing focus in the US on the collection and synthesis of “big data” that is made
available through the growing number of smart monitoring devices in use today (automated meters,
acoustic monitors, etc.). Analytics, machine learning, cloud computing, mobile solutions, robotics,
sensors are some of the current hot topics, where technological advancements are opening up new
opportunities for utilities.
This wealth of data presents new opportunities to optimise asset management. There are similar
opportunities in the UK, for example as smart metering grows in popularity and level, pressure and
transient monitoring increases. The move towards ‘intelligent assets’ is already being paralleled in
the UK. In the case of networks were have heard about near real time monitoring of pressure, flows
and levels and pressure transients that inform asset leakage and burst potential. We heard less
about big data and its analysis with respect to non-infrastructure assets so perhaps there is learning
potential from elsewhere and it would be interesting to look at the current state of the art in other
sectors e.g. oil and gas, where there is potentially greater risk and cost associated with asset failure
and system downtime.
Making use of data and smart networks is an area for development and potentially the focus for
considerable research and opportunity.
However, it is important to appreciate that as our networks and assets become smarter and we rely
more and more on remote sensing and control, the need for cyber-security and resilience will
increase.
CH2M 62
8.6 Deterioration and failure forecasting
We found little in the way of advanced techniques for deterioration and failure forecasting that are
not already used or known to our UK water companies. The most typical approach found was to
base forecasting on historical failure data, much as the UK companies have historically done. This
approach has limitations though, in particular for low probability, high consequence assets. It is likely
that in the water industry at least, the UK is leading the field in this area. For example, the UKWIR
report ‘Deterioration Rates of Long-life, Low Probability of Failure Assets’ considers long life, low
probability / high consequence assets in detail, developing deterioration models and a toolbox
covering key asset groups and failure modes.
8.7 Encouraging innovation

No particular beacons of innovation stood out as an example for UK companies to follow, although
there were some interesting comparisons and opportunities to share learning. The US appears to be
in a similar position in terms of trialing pipeline inspection technologies.
The UK companies sometimes referred to the twin factors of cost and risk as barriers to innovation.
Some of the UK companies also mentioned competition as a barrier to cooperation and knowledge
sharing. This is perhaps more perceived than real, however, since the UK companies have a history
of successful collaborative research through UKWIR. In any case, competition should not be a barrier
to working together for the common interest to understand the basics of asset health.
By way of comparison, the US Water Environment Research Foundation (WERF) synthesis report
‘Condition Assessment for Wastewater Pipelines’ notes that in the US:
“Information about condition assessment, renewal engineering, and underground infrastructure
locating technologies; models and tools; best appropriate practices; experience with technologies
(both positive and negative); cost of technologies and available vendors and contractors exists
within the institutional knowledge of individual utilities, but is often not readily shared or accessible
from outside the utility.”
This presents a very similar picture to what we heard from some in the UK.
However, the UK water industry offers a great opportunity to encourage innovation. We have large
utilities, whose private status enables them to make independent decisions about risk; companies
have mature alliance and supply chain relationships; it is a relatively close-knit industry facing
common challenges and an established path for single voice R&D (UKWIR). These factors lay the
foundations for a collaborative approach to innovation which would bring down the cost and risk of
research and development for individual companies.
CH2M 63
SECTION 8
Figure 8-2: Overview of pressure pipe assessment tools used in the US (internal source)
Gravity Pipe Assessment Tools
Multiple Sensor Data Pressure Testing Infiltration Detection by Spark Testing

Flow Monitoring Smoke Testing Dye Testing Isolation Closed Circuit Television (CCTV) Laser Profiling Sonar Profiling Manhole Inspections Lateral Inspections Vacuum Testing Manholes
Colletion Joints Electroscan (ASTM F2550) Manholes
Diameters 6 inches and greater 4 inches and greater 4 inches and greater Any 4 to 120 inches 18 to 144 inches 18 to 144 inches 18 to 144 inches Any 4 to 6 inches 6 to 42 inches 6 to 60 inches Any Any
Done during low flow times (for example

Inject non-toxic smoke into an isolated area of
midnight) measure flow rate at the most
Sensors installed in key manholes to sewer system by large blower defects should be Inject dyed water to assist in Digital Cameras have the abiltiy to Can be used to assess pipes Inspection can be preformed using Electrical currents can be
downstream location, the flow at this time is Can provide CCTV, Laser, Requires specialized CCTV Integrity of Manhole
Description measure depth of flow or depth of photographed and logged (note certain verifying leaks, illegal cross quantify defects, pan and tilt on Above water inspections that have a continuous high pole camera technology and can Test joints for I&I used to determine leakage Potential I&I
generally attributed to infiltration, will need to Sonar at one time equipment through high voltage
flow and velocity conditions this type of testing will not help such connections camera can focus on defects flow provided photos of defects potential
work way upstream until infitration area
as a high water table)
located
Output can be used to determine

Can calculate volumes and
wet weather I&I flows, diurnal flows Can assist in identifying Can identify cracks (longitudinal and Can identify pipe wall loss Can assist in the
Can indirectly identify inflow points, uncapped levels of debris in pipe and Structural integritys, historic sucharge, Can indicate where joints
(dry weather), can be used to assist where flow goes (in illegal Can identify areas of high flow rates of circumferential) fractures, offsets, (loss can happen through prioritization of Will provided
or broken cleanouts, potential illegal determine potential pipe CCTV, Laser, and Sonar I&I, roots, stains, active water and Similar to CCTV of mainlines, and other portions of pipe Can indicate which manhole is not
Output in projecting capacity requirments, connections, in it's own infiltration, groundwater in areas of significant misalignments, roots, broken pipe, corrosion) and can compare rehabilitation and information on lining
connections, manhole covers that don't fit well, capacity issues and and output could also assist in identifying determines defects of lateral that is electronically completely sealed
master planning and could identify system, etc.) can also assist in flow poor lateral connections, protruding the pipe defleciton versus the extent of joint integrity
broken sewer lines identify major structural blockages connected to ground
issues with upstream pipe in a mass finding voids taps, vermin, etc. design leakage
issues
scale
CCTV operators should be trained Manhole inspector should be trained Operators should be trained and Generally required on new
Temporary flow stoppage
and certified using an established Could also measure the and certified using an established certified using an established Done generally manholes, if test needs to be done Man entry, and
Technology is improving and can required (could require
Comments Should notify public in advance of testing standard such as NASSCO's pipeline hydrogen sulfide levels, standard such as NASSCO's manhole standard such as NASSCO's lateral with Chemical on existing manhole flow stoppage generally flow
provide more reliable data bypass) need to fill the test
assessment and certification temp and other things assessment and certification program assessment and certification Grouting would be required (could require stopped
section with water for test
program (PACP) (MACP) program (LACP) bypass)
Figure 8-3: Overview of gravity pipe assessment tools used in the US (internal source)
CH2M 64
Discussion and conclusions
There were a number of key areas pertaining to asset health that we have explored through our
interviews with the companies. In our discussion and conclusions below we have drawn on the
responses provided by the companies and our wider knowledge of the sector.
9.1 Understanding asset health

9.1.1 What does asset health mean?
The responses to our questionnaire provided valuable insight into prevailing concepts of asset health
being used in the water companies of England and Wales. These responses helped us understand in
what ways companies are working to develop their measurement and management of asset health.
We heard that some are focussing on the specific asset groups (e.g. waste-water systems) that they
consider need to be understood better, whilst others are focussing on those assets that, if they
failed, would impact significantly on the risk of service failure.
What we were told about was more nuanced than what we expected to be told, for example:
1. although many companies infer the health of their assets from serviceability and reliability
indicators, these so called ‘lagging’ indicators (which do not focus explicitly on the condition
of assets) can make it difficult for companies to make proactive asset-management decisions
2. some companies are therefore looking at developing ‘forward-looking’ indicators that are
more focussed on the actual condition of assets, and also take into account the ‘criticality’ of
their contribution to the resilience of water and waste-water services within which they
function, in order to make decisions that are more ‘forward-looking’
However, there is broad agreement across companies that ‘asset health’ is a much richer concept
than ‘asset condition’ in that it needs to consider not only the physical state of the asset, but also
the performance, role and importance of the asset in ensuring that service performance targets and
customer expectations can be met.
We could not distil a definitive definition of asset health out of this consultation and review; there is
currently debate and diversity of view. Although there is no consistent working definition of asset
health across all companies, most companies associate good asset health with achieving good levels
of service and observed performance that compares well with expected performance. This general
view held by most companies is that good, stable, measured performance is a strong indicator of
good asset health and conversely, that inadequate or declining performance may indicate poor asset
health.
9.1.2 Asset health and resilience

Asset health is difficult to define, but it is recognised that asset health is part of the wider concept of
resilience. Resilience is currently defined (Ofwat’s Resilience Task and Finish Group: Final Report,
December 2015) in the UK water sector as:
…the ability to cope with, and recover from, disruption, and anticipate trends and variability
in order to maintain services for people and protect the natural environment now and in the
future
CH2M 65
The characteristics of assets and asset systems that determine resilience have been defined by the
Cabinet Office (Keeping the Country Running, 2011) as:
 Reliability
 Resistance
 Redundancy
 Response and Recovery
We can potentially make a distinction between asset health (a function of asset reliability and
resistance?) and the other elements of resilience (which also includes system configuration,
redundancy and response/recovery dimensions).
It may be helpful to consider that the purpose of the asset base is to deliver the desired/required
level of service to customers and the environment. This will be determined by consideration of:
1. The state of the assets
2. The way they are configured
3. The way they are operated
In combination, these factors will determine the resilience of services and we can consider issues
around the state of the assets to be issues of ‘asset health’. These three parameters can be
optimised for specific situations and this requires an appreciation of their relative costs and benefits
and synergies.
Several company responses indicate a need for greater clarity in understanding exactly how asset
health influences asset resilience and systems resilience.
9.1.3 Benefits of the current measures

There is currently widespread use of the original (and variants of) Ofwat serviceability indicators as a
surrogate or proxy indicator of asset health. Most companies are satisfied with the use of these
serviceability measures as headline indicators of asset health and as a MoS. In addition, these are
variants of long established parameters and companies continue to collect data in a consistent way,
there are certain benefits of retaining these measures:
“The greatest benefit of the current measures of asset health is that they have been established for
some time, providing a useful dataset for trend analysis of asset health and rates of deterioration”.
9.1.4 What would a good measure look like?

Answering this question requires some clarification of the definition of asset health. We could
assume that asset health is a measure of asset failure probability (rather than service failure
probability) associated with the state of the asset and typical operating conditions. In this case, the
significance of asset health can be influenced by a range of factors that may mitigate the effect of
the asset failure on service e.g. operational response, redundancy.
Asset health is therefore likely to be very important in some instances and less so in others, so it
essential to understand asset failure modes (how an asset can fail) and effects (the impacts) so that
investment in asset health measurement and monitoring, and development of appropriate
indicators, can be determined.
In terms of developing a good measure, we can speculate that improvements would be more
forward looking, predictive measures. Emerging ideas for effective asset health measures in the
water sector include combining data e.g: on the physical condition of assets; the environmental and
weather conditions to which the asset is exposed and the operating conditions within which the
CH2M 66
asset is expected to function. This would potentially be more balanced and predictive. These data
have a focus on individual assets or sets of assets (thus providing a ‘bottom-up’ view of asset health)
and can be combined with information on their criticality, the resilience of the network or system
and the overall performance of these systems.
By applying data analytics, statistical analysis and visualisation with GIS outputs and ‘dashboard’
displays, companies are also exploring ways of integrating ‘top-down’ and ‘bottom-up’ information
in order better to infer how healthy their assets are. Some are combining these strands of
information into indicators that are then used (via a form of scorecard) to create an index of asset
health.
We also heard that some companies are concerned that using serviceability derived measures has
conflated the issues of asset health and their performance in providing a good service to customers.
In this context, asset health is more of an asset centric concept e.g. the physical state of the asset,
and these companies feel that asset focussed indicators may need more prominence.
It is possible to maintain an adequate level of service to customers even though individual assets are
in poor health (a poor physical state) when redundancy and spare capacity within networks and
systems and operational responsiveness can compensate for inadequate asset health. As a result,
companies are looking closely at resilience and the role of asset health in this context.
It is also recognised that asset performance is a lag indicator – the failure event has occurred.
Understanding the state of the asset before failure can be used as a lead indicator, which may help
to avoid significant repair costs and undesirable consequences if the asset is service critical.
Decisions to gather such data need to take account the potential costs and benefits of doing so.
In response to this concern, some companies are developing and testing new definitions for asset
health. These developments are initially internal to the company and are intended to shadow
serviceability and reliability measures during PR19. It is too early to say if these will be significantly
better that the current measures, but we can say that more forward looking measures (lead
indicators) should be of additional value. However, we also note that companies are already using
deterioration model outputs and trend data as part of their business planning approach.
In conclusion, the term ‘asset health’ has yet to be defined fully and agreed within the UK water
sector. There is probably not a single measure that works well for all assets in all circumstances.
There is potential merit is looking to develop improved lead indicators and a basket of indicators,
which can be tailored for different assets and different levels of criticality of the assets.
9.2 Measuring and monitoring the health of our assets

We wanted to find out what ‘detailed level’ data and indicators were used to understand asset
health and inform the headline serviceability type measures. We also wanted to understand if good
use was being made of technology.
9.2.1 Data
Water companies collect and store and analyse data that supports their understanding of asset
health and service. This can be the same data that is used for reporting on their MoS, but resolved to
a greater level of resolution and it can be additional data about the condition of the assets, their
performance, location and operating environment.
Companies generally felt that the information they already collect is sufficient for managing risk to
service, although there was some acknowledgement that there are some uncertainties in data
records, which can include information on location of buried assets.
We agree that on balance that the companies appear to have invested in improving data (confidence
grades are typically assigned to asset register data and efficient processes have been developed for
CH2M 67
updating records) and recognise the benefits of good data; however, ongoing efforts need to be
made, in particular, with respect to the critical, high consequence of failure assets.
Infrastructure asset data
Typically, bursts, leaks, water quality, collapses and blockages are recorded and resolved to
individual pipes. GIS is used extensively to visualise the data.
Flow, pressure and surge are being recorded more extensively and in real time, to help understand
network dynamics and potential risk.
Condition, based on CCTV inspection, is used to support understanding of the health of sewer pipes.
Non-infrastructure asset data
Maintenance activity (particularly reactive and repeat events), is used to help understand asset risks.
Monitoring of asset health using, for example, vibration, acoustic and thermographic sensors
appears to be relatively well established, albeit with some mixed feelings regarding effectiveness.
We believe that as systems improve for recording maintenance event data that it will be possible to
develop more effective and optimised maintenance plans for these assets.
Asset condition data
In the past, visual condition grading was used to inform capital maintenance need for non-
infrastructure assets and this has been challenged on the grounds that this is not a risk based
measure. Therefore, if condition data is used, there needs to be evidence that it is a good indicator
of asset failure probability and service impact.
Due to the inappropriate use of some condition information in the past (primarily visual assessment
data) and the desire to move to a risk-based approach, companies are no longer collecting asset
condition data for all assets on a systematic basis.
We agree that condition data is not necessarily a good indicator of asset performance or failure
probability, but it is of value where it can be shown that it does correlate well and is the only viable
indicator (this point is made in the context of being able to link condition to service risk).
9.2.2 Technology
In general, the water companies told us that they support the use of new technologies for
inspection, monitoring and control of their assets. There have been several research projects in the
past that have sought to appraise and develop technologies (particularly for pipeline inspection) and
organisations such as UKWIR and WRc have previously undertaken work on behalf of the sector.
There are examples of advanced technologies being applied, but companies were quick to point out
that the costs and benefits are a key consideration.
There is considerable interest in technology for collecting and processing data and the topics of big
data and analytics were raised frequently. Some companies are progressing rapidly in exploration of
analysis and visualisation tools – an important enabler being the ability to mine data from their asset
information systems.
We were a little concerned to hear that some considered the 5 year planning cycle to be a potential
barrier to innovation, in the belief that it meant that investment returns need to be recouped in the
5 year period and this does not encourage taking a risk on new technology. This viewpoint requires
further understanding, consideration and resolution.
Note: it is important that all technology and innovation, whether ‘hard’ or ‘soft’ needs to be
evaluated and appraised in terms of the costs and benefits that it provides and that we do not
advocate any specific solutions.
CH2M 68
9.3 Asset health and decision making
The purpose of the questions on decision making was to better understand the how asset health is
used to influence and support expenditure decision making.
9.3.1 Linking asset health and service

All companies appreciate the linkages between asset health, service and expenditure and recognise
that in certain cases they can be difficult to quantify due to the effect of factors such as the
environment and non-asset health related resilience.
Companies develop a framework of indicators and measures that enables forecasting of outcomes
and estimation of the costs and benefits of expenditure. These frameworks typically include a set of
service measures that can be monetized (typically based on customer research) and hence used to
appraise the costs and benefits of expenditure. The estimation of probability and consequence of
failure is used to determine risk to service.
All the water companies use a framework such as this and nearly all use specialist investment
planning tools to predict asset performance, service and the effects of investment on risk. These
tools are used to estimate the costs and benefits of expenditure and to prioritise and optimise the
plan based on risk reduction and value.
From the review activities, it was evident that some of the strongest causal links between asset
health and service can be made where asset failure data is plentiful, such as for water distribution
infrastructure bursts and leaks and the impact on interruptions. In these instances, the cause and
effects are relatively well researched and understood and most companies described ‘mature’ and
continually improving infrastructure deterioration models where they are able to validate forecasts
by comparing observed and predicted data.
There was general agreement that creating the link between asset health and service in non-
infrastructure investment models was more difficult, because these asset systems usually include
standby and redundancy, which makes the link between asset failure and service more difficult to
quantify. Companies described tools and techniques like FMEA (failure modes and effects analysis)
and RCA (root cause analysis) being used to support the modelling activities.
All companies recognised the need to better understand the role of the asset ‘system’ in delivering
service. Companies recognised the benefit of understanding and planning for risk at a system level.
As mentioned, at least one company is actively engaged in a programme of work to quantify system
resilience using models that can simulate conveyance, asset reliability, system control and resilience
hazards underpinned by a 4Rs type assessment.
All companies articulated the view that asset health and resilience are related; most see asset health
as a component of resilience. Service risk assessment needs to consider operational and system
resilience as well as asset health. One or two companies indicated that they are exploring the
possibility of identifying whether there is an economic level of resilience and whether this could be
quantified in a similar way to an economic level of service. One company that was developing
thinking along these lines identified the need to understand the relationship between a ‘healthy’
asset system and one which has built-in resilience. To do this they recognised the need to have good
granularity in direct and indirect cost of consequence models.
It is noted that if the system is inherently resilient (due, for example, to spare capacity, alternative
routing options in networks) the consequences associated with the failure of an unhealthy individual
asset may be mitigated.
In summary, the ability of companies to predict asset health impact on service will vary depending
on:
CH2M 69
 The precise measures of asset health being used;
 The level of resilience (ability to mitigate failure consequences through operational
or other means); and
 The availability of data to build the models.
The companies generally recognise that an increased knowledge of intrinsic asset health will help
ensure long term reliability and build linkages between asset health, deterioration and failure, based
on statistical analysis for below ground assets and risk-based discussions on above ground assets.
9.3.2 The tools for linking asset health, service and risk
There are a number of tools and approaches that companies use to link asset health, service and
risk. We have already described event tree, fault tree and failure mode and effects type tools for
mapping the linkages. There are other approaches that quantify relationships and target investment,
and a framework that provides an integrated structure for risk-based planning.
As mentioned previously, monetization of service measures is used to quantify the risk.
Asset performance scorecards and GIS
Virtually all the companies reported that they use scorecards/dashboards and GIS to monitor asset
performance and service to customers. As part of this approach, different perspectives are used
(strategic, tactical and operational) to help ensure appropriate investment to keep service risk at
appropriate levels.
It was very apparent that companies are continuing to develop dashboards to track asset health,
performance, service and outcome data to manage service delivery. These dashboards are typically
linked to corporate data systems and create important lines of sight through the business; some
companies described the role of asset health data in this way. We saw extensive use of asset health
to support planning in near and real-time, short term, medium term and longer term. All companies
described activities to improve, consolidate, and create consistent asset health data and maintain a
single version that could be served to corporate applications such as dashboards.
Deterioration models and consequence models
Deterioration models, observed data and expert elicitation are used to determine failure probability
and link this to likelihood of loss of service. Consequences of failure are estimated using hydraulic
models, flood models, observations and estimations and Service Measures and KPIs are used to
value service. Expenditure is optimised to balance risk, cost and service and derive maximum value
by using an ‘integrated asset management’ toolkit.
The Expenditure Planning Framework (the new Common Framework), asset health and resilience
The water companies in England and Wales use a risk-based planning framework to develop their
business plan.
We believe the Framework is a very positive tool and helps provide a solid foundation and common
language for asset management. We consider that the Framework is consistent with (and exceeds)
the expectations for risk-based planning as laid out in ISO 55000. Furthermore, it is not prescriptive
in terms of the tools and techniques that companies should adopt and it encourages companies to
have a forward-looking and integrated approach. The risk-based approach, in our opinion, provides a
basis for companies to evaluate resilience and asset health; understand how these factors influence
service outcomes to customers and the environment and provide the ground rules for developing a
compelling business case, based on an understanding of cost, risk and service now and in the future.
CH2M 70
9.3.3 Using asset health to inform expenditure
Linking health to service and service to expenditure is done using the service measure framework,
which has already been discussed. The relationships between asset health, performance and service
are typically established at a general level using expert understanding and statistical analysis of data.
For planning purposes, companies use optimisation tools to estimate the costs and benefits and
prioritise expenditure.
When the plan is being delivered, and at the local asset level, the actual performance may differ to
the predicted and when the companies target their investment it is typical to examine asset specific
data to inform the expenditure planning/response. To this end, for infrastructure assets, all of the
companies can use their GIS systems to resolve infrastructure failures and events to the pipe level,
cohort level and zonal level.
Increasingly, non-infrastructure data can be mined at the asset component level and this informs
failure investigation and understanding of typical asset life and can support the optimisation of the
maintenance regime. There could be major opportunities in this area.
Reference was made to the collection of post event information and a logging of not only the
characteristics of the failure, but also the associated consequence. There was a lot less emphasis on
quantifying the consequences associated with above ground asset failure.
In our experience, more work needs to be done to demonstrate the effectiveness of the expenditure
programme in terms of service measure improvements and risk reduction and this data should be
used to validate and calibrate the planning tool model assumptions.
Another important issue that arose was how best to optimise the balance of operational resilience
and asset health. The concept of an economic level of resilience was raised and we believe this is
about achieving the most cost beneficial balance of resistance, reliability, redundancy and
response/recovery, where the resistance and reliability dimensions define asset health and the
redundancy, response/recovery dimensions relate to operational and configurational resilience.
9.3.4 Criticality
The concept of ‘criticality’ was explored with all companies. From the discussion it was evident that
standard definitions are not being used across the industry, perhaps with the exception of
wastewater networks where terminology from WRc’s SRM documents provides definitions of critical
sewers.
Most talked about criticality in terms of essential assets that have a high consequence of failure.
All companies described the use of criticality in investment planning. These use included prioritising
monitoring and inspection activates, specifying equipment (e.g. in built vibration monitoring on the
most critical pumps), defining response plans and prioritising assets within investment models. Most
companies indicated that a measure of ‘criticality’ was commonly stored in asset databases and
systems against specific assets, e.g. at pipe level. The unit measures of criticality can vary by asset
group, which suggests that cross-asset group comparisons could be problematic.
One of our concerns in terms of the current ‘serviceability’ and performance dominated approach to
asset health is that of highly critical (high consequence) assets that have a low or non-existent failure
history. There may be critical, high consequence of failure assets, such as major trunk mains, where
companies are not fully aware of the state of these assets and hence the risk of failure. It was not so
evident from responses as to the level of confidence that companies have that failure probability of
the critical assets is fully understood.
CH2M 71
In these cases, it is important to know where these assets are, what the consequences of failure will
be and to consider the value of gathering additional asset health data, such as physical condition, to
help inform the asset management process and plans put in place for dealing with potential failures.
9.4 The views of the customer

We felt that it was important to understand if and how the companies were eliciting the value
placed on asset health by customers.
9.4.1 Is there an active dialogue with customers?

Based on the findings of our interviews under a third of water companies have undertaken explicit
dialogue on asset health or asset health issues with customers at this time.
9.4.2 Has a language for conveying asset health concepts been developed?
The water companies told us that customers have difficulty in understanding the concept of asset
health where they have consulted with customers, but only a small number (< 30%) of water
companies have undertaken explicit dialogue on asset health with customers. As a result, most
companies have not been able to develop an effective language for an asset health dialogue with
customers. Some companies, such as Yorkshire, South Staffordshire, Southern and Wessex Water
are more receptive to a dialogue with customers specifically on asset health and therefore we
believe that a dialogue can be had providing it is framed appropriately and uses the correct language
to convey messages. There are specific examples that illustrate customers wish to be consulted on
asset health issues.
Where companies have not undertaken explicit discussions on asset health, companies informed us
they were exploring alternative methods and more effective language to discuss asset health; others
retain the approaches used in PR14, which valued or prioritized investment options based on service
and performance.
In the absence of an explicit dialogue with customers, most companies told us that asset health
could be proxied using other measures. These could include a basket of different output or outcome
measures that are valued by customers or from an engineering perspective by looking at specific
types of asset failure metrics.
9.4.3 Do customers value asset health?

Our reviews did not show at this stage that any Company would use an asset health concept to
obtain an explicit valuation from customers as part of the business plan processes.
Methods for valuation (although they may not relate specifically to asset health) are very much
similar to the approach used for PR14 to elicit values for outputs or service levels.
Other methods such as simulation or gamification are being, or will be tested by some companies, to
enhance customer experience or understanding of asset health and/or resilience. Yorkshire Water
appeared to have good understanding of customer communication preferences and had harnessed
the use of social media.
To understand and strengthen the robustness of customer engagement findings for asset health and
resilience some companies said they would use (more) triangulation, which draws on findings from
multiple approaches and, or studies.
Specific examples provided by companies, such as the Butterley reservoir controversy illustrate that
some customers do want to be consulted and do value asset health.
CH2M 72
Ofwat has consulted on six draft asset health expectations for PR19, three of which relate to
customer engagement. Companies indicated that guidance for customer engagement on asset
health from Ofwat for PR19 would be welcomed.
Our analysis also found that there was:
 Limited mention of continuous dialogue;
 Some mention of big data more generally, but not specifically to customer services; and
 Some use of triangulation and gamification but there is scope to extend the use of these
methods.
9.5 Asset health assurance

The assurance framework under which companies operate underwent
significant change during AMP6, with Ofwat requiring companies take
responsibility for assurance that they provide across the whole
business.
9.5.1 Delivering asset health commitments

Universally, companies state that they consider they are
meeting their target measures for asset health. The
performance is generally substantiated with reference to
meeting serviceability and customer service target levels of
performance. Some companies referenced the case of weather
dependent failures, outside their control and others unexpected
failure of a single asset that had had significant impact on an
otherwise stable trend.
Companies told us that there were several reasons for this, including
improved risk based planning, better data, tools for targeting and developing historical knowledge of
the assets and their performance.
Generally, we concur with this view. It is also important that companies develop good
methodologies for exploring climate and weather impact on service (work is ongoing in this area and
models have been developed) and develop an appreciation of critical assets whose failure can have
profound impact on service levels.
9.5.2 Assurance processes

We expected to see that companies had implemented rigorous assurance at three levels; the
operational management level, in the internal risk management and compliance function of the
company, and with the internal audit function. However, given the significant variation in size,
circumstances and risks across the companies, we also expected a wide variation in assurance
provision, but expected there to be independent and specialist expertise in all companies.
We found that the deployment of an independent assurer in a reporter role for the assurance of
annual reporting was universal across all companies. A three line of defence model was only
described by a minority of companies.
We expected to hear that companies could identify systematically which assets were critical to
service delivery, that the health of these assets was monitored and the company had response
management plans to cover a range of failures of critical assets.
We heard that risk based approaches have been employed, and that criticality of assets had been
reviewed with modelling, and this gave the companies an understanding of those parts of the
CH2M 73
system that would leave the company vulnerable if assets were to fail. However, it was not clear
from the responses that the companies understood the level of confidence in the failure probability
of critical assets.
There is evidence that companies are continuing to improve their understanding of consequences
through, for example, critical links analysis and extending modelling coverage. In some cases, their
ability to do this is limited by the incomplete knowledge of their assets. This is more an issue with
below ground infrastructure assets than for non-infrastructure assets.
We heard that lessons learned analysis is common across most companies, but that the formal
assurance of lessons learned is less clear.
There is a shift towards the collation of asset health data in a single source, after this data has been
validated and assured. Information based on this data is then served back in a variety of formats for
different users.
Although we did not ask the question directly, one or two companies told us that their ISO 55000
certification/alignment played a role in improving assurance. ISO 55000 is a standard for asset
management good practice and it provides a check list of expectations against a framework (a
system) for asset management. The standard does not tell companies how to meet the standard and
does not prescribe a specific approach to asset health. What is does do is encourage consistent
approach and continuous improvement. We consider that companies that have adopted the
principles of ISO 55000 are likely to have an additional level of assurance and scrutiny in place and
we also saw that one of the companies we spoke to had enhanced their emergency response
process as a direct result of their ISO 55000 audits.
ISO 55000 will not guarantee healthy assets, but it is a systematic framework that helps to improve
transparency and accountability and should be of general benefit accordingly.
9.5.3 Asset failures

The companies all have emergency management plans in place to deal with major asset failures.
More routine failures are dealt with using business as usual reactive maintenance processes and
companies recognise that effective response plans are a key element for maintaining service levels.
The feedback of lessons learned, post incident reviews etc. are a key element for good asset
management and the improvement of long term asset performance; it is an area of requirement for
ISO 55000 certification. All companies responded that they have lessons learned processes in place,
though we were not able to test the effectiveness of these processes as part of this review.
9.5.4 Commissioning new assets

We heard that companies are approaching assurance of new assets from three directions: design
and technical standards of the assets, comprehensive commissioning plans and processes, and post
project reviews.
We believe from our discussions that all companies have these processes in place, but the detail
provided varied and testing the effectiveness of these assurance procedures was well beyond the
scope of this study.
There is reliance on the supply chain in many cases and, as such, this may merit further scrutiny
along with other elements of the assurance process around the issue of asset health.
9.6 Incentives and barriers

We asked questions around current drivers (TOTEX and outcomes) and innovation to try and
understand how incentives and barriers may be influencing asset health.
CH2M 74
9.6.1 TOTEX and outcomes
We did not find a significant difference between what we were expecting to be told and what we
were told.
For most assets, for most of the time, companies are broadly happy that the current approach and
emphasis is maintaining adequate asset health to maintain serviceability. It was noted by some
companies that the focus on achieving annual performance commitments and outcomes at the end
of each AMP can focus investment on operational patches rather than long term performance.
Some companies expressed a view that this could lead to a deteriorating asset base in the longer
term and more volatile asset capability.
This type of behavior has been talked about in terms of a ‘bow wave’ of deteriorating asset health
(backlog of poor condition assets’), the impact of which is masked by operational expedients and
reliance of asset redundancy and which will eventually lead to a ‘cliff edge’ of failure.
However, there was no declared evidence that this was currently the case and it was accepted that a
properly designed planning approach should incorporate long term aspects. Furthermore, it is our
(CH2M’s) view that because assets age at significantly different rates, even if asset renewals are not
keeping pace with aging, the eventual rate of failure increase should not be a dramatic ‘cliff edge’.
Nonetheless, this merits ongoing investigation and monitoring.
It was widely accepted that good, long duration trending data on asset health is itself an important
asset and essential to monitor the impact of TOTEX / outcome management approaches; there are
no inherent barriers to stop companies from collecting this data. Because companies recognise the
importance of these data, most have been improving the capture, management and assurance of
the data.
Ongoing discussion is required to better understand the assertions from some that the outcomes
approach could inadvertently result in short-term behavior and ensure that long-term asset health
and resilience is encouraged, where it is economic to do so.
9.6.2 Innovation barriers

Many water companies are trialling and piloting innovative methods to capture, analyse and use
asset health data, so there seem to be few barriers to piloting innovation. However, the slow and
low rate of conversion of these technologies to business-as-usual provides some support for the
companies’ concerns that the 5 year planning framework and the low RoI (return on investment) on
innovation provide a barrier for effective implementation.
A number of companies (WOCs and WASCs) claim that a ‘league table’ approach is preventing
effective sharing of data and best practice and that sharing of data and communication between
companies is getting worse. The smaller companies claim this has a disproportionate impact on
them because of their already limited population of data on asset health.
Some WOCs were concerned that their size was a barrier to innovation, whilst others suggested that
it facilitated innovation.
Our thoughts on innovation are that the reported barriers can be surmounted and that the water
companies of England and Wales and the UK as a whole are in a good position to be world class
innovators for the sector on the measurement and management of asset health.
We have sound asset management tools and systems and have been taking a performance, service
and risk based approach to managing our assets and the service they deliver for longer than most.
Our privatised water sector is able to set its own risk appetite and we have been leaders in trialling
technologies in the past and adopting innovations from other sectors.
CH2M 75
We also have effective communications and a single voice R&D platform. Our networks are relatively
large and our supply chain management and partnerships sophisticated and efficient.
We conclude that it is important that there is ongoing UK water sector debate and sharing in the
efforts to develop thinking in the areas of asset health, resilience and technology.
9.7 Lessons from international practice

We drew on our international experience to see if there were any major leanings from outside the
UK.
Our conclusion is that whilst there may be no ‘game changing’ water sector beacons of innovation
for UK companies to emulate, there are some interesting opportunities to share learning.
We note the use of:
 ‘State of the Assets’ reporting in Australia and New Zealand, which is asset condition and
asset performance focussed and generally driven by company Boards’ requiring greater
understanding of how the operational side of the business is tracking
 ‘Big data’ and the innovation around information and data in general (smart systems,
sensors, control, analytics, robotics, cloud computing etc..) which appears to be influencing
change in a number of sectors
9.8 Summing up
We sum up this review by revisiting some of the key issues.
We have not established precisely what asset health is and the best measures of asset health. There
are views on measures ranging from asset specific parameters (e.g. condition, remaining life)
through to composite index type measures that include asset performance and service elements. It
is also recognised that asset health is integral to the wider debate on resilience and that asset health
needs to be considered as part of a bigger picture that incorporates operational factors, criticality,
consequence and risk.
The changing focus in the sector in terms of the evolution of asset management planning has seen
an evolution from condition, to performance, to risk and serviceability, to customer and
environmental outcomes based approach. We (CH2M) believe that asset health is most
appropriately viewed as a measure of asset condition and performance and that it is important to
understand the impact of asset health on potential risk and outcomes. This requires good
appreciation and understanding of how assets can fail (failure modes analysis) and the potential
consequences.
On balance, we believe that the water companies appreciate these issues and generally do gather
appropriate asset data based on consideration of risk and value of the data, though we think more
can be done and needs to be done to ensure that our services remain resilient in the future and that
critical assets are managed effectively.
There is debate on the perceived value of asset condition data and some argument that
performance and service data reduces the need for asset condition data, which can be difficult and
expensive to obtain, and, in the case of simple visual data, may not be a particularly good indicator
of asset failure probability.
However, we strongly believe that it is important to understanding asset condition and deterioration
mechanisms in the case of low failure probability, high consequence assets, in order to help to be
able to pre-empt failure and intervene in a timely fashion. In turn, this means it is important to know
which assets are critical (high failure consequences) and where they are. It is important to also have
CH2M 76
good contingency measures in place because there will always be unexpected asset failures and it is
important to be able to detect, respond and deal with these events in an effective and timely way.
It is also important to consider the wider resilience of the system and how this can help to mitigate
failure impacts and understand the most cost-beneficial mix of options for managing risks relating to
asset failure.
In the case of complex non-infrastructure assets, better understanding of asset lifecycle and
maintenance impacts could enable more efficient asset operation.
We heard that there could be situations where short-term drivers detract from long-term asset
health and resilience and some concerns regarding ‘bow waves’ and ‘cliff edges’ of failure. Whilst we
do not think that any ‘cliff edges’ are imminent, this does merit ongoing investigation.
With regard to assurance, companies have processes in place, but we were not able to look at this
subject in detail within the scope of this project.
Innovation is a key area of focus and some companies suggested that elements of the regulatory
process could discourage risk taking with regard to novel technologies and may have discouraged
communication. It is important to understand why these views are held, but do not think there are
any barriers, real or perceived, that cannot be overcome.
Whilst we found no evidence that the water companies in the UK are behind the curve in trialing and
technology, compared to water companies elsewhere, we believe that the UK has potential to be
leading innovators and should embrace benchmarking of sectors which have a more progressive
approach to innovation.
CH2M 77
Key Report What is asset health? Measuring and monitoring Making a business case for Understanding the views of Assuring our approach to Incentives and barriers
Theme (see Q 1 - 4) the health of our assets asset health (Q7 and 9) the customer (Q11) asset health and maintaining (Q10, Q15)
(Q3, 5, 6) service (Q8, 12, 13, 14)
Company
Affinity  Serviceability definition: Asset  Believe they have a good  Risk-management system  Not asked customers formally  ARM – operational risk capture  Suggested that there needs to
Water health being the ability of the understanding of network - balances investment based on about asset health system be more incentivisation of
assets to meet service less confident on non-infra potential impact on customer  Some conversations about  Monthly meetings of asset collaboration
requirements now and in the assets of failure – this is used to asset health, but prompted by engineers with operational TLs  Investment in solutions can be
future.  Pipe sampling programme - prioritise investment in the customers raising specific and engineers to review risks covered retrospectively
 About to refresh their asset approx. 1600km of trunk main programme concerns about service and identify any changes – through ODI reward, if specific
health policy and planning to in total, doing around 200km  Target investment based on impacts review assessment and scores issue is built into an ODI, but
include a definition in this per year on a criticality based consequences of failure to  Navigate tool layers to make sure these are concerned about the level of
 Mapped various elements of prioritisation approach customer everything in one picture, so consistent across the board. risk expected to be carried by
asset health to resilience  QR coding approach  PIONEER investment customers can visualize how  Each project has a funding the business
successfully trialled and rollout optimization system (since they are spending their money program manager, supporter  Limited by standards and
in progress. Standardising 2012) based on failure modes, – aiming to make this more and a program user to decide approval; there is no DWI
scripts to improve consistency. models impact of failures on available so that they can use on project goals and tasks. approved reservoir lining
 Developing new burst model performance commitments. it to inform WTP  “Trackdown” and “Trace” though a lot of the equipment
 Asset Care Optimisation and conversations with customers software satisfies EU standards
Innovation and technology ARM  ISO55001 compliance by
 “ICE PIGGING” – scouring out September 2017
large mains by isolating and
flushing with slush puppy
water – brought in to address
manganese
 Innovation strategy based
mainly around development of
existing technologies
 Intending to apply the UKWIR
‘leading API’ work to provide
structure and linkage between
APIs, MoS, performance
commitments and outcomes
Anglian  Asset health + asset operation  Asset deterioration  Use new system Copperleaf  Do not specifically talk to  ISO550001 accreditation –  TOTEX approach is not making
Water = asset health. Health and determined using observed C45. Run complex business customers about asset health early adopter business case of asset renewal
operation not easily measured, data to produce deterioration scenarios  New approach for PR19 cycle –  Serviceability indicators cf predicted 1000 year old
therefore focus on curves and detail probability of  Service impact measures making use of the 4500 people subject to formal assurance sewers. AW believe that a
performance and service. failure linked back to asset health and working for AW – asking more  Value of very mature step change to increase
 Asset health key component  Moving to forward looking performance/risk of failure questions around resilience alliancing relationships with renewals is ultimately
of resistance to hazards – but rather than lag indicators  40 year life to assess benefit especially supply chain emphasised – necessary
using lag indicators  Basket of serviceability cost maintenance and new capital  TOTEX approach has
performance measures rather encouraged Planned
than asset health measures. Preventative Maintenance
But some are good indicators programs and customer FOG
of asset health (exceedances, engagement
CH2M 78
time to repair, % planned to  Believe that service measure
reactive interventions) framework is delivering for
 Combination of ‘expert customers
analysis’ of alarms/alerts, and  Innovation – no barriers that
predictive analysis using lag cannot be managed. Materials
data. Although moving to in contact with water cited as
predictive with WISPA a barrier to innovation –
 ‘ShopWindow’ innovation especially for overseas
area Ely/Newmarket to test suppliers.
/pilot new technologies
 Recognise ‘big data’ challenges
and opportunities –
established a new IT/data
alliance to support this.
Bristol  Performance in line with  BW believe that they  For water pipes, BW can  Do not refer to asset health  Emergency Response Manual  Should not be barriers, but
Water standards, LoS and Outcomes. understand the network very determine general impacts on with customers. and SEMD contingency plans current outcomes are focused
 Sub-set of the old serviceability well. Have rigorous processes service using SEAMS WILCO  BW convey asset health as  Process around lessons learnt on service to current
indicators. to update mains records deterioration analysis tool. terms such as reliability of seem robust. customers based on their
 Resilience and service are the  GIS analysis identifies local supply, leakage & roadworks.  Ongoing provision of ‘reporter priorities and preferences
Innovation and technology:
focus for managing assets. impacts and relationships but grade’ assurance is considered identified via WTP research,
Asset health is a building block.  IWLive which will give a near a new process is being by CH2M to be good practice. and therefore the priority
real time view of the network developed for AMP6. remains to retain the high
 On a complex asset like a  Measures of health based on
treatment works, the inherent in terms of valve operation  Burst deterioration modelling level of service and
etc. BW one of the first to the established serviceability performance now, leading to
redundancy, stand by and provides a solid link to service.
indicators have a long history
duplication means that asset implement this.  For treatment works, the shorter term horizons when
and have been subject to considering asset health. This
failure due to poor health  Installing Syrinix PipeMinder-S direct link is not so clear and a
assurance/governance may lead to a deteriorating
does not typically result in an units across the network to risk workshop approach is
procedures for some time. asset base for the longer
output failure. understand the impact of being developed.
transients.  There is a potential gap in the  New approach and tools will term.
 The burst frequency indicator not be assured yet.  Innovation which leads to
understanding of how asset
is the main asset health  Installing Syrinix PipeMinder-T efficiencies may be beneficial
failure at the component level
indicator for water mains. units on seven trunk mains, ultimately to customers,
could affect the service
identified through operational however, there is no major
output for non-infrastructure
experience to present a risk to return on investment for
assets. Developing a process
the business Shareholders, and leads to an
to understand criticality of
 ‘DIGDAT’, as an information trunk main estate. industry waiting for other
and communication tool for companies to take the risk.
our customers.  There are difficulties in
 Have an innovation implementing innovations
programme. were the benefits cut across
multiple AMPs.
Dee Valley  For DV asset health represents  DV have a three tiered KPI  DV claim a clear line of sight  DV believe it is difficult to  DV state they have emergency  They do not believe there are
Water how hard the assets are scorecard reporting structure between the measures that engage with customers on response plans in place which significant barriers to
working in order to deliver  Two scorecards in place on are used to manage the day to asset health and customers do include support from 3rd innovative approaches
service to customers – they drinking water quality and day operations and the not understand this concept parties  They believe there is a need to
aim to keep risk levels stable. treatment works performance. company level performance Language and terms used are  Post incident reviews ensure an appropriate balance
 AH is part of resilience but also Below ground asset commitments. generally things customers undertaken as standard and between risk and reward in
must consider employee skills, performance scorecard in  Prioritise interventions on may be able to relate to then measures put in place to the way asset health measures
training and succession development above ground assets through prevent future service failure are integrated into the
planning  Recently introduced an asset availability, reliability and  Feel overall performance on outcomes framework.
management system called maintainability (ARM) studies asset health has improved due Specifically Ofwat could retain
CH2M 79
 DV have ten performance RAM (Real Asset on key Water Treatment to introduction of RAM asset the ability for companies to
commitments that relate to Management) for above Works. management system and KPI decide which asset health
asset performance. Includes ground assets. 100% coverage  PIONEER (Tynemarch scorecard system measures are used
direct measures (e.g. number of non infra assets with bar deterioration model) used as  Internal and external audits in  Issues around the TOTEX
of burst mains) and indirect coding in place long term planning tool for place to assure the data that approach and lumpy
measures (e.g. interruptions).  Teams can access and update below ground assets. Does not forms part of performance investments
Performance commitments are all data via smartphone/tablet. integrate with GIS or Netbase commitments and ODI
supplemented with other Believe they are industry therefore manual updates performance
measures including call outs to leader in this aspect required for business planning  Tools used are industry
provide an overall picture.  GIS database for below ground purposes standard so believe this is a
assets. 90% complete for  Risk registers held for all asset form of assurance
age/material groups (e.g. service reservoirs  Size of asset base means sense
 Focus for improvement is in risk register, drinking water checks possible while new
system integration quality). Same group of RAM system easily identifies
 Traditional inspection people review all risks to data outliers for investigation
methods combined with understand high priority risks.  New asset assurance
telemetry data help  They prioritise investment undertaken by standardisation
understanding of assets and based on criticality of equipment where possible
targeting investment. and having processes in place
 Claim they make 75% of for factory acceptance testing
interventions proactively so and commissioning
know their assets
Northumbrian  Asset stewardship. Asset  80% confidence in GIS pipe  SEAMS WiLCO.  The term asset health is not  NW said that the need to  Depends on how asset health
Water health term not really used. positional accuracy.  Deterioration models for really used in general improve the emergency is defined.
 A balance of asset  Scorecards for asset service pipes. consultation exercises. NW response and capture of  NW are aware of the need to
performance and service and performance: NW operate  MEICA asset life models. said that they had consulted lessons learnt was identified as have a balance of investment
metrics at the headline level. a “Trigger” approach with  Non-infra – scorecards and on resilience in the past and part of their ISO 55000 audit types to maintain resilient
monitoring of measures. trigger values. customers see resilience as process (these issues are services now and in the future
Whenever a trigger level is provision of a strong, reliable essential requirements for ISO and believe that their tools for
exceeded and an issue raised service that can stand the test 55000 certification). forecasting asset health and
the operator follows processes of time, cope with change and monitoring service are suitable
to immediately resolve the bounce back from difficult for managing the issue.
issue. situations.
 Encouraging staff to think  NW want to pitch complex
about asset health in terms of issues such as risk at a more
human health. understandable level. They are
 NW say they are industry looking at a range of measures
leaders in the deployment of to help assess risk including
sewer level monitoring (SLM) some risk scales used in the
technology with nearly 90% NHS.
coverage of our combined
sewer overflows (CSOs).
 Active in innovation e.g. we
commission and support grant
funded research focused on
early stage development and
proof of concept of ‘asset
health’ tools, including EngD
with Sheffield investigating
CH2M 80
using pressure transients for
leak detection. MINT.
Portsmouth  No definition of “Asset Health”  Company knows its assets well  ARM provides a sound basis  The extent of dialogue with  Continued sense checks and  Concept of asset health is not
Water and an asset health measure is and confident of GIS accuracy. for infrastructure investment. customers is described as validation of ARM for well understood nor defined in
not used directly. Has used a consistent  Non-infrastructure business limited by the company. infrastructure assets. terms of assets, the “woolly”
 Infrastructure AH gauged from approach for many years. cases harder to develop due to  Customer opinion is reckoned  Independent assurance to the nature of asset health as an
burst rates measured by  A relatively simple network resilient network lading to to be understood and this Board provided by Atkins. indicator has been considered
cohort and interruptions to benefitting from having difficulty in linking outages to revolves around outcomes.  Asset Management processes a barrier.
supply (an ODI measure). resilience built in over time customer impact. Customers expect a safe and are aligned with, but not  Cost of innovation and the
 Non-infrastructure assessed therefore has not been  Historically the company secure supply of drinking certified to ISO or PAS 55000. lack of incentive to adopt
through detailed site survey necessary to focus on asset believes it has applied a water, this is what they pay for  Diligent use of quality plans for innovation is considered a
and asset replacement health. consistent approach to asset and they value this highly. programmes of works and major barrier.
modelling based on assessed  Non-infrastructure assets replacement and has invested projects.  Innovation frequently involves
life expectancy. historically monitored through in capital maintenance at  Lessons learned process increased risks, there is no
 Aquifers are key asset to the number of unplanned work every AMP to mitigate specific applied and subject to incentive to adopt increased
company; a copious supply of orders, and focused site risks. management scrutiny. risk given the present view of
high quality water, health of surveys. asset performance and
this asset is often beyond  New works management customer preference.
company control. system will provide greater
granularity and improve asset
health knowledge, for non-
infrastructure in particular.
 Future plans for decision
support tools using data from
new WAM system.
 Not an early adopter of
innovation but invest when
necessary for specific risks and
prudent for a small company
to do so.
Severn Trent  Asset health is used to  A range of internal  Standard modelling  Consultation to date suggests  Self-Assurance status with Barriers:
Water describe the overall concept of dashboards used to regularly approaches are used where that customers are most Ofwat.  Improved granularity of
understanding the communicate a hierarchy of assets are homogenous, there interested in ODIs, service and  Three lines of defence. consequential cost models.
state/performance/condition. information to different levels is lots of data history and the leakage. Asset health is  Asset data improvements and  Increased presence of asset
It is not a single value or in the business. past is considered a good expected to be taken care of opportunistic data affirmation health PCs in regulatory
defined indicator.  Golden Measures for predictor of the future. during business as usual. processes. contract in areas which may
 Asset health relates to the treatment works.  Probability theory is used to  Usually engage with  Updating and analysis of data not be directly linked to
likelihood side of the risk  Internal measures used as part account for parallel assets. customers in terms and regularly to inform models. customers highest priorities.
equation. of the performance contract.  Criticality used to target concepts they can relate to:  Extensive assurance and  Exploring the economic level
 The basket of if measures  Water infrastructure pressure monitoring such as CCTV WTP for service improvements testing of new assets and of resilience.
include those that reflect the transient monitoring. programs on critical sewers or that can be a manifestation of suppliers. Opportunities:
effectiveness and efficiency of  Critical pump vibration and reservoir inspections. Also changes in asset health;  Possible use of innovation
maintenance. acoustic monitoring. consider the criticality of understanding attitudes to risk incentives such as in energy
 Asset health is a short to  Waste infrastructure ODM systems such water treatment and resilience via WTP for a sector.
medium term measure of asset software on CSO to help works. reduction, general views on  Better alignment between
performance and prevent spills.  For other assets which are maintenance and protecting TOTEX and LoS being offered
serviceability. Asset health is  Large structures destructive more critical, such as the Elan services for future in plan.
not a single measure but needs and non-destructive testing, Valley Aqueduct, other generations.  How to ensure an appropriate
to be monitored by trends in e.g. drones, movement modelling and monitoring  Developing enhanced balance between risk and
appropriate indicators for the sensors and imaging. approaches are used. approaches to consultation for reward around asset health
asset type.  Smart network trail sites. PR19.
CH2M 81
 Asset heath measure capture measures integrated into an
resilience issues that occur outcome framework.
frequently (e.g. many times
per year like bursts or
collapses.
South East  The actual performance of an  Asset Dashboards regularly  A range of statistical and  Trialling a company magazine  Maintained PAS55  Above ground measures of
Water asset compared to the updated with a range of FMECA based tools. and opening WTW for accreditation. asset health should not be
expected performance. metrics which relate to ODIs.  Models are run through visits/tours.  Self-assurance status. about quality and this can be
 TOTEX costs of maintenance  Separate asset health PIONEER.  SEW also hold focus groups  Independent assurers for managed by disinfection.
against repairing, renewing or dashboard for those measures  Asset relationships are with customers to give them a highest risk elements.  Measures need to be longer-
replacing. that do not relate directly to tailored by geography to better understanding of the  Asset data improvements and term.
 Performance against targets. ODIs. reflect regional variation and working network. When opportunistic data affirmation  Strength and direction if
 AM monthly report. local factors that affect the discussed they appear to be processes. movement needs to measured
 A number of Business Objects ratios of asset failures to more concerned with short-  Extensive assurance and rather than absolute values.
reports are available to serve customer events and also use term response issues such as testing of new assets and  Linking residence to return
data to investment planners. knowledge from local people reducing leakage or suppliers. period.
 100 PRBs have condition to better understand their interruptions. However, they  SEW are developing further
monitoring pipe connectivity to ensure can be receptive to longer asset health measures and
 DMA availability metric based the models. term solutions should time be dashboards.
on availability of metric data.  There is a large dataset for the given for explanation with
 Acoustic leakage loggers. assessment of mains bursts case study examples.
 CCTV inspections. and maintenance, indicating Resilience focus groups being
 Some pumps with vibration recent failures, which drives considered.
and temp monitoring. towards replacement.  Consultation suggests that
 Service Reservoir asset health customers are most interest in
monitoring focus. outcomes and leakage and
not asset health – the concept
is not easy to articulate.
South  Maintaining service now and in  Resilience lens (a scorecard)  Deterioration and cohort  PR14 customer panel where  ISO 55000 certification for our  Affordability could be a barrier
Staffordshire the future. Based on  Significant program of testing tools for water pipes. engaged on asset specific asset management processes to bigger aspirations.
Water serviceability indicators. and trialling technology (for a  Risk based process for WTWs. schemes which directly  Monson provide independent
WOC). impacted on asset health. ODI audit.
 Acceptability of water and  As part of the annual review  Cohort tools well established
‘calm network’ – South Staffs we discussed our outcomes and assured previously.
Water have provided training performance with our CCG last  Models have been subject to
to help prevent adverse summer, which included asset peer review.
impacts from poor operations. health ODIs.
 Participate industry innovation  Recent CCW/customer panel
forums (TAG). engagement re asset health
 Surge/transient analysis. based decisions.
 WQ profiling.
 Portable water quality loggers.
South West  Based on serviceability and  Asset Reliability ODI (based on  Make TOTEX investment  Customers were very clear  OPRA and developing EPR  TOTEX and outcomes
Water asset reliability measures. serviceability indicators). decisions using risk-based that they did not want to management systems. incentives allow companies
 Additional measures needed to  Seeking to link serviceability assessments to achieve burden any particular  Operational centre provides the flexibility to deliver the
(1) address DWI risk concerns, measures with physical performance commitments, generation with unfair bill 24/7 cover to telemetry and right outcomes against
(2) EA desire for companies to condition of assets and asset reliability metrics and rises – it is important to Scada alarms and ensures that customer priorities.
be more risk averse (SW Water informing this by monitoring. outcome delivery incentives. deliver an affordable and these are prioritised through a  WATERSHARE – if SWW
beach issues) (3) isolated  Real time information from  Deliver this programme based reliable service to customers risk based approach and then outperform our business plan
communities – exploring these PUROS (Phased Utilisation of upon a planned programme now and in the future. acted upon through the there are financial benefits
Remote Operating Systems), reviewed annually and a RIO which are shared with
CH2M 82
issues with criticality and iScada, network and EDM (Rapid investment) operational teams and standby customers. If we
resilience. monitors. programme reviewed rotas. underperform, customers are
 MOSS program uses monthly.  Report the number of protected.
SewerBatt to quickly and  An Investment Manager suite attempted IT ‘hacks’ to the  Environment Agency is seen to
efficiently look for ‘hot-spots’ of programmes will optimise board. Test ‘hacking’ scenarios be unclear over processes that
in the wastewater network. on a cost benefit analysis and review ‘data theft’ water companies should
 Looking to predict where and linked to willingness to pay. reports. follow for having innovation
when where and when man-  Established Governance approved.
hole covers could ‘pop’. Committees (Audit,  Water companies tend to get
Sustainability, PR19, ‘punished’ by Ofwat for
Remuneration and failures if innovation does not
Nominations) chaired or work – this places a
attended by independent Non- disproportionate part of the
Executive Directors. risk associated with innovation
on water companies. SWW
would prefer there to be some
sharing, across water
companies, of risks (and
successes) associated with
innovation.
 UKWIR Framework for
Expenditure Decision Making
with a goal of how to
incorporate ‘future’ issues into
investment decision making
(including intergenerational
finance) concern over the low
level of support from Ofwat
for the outputs from this
project.
Southern  Asset health is how the  Headline asset health  Low criticality and low failure  Soon to be published Talking  Rapid response and  An inadvertent consequence
Water physical asset condition and measures currently defined may be a replace on fail capital Water Part 1 – which is a emergency planning, strategic of a TOTEX driven approach is
performance enable service to within our measures of service maintenance policy. precursor to our Strategic spares management and out- that it could drive in-AMP
be provided to customers and (MoS) and performance  Low probability and high Statement (Talking Water Part of-hours supply chain efficient maintenance
/ or the environment, now and commitments. impact or consequence 2) – we touch on asset health arrangements. Emergency solutions, at the risk of
in the future. Given the inter-  Headline measures are requires contingency planning in a number of ways. This response is coordinated impacting long-term asset
generational impact, asset complimented by a hierarchy and/or remote monitoring. includes: the disruption impact through our operational health.
health is not about measuring of asset / performance data  Prioritise investment, based of replacing ageing assets; and control centre and is  If there is too great a focus on
one off performance, but use and information. on declining health indicators the different options of supported through 24/7 the achievement of five-year
of trends in order to inform  Increased number of logging and the impact of failure maintaining or replacing assets arrangements with mobile outcome delivery incentives in
the longer term risk to service. devices and flow which in turn can be related to and the impact on costs and water providers and react and the investment decision
measurement instruments critical assets. Therefore for bills. maintain contractors. process, there is a risk that
complimented by LAMARS. less critical assets, the  A sample of 100 customer  Water supply bacterial companies might ignore the
 Technolog Regulo PRV measurement of asset health interviews show they most compliance performance longer term or more strategic
controllers. is monitored less frequently. value affordable, stable asset improved via deployment on and multi-benefit resilience
 R&D in information analytics health and service resilience an automated control and options in pursuit of shorter
and modelling, sensor and for future generations. shutdown system on the term performance confidence.
detection tools and methods works.  The estimation of monetized
and technology / service  Assure with 3 lines of defence; figures for the cost benefit
delivery technologies. business operations, oversight analysis is likely to undervalue
and independent assurance. the true value of the health
CH2M 83
 Customer advisory panel can and resilience of new assets,
scrutinise performance especially where there are
monthly. intergenerational benefits to
 Gateway challenge and review be had.
sessions for new assets.
Sutton & East  Using risk of failure as measure  CMMS.  PIONEER to test service  Asset health not used in  Detailed emergency  Concern that TOTEX approach
Surrey Water of asset health from CMMS;  GIS mapping of bursts. impact of asset deterioration. public communications. Focus management/contingency incentivises patch repair on
historic outage, performance  CALM.  Failure curve of each asset on performance commitments plans. long lived assets.
and reliability. CMMS  Performance (bursts) is generated from age, and and outcomes.  Modelling used to predict  Small company can be good
identifies likelihood of asset primary indicator of linear historic failure rate. failures and service outages. test bed for innovation. Can
failure based on range of asset health.  Failure rate used in CMMS. Outputs of this and criticality Ofwat incentivise innovation?
parameters.  Vibration monitoring and oil  Distribution model inform interventions; whether  Communication/sharing best
 For long life assets use monitoring for above ground improvements risk based. contingency or investment. practice between companies
(reservoirs and tanks) physical rotational assets.  Proactive replacement/  Self-check/data modeller is getting worse (innovation
inspection and grade.  Infrared thermography refurbishment of assets likely check of data validity. manager).
 For major pumps use programme being to fail/critical assets.  Independent assurance
efficiency and duty point as implemented. (annual returns and business
measure of asset health.  SES being smaller is happy to plan).
 No single measure of AH, but test new innovation and its
moving to this. smaller size makes it
 Good contingency planning. favorable.
 Overreliance on experience of
data analytics.
 Acknowledge that small
customer base/asset base
impacts on availability of
trend/performance data.
Thames  Asset health measured using  Lag indicators used in MARM  AIM for WWTW and WTW.  ‘Customers do not understand  The level of reporting and  Outcome based rather than
Water Output/serviceability to trade reactive and proactive  Deterioration modelling. (or want to understand) the assurance inside Thames output approach allows them
measures. capital investment  Cohort analysis. term asset health’. Water and external assurance to decide how best to manage
 Different measures for programme.  MARM.  Asset health not used in has increased. In the last year their assets.
Infrastructure / non  System scale, or major asset  Investment based on service public communications. Focus we have employed an extra 5  Cross industry research and
infrastructure and water / scale indicators (e.g. WwTW and outcomes – to reflect on performance commitments people to ensure that data is intercompany collaboration is
wastewater. discharge compliance, sewer customer priority and risk, not and outcomes. correct . likely to be better value for
 Good asset health improves flooding, and water quality asset condition.  Assurance by management customers.
resilience. complaints). consultants KPMG and sub-
 Real time data from depth contracted specialists.
monitors used for wastewater
network in ‘Floodworks’.
 Proactive condition
assessments of critical linear
assets in place or planned.
 Basket of techniques used
across the different types of
asset. Seek to identify the
best tool for the job.
United  Calculation of remaining life  An asset health Index  A consistent dataset is used  Concerns about raising  Test emergency response with  TOTEX allows flexibility in
Utilities and expected economic life combines age, cost and with: questions regarding asset Exercise Triton and develop intervention choice.
weighted by asset condition, last 5 years  PIONEER to identify long term health alone, without approach with stakeholder  Assessment needs to be
replacement costs (GMEAV). performance of mechanical investment requirements providing an appropriate engagement – e.g. with undertaken on a long term
and infrastructure and failure based on asset health Cumbria Flood Forum. basis and not just 5 years and
CH2M 84
modes analysis into a RAG  OPTIMUS to prioritise short context of service levels and  Post Incident Review process. that service risk should take
categorisation of Healthy term and capital investments willingness to pay.  3 lines of assurance, with account of both high
Asset, Minor Health Issues or based upon impact to service. Level 3 manager, director and probability low impact risks as
Poor Health.  Management information external auditor for data. well as low probability high
 This is a consistent risk-based reports and dashboards to  Working towards ISO 55000 impact risks.
measure applied across the support and target operational accreditation.  Could be a steer on what level
asset base that can be and maintenance activities.  Asset Integration service on of resilience to plan for, i.e.
aggregated to site level,  Range Management: assign new assets includes Technical common threshold for future
planning areas and a strategic expert engineers to monitor Approval Points, and 3rd party flood risk.
overview. and manage similar assurance for example.  Recognition that companies
components across the asset  D&B approach backed up by face different challenges that
base – e.g. screens or pumps. Technical Assurance and could influence an assessment
 A sewer incident RAG heat include ‘fitness for purpose’ of cost impact (i.e. resilience
map identifies areas of the and ‘performance guarantees’. to drought in the south east
network that are most critical and resilience to flooding in
for improvement on a risk- the North West).
based spectrum.
Welsh  Currently based on lag  Most forward looking vision  Have Investment Manager  Active customer engagement  Currently have independent  Not really – it used to be data
Water indicators – serviceability for the new asset health MoS toolkit, which links service and programme – in progress. assurance – signed off by and tools a few years ago but
measures. In future  lead – lead indicators around investment.  Work conducted on wider board, as part of annual have had a massive
measures of asset deterioration model  IM deterioration models for topic of resilience, not asset reporting. investment in this and has
performance, deterioration predictions. infrastructure and some non- health. increased confidence in the
 Governance occurs monthly
and asset risk.  Data analytics and infrastructure assets.  Do not expect to talk about new approaches which now
internally (Monthly
 Appreciate value of asset visualisation (Tableu and  WTW risk based (DWSP) asset health sub-measures in make an asset not a barrier.
Management Report, Capital
health and not allowing an Igloo). rather than health based. consultation. Programme Board).
asset health ‘bow wave’ to  Have an innovation
develop. programme and have trialled a  Tableau reports also expose
number of inspection and issues (Visualise data
monitoring technologies. relationships).
 Will need to enhance
assurance on the
deterioration models when
they become part of the MoS.
Wessex  Focus is on service delivery to  Measure asset health through  Deterioration models linked to  Active customer engagement  Stable asset health metrics  Accounting treatment of
Water customers – ‘asset health is lagging health indicators of consequence of failure for program – now looking to (bursts and collapses) infrastructure renewals being
the ability to deliver high water pipe bursts and sewer infrastructure assets – permits explore resilience issues more demonstrate effective the same as OPEX is a
quality services … now and in collapses. understanding of benefit-cost carefully. stewardship of assets. disincentive to large scale
the future’.  Established condition grade relationships.  Difficult to address issues  Processes and data subject to renewal.
system but also sophisticated  Links asset health to about AH as distinct from independent assurance  Common industry targets for
sewer risk model – taking investment and service. service.  ISO550001 accreditation for bursts and collapses are not
account of condition and  Difficult to address issues of asset management approach welcomed because they don’t
consequence of failure. future service (which may be generally. take account of local
 Leakage is an indirect and dependent on good asset conditions and performance
useful indicator of asset health). history.
health.
 In sewers, the ‘risk score’
directs proactive inspection
(CCTV) programme,
successfully focussing on
CH2M 85
problem sewers for rehab or
replacement.
Yorkshire  Service & reliability (S&R).  S&R scorecards supporting  LEADA strategic investment  Triple line assurance  The companies may have very
 Examples of recent
Water Don’t use asset health term. awareness of asset health in planning tool – first of type for processes. different historic expenditure
communications campaigns on
Currently based on key decision processes. PR04 – but needs refreshing. o Board profiles within the asset group
asset health related subjects.
performance and service, Wastewater non –  Service Measure framework o Internal audit and being considered by the
derived from serviceability infrastructure utilise these for assessing impact of asset o ‘This is Water’ DAGs performance commitment
indicators. scorecards most rigorously in health. campaign – incl. water o External target. This can lead to
this context.  Asset level scorecards and efficiency, sewer  Emergency response, lessons companies opting for
 Detailed process working with trigger values. blockages and FOG. learnt, new asset operational interventions such
SAP data to ensure data  New DMF in development. o Water efficiency
commissioning seem as flushing to meet short term
accuracy for determining  5 capitals (Natural, Social, campaign, ‘Save a
comprehensive. targets, whilst deferring
failure data, which then feeds Human, Manufactured,  YW use a Minimum Technical significant investment on
little, Save a lot’ was
into the asset deterioration Financial) for the new DMF Specification to help ensure asset replacement which may
recognised at the CIPR
modelling. and 30 service measures to be assets are installed and have a longer term benefit on
Pride Awards 2016.
 BS1192/BIM processes and incorporated in DMF. commissioned appropriately asset health and resilience.
the benefits offered by the  Tailored media and and this is coupled with
SAP refresh programme will consideration of requirements relating to
enable improved data demographic. competence and training and
collection, update and  Regular customer elements of site supervision
governance. communications campaigns, and inspection.
capital programme
communications, incident
management activity, media
management, customer
insights activity, wider
stakeholder contact
programme.

Technology Inspection technology Monitoring and control Data analysis
type
Company Sewers Water mains Treatment & Other Sewers Water mains Treatment & Other Sewers Water mains Treatment & Other
pumping pumping pumping
Affinity Water NA  Distribution pipe Not specified Not specified NA  Syranix trials (cost  QR asset tagging Not specified NA  Deterioration  Asset Care: Not specified
CA issue)  Energy models Trace and
 Critical trunk monitoring (trial)  GIS visualisation Trackdown
mains NDT (cost  Pump vibration (capture ops
issue) (trial) data)
 Thermal imaging
(limited value)
Anglian Water  CCTV (planned For trial (Shop  Inspections by Not specified Not specified For trial (Shop Not specified Not specified  Deterioration  Deterioration  Failure models Not specified
and reactive) Window): staff Window): models models and inspection
 Echologics For trial (Shop  Syrinix – Transient  GIS visualisation  GIS visualisation (some linked to
ePulse (acoustic Window): Monitoring  Hydraulic  Historical data condition grade)
condition models overlay
assessment)
CH2M 86
 FLIR / TI –  Artesis -Motor  Inflowmatix –  CCTV validation For trial (Shop  Models
Thermal Imaging Condition Transient of model Window): validated and
Leak Detection Monitoring Monitoring predictions  I2o Pressure linked to SAP
 Gutermann – Management
Permanent Leak
Noise Monitoring
 I2o Pressure
Management
Bristol Water NA  WQ sampling  Inspection and  (Civils)Inspection NA  IW live (near real  Acoustic and  Drones NA  DIGDAT customer  SAP database  MeMboard
condition and condition time valve ops) thermographic (catchment info. tracks key info software
grading grading  Syrinix-S, T, RADAR monitoring management)  GIS (Smallworld)  Risk assessment (submerged
(selective)  HV assets (oil implementation (mixed results)  MINERVA (trunk process and membranes)
analysis)  Hi speed logger mains criticality) asset life models
(trials)  SEAMS
Deterioration
models
 Live hydraulic
models
Dee Valley NA  Ice pigging mains Not specified Not specified NA Not specified Not specified Not specified NA  GIS  Level 1 – 3 Not specified
Water cleaning (good  PIONEER performance
results) deterioration scorecards (WQ
 Standard models and TWs
methods (not performance)
specified)  RAM asset
management
system captures
events data
CH2M 87
Technology
Inspection technology Monitoring and control Data analysis
type
Treatment & Treatment & Treatment &
Company Sewers Water mains Other Sewers Water mains Other Sewers Water mains Other
Northumbrian Not specified Not specified Not specified Not specified  Sewer level  Flow cytometry  SPS intelligent  IT asset health  Project SNIPeR (data  GIS analysis of  Scorecards for  Civil structure
Water monitoring (bacteria in DW) control monitoring analytics/ early failures all major sites condition
(90% CSO  Mass balance warning)  SEAMS WILCO  Root Cause surveys
coverage - optimiser for  ESRI Collector (easy deterioration Analysis
EDM) water transfer positional correction) models  Trigger point
 Water Rangers  GNSS (positional  ESRI Collector analysis
accuracy e.g. (easy positional  Strategic MEICA
manholes) correction) asset life models
 GIS analysis of
failures
 SEAMS WILCO
deterioration models
Portsmouth NA  Trunk mains CA  Site inspection Not specified NA Not specified Not specified Not specified NA  Cohort  IFS data Not specified
Water (historic) (visual) deterioration collection (in
 Drones (leaks), models infancy)
calm networks,  GIS analysis
pressure
transient
monitoring
(being
considered)
(Note: late adopter

policy declared)
Severn Trent  CCTV  Satellite imaging Not specified  Destructive  EDM software  Pressure transient  Vibration and  Telemetry alarm  GIS analysis of  GIS analysis of  SAP project  BIM WG to
Water  Acoustic for leaks (future) and NDT on CSOs modelling PhD + acoustic rationalisation failures and data failures and management improve data
blockage sensors  Next gen. e.g. siphons,  HS logger trial monitoring to improvement data  Suite of decision flows through
– SewerBatt pipeline CA: JD7, masonry (burst prevention reduce energy and  Suite of decision improvement support tools asset lifecycle
(planned trials) Echologics E- crossings & management) PM costs (in support tools  Suite of  Reliability and
Pulse (planned  Laser scans on progress) decision Monte Carlo
trials) structures  SMART networks support tools deterioration
 Movement (trial) modelling
sensors  FACTORY concept
 Drones (future)
 GPR  Intelligent SPS
(future)
 Contact tank bacti
monitors (future)
South East NA Not specified Not specified Not specified NA  Acoustic leakage  Asset tagging  PRV condition NA  GIS analysis of  Dashboards  Working
Water monitoring (80%) monitoring mains failures  MAXIMO data towards BIM 1
 CCTV for mains  Vibration and  Customer call  Dashboard  FMECA or 2
condition (priority temperature on monitoring  Deterioration
areas) some pumps  Met office models
 Trunk main loggers  NDT on pump weather data
and WQ sensors spindles (large (leaks and
(planned trials) ones) bursts)
South NA  Trunk mains CA  Condition  Rebound NA  Real time  Vibration and  Do mobile and NA  Dashboards  Dashboards  Want to invest
Staffordshire (some by NDT) reported by hammers to customer data thermography fixed monitoring  Resilience lens  Resilience lens in analytics
Water operatives test concrete  WQ monitors (some) tool (in tool (in
structures Chloro/Hydra  Condition reported development) development)
(new) clams and by operatives  GIS analysis of  Risk based
Nephnets  Asset tagging failure clusters assessment of
 Trunk mains need
‘conditioning’ 
CH2M 88
Technology
type
(raise velocity to  Cohort
remove sediment) deterioration
 Ice pigging models
(successful but  Hydraulic
expensive) models
 Surge/transient (consequence)
analysis
South West  Sewer- batt Not specified Not specified Not specified  EDM monitors  Flow and pressure  iScada (monitors,  GIS analysis of  GIS analysis of  Dashboards  Exploring IT
Water (blockages) monitors telemetry, Scada) failures failures  FMECA system
 CCTV  Surge monitoring,  Cohort models  Cohort models  Weibull analysis resilience
acoustic leakage  Trend analysis  Trend analysis  Root cause
detection (Calm  Hydraulic models to  Hydraulic analysis
network) predict models to
consequences predict
consequences
Southern Water  CCTV  Trunk &  Thermal imaging, Not specified Not specified  Increasing use of  Thermal imaging, Pilots on variety of  ODI dashboard  ODI dashboard  ODI dashboard  Exploring
distribution oil sampling, flow logging oil sampling, sensors and  GIS analysis  Age and  Ellipse FMECA,
mains CA (NDT & ultrasonic devices ultrasonic detection tools deterioration  Root cause condition
DT) detection  Remote control detection models analysis based
pressure  GIS analysis monitoring
management  MTBF an AH
viaTechnolog metric?
Regulo PRV
controllers
Sutton & East NA Not specified  Oil analysis  Impounding NA  Pressure  Oil analysis Not specified NA  GIS analysis  PIONEER Not specified
Surrey Water  Hand held reservoir management: flow  Hand held  PIONEER deterioration
vibration monitors inspection by and pressure vibration monitors deterioration models
 Infrared panel engineer loggers throughout models
thermography network  Hydraulic
being considered  ZONESCAN noise models
(electrical assets – loggers (leakage)
faults, load
imbalances etc.)
Thames Water  Floodworks (RT  DT and NDT on Not specified Not specified  Telemetered  Vibration analysis,  SIMEOTM4D  AIM decision tool  AIM decision  MARM (CM Not specified
sewer depth high risk assets network (leakage thermography and ambient noise and detn. Models tool & detn. tool)
data analysis) monitoring) oil analysis sensors for civils  GIS analysis Models  RCM and
 CCTV (risk  Pressure monitors  WWTW: quality, condition  Rising mains RCA  GIS analysis predictive
prioritised)  Trunk mains flow, spot (proposed trial) maintenance
 Auto defect acoustic monitors sampling, power  H2S sensors (proposed trials)
coding (trial)  Surge monitoring usage, portable (proposed trials)  Process models
 SewerBatt  Asset tagging and vibration sensors (BIOWIN)
(blockages) GPR (proposed etc.
 Electroscan trials)
(infiltration)
United Utilities  CCTV Not specified Not specified  Civil: visual (1 –  Smart Ball Not specified  Condition based Not specified  GIS visualisation  GIS  MARS asset Not specified
5 grading) technology maintenance  PIONEER visualisation register
that assesses  Vibration and deterioration models  PIONEER  PIONEER
rising main thermographic deterioration  Multi level
condition analysis models scorecards
whilst  Real-time  MISER (system  FMEA
operating monitoring of ‘run- model)  Reactive work
within the hours’  Age and analysis
main  Telemetry alarms condition  Maintenance
optimisation
CH2M 89
Technology
type
 Some remote  RCM
control
Welsh Water  CCTV  Sahara leak Not specified Not specified  EDM monitors Not specified Not specified Not specified  GIS analysis (near  GIS analysis  DWSPs  Igloo 3D
detection real time) (near real time)  Tableau data immersion
(trialled)  SDPs  Zonal studies analytics and technology
 Ice pigging  Deterioration models  Deterioration visualisation  Tableau
(trialled)  Connectivity & models  PS deterioration
 Portable hydraulic models to  PODDS (water models
pressure look at consequence quality in  Reliability blocks
monitors distribution (promising but
predictive £££)
model) Flood  Analyse
models to look unplanned
at trunk burst maint.
consequences
Wessex Water  CCTV Not specified Not specified Not specified  EDM monitors Piloted technologies Not specified Not specified  SRM type risk based  Deterioration Not specified Not specified
 Laser and sonar for active leak control approach – sewer models
scanning for – Takadu and risk model
large sewers Gutterman  Deterioration models
 Electroscan
(sceptical)
 SewerBatt
(accuracy
concern)
 Cut outs tested
when failures
occur (rising
mains
 Petro concrete
testing (H2S
attack)
Yorkshire Water  CCTV  Ice pigging  NDT: radiography,  Outfalls: CCTV,  Wide scale  rtNet pressure and  Pumps: modulated -  Scorecards  Scorecards  Scorecards
mains cleaning ultrasonics, EM sonar, telemetry flow telemetry pressure  Deterioration models  Deterioration  Deterioration
(mediocre condition ultrasonics and (levels) management (BGASP) models models (AGASP)
results) inspection MFL  GIS visualisation (BGASP)
 Whirlwind (selective: pipes,  Dams & res:  GIS
vortex mains tanks, roofs) sonar, radar, visualisation
cleaner  Visual CA (value GPR
uncertain)  Water towers:
drones
CH2M 90
Water Tools for linking AH, service & Predicting AH impact on Predicting consequences of The role of criticality Future plans
Company investment service asset failure
Water & Sewage Companies
Anglian Established consistent valuation - Risk, Prediction modelling below and above  A clear line of sight between asset  Asset Criticality ≡ Consequence 1. Understand the asset base and the
Water Opportunity and Value approach (ROV), ground. deterioration, asset performance of Failure areas where they would like to have
supplemented by willingness-to-pay (WTP). and service impacts. better data and tools
Pipes assessed for the consequence of  Service measures that enable the
2. Have plans for when to improve
Likelihood of component failure X Likelihood of failure including deterioration,  All assets have a consequence of business to value the cost of
Service Failure if Component Failed X Cost of discoloration and flooding (CCTV failure linked to a probability, which failure
Service Failure (£) = Risk (£) validation) is monetised to create a common
 linked to the outputs of societal
• Input risk valuations into new system Above ground asset failure models currency of risk throughout the
valuations
Copperleaf C55 to optimise based on complex some linked to condition grades, business.
business scenarios impacts on connected assets modelled
with Monte Carlo
• Service impact measures linked back to asset
health and performance/risk of failure. Pollution App predicts events and Blue
Light System and Eventus software
assesses scale and response
Northumbrian SEAMS WILCO integrated investment planning Water and sewer pipes: WILCO  Rolling ball analysis is used to  Criticality of assets is used to 1. Continue to use the WILCO strategic
Water tool statistical deterioration models estimate impacts of water and assess investment needs (non- tool.
sewage flooding from failed pipes infra) and inform impact
 CMPCF risk based MEICA: asset life models and FMEA 2. Enhance understanding and use of
appraisal. For infra assets the
 Hydraulic analysis is done to assess scorecards and trigger values
 Used at PR14; third party assurance Civil structures: concrete deterioration appraisal of criticality is done on
models effect of pipe failure on customer 3. Explore understanding of risk for low
a reactive basis.
 Forecasts effect of investment on service supply probability, high consequence assets
and AH and vice versa Trunk mains: condition assessment  Investment decisions are made
linked to probability of failure on assessment of probability and
Asset level scorecards that enable monitoring of
sub-threshold trigger values and prioritisation of consequence.
actions
CH2M 91
Water Tools for linking AH, service & Predicting AH impact on Predicting consequences of The role of criticality Future plans
Severn Trent Links between health measures and service Infrastructure: Infrastructure models are aggregated to Standard modelling approaches are 1. Water network model based on failing
Water demonstrated through use of Driver trees. zone or appropriate spatial area used where assets are homogenous, all pipes individually
 Wastewater mature model in
there is lots of data history and the
Generally AH information used to create development since PR09. Totex Non-infrastructure models are rolled up 2. Improved life distribution, repair rate
past is considered a good predictor
deterioration models to forecast the AH will model 25 year+ planning to process stream. and repair time models
of the future.
change with time. horizon. 3. Improved cost models
Probability theory is used to account
Service impact relationships used to link  Water (WiSDM) in 4. Improved consequence models for
for parallel assets.
probability of failure to service. development since PR04. Totex wholesale
model 25+ planning horizon. Criticality used to target monitoring
Several mature tools which have been
such as CCTV programs on critical 5. Ability to constrain spend and set
developed over successive AMPS. Non infrastructure:
sewers or reservoir inspections. Also targets for service levels
 Expected life distribution consider the criticality of systems 6. WLC calculations within NI models#
models with repair rates and such water treatment works
repair times 7. Including process performance in
For other assets which are more DWSP EoC
 Probability of failure, quantity critical, such as the Elan Valley
of impact, severity of impact, Aqueduct, other modelling and 8. Application of linear solver to optimise
probability of consequence monitoring approaches are used. solutions
South West Asset health and performance, and service Assumes that past performance is an Identify and assess the implications of Consider assets are ‘critical to 1. SWW is developing an ‘index of
Water performance are modelled with a risk-based indicator of future performance. potential asset failures through the performance of the system as a criticality’ based on aspects that affect
approach (e.g. DEAM) to predict future application of a risk based approach whole’ and ‘critical in terms of ‘vulnerability’
As part of business planning this year,
performance using either FEMCA, Weilbull which informs our investment planning outcomes on customers and the
future trends resulting from climate,
analysis or network risk/performance curves approach. environment’. For example:
population growth, social or regulatory
based upon cohort analysis of pipe groups such
change will be explored to understand Failures are captured through either  The threshold for vulnerable
as rising mains.
uncertainty related to asset health. telemetry and scada system alarms or communities from a COBRA
Stochastic hydraulic modelling is used by operator contact. perspective is a community
Report the number of attempted IT
understand the effects and risks of climate larger than 10000 inhabitants,
‘hacks’ to the board. Data analysis to provide early warning
change and future system needs. but SWW serves numerous
of asset failures, e.g. analysing the type
Make TOTEX investment decisions using risk- and frequency of pressure transients isolated communities that have
based assessments that occur in the water distribution fewer than 10000 inhabitants
networks as these can signal a potential and are supplied from a single
An Investment Manager suite of programmes source and are therefore
pump failure.
will optimise on a cost benefit analysis linked to regarded as ‘critical’
willingness to pay. communities.
Input data is based on asset deterioration
modelling over a 25 year time horizon to help
identify the need for interventions within any
5year period.
CH2M 92
Water Tools for linking AH, service & Predicting AH impact on Predicting consequences of
The role of criticality Future plans
Southern Pioneer system modelling used to prioritise Apply principles of the common Ability to predict the performance of  Asset criticality is assessed using Improve asset health measures to provide
Water investment, based on declining health indicators framework for maintenance planning: assets depends on the quality and a number of approaches more forward-looking, lead measures,
and the impact of failure, which in turn can be maturity of the required data: depending on asset type. These aligned to specific performance / delivery
 Measures of asset health will trigger
related to critical assets. approaches including water and levers and ultimately our outcomes and
a ‘need’ to be raised. Needs are  data for networks more mature and
wastewater hydraulic modelling performance commitments. Examples
Asset health is monitored less frequently for reviewed, challenged and approved be more readily trended
for infrastructure assets and include: control and shut down measures,
less critical assets. though Asset+1 meetings.
 factors that affect processes at FMECA studies for non-infra Failure Mode, Event and Criticality Analysis
 Solutions are then developed and water supply and wastewater assets. (FMECA), condition based monitoring
challenged on their cost treatment works can be more
 Also use ‘hot-spotting’ i.e. geo- Review measures to determine the
effectiveness at Asset+2 meetings varied
spatially plotting of asset and effectiveness of asset maintenance
Currently developing a forward looking service risks identified, to arrive standards. Mean Time Between Failures
blockage prediction model. at a consolidated view of issues (MTBF) may be used as a specific asset
and impacts. health measure.
Thames APS system collates service risk information System models used to model strategic Use deterioration models to predict Network criticality is based on a View risk consequence and criticality more
Water against a set of detailed service impacts that performance and impacts throughout Service risk over time (in conjunction score for condition, age, its dynamically and have clear visibility to see
map to the Performance Measures. the system: WARMS – security of with asset downtime models; system environment, probability of failure how criticality of remaining assets changes
supply, Voyage – resilience. failure models; and service and consequence. Managed in AIM. or escalates when assets within a system
Detailed service impacts are collected in a Risk
consequence models). fail.
Framework. IBM flooding and pollution risk tools Maintainable assets have criticality
learn from previous incidents to scores stored in SAP.
Decision to move to a new level of service is
highlight risk properties or
based on optimising the costs and benefits. Developing models for treatment
watercourses.
Hold monthly review meetings. work and pumping stations linking
assets to risk and cost of failure
The link between asset health, service and
assessing consequence.
investment is better seen over a long time
period
Use deterioration models to predict: spend
requirements over time (in conjunction with
cost models).
Live system capability models enable production
optimisation and hydraulic models and ICM live
wastewater network models which predict
network capability in real time.
All models use cost of service data and cost of
failure data which is aligned with APS.
CH2M 93
United A consistent dataset is used with: Consistent asset health calculations are  Failure modes analysis, is a key Criticality is defined by the To improve the Asset Health metrics going
Utilities applied on point and linear assets: contributor to the Asset Health consequence in terms of impact to forward UU are looking at:
 PIONEER to identify long term
metric the customer, impact to service and
investment requirements based on PIONEER takes the output from data  Further improvements to data quality
the cost to recover.
asset health systems (MAMS plus MARS) and  A review of all failures identifies the to ensure they meet data requirements
deterioration models at the asset type why and the associated Criticality drives investment in so that the information can inform
 Optimus to prioritise short term and
level consequence. PIONEER and influences priority robust business planning.
capital investments based upon impact
ranking in Optimus and maintenance
to service. In AMP6, outcomes and performance  Using scenario modelling to inform the
scheduling systems.
commitments measure the customer effect of Totex interventions on the
 Management information reports and
impact of the performance of assets. A For example; a sewer incident RAG health of the asset base.
dashboards to support and target
number of these measures are heat map identifies areas of the
operational and maintenance activities.  Developing direct linkages of asset
however, indices, which have asset network that are most critical for
health to serviceability measures of
A Range Management practice assigns expert performance and condition metrics as improvement on a risk-based
performance.
engineers to monitor and manage similar sub measures. For examples, the spectrum.
components across the asset base – e.g. screens Network Performance Index has  Developing our focus and
or pumps. Blockages, Collapses, Rising Main Bursts understanding of leading measures for
and Equipment Failures as sub asset health to enable targeted
An Integrated Logistics and Support practice
measures. intervention plans.
ensures improvement of asset designs and
procurement). Serviceability indicators and their sub-  Developing a clearer linkage to our
measures are still reported on as resilience measures.
previously, to ensure a consistent
appreciation of historic trends.
Welsh Water ICS’s Asset Investment Manager (AIM) for The deterioration models link asset to  Workshopping with risk maps  Decision making is risk based and DCWW are investing in data analytics and
integrated investment planning service, via the SMs. (FMEA) informed by CBA visualization techniques to provide a better
understanding of the asset base generally.
 CMPCF risk based; Service Measure Deterioration models are part of the  Statistical analysis of events and  Non-critical asset health is also
based Investment Manager toolkit and cover impacts important as this represents a The igloo 3D visual asset immersion
infrastructure assets and pumping large portion of the asset base technology is finding increased usage and
 Used at PR14; third party assurance  Connectivity/hydraulic models
stations. DCWW continue to explore novel ways of
 Forecasts effect of investment on  Flood modelling using this technology.
service and AH and vice versa
* DCWW state that their WTWs and  RBDs (reliability blocks - trialed but STWs are not well represented by current
SDPs and zonal studies for developing too expensive) deterioration and service forcasting
STWs do not demonstrate a good link
awareness of area level environmental and models. There are many small assets so a
between asset failures and output
water quality risks risk based approach is not cost effective.
performance so a risk based rather than
This is under investigation to identify
Visualisation and data analytics reporting tools health based investment planning
appropriate methodology for expenditure
for identifying relationships in data approach is used
planning.
CH2M 94
Wessex Water mains & sewerage deterioration models Sewer deterioration models indicate Sewer Risk Model contains In depth criticality assessments • Moving to forward looking rather than
Water and sewer risk model link asset health to that renewal rate of 0.32% is needed for consequence information embedded in based on likelihood and lag indicators
investment. stable asset risk – with inferred impact risk score – proximity to rivers, consequence of failure - applied on
on service. environmental sites, highways, railways drinking water trunk mains.
Intel and knowledge captured within Drinking
etc.
Water Safety Plans.
Water mains deterioration model
Sewer with high risk scores are prioritized for
linked to mains failure consequence
pro-active sewer renovation.
model.
Yorkshire In house developed LEADA integrated planning AGASP: above ground asset  Proximity analysis to receptors  An understanding of risk 1. Major upgrade to LEADA system
Water tool deterioration models and service (probability of failure and planned for PR19 (DMF)
 GIS and hydraulic analysis
prediction. Asset life and Weibull curve consequence of failure) rather
 CMPCF risk based; Service Measure 2. 5 capitals benefits valuation
based  Traffic disruption potential than criticality informs the
based 3. Refinement of current
investment process
BGASP: below ground asset  Contact data
 Used sing PR04; third party assurance deterioration and service
deterioration and service prediction.
and external review  Leakage volume loss estimation forecasting tools and actual data as
Pipe level forecasting based on
a basis for estimating equipment
 Forecasts effects of investment on risk statistical analysis. Bayesian tuning to  CBA in LEADA based on estimated failure probability
integrate modelled vrs actual failure risks for options
S&R asset level scorecards using wider range of
data. 4. Refined benefits assessment
AH information to support targeted decision
making 5. Enhanced Service Measure
Framework
CH2M 95
Water Only Company
Affinity Through Pioneer optimization modelling Using Innovyze hydraulic models and Pipeline failure modelling includes: Impacts of consequence in A goal is to understand the failure of TMs
Water individual assets are linked to MoS. Pioneer optimization tools likelihood of investment optimization and in short better (why do TMs fail when in good
 Condition relating to failure
failure is linked to service term business cases are captured in condition)- linking measure with long term
Individual Assets are characterized by an Asset understood (impact of weather
consequences. (ARM) asset plan with more certainty -
Care Optimisation approach. etc.)
development of existing research to target
Distribution Yearly review as part of the
The complete investment plan is linked in the  A burst model to evaluate health of large trunk main renovation strategies
“hotspots” programme uses
Business Plan using Pioneer models. Deterioration analysis through separate pipes, backed up by pipe lab with
information received through The Asset Health Index is being developed
models (Pioneer) for distribution mains 5000 samples organised by pipe
operational modelling to validate the further.
and trunk mains- linked to risk and material and ground conditions.
Protection programme. Protection
service consequence against all asset An intention is to apply the UKWIR
Asset Care Optimisation workshops link has moved towards reduction in the
groups and modelling capability to research leading Asset Performance
non-infra defect data with criticality, number of properties affected well
rezone. Indicators work to provide structure and
future maintenance and long term below 2000 as originally planned.
linkage between APIs, MoS, performance
Production investment. Future targets move towards 900,
commitments and outcomes.
Above ground deterioration models by then 600 and 300 properties.
Customer impact is now increasing in The Asset Care Optimisation program will
equipment group linked to operational
importance especially with customer improve linkages with criticality and
maintenance.
hours as the new measure. maintenance.
FMECA via production asset hierarchy.
ARM risk management tool.
Dee Valley  Non Infrastructure: ARM (availability and  Non infrastructure: ARM  Drinking Water Safety Plan  Non infrastructure: ARM Uncertain in the context of the merger
water reliability studies for above ground assets (availability and reliability studies understanding of risk used to (availability and reliability with Severn Trent Water.
at key WTW. Provides an understanding of for above ground assets at key develop an understanding of studies for above ground assets
service impact of asset failure and WTW. Provides an understanding of critical asset risk and inform critical at key WTW. Provides an
therefore which are high priority. RAM service impact of asset failure and spares holding. understanding of service impact
includes data may inform deterioration therefore which are high priority. of asset failure and therefore
modelling. which are high priority.
 Service impacts assessed at water
 Infrastructure assets modelled in PIONEER resource zone level.  Drinking Water Safety Plan
(Tynemarch) which includes asset understanding of risk to develop
deterioration models. Used as long-term an understanding of critical
planning tool with integration into other asset risk and inform critical
corporate systems. Age and material are spares holding.
key drivers on deterioration modelling.
CH2M 96
Bristol Water  For PR14: SEAMS WILCO integrated asset Water pipes and SRs: WILCO failure  The strategic level water pipe Consequences are currently assessed Abandoning the SEAMS model as it is too
management tool was used to link impact and statistical deterioration model uses an average of in an ad hoc manner using a variety generic. Looking now for a benefits
investment, service and risk. models (For WTW had SEAMS process interruption/burst. of techniques including engineering quantification approach and a standard
level models, but BW never supply bad judgement, GIS, network models. method.
 For water mains, AH (burst rate, condition)  Mains rehab. Tool identifies no
water (would shut down works first)
enables estimation of failure probability and properties connected to the main. This is linked to the criticality process Quantification and realization is of benefit
What is modelled never happens and is
a direct link to interruptions based on (population affected) is also really important – how will the
always mitigated operationally. So, the  Have piloted the use of the bow-tie
observed events (strategic and local benefits be measured and demonstrated?
linkage for non-infra cannot be approach which links asset events Infra: BW said that they are doing a
analysis). Are we getting the expected benefits? This
validated from actual service failure through mitigations to service lot on criticality to understand which
is also a focus for development activities
 DOMS approach to manage water quality data impact. Future development of this assets leave us vulnerable. The
risks in the network. This allows us to approach will be considered as part MINERVA project is planned. BW said that going forwards they will
MEICA: risk assessment and cost of
identify flushing activities in the network of our continued business as usual Hydraulic models and resilience identify critical assets and consider more of
ownership
improvements modelling has been done in the past. a planned maintenance approach to
 Generally non-infra investment used a risk we do not follow a formal FMECA mitigate risks associated with breakdowns;
assessment process due to weak observed Non-infra: BW said that they have a
process the principles underpin our this also informs the spares policy
link between AH (e.g. equipment failures, criticality assessment for every non-
approach to the population of the risk
asset age, visual condition) and service. infra site belonging to Bristol Water  The intention is that criticality will
register and for some assets we have
(sources/treatment works/pumping define the routine maintenance,
 BW’s Information/Performance/ Risk/ held specific workshops investigating all
stations/service reservoirs etc.)9 and will define the spares policy
Solutions processes uses asset performance possible failure modes (e.g. Line of
which is used to understand the such that like for like tactical asset
data alongside other information i.e. Works)
service impact associated with replacement will either be
probability and consequence of asset changes in asset health performance based or fix on fail.
failure, to quantify the risks associated with
an asset. To support these new approaches we will
be developing more nuanced maintenance
performance measures.
Portsmouth Site surveys include a risk score to define when Deterioration modelling by WRc has Critical Links analysis provides a Criticality is identified through: Develop an assessment of the criticality for
Water asset repair or renewal gets into the investment been running for 3-4 AMP and so powerful tool as it looks at the impact non-infra equipment, which will come out
 Hands on review and
programme. This is ‘impact’ (on customers) x provides consistency. of number of customers likely to be of MARM. This is a complex due to the
HydroCo network models
‘priority’ = risk score, this is an informed score effected given the loss of all link across resilient nature of the network.
Mains selection uses a risk based
from engineering team. the system. The spatial platform  The new MISER model
strategy using spatial platform, with
identifies locations of ‘special need’ and highlights which
Promote replacement over repair – we won’t segment tracing of whole network to
‘vulnerable user’ and critical crossings combinations of failure are
sweat the critical assets. Criticality overshadows identify potential mains renewal groups,
(e.g. Motorway, Rail, River). critical, this leads onto work
health. these are attributed past operational
looking at the criticality of
data (bursts, customer contacts, repairs,
A Monte Carlo risk based analysis is applied to equipment on site,
special need users etc.). The scoring is
the non-infra Asset Register and is used to
supported by a critical links analysis to A recent MISER critically study
support the long term Asset Renewal Modelling.
identify customer impact from across demonstrated that all of our
the network given the isolation of ‘Valve single site source failures could
closure areas’ be sustained with exception on
the largest treatment works at
Farlington, however this has also
highlighted the impact of
multiple source failures
CH2M 97
South East Models are run through Pioneer: Above Ground = FMECA, Below Ground Tool that also uses MET weather office Within Maximo the criticality of SIRA (an Investment Optimization model)
Water = Statistical data to help make ‘predictions’ of asset assets is recorded and used this in takes outputs from Pioneer and other
• Asset relationships are tailored by
failure for operational preparedness. Pioneer modelling to determine the models and links the combined list to ODI
geography to reflect regional variation and local
potential impacts of failure. and impact of the project enabling accurate
factors that affect the ratios of asset failures to The consequences of asset failure
prioritization of schemes to meet the
customer events and also use knowledge from depends on asset type and a The consequences of asset failure
business needs at that time – this is in
local people to better understand their pipe combination of workshops, statistical depends on asset type and a
development with the intention of
connectivity to ensure the models data capture and tools such as ARM (to combination of workshops, statistical
capturing some of the less tangible cross
input asset specific risks) are used. data capture and tools such as ARM
• There is a large dataset for the asset group benefits.
(to input asset specific risks) are used
assessment of mains bursts and maintenance,
to identify these e.g.
indicating recent failures, which drives towards
replacement • GIS enables calculation of
properties , vulnerable
customers and public highways
affected by an incident
• Models such as our all mains
Hydraulic Models and MISER
enable the prioritization and
calculation of the criticality of
our above ground sites.
Where the data is not held within
Pioneer and/or to ensure the data is
accurate we use workshops with
operational teams to identify the
impacts of a failure.
Asset criticality as:
 Red: instantaneous impact of
failure
 Amber: partial impact of failure
 Green: no impact
CH2M 98
South Use CBA based on monetizing service risks Deterioration models for small diameter  Hydraulic modelling techniques Failing infrastructure is prioritized for It was noted that South Staffs also want to
Staffordshire (based on WTP) mains, reservoirs, boreholes. Burst have been used to look at investment above infrastructure that build WQ in more strongly into the asset
Water deterioration model includes climate impacts for some critical assets might fail. health related investment decision making.
 Clearest link is with the burst indicator
factors. Statistically linked to service
which can be readily forecast and linked to  Water resource planers and the   It would be beneficial to capture
and investment. Verified through
service (e.g. interruptions) generally and outage assessments may current knowledge of potential
observation
observationally using the GIS data provide some insights into the failure impacts
Non-infrastructure is less clear. Use risk effects of asset failures
 Mains bursts the key failure mode in South Staffs Water would like to set
based methods and asset condition is
deterioration model, this model determine up/participate in an asset management
not a primary driver. Driven through risk
the envelope of investment. Bottom up forum for WOCs to share best practice.
assessments :
scheme selection optimized on asset
criticality/ security of supply and growth  Statutory compliance
 Non infra uses risk approach to link  Health and safety
expenditure to service. AH in terms of
 Water quality
equipment failures does not correlate well
to service  Distribution impact
 South Staffs state that asset health decisions  Environment
at site level link into wider resources and
 Other service measures
programmes of work and they link together
to avoid revisiting issues.
Sutton & East Have updated the below ground asset models ‘Pioneer’ models used to explore a The consequence of any failure is  Criticality defines the
Surrey Water used in the PR14 price review to take account of number of scenarios that evaluate the logged to provide more realistic acceptable failure rate
the latest data various potential service impacts of measures with failure probability
 Interventions targeted on
asset deterioration. analysis.
Asset health of our above ground assets is critical assets
linked to water quality and service disruptions The cost of consequence is included in
 Critical pressure points are
Pioneer modelling.
Use the number of reactive (failure) work orders an emphasis for monitoring
and the consequence of their failure to link efforts
asset health, service and investment.
 Criticality of above ground
This information is input into modelling to assets is assessed by knock
inform investment planning. on effect.
CH2M 99
Appendix D: Inspection and monitoring
technology
CH2M 100
Water infrastructure
Our water pipes are made from a diversity of materials and can deteriorate due to the effects of
external corrosion or attack, as well as due to internal corrosion. Inspection and monitoring can
involve connecting sensors to the outside of the pipe, inserting devices into the pipe and taking
samples of the pipe and the conveyed water.
Devices inserted into a live main must not be deleterious to the quality of the conveyed water and it
must be ensured that they can be recovered from the pipe. In addition, devices inserted into the
water pipes must not damage coatings or disturb deposits. They must be capable of traveling
through the pipe and dealing with physical constraints such as tapers, joints and valves.
Water pipes present a technically challenging environment and pipeline inspection technology has
been subject to many years of research and development.
Monitoring
Flow and pressure are monitored across the network as a matter of routine and this supports
operational activity on the network. Water quality is monitored to ensure that outputs from
treatment works are satisfactory. These are not strictly speaking asset health measures, but
problems with the health of the assets (bursts, leaks, occlusions) could impact on quality and
quantity of conveyed water, so these are potential ‘symptoms’ of poor asset health.
South Staffordshire Water referred specifically to the use of Nephnet, Hydraclam and Chloroclam
devices.
Figure 9-1 Nephnet device
Nephnet is a portable, battery powered
turbidity monitor shown in Figure 3.3.
They can be used in temporary or
permanent installations with their 1 year
battery life. The system can provide
wireless alarms, and can be attached to
existing logging devices.
Hydraclam and Chloroclam are water
quality and pressure monitors. The web-
based user interface allows companies to
analyse their data through automated
visualisation tools.
Once sensor and diagnostic data is
uploaded to the cloud based system, nodes are mapped geographically, and data is presented as
time-series for each instrument. The data is also presented in different visual forms such as turbidity
event size, frequency, and duration. For chlorine, disinfectant effectiveness is presented using a
concentration size/class analysis tool.
As well as routine monitoring and control, companies including Severn Trent Water, South East
Water, South Staffordshire Water, South West Water, Thames Water and Bristol Water told us
about existing trials and plans to utilise acoustic monitoring of pressure transients to better
understand burst and leakage risks.
The Syrinix PipeMinder transient monitoring tool was mentioned specifically in several cases.
PipeMinder is stated as being a high-resolution pressure transient logging device for monitoring
anomalous behaviour in the network. Syranix produce tools for monitoring networks as well as trunk
mains.
CH2M 101
Routine inspection
It is probably fair to say that there is limited opportunity for routine inspection on water mains
because they are not accessible. We know that there have been trials in the past using a range of
devices including cameras that can be inserted into hydrants, and devices for sampling sediments in
trunk mains. There was also a device known as the ALFI (automatic liquid filtration instrument) that
could take timed water samples from a suitable taping or hydrant. As far as we know, these have
fallen out of use and no such devices or their relative merits were reported during our consultation.
Water quality spot sampling is carried out by all water companies and can be used to help identify
where corroding ferrous water mains and sediments present a water quality risk. Water quality
sampling is also used to identify sources of contamination which may arise due to poor integrity of
the network; issues with treatment or problems with service reservoirs.
In the 1990s, there was significant investment in water infrastructure to address the effects of the
aging and corroding network and pipe sampling (taking short cut outs) was commonly deployed to
help understand the state of the network and inform the rehabilitation programme. Today, our
networks are in a better state and service is good and unsurprisingly destructive testing to ascertain
state and condition of water pipes is greatly reduced.
Only Affinity Water and Southern Water mentioned distribution pipe condition assessment being
part of their approach, although several others including Portsmouth Water, South Staffordshire
Water and Thames Water talked about trunk mains condition assessment using primarily non-
destructive testing (NDT).
The reality is that the water companies have good asset performance data for their distribution
pipes. Burst rates and cluster/hotspot analysis (mapping failures and identifying the pipes that are
repeatedly failing) tend to be used to inform where distribution pipes may need replacing.
Distribution pipes tend to have a lower criticality and consequence of failure so this reactive type of
approach is considered effective.
Nonetheless, removing cut outs and undertaking physical and chemical analysis can be helpful in
confirming failure mechanisms and determining the most appropriate rehabilitation solution. For
ferrous pipes, shotblasting is sometimes used to assess corrosion (and corrosion rates if the pipe age
is known). Figure 3.4 shows corrosion after the pipe surface has been cleaned.
Techniques also exist for assessing the state of cementitious and plastic pipes. In particular, in the
case of plastic pipes, an historical problem has been the integrity of the joint (see Figure 3.5) and this
has tended to be a result of poor installation practice. This is a case of the asset health issue being
present when the asset was installed, rather than as a result of ‘asset aging’.
Figure 9-2 Condition assessment of ferrous pipes
CH2M 102
Figure 9-3 Failed electro-fusion joint
Testing of plastic pipes tends to involve taking
samples and undertaking mechanical testing;
however, we believe that there are also a number
of spectroscopic and chemical tests that can be
conducted to ascertain problems relating to
manufacture and composition.
We also know from experience that PVC pipe
installed in the 1970s sometimes was of inferior
quality and these pipes are prone to bursting
under cyclic pressure. These issues become
evident when the pipes fail and companies are
probably already aware of these historical
problem populations of pipes.
This issue was not raised during the consultation
so we assume that it is no longer considered a
major risk.
Figure 9-4 Phenolphthalein testing on cement pipes

Asbestos cement pipe was installed by some
companies in the 1950s and 1960s and notably in
soft water areas the cement leaches and
degrades. This severely affects the residual
strength of the pipe making it prone to bursts and
leaks. There are simple chemical tests that can be
used to measure the extent of lime leaching from
the pipe, which can be done on cut outs or core
samples. This is a simple chemical test and the
phenolphthalein is an indicator which turns bright
purple at high pH levels. The bright purple core in
Figure 3.6 shows sound cement (highly alkaline)
but the brown outer layer is lime depleted
cement and is very weak.
Trunk main assets tend to be far more critical with a high consequences of failure. The burst rates
are typically very low compared to distribution pipes. Trunk mains can be assessed using discrete
sampling. This involves digging down to the pipe and analysing the condition. The sample points
should be targeted based on an appreciation of risk factors, which can include soil agressivity and
proximity to population centres and roads. This can be quite an expensive approach and sometimes
the pipe is not readily accessible.
However, assuming access can be gained, the pipe surface can be prepared and internal and external
condition can be assessed. Figure 3.7 shows a trunk main section that has been cleaned and ready
for inspection.
CH2M 103
Figure 9-5 Trunk mains condition assessment
If the pipe has been shotblasted, the external pitting due to corrosion can be measured directly. The
internal corrosion can be assessed using ultrasonic thickness measurement devices.
We understand that the technology for measuring pit depths has progressed significantly and that
there are tools that enable easier and core comprehensive analysis compared to use of the discrete
point assessment devices.
Our view is that discrete point condition assessment can provide useful information on the asset’s
useful remaining life and failure probability. We acknowledge that this can be quite expensive and
that condition along a pipe can be quite variable so there is uncertainty in extrapolating the data and
in relating condition to failure probability.
However, in the case of a critical, high consequence asset, we believe that there is value in gathering
this type of data.
Advanced condition assessment
The UK has been at the forefront of developing in-line insertion leakage detection technology.
Sahara (Figure 3.8) is an acoustic hydrophone based technique, first developed and trialed by WRc
over 10 years ago and now available through Pure Technologies Ltd and WRc’s Assess & Address
business. This is reported to be a very precise leakage detection system and the pipe remains in
service. It is aimed at larger water mains and it has been used effectively by several UK water
companies. Pure also provide a tool called SmartBall, which is a free floating alternative for leak
detection. Both these techniques are available in the UK.
Figure 9-6 Sahara leak detection
CH2M 104
There have also been a number of investigations into devices that can be passed through water
mains that determine the structural condition of the pipe wall. These have tended to be based on
ultrasonic or magnetic flux leakage, though there have even been trials using X-ray techniques.
Figure 9-7 Magnetic flux leakage device
None of these have become established due to technical
constraints and costs. Outside the UK, some of these
techniques continue to be explored. For example,
magnetic flux leakage (Figure 3.9) devices offers the
advantage of:
 Precise comprehensive inspection;
 Identifies remaining wall thickness; and
 Identifies size and location of defects.
However, the disadvantages are:
 Pipe must be dewatered, and cleaned;
 Still emerging as technology for water pipelines;
 Ferrous, unlined pipes only (some exceptions);
and
 High cost.
Figure 9-8 Schematic of ePulse device

Another device that was
mentioned for UK trials by
Severn Trent Water was
Echologics ePulse shown in
Figure 3.10. This is one
example of an acoustic
pipeline condition
assessment application.
Echologics' condition
assessment technology
claims to quickly determine the structural strength of buried assets and optimise rehabilitation and
replacement programs. Echologics say that the ePulse® technology uses acoustic signals and
algorithms to assign a grade of good, moderate or poor based on the actual condition of the pipe
segment.
Acoustic sensors are attached to existing contact points, such as fire hydrants, valves or directly in
contact with a pipe. A sound wave is induced in the pipeline and travels along the pipe. The acoustic
sensors capture the time it takes the sound wave to travel between two sensor stations. The speed
at which the sound wave travels is dictated by the condition of the pipe wall.
As the sound wave travels, the water molecules push outward on the pipe wall. This causes a
microscopic flex on the pipe wall — and the greater the flex, the weaker the pipe. Through this
method, ePulse measures the actual strength of the pipe wall which is an indicator of pipe condition.
The device is reported to be suitable for distribution and trunk mains and for a variety of pipe
materials, (not plastic).
CH2M 105
Cleaning
Although mains cleaning is a rehabilitation option, rather than an asset health measurement
technology, several companies mentioned ‘ice-pigging’ as an example of an innovative technology
they had trialled. This technique was described as a large slurry of ice (a slush puppy) that is
propelled along the pipe and it removes sediments and deposits that can lead to discolouration
events. The density of the ice ball can be controlled to effect a variable level of cleaning.
Ice pigging claims to use less water and require less clean up compared to traditional underground
pipe flushing or other pigging techniques. It is claimed that the application leaves the pipes
undamaged after cleaning. The suppliers state that when the Ice Pig is inserted into the pipe, it
forms a semi-liquid form and adapts to the topography of the pipes as well as any changes in
diameter, making it an effective and ideal solution to common pigging problems.
It was interesting to note that the perceived results were quite varied. For example, Yorkshire Water
had found results from their trial to be rather mediocre, whereas Dee Valley cited good results.
We speculate that this may be due to local circumstances and the nature of the deposits being
removed. For example, we know that in the past efforts to clean water pipes using air scouring can
have detrimental results if the existing corrosion products are damaged and a fresh corrosion
surface is exposed.
Wastewater infrastructure
Monitoring
Sewer level monitoring has become well established for near real time monitoring of sewers.
Northumbrian Water cited EDM (event duration monitoring) coverage of 90% of their CSOs
(combined sewer overflows). EDM and sewer level monitoring have become established among the
WASCs.
Figure 9-9 Hawkeye installation
Sewer level monitoring can be used
to help understand issues relating to
sewer hydraulics and combined with
weather data can help anticipate
flooding events. Arguably, this is not
an asset health issue – it is one of
system capacity. However, we
believe that many of the companies
are also exploring whether level
changes during dry weather could
be indicative of developing blockages or collapses, and which is a more direct asset health issue.
Figure 3.11 shows a typical ‘Hawkeye’ installation used by Welsh Water.
Routine inspection
A well established and reliable inspection technology is the use of CCTV, coupled with a grading
classification based, for example, on WRc’s condition grading system. However CCTV inspections of
wastewater infrastructure assets are typically not undertaken routinely for all assets. The aim of
consistent and structured survey programmes is twofold:
1) identifying critical sewer assets that have a high propensity to fail or are already failing; and
2) to build up an understanding of time related deterioration and so inform predictive models.
Ideally, companies would proactively plan a CCTV inspection programme, with the highest risk
assets being surveyed most regularly, but with a plan for other assets to be surveyed. We cannot say
whether this approach is used by all, but we were given one example, where the company applies a
CH2M 106
criticality/consequence approach (e.g. sewers crossing railways and large diameter sewers) to
identify where survey and engineering assessment should be undertaken to better evaluate the risk
of failure and develop an investment case. They also undertake proactive CCTV monitoring of ‘high
risk critical sewers’, and use software solutions to identify changes over time. The use of automatic
recognition of sewer defects is being tested by a number of companies, with the intention to use
system based deterioration reporting over time.
As well as CCTV inspection, we are aware that simple non man-entry manhole photography and
video has been used to assess the sewer condition. This was not mentioned by companies during
this survey so we do not know if this approach is currently used.
Advanced condition inspection
As well as the more routine CCTV mentioned in the previous section there are a variety of other
options, including:
 Sonar;
 Laser scanning;
 Ground probing radar (GPR) and pipe penetrating radar (PPR); and
 Core sampling.
Figure 9-10 Sonar device

Sonar has been around a long time and is useful for
inspection through siphon sections, but relies on the
sewer being half full with liquid. Sonar was not mentioned
as a typical inspection device in our consultation.
Ground probing radar has also been around for some
time and can be used to locate buried pipes and
apparatus and to find voids around pipe joints caused by
infiltration/ exfiltration. Little was said regarding GPR, but we are aware that it has been used by UK
companies.
We do not believe pipe penetrating radar has been used in the UK, but it has been used in North
America to look from the inside out at voids around PCCP (pre-cast cylinder pipe), which is not a
material used in the UK.
Figure 9-11 Laser scanning

Laser scanning offers high precision scans of the pipe
interior which can facilitate design of sewer lining
options. The limitation (aside from cost) is that it only
functions above the water level and cannot distinguish
pipe materials.
A number of the UK companies mentioned the use of SewerBatt and Electroscan. There were mixed
opinions as to the effectiveness of these non-destructive methods of wastewater infrastructure pipe
CH2M 107
condition assessment, with some companies having had successful pilots, and others having less
success. Thames Water claim to have reasonable success with their low-cost temperature sensing,
but have found that SewerBatt and Electroscan do not give accurate enough results for use on asset
health indicators.
Figure 9-12 SewerBatt™ tool
SewerBatt™ uses acoustic signatures assess the
condition of drains and sewers. It claims to
identify defects, breaks or blockages using only
acoustic signals, and can record pipe diameter
and sediment depths. It also claims to uniquely
be able to see around or through bends and
blockages and identifies pipe junctions.
It’s reported to be easy to use with operators
just needing to insert a sonde on a telegraphic
rod to the bottom of the manhole chamber.
Outputs are provided as a simple 5 level categorisation. Because of its speed (100m covered in 3
minute) it’s recommended to be used as a first pass tool across the whole network to identify
category 4 or 5 sewers where further investigation by CCTV may be required.
Figure 9-13 Electroscan schematic signal output

Electroscan uses EM (electromagnetic) waves from a sonde pulled through the wastewater network
to identify defects. Electroscan claims to find all pipe defects that can leak. It shows the defect
location within inches, the size and the type. It’s a quantitative application supplying metrics that
describe the leak potential of a pipe. Electroscan metrics are reported to be suitable for comparing
the leak potential of sub-catchments, manhole-to-manhole pipe segments, or individual joints, and
outputs are provided in a GIS format.
Results are transmitted, processed, recorded and displayed on a smartphone app. The scan data is
exported to a cloud computing app for post-processing and re-transmission to the smartphone.
These techniques appear worthy of further investigation, though the mixed results reported by the
companies need to be considered.
Water and Wastewater Non-infrastructure

Above ground assets are perceived to be easier to monitor for asset health because they are more
accessible and visible (and audible). Most companies have, at some point, undertaken a condition
grading for their civil, mechanical and electrical assets. However, there has been controversy as to
the value of visual condition grading, therefore it is important to ensure that any inspection based
grading and assessment, is a reliable measure of asset failure probability or performance risk.
Improvements to the asset register and maintenance management system mean that it is
increasingly possible to track events at the asset component level and analyse failure events.
It is encouraging to hear that a number of water companies are seeking to better understand
maintenance and endeavor to establish the optimum balance of planned and reactive maintenance
for their non-infrastructure assets.
Monitoring and inspection
We have combined the inspection and monitoring sections, because the techniques and tools for
example ultrasonic detection, are used for both activities.
We were told that the key focus for managing complex assets is the output quality and that the
complexity and duplication on a typical plant means that the inherent redundancy enables the
CH2M 108
effects of asset component failures to be mitigated operationally. This may be true, but this does not
take account of the potential for multiple asset failures and it is important to ensure that the level of
redundancy, and the ability to detect and repair failures in a timely fashion is ensured. The idea of
‘run to fail’ for non-critical assets is acceptable, but it is important to understand the tradeoff
between asset health and resilience for complex assets.
Whilst there is potentially more work to be done in this area, we were nonetheless encouraged to
see that all companies collect asset component data and use this to inform maintenance activity.
A number of companies, including Affinity Water and Thames Water, reported using or planning on
using asset tagging (QR or barcodes) for non-infrastructure and claim that this reduces errors in
asset data, and facilitates more effective maintenance and operation reducing the risk of
deteriorating assets.
Asset tagging enables basic asset information to be associated with the asset being maintained or
inspected. It is possible to associate other information with the asset through the tag, including
maintenance activity and history and maintenance manuals. All of this could be made available at
the point of inspection to the maintenance engineer. However, we do not know what the “state of
the art” is in this area and detailed discussion of this topic did not arise during consultation.
In terms of control, Northumbrian Water and Severn Trent Water told us about intelligent control on
sewage pumping stations and Yorkshire Water about modulated pressure management on direct to
supply pumps.
Our consultation revealed that the UK water companies are aware of the typical condition
monitoring tools and techniques used for MEICA assets. These include:
 Vibration monitors;
 Thermal imaging/thermographic analysis;
 Infrared thermography for electrical assets;
 Acoustic sensors;
 Power usage analysis; and
 Oil analysis.
These devices are aimed at detecting signs of asset failure at an early stage (before it becomes a
problem) so that maintenance can be planned according to need and anticipated benefit. This is
condition based maintenance. The basic assumption is that early intervention can avoid costly repair
or failure and timely maintenance could extend asset life in a cost beneficial way.
Maintenance engineers often refer to the P-F curve when explaining this concept (shown in Figure
3.16).
It is worth noting that the following maintenance optimisation concepts are typically applied outside
the UK water sector for non-infrastructure assets and typically for MEICA assets. However, the
principles are generic and we (CH2M) believe that they could be developed for wider use.
The point P is the point through the asset life cycle where it becomes possible to detect the first
signs of failure. F is the point of ‘hard’ failure and the P – F interval is the window of opportunity to
try and detect and resolve the problem. Using basic detection methods e.g. visual inspection, means
that the P – F interval is typically very short. The damage done and likely cost of repair may mean
that it too late to intervene effectively. Advanced condition monitoring tools, such as the techniques
listed above, are aimed at increasing the P – F interval by enabling earlier detection and informing a
more optimal maintenance strategy.
We note that the companies we spoke to did not make reference to this concept, though RCM
(reliability centered maintenance) and FMEA (failure modes and effects analysis) and other
maintenance planning approaches were discussed.
CH2M 109
Figure 9-14 P-F curve example
We are aware that the IT systems for managing asset information that are now being implemented
by many water companies offer better capability for capturing failure data and maintenance
information at better levels of resolution than in the past. We believe that as the understanding of
asset component level failure and maintenance activity improves, there will be a good opportunity
for the UK water companies to explore and adopt approaches that will support maintenance
optimisation. An area that is often raised by maintenance specialists is that of failure characteristics
of equipment. There is a tendency in the UK water sector to assume that failure will follow a typical
‘bathtub’ curve but maintenance specialists know that this is not a typical failure pattern. The
following figure shows a number of typical patterns of failure against time and, as the water
companies build up experience of equipment failure characteristics it would be hoped that such
insights become clearer and maintenance planning can be tailored accordingly.
Figure 9-15 Failure curve examples
CH2M 110
The focus of attention during our consultation tended to be on MEICA assets (mechanical, electrical,
instrumentation, control and automation). There are also many civil assets to consider. There are
plenty of tools and techniques for assessing the condition of these assets, including a variety of
destructive and non-destructive techniques for condition grading and structural survey. There are
also specific corrosion models and tools available and then have been subject to research by UKWIR
(Deterioration Rates of Long-life Low Probability of Failure Assets – 11/WM/13/2) and a process to
inform risk based planning.
Figure 9-16 Physical inspection of concrete structure
Techniques that can be considered for
assessing the health of civil structures
include:
 Ultrasonic pit depth measurement
(ferrous structures);
 Chemical testing (concrete
structures);
 Strength testing and crush tests
from core samples;
 Movement and load sensors;
 Rebound testing (integrity of
concrete); and
 Cover depth tests (depth of
coverage of rebars).
Advanced condition monitoring and inspection
Some of the techniques already noted could be classed as advanced condition monitoring tools, in
that they enable more timely detection of the onset of failure. However, they are not panaceas and
some companies stated that the benefits were not necessarily clear enough to merit the costs and
CH2M 111
further investigation is required. During our conversations we also heard that there were other
examples of asset health monitoring being considered. For example:
 United utilities reported their condition based maintenance programme and real-time
monitoring of ‘run hours’;
 South East Water told us about NDT (non-destructive testing) on pump spindles;
 Severn Trent told us about future plans for contact tank bacti monitors; and
 Anglian water told us about a planned trial using their Shop Window process to examine
Artesis - Motor Condition Monitoring.
We have not heard of Artisis Motor Condition Monitoring, but the manufacturer states that the
approach works for any asset driven by an electric motor:
Figure 9-17 Motor Condition Monitoring
MCM monitors the condition of
equipment driven by an electric
motor, effectively using the motor
itself as a sophisticated transducer.
When connected to the power supply
of the motor, MCM carries out an
automatic self-learning process
during which the normal operating
condition (speed, load etc) of the
equipment is established.
MCM constantly takes measurements
and compares them with its
reference condition, in order to
assess the severity and type of any
developing fault.
When MCM detects a fault, it

presents the results of its analysis to
the user in a traffic light display. This
provides local staff with an
immediate indication that a problem
is developing. More detailed analytics
allows the operator to understand
the type and cause of failure and take
action to intervene.
Another technology that we were told was being used by some companies, including Yorkshire
Water was the use of drones/unmanned aerial vehicles to undertake visual inspections of remote or
inaccessible assets. They’re used widely for inspections of wind turbines, chimneys and stacks, and
dam walls. They can be fitted with standard 4K colour and high resolution video cameras, infrared
cameras, or even LIDAR.
CH2M 112

Targeted Review of Asset Health

Uploaded by

Copyright:

Available Formats

Targeted Review of Asset Health

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

Targeted Review of Asset Health

Uploaded by

Copyright:

Available Formats

FINAL REPORT

Targeted Review of Asset Health and

Date: September 2017

Understanding of asset health and risk to service

Asset health, expenditure planning and decision making

Understanding the views of the customer

Assuring our approach to asset health and maintaining service

Incentives and barriers

Lessons from international practice

1.2 Objectives and approach

1.2.1 Working definitions of asset health and its context

1.2.1 Measurement of asset health

1.2.2 The influence of asset health on expenditure planning and decision

1.2.3 Communication and the views of the customer

1.2.4 Asset health assurance

1.3 Report structure

2.2 The questions we asked and their purpose

2.3 What were we expecting to be told?

2.4 What were we told?

2.4.2 The relationship between ‘asset health’ and ‘resilience’

2.4.3 Benefits of the current measures of Asset Health and improvement

3.3 The questions we asked and their purpose

3.4 What we were expecting to be told

 Discoloured water  Reactive maintenance  Repeat sewer flooding  Reactive maintenance

3.5.2 Asset data

Expenditure Failure Asset Network

Probability of asset failure Asset failure Probability of

3.5.2.1 Infrastructure asset data

3.5.2.2 Non-infrastructure asset data

3.5.2.3 Condition data

3.5.3 Data quality

3.5.4 BIM (building information modelling)

3.5.5 Analysing and presenting data

3.5.6 How data is used

3.5.7 Inspection, monitoring, control and analysis

Trunk water mains

3.5.7.4 Treatment works and pumping stations

3.5.8 Innovation in technology

4.2 The questions we asked and their purpose

4.3 What we were expecting to be told

Level of Risk = (the likelihood that a service event

4.4 What we were told

4.4.1 Linking asset health to service

Figure 4-2: Elements of a risk-based approach to asset management

Figure 4-3: Illustrative relationship between asset health and resilience

4.4.4 The role of criticality

5.2 The questions we asked and their purpose

5.4 What we were told

Direct Dialogue No or Little Dialogue

5.4.2 Effective language to convey asset health concepts to customers

5.4.2.1 Asset health complexity and comprehension

5.4.2.2 Risk and risk perception

5.4.2.3 Targeting and appropriate communication

5.4.2.4 The use of scenarios

5.4.3 The value customers place on asset health

6 Water 2020 Customer engagement: lessons and opportunities, Southern Water

7 conjoint analysis to assess service attributes

6.2.1 Delivering on asset health commitments:

6.2.2 The main factors for success: