You are on page 1of 148

Smart Grid, Smart City:

Shaping Australias
Energy Future
National Cost Benefit Assessment

July 2014

Part 1 Overview of Australias energy sector


Part 2 The business case for smart grids in Australia
Part 3 Conclusions and Recommendations

7
149
237

Appendix 1
Appendix 2
Appendix 3
Appendix 4

289
349
437
483

Overview of Australias Energy Sector


Methodology
Table of Results
Customer Bill Impacts

ABOUT THE AUTHORS

ACKNOWLEDGEMENTS

In March 2013, a consortium of Australian-based


consultancy firms was commissioned to utilise the
Smart Grid, Smart City trial results and learnings to
develop an integrated cost benefit assessment for
smart grid technologies in a national context.

AEFI would like to thank Ausgrid and Australian


Government staff for their detailed input in providing
background information, Smart Grid, Smart City trial
data and feedback on different elements of the cost
benefit assessment and final report.

TheAEFI consulting consortium included:


Arup
Energeia
Frontier Economics
Institute for Sustainable Futures
(University of Technology Sydney)

This report was prepared by Arup in connection with


the Smart Grid, Smart City Program in 2014. Ittakes
into account our client's particular instructions,
requirements and priorities at the time. This report
is based on specific assumptions which have
been outlined in a number of supporting technical
documents published in conjunction with this report.
No responsibility is taken by Arup to any third party
including in relation to their use of the findings in this
report and third parties are advised to undertake their
own assessments as required to satisfy themselves
as to the adequacy or otherwise of the contents on
this report.

WRITING AND PROJECT


MANAGEMENT TEAM
Arup
Project Director Richard Sharp,
Project Manager Ranelle Cliff
Michelle Norris (lead writer), Ranelle Cliff,
Richard Sharp, Samuel Koci, Hugh Gardner
Energeia
Melanie Koerner, Ezra Beeman, Geoff Erder
Frontier Economics
Andrew Harpham, James Allen
Institute for Sustainable Futures
Edward Langham, Jenni Downes

1 Wayfinding
This document presents the findings of the national cost benefit assessment for smart grid technologies
and products based on the outcomes of the Smart Grid, Smart City program. The document is presented in
three distinct parts which, when combined, provide the reader with the required context and information to
understand the results and the corresponding conclusions and recommendations. In this sense, the document
is intended to be accessible to readers with a reasonably technical background but no experience with either
the electricity sector or smart grid technologies. Readers who are already familiar with the electricity sector
and the history of the Smart Grid, Smart City program, may wish to skip straight to the findings in later parts.

Overview of the national cost


benefitassessment.

Provides the background to the


Smart Grid, Smart City program and
a description of the technologies
and products trialled as part of the
Smart Grid, Smart City program.

Presents the cost benefit


modelling results. This part
alsodocuments the stakeholder
engagement process. The
modelling approach and
assumptions is detailed at
Appendix 2.

0 1 2

PART TWO:
The Business
Case for
Smart Grids
in Australia

PART ONE:
Smart Grid,
Smart City
Trials

Executive
Report

Start here if

Start here if

Start here if

Have limited time and require


an overview of the assessment.
You may need to refer to other
parts where you require further
information, particularly if you are
new to the energy sector or the
Smart Grid, Smart City program.

You have a good understanding of


the Australian energy sector, but
have had limited exposure to the
Smart Grid, SmartCityprogram.

You have detailed knowledge of


the Australian energy sector and
have been following the Smart
Grid, Smart City program. You
would like to understand the
detailed quantitative results of the
national cost benefit assessment.

Available from the ICH

Go to page 07

Go to page 149

Provides a strategic discussion of


the conclusions and associated
recommendations based on the
results of the national cost benefit
assessment. This part provides an
indicative assessment of how smart
grid technologies could be deployed
in Australia in the context of the
existing energy policy and regulatory
reform agenda and an appreciation
of the forward work program which
will be required to achieve the
benefitsidentified.

These are a series of technical


documents which are separate
to this national cost benefit
assessment report. These
technical compendia provide
detailed results and lessons
learned in deploying individual
smart grid technologies and
products over the course of the
Smart Grid, Smart City trials.
These documents can be
found on the Ausgrid Smart
Grid, Smart City Information
Clearing House (CH) http://ich.

3 4
PART THREE:
Conclusions and
recommendations

smartgridsmartcity.com.au/

Technical
Compendia

Start here if

Start here if

You have detailed knowledge of the


Australian energy sector and the Smart
Grid, Smart City program. You are
not so concerned with the technical
detail of the assessment but require
an understanding of the strategic
implications of the findings of the
assessment on the deployment of
smart grid technologies and products
inAustralia.

You require detailed technical


information about deploying smart
grid technologies and products
within a distribution network and/or
to individual customers.

Go to page 237

Available from the ICH

You are also likely to have a


detailed appreciation of electrical
engineering and the safety
standards and operations of an
Australian electricity network.

Part 1
Smart Grid,
Smart City Trials

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 7

Part 1 Contents
1
1

Wayfinding
The Smart Grid, Smart City Program

4
12

1.1

Introduction

12

1.2

Funding for the Smart Grid, Smart City Program

12

1.3

Location of smart grid trial activities

13

1.4

Scope of the Smart Grid, Smart City Program 17

1.5
Smart Grid, Smart City Information Clearing House

19

1.6

20

1.7 The Smart Grid, Smart City: National Cost Benefit Assessment Report

23

An introduction to transitioning Australias electricity


distribution networks

24

2.1

Defining a smart grid

24

2.2

Is a smart grid right for Australia?

25

2.3

Overview of Part One of this report

26

The global evolution towardssmart grids

27

3.1

29

Reporting the findings of the Smart Grid, Smart City Program

Global smart grid supporting policies and programs

3.2
Smart Grid, SmartCity work streams and trial technologies

32

Common platform study and NBN interoperability

33

4.1

What is a commonplatform?

33

4.2

The role of a common platform inasmart grid

33

4.3 The Common Platform and NBN Interoperability Project and the Smart Grid, Smart City
trial objectives and timelines

35

4.4
Smart Grid, Smart City common platform and NBN interoperability trials

36

4.5

38

5
6

Grid applications workstream an overview


Active Volt-Var Control(AVVC)

41
43

6.1

What are Active Volt-Var Control (AVVC) technologies?

43

6.2

AVVCs role in a smart grid

43

6.3

AVVC Smart Grid, Smart City Project objectives and timelines

44

6.4

Where are AVVC technologies currently deployed?

45

High level lessonslearned

6.5
Smart Grid, SmartCity AVVC trials

45

49

6.6

High level trialresults

8 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Fault Detection, Isolation and Restoration (FDIR)

52

7.1

What are fault detection, isolation and restoration (FDIR) technologies?

52

7.2

FDIR role in a smart grid

53

7.3

FDIR and the Smart Grid, Smart City trial objectives and timelines

54

7.4

Where are FDIR technologies currently deployed?

55

7.5
Smart Grid, SmartCity FDIR trials

56

7.6

58

Substation and feeder monitoring (SFM) technologies

61

8.1

What are substation and feeder monitoring (SFM) technologies?

61

8.2

SFM role in asmart grid

61

8.3

SFM and Smart Grid, Smart City trial objectives and timelines

62

8.4

Where are SFM technologies currently deployed?

64

High level trialresults

8.5
Smart Grid, SmartCity SFM trials

64

8.6

68

Wide Area Measurement (WAM)

73

9.1

What are Wide Area Measurement (WAM) technologies?

73

9.2

WAMs role in a smart grid

73

9.3 WAM Smart Grid, Smart City trial objectives and timelines

75

9.4

76

High level trial results

Where are WAM technologies currently deployed?

9.5
Smart Grid, SmartCity WAM trials

78

79

9.6

High level trialfindings

10 Electric Vehicles (EVs)

81

10.1 What are electric vehicle (EV) technologies?

81

10.2 The role of electric vehicles in a smart grid

82

10.3 The electric vehicle Smart Grid, Smart City Trial objectives and timelines

83

10.4 Where are electric vehicles currently deployed?

84

10.5
Smart Grid, SmartCity Electric Vehicle trials

88

10.6 High level trial findings

91

10.7 Further considerations for Electric Vehicle deployment for Australia

93

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 9

11 Smart Meter Infrastructure (SMI)

95

11.1

95

11.2 Smart meter infrastructures role ina smart grid

95

11.3 Smart meter infrastructure and the Smart Grid, Smart City trial objectives and timelines

96

11.4 Where is smart meter infrastructure currently deployed?

98

11.5
Smart Grid, Smart City smart meter infrastructure trials

102

103

What are smart meters and smart meter infrastructure (SMI) technologies?

11.6 High level trialfindings

12 Customer Applications

109

12.1 What are customer application technologies?

109

12.2 The role for customer applications in a smart grid

110

12.3 Customer applications objectives and timelines in the Smart Grid, Smart City Project

110

12.4 Where are customer application technologies currently deployed?

112

12.5
Smart Grid, Smart City customer application trials

113

127

12.6 High level trialresults

13 Distributed Generation and Distributed Storage (DGDS)

134

13.1 What are distribution generation and distributed storage technologies?

134

13.2 The role for distribution generation and distributed storage in a smart grid

136

13.3 Distribution generation and distributed storage objectives and timelines


in the Smart Grid, Smart City Project

137

13.4 Where are distribution generation and distributed storage technologies


are currently deployed?

139

13.5
Smart Grid, Smart City distribution generation and distributed storage trials

139

143

13.6 High level trial results

10 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 11

1 The Smart Grid,


Smart City Program
1.1 Introduction
The Smart Grid, Smart City Program was part of
the (then) Australian Governments Energy Efficiency
Initiative. Australian Government funding allocated to
the project was intended to develop an understanding
of the potential opportunities and issues in deploying a
21st century electricity grid that uses information and
communications to improve the efficiency of electricity
production, delivery and use. It was intended to deliver
Australias first demonstration and / or commercialscale rollout of smart grid technologies.

1.2 Funding for the


Smart Grid, Smart City
Program
The Ausgrid consortium was selected as the preferred
tenderer following an open tender process in which the
Australian Government received four applications.
In 2010, the Australian Government committed
approximately (AUD) $100 million in funding to the
Ausgrid consortium: Ausgrid (one of the three New
South Wales electricity distribution entities); IBM
Australia; GE Energy Australia; Grid Net; CSIRO;
Transgrid; EnergyAustralia; Landis+Gyr; Sydney
Water; Hunter Water; the University of Newcastle; the
University of Sydney; Lake Macquarie City Council and
the City of Newcastle. Ausgrid and its trial partners
provided additional funding resulting in total funding
for the Smart Grid, Smart City Program of more than
(AUD) $490 million in cash and in-kind support.
The Smart Grid, Smart City Program represents one
of the largest and most ambitious commercial-scale
trial deployments of smart grid infrastructure in
the world.

12 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

The Australian Government objectives of the Smart


Grid, Smart City Program were to:
Deploy a demonstration and/or commercial
scale rollout, as specified in the Guidelines1, that
informs a business case for key applications and
technologies of a smart grid
Build public and corporate awareness of the
economic and environmental benefits of smart
grids and obtain buy-in from industry
and customers
Gather robust information and data to inform
broader industry adoption of smart grid
applications across Australia
Investigate synergies with other infrastructure
(suchas gas and water) and the National
Broadband Network

1.3 Location of smart


grid trial activities
The trial was largely focussed on the greater
Newcastle and Sydney CBD areas with some
additional areas selected to test specific smart grid
applications focussed on long rural
network topologies.
The different trial work streams implemented
technologies and activities on one, multiple, or all
of the selected zones depending on their specific
requirements. The selection of appropriate geographic
locations for the trial was considered critical to
producing reliable data that could be accurately
extrapolated to assess the viability of a large scale
smart grid roll-out in Australia.
The greater Newcastle area was selected as one
of the focal points for the trial due to its mix of
regional and suburban characteristics that result in
representative geography, climate, socioeconomic and
demographic factors. The customer demographic and
socioeconomic indicators in Newcastle closely reflect
the demographic attributes of a typical Australian
city. Newcastles close relationship to the Australian
average for customer demographics is widely
accepted and has resulted in the city being used as
a test market for products and services prior to their
rollout across Australia in the past.

1 The Guidelines refer to the Smart Grid, Smart City Program


Funding Guidelines issues by the Australian Government
and agreed to by Ausgrid
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 13

The greater Newcastle and Sydney CBD areas, in


combination with the four additional zones, provided
a sound representation of the geographic, climate,
customer demographic and electricity network
characteristics of a number of regions throughout
Australia. It was felt that the trial would produce
nationally transferable results. Importantly, the
trial locations:
Included a mix of both urban and regional areas

Given the different network types captured by these


varying geographic, socioeconomic, demographic and
electricity network conditions, it was felt that these
areas were broadly representative of a large portion of
the Australian population.
Each of the Smart Grid, Smart City trial districts and
their specific characteristics are discussed in Table1-1
and shown on the map presented in Figure1-1.

Had demographic characteristics closely reflecting


the national average in terms of household income,
household occupancy, English proficiency, housing
types, tenure types, energy sources and appliance
stock. In addition, the trial locations contained
sufficient variability in these characteristics to test
their impact on measured outcomes
Had similar climates to a large portion of the
Australian population including both Climate Zones
5 and 62, in which 60-65 per cent of Australias
population is located
Demonstrated energy consumption patterns
reflective of the Australian population, including
both summer and winter peaks in energy demand
Had sufficient variability in topographic and
terrain characteristics to allow accurate testing of
alternative technologies
Demonstrated a range of different overhead and
underground network configurations, both radial
and meshed networks, and had rural, urban and
CBD characteristics typical of Australian networks
Contained several areas with high network
utilisation making them good locations to
demonstrate energy efficiency and demand
management initiatives.

2 The climate zones used for the purposes of the Smart Grid,
Smart City Project were based on those produced by the
Australian Building Codes Board, http://www.abcb.gov.au/
en/major-initiatives/energy-efficiency/climate-zone-maps last
published in December 2012.

14 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Table 1-1 Zones of the Smart Grid, Smart City trials


Location

Characteristic useful for the Smart Grid, Smart City trials

Greater Newcastle area

The greater Newcastle area was selected as the focal point for the trial, due to
its good mix of regional and suburban characteristics, representative geography,
climate, socioeconomic and demographic factors of the broader Australian
population. The customer demographics and socioeconomic indicators of the
area closely reflect the demographics of a typical Australian city.

Sydney CBD

Part of the City of Sydney Local Government Area (LGA), this area provided
additional high density residential buildings and large scale co-generation.

Ku-Ring-Gai area

Situated on the north shore of Sydney, this area provided additional testing of
high income demographics. The area also provided a high number of customers
with swimming pools (approximately 36 per cent in some areas of the Local
Government Area) for testing new products.

Newington area

Part of the Auburn LGA, this area provided a typical Western Sydney climate
zone and suburban environment for broad testing. The suburb added a high
multicultural population and contains the highest penetration of customer
photovoltaic energy generation, assisting in the trial of renewable generation and
storage applications.

Scone area

Part of the Upper Hunter LGA, this area provided additional rural characteristics
and a more extreme climate zone. The area provided a rural network to perform
end of feeder trials and much lower levels of internet use representative of more
rural geographies.

Nelson Bay area

An area north of Newcastle in NSW, included two zone transformers with


eight feeders exhibiting signs of moderate voltage constraint, supporting
around 10,000 customers spread across 210 distribution transformers. This
area is typical of older (brownfield) distribution zones with existing constraints
which made it amenable to testing the potential benefits of grid application
technologies.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 15

Figure 1-1 Map of Smart Grid, Smart City Trial Locations

16 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

1.4 Scope of the


Smart Grid, Smart City
Program
The aim of the Smart Grid, Smart City Program was to
gather robust information about the costs and benefits
of smart grids to help inform future decisions about
smart grid technologies by government, electricity
providers, technology suppliers and consumers
across Australia.

A broad group of participants and stakeholders


were engaged in the trial including residential and
business customers, network (Ausgrid) and retail
(EnergyAustralia) businesses, technology providers,
research institutions and utilities.
The schematic in Figure 1-2 provides a high level
overview of the main components of the trial and
network location in which the trial activities
were targeted.

The trial involved the development, deployment, trial


and analysis of smart grid infrastructure, products and
solutions. For the purpose of the trials, the smart grid
technologies and applications were categorised into
work streams described in Table 1-2.

Table 1-2 Work streams for the Smart Grid, Smart City trial

Name of work stream

Description of work stream

Grid Applications

Investigated the ability of grid-side monitoring and control technologies


to reduce network operating costs and support the future planning and
implementation of lower cost networks.

Customer Applications

Focused on residential electricity consumption, reliability, customer behaviour


and responses to feedback technologies and pricing models. It also included
an electric vehicle trial and investigations into the interoperability of electricity
metering with gas and water metering.

Distributed Generation
and Distributed Storage

Investigated the feasibility and potential benefits of distributed generation and


distributed storage within electricity grids.

Supporting Information
and Communication
Technology platforms
(Common Platforms)

Investigated the feasibility of various high-speed, reliable and secure data


communications network and associated IT systems which integrate with the
electrical distribution network. It also examined interoperability with the National
Broadband Network.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 17

Figure 1-2 Overview of Smart Grid, Smart City trial technologies and participants

18 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Consumer participation was identified as a key aspect


of the trial and therefore consumer involvement
included:
Participation in product offers in which various
combinations of customer applications, smart inhome devices and pricing incentives were offered
through network-focussed and retail-focussed trials
Participation in control groups for the trial, where
participant energy consumption and pattern of
consumption were collected to provide reference
data and baseline information for other consumers
participating in network and retail trials
More than 60 different community engagement
activities that included monthly public information
seminars and attendance at community events.
Trial participants were selected through a
pre-determined sampling framework. The framework
was designed to deliver statistically robust trial results
for the national business case and to capture the
impacts of various physical and socio-demographic
factors on energy consumption. Further details on
the sampling design framework are provided in the
Customer Applications Technical Compendium.

1.5 Smart Grid,


Smart City Information
Clearing House
The Smart Grid, Smart City Information Clearing House
(ICH) is a web portal designed to make reports and
data from each of the Smart Grid, Smart City projects
readily available to registered parties including electricity
industry members, governments, individual researchers,
suppliers and members of the public.
The Smart Grid, Smart City ICH provides registered
users with a user-driven online data selection,
visualisation and download facilities. The intent is that
registered users will be able to download and use the
Smart Grid, Smart City data and research to further
their own knowledge and research and in so doing
contribute to the global knowledge base of smart grids
and associated technologies. In addition users will be
able to access a public version of the model used by
AEFI in the cost benefit assessment.
The Smart Grid, Smart City reports are categorised
for easy access, searching and downloading including
this final report, trial technical compendiums and
supporting documentation.
This site provides registered users with access to
the Smart Grid, Smart City reports and data. Once
registered, users can browse the reports to gain a
detailed understanding of the trials and their findings
as well as access the public model (including key
functions and assumptions). Registered users can
analyse the Smart Grid, Smart City trial data by
downloading relevant sets of data from the ICH for
their own independent analysis or use the ICHs
business intelligent system to analyse the data online.
Registration to the Smart Grid, Smart City ICH is free
and available to all members of the community and
can be accessed at: https://ich.smartgridsmartcity.
com.au/
The Smart Grid, Smart City ICH will be maintained until
30 September 2015.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 19

1.6 Reporting the


findings of the Smart
Grid, Smart City
Program
The Smart Grid, Smart City Program, trial design
and reporting was governed to a large extent by
the objectives and priorities set by the Australian
Government. The objectives and priorities were a
condition of the funding agreement.
Significant effort was invested in the original design of
the Smart Grid, Smart City Program to endeavour to
meet the key objectives and priorities of the Australian
Government and its stakeholders. It was the intention
that the Smart Grid, Smart City trial elements produce
meaningful data that could be extrapolated in a
national context and utilised by industry and other
interested stakeholders.
With this in mind, the Australian Government specified
a range of activities, outcomes and data priorities
for each of the program / trial work streams. The
activities and data priorities were designed to prove (or
challenge) a number of trial hypotheses (assumptions).
The hypotheses were focussed on determining
whether different smart grid technologies could
achieve an economic or other benefit for Australian
electricity consumers.

In March 2013, a consortium of Australian-based


consultancy firms was commissioned to utilise the
Smart Grid, Smart City trial results and learnings
to develop an integrated business case for smart
grid technologies in a national context. The AEFI
consulting consortium included:
Arup
Energeia
Frontier Economics
Institute for Sustainable Futures
(University of Technology Sydney)
While undertaking an integrated financial and
economic assessment and developing conclusions
and recommendations, AEFI considered the broad
requirements of the program, including:
The key questions and assumptions that the
Australian Government were seeking to validate,
or challenge
Existing energy market reforms under investigation
or in progress
Australian Government energy policies, programs
and its strategic intent for the energy sector and
energy consumers
Stakeholder priorities and expectations including
those identified during the stakeholder engagement
processes completed as part of AEFIs assessment
of the Smart Grid, Smart City Program (discussed
in Part Two of this report)

Progress reports have been published by the


Smart Grid, Smart City project team over the past
three years. The progress reports have allowed the
Australian Government to oversee and assess trial
progress against the contractual obligations for the
duration of the project. The Australian Government has
made available the final series of reports describing
the results of all of the Smart Grid, Smart City trial
analysis and these can be found on the Smart Grid,
Smart City Information Clearing House at https://ich.
smartgridsmartcity.com.au/

20 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

In addition, a clear expectation of the Smart Grid,


Smart City Program was that there would be
significant effort focussed on presenting the results of
the trial for different audiences. To this end, there are
several levels of reporting being delivered as part of
completion of the program:
The National Cost Benefit Assessment Report
(this document) which provides stakeholders
with an independent assessment of the potential
integrated national cost benefit assessment.
Data (predominantly) produced by the Smart
Grid, Smart City trials was used to determine
the potential national net benefits of deploying
economically efficient smart grid devices,
customer feedback technologies and dynamic
(cost reflective) electricity tariffs. The integrated
methodology used to develop the analysis and
costings within this report ensures that there is no
double counting of benefits provided by two or
more technologies and any under or overestimation
of the value proposition achieved from technology
interdependencies is minimised.
The Technical Compendia prepared by Ausgrid
and partners that provide a detailed overview of
the results for each of the work streams. These
documents are technically complex and aimed
at an audience that is familiar with energy sector
terminology and technology. The intended
audience includes network operators, energy
retailers, technology providers, universities,
regulators, governments and energy sector
personnel. In addition these documents are
intended to have global reach and contribute to
the global knowledge base of smart grids and the
associated technologies.

The Modelling Inputs Compendium is also


part of this suite of technical documents. This
compendium provides detailed information as
to where data and information generated from
the Smart Grid, Smart City trials have been used
either as direct inputs or to validate and inform the
modelling inputs to the national cost
benefit assessment.
The Modelling Inputs Compendium also provides
a series of stand alone cost benefit assessments
for individual smart grid technologies which could
be applied in the instance that only one device or
technology (or a series of discrete technologies
or devices which do not have any interaction) is
deployed in the Australian context. These stand
alone cost benefit assessments represent a less
sophisticated approach than the integrated cost
benefit assessment, but remain a valid estimate
ofcosts and benefits under some circumstances.
Supporting documentation prepared by Ausgrid
and partners that provides the detailed information
and data upon which the Technical Compendiums
are based.
An Information Clearing House (ICH) that provides
the communication mechanism for the data,
analytic tools and results from the Smart Grid,
Smart City Program and will be maintained until
30September 2015. This site can be found at
https://ich.smartgridsmartcity.com.au/
The breadth of the subject areas investigated as part
of the Smart Grid, Smart City Program is shown in
Figure 1-3.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 21

Figure 1-3 Structured reporting and access to information for the Smart Grid, Smart City trial

National Cost Benefit Assessment Report

Part Two

Part One

Smart Grid, Smart City trials

Part Three

The business case for


smart grids in Australia

Conclusions and
recommendations

Supporting Documents

22 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Modeling Inputs
Compendium

Wide Area
Measurement

Smart Meter
Infrastructure

Active Volt
VAr Control

Common Platform
Study

Fault
Detection
Isolation and
Restoration

Electric
Vehicles

Substation
and Feeder
Monitoring

Customer
Applications

Grid Applications

Distributed
Generation and
Storage Applications

Technical Compendia

1.7 The Smart Grid,


Smart City: National
Cost Benefit Assessment
Report
This is the final report to the Australian Government
for the Smart Grid, Smart City Program and differs
from previous progress reports. Importantly, it is an
independent assessment that provides the analysis
and discussion necessary to form:
The business case (potential net economic
benefit) associated with a future implementation of
individual smart grid technologies, compared to a
business as usual approach using the results to the
integrated assessment3
The business case (potential net economic benefit)
associated with a future implementation of an
optimised group of smart grid technologies (an
integrated assessment), compared to a business
as usual approach
The business case (potential net economic benefit)
for smart grid technologies for electricity networks
and electricity generators (on a state-by-state4 basis)

This report is broken into a number of sections


including:
Part One An overview of the Smart Grid, Smart
City Program and the high level findings of the
individual technology and customer trials
Part Two A cost benefit assessment for the
deployment of a national smart grid
across Australia
Part Three A discussion on the potential
benefits, barriers and opportunities to implement
a smart grid in Australia including the independent
conclusions and recommendations from
this assessment
Appendix One Provides an overview of
Australias energy sector including the policy,
regulatory and investment drivers
Appendix Two Describes the cost benefit
assessment methodology that AEFI used to
complete the cost benefit assessment
Appendix Three Tables of results for each of
the three macroeconomic scenarios and for each
Australian state
Appendix Four Customer electricity bill impacts

How the costs and benefits accrue to electricity


customers and the potential impact on retail
electricity prices for a smart grid scenario
compared with a business as usual approach.

3 Reporting the analysis in this way prevents double


counting of benefits which can occur if assessments
for different technologies are competed individually
4  State assessment has been limited to Western
Australia, South Australia, Queensland, New South
Wales, Tasmania and Victoria.
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 23

2 An introduction to
transitioning Australias
electricity distribution networks
2.1 Defining a smart grid
This need for more advanced decision-making
abilities for network operators is reasonably consistent
throughout the world. As a consequence a range of
innovative information and communication systems
have been developed and are now driving an electricity
grid modernisation often termed a smart grid.

Smart grid technologies offer the potential to:

Standards Australia define a smart grid as an


electricity system incorporating electricity and
communications networks, that can intelligently
integrate the actions of parties connected to it.

More efficiently utilise labour and materials

It also notes that The implicit assumption in smart


grids is that energy may be generated and used
anywhere on the grid, the behaviour of both generators
and users is much more variable and less predictable,
and that both will continuously react to the other5.

Better predict electricity supply and demand at


specific locations in the grid
Continuously monitor the condition of the grid and
major assets
To dynamically reconfigure the network
Smart grid technologies also provide the opportunity
for network operators to interact with customers
to actively manage demand on different parts of
its network. Standards Australia in its Smart Grid
Vocabulary also notes that the boundary between
a conventional grid and a smart grid is not easily
identified:
There is no absolute level of capability or set of
attributes which marks the transition of a specific
electricity network from conventional to smart.
Grids may become more smart over time as they
acquire more of the transactional, operational and
technical features defined in this Standard.
In general, a smart grid is distinguished from a
conventional grid by the deployment of enhanced
information and communications systems,
in order to manage equipment in a way that
achieves outcomes remotely, automatically, more
rapidly and more precisely6.

5 Standards Australia AS5711 Australian Standard: Smart Grid


Vocabulary, 2013 ISBN 978 1 74342 585 5

6 Standards Australia AS5711 Australian Standard: Smart Grid


Vocabulary, 2013 ISBN 978 1 74342 585

24 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

2.2 Is a smart grid


right for Australia?
Smart grid technology has the potential to provide
networks and customers alike with information and
automation processes to optimise investment and
improve the environmental performance and economic
efficiency of the system. In this sense, the smart grid
marks a transition of the existing grid from a physical
electricity distribution system to a transaction enabler,
underpinned by distribution of information and data to
all stakeholders.

It is these questions and others defined in the Smart


Grid, Smart City Program objectives that Ausgrid, its
trial partners and the Australian Government sought to
answer during the trials. This independent cost benefit
assessment contributes to answering a number of
these questions.

Current smart grid literature predominantly supports


an integrated approach to deploying smart grid
technologies as being optimal to deliver a range of
benefits including:
Assisting in the reduction of transmission and
distribution losses
Optimising the use of existing infrastructure by helping
to regulate power flows and meet peak demand
Accommodating significant volumes of
decentralised and renewable energy into the grid
and improving energy efficiency7
Consistent with other smart grid technology
trials and deployments across the world, the key
questions include:
Which technologies work best under Australian
network conditions?
Which technologies offer the best return on
investment in Australia?
What is the best way to integrate these
technologies in an existing electricity network?
What is the best way to deploy the
technologies?

7 World Energy Council, Smart grids: best practice


fundamentals for a modern energy system, 2012.
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 25

2.3 Overview of Part


One of this report
This part of the report examines the evolution of smart
grid technology funding over the past 20 years, the
drivers for global smart grid programs and a high level
description of the Smart Grid, Smart City Program
technologies, trial objectives and results.
Given the complexity of many aspects of the
technology trials this part of the report only provides
ahigh level overview of the key trial elements
and findings.

In particular the Modelling Inputs Compendium


demonstrates where data and information generated
from the Smart Grid, Smart City trials have been used
either as direct inputs or to validate and inform the
modelling inputs to the national cost
benefit assessment.
The Modelling Inputs Compendium should be read
in conjunction with the business case methodology
presented in Appendix Two and the national cost
benefit assessment results presented in Part Two
ofthis report.

2.3.1 Additional information


sources
More comprehensive and technical discussion on
each of the trials described in this part of the report
can be found in a series of Technical Compendia
and in supporting documentation developed by
Ausgrid which is available from the Smart Grid,
Smart City Information Clearing House at https://ich.
smartgridsmartcity.com.au/
There is one Technical Compendium for each of
the Smart Grid, Smart City Projects as well as a
Modelling Inputs Compendium 8 which describes the
quantitative inputs used to determine the net national
benefits from an integrated approach to deploying
smart grid devices, consumer feedback technologies
and dynamic tariffs. In addition, the Modelling Inputs
Compendium provides a series of stand alone
assessments of individual smart grid technologies and
devices (as previously described in Section 1.6).

8 The Modelling Information Compendium was developed by


Energeia with input from Ausgrid

26 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

3 The global evolution


towardssmart grids
Over the past five years there have been significant
funds and program funding sources made available for
smart grid technology demonstration and deployment
across the world. These programs vary considerably
in scope and objectives with some focussed on
specific technologies, whilst others provided broader
perspectives. Based on publicly available information,
some of the key global programs have been listed
in Table 1-3.

Table 1-3 Summary of global existing smart grid trials including key technologies and Government funding (USD)

Technology Focus
Country/

Australia

National Energy

China

Strong and

EU

FP7

France

Region

Program

Smart Grid

Efficiency

Start

End

SMI

DG

DS

EV

SFM

FDIR

AVVC

Programme

2013

$100
million

2011

2020

$100
billion

2007

2013

$520
million

2010

2014

2010

2015

$1.2
billion

2011

2013

$50
million

2009

2015

$200
million

2010

2015

$480
million

2009

2015

$3.4
billion

2009

2015

$325
million

(CAN)

Smart Grid

Korea

Koreas Smart

UK

Ontario

Demonstration

South

Japan

Smart
Community

Fund

Grid Project
Low Carbon
Networks Fund

US

Smart Grid
Investment
Grants

US

Smart Grid
Demonstration

(USD)

2010

Initiative

Smart Grid

Funding

$600
million

Program

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 27

This information shows that the majority of current


funding sources are scheduled to conclude over the
next two years, with the exception being China where
the Strong and Smart Grid Program is providing
investment funding of (USD) $100 billion through to
2020. In contrast to other smart grid programs and
trials, the Chinese program provides funding for a
broader range of (conventional) grid infrastructure
and projects than those associated with common
smart grid technology definitions. The range of smart
grid technologies being deployed as part of global
funding programs (with the exception of the Smart
Grid Programme in France) are very similar to those
investigated by the Smart Grid, Smart City Program9.
The key difference in many cases is that the programs
are more narrowly focussed compared to the breadth
of smart grid technologies and customer applications
tested as part of the Smart Grid, Smart City Program.

Figure 1-4 shows the current committed investment


for smart grid trials internationally, excluding Chinas
Strong and Smart Grid Program, grouped by three
global regions. It shows that there is a clear drop
in global funding when current programs in the US
conclude in 2015. The European Union FP7 program
will transition to FP8 and extend to 2020. Some
continuing investment in parts of Asia (Japan, Korea
and Taiwan) will also remain in place until 2020. There
remains the potential for new funding programs to be
announced following the assessment of trial results
from the current program areas.

1.80

8.00

1.60

7.00

1.40

6.00

Annual (USD billion)

1.20

5.00

1.00
4.00
0.80
3.00

0.60

2.00

0.40

1.00

0.20
0.00

2013

2014

2015

NA

2016

EU

2017

JTK

2018

Cumulative

9 Funding for the Smart Grid, Smart City Program is


described in section 1.2

28 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

2019

2020

Cumulative (USD billion)

Figure 1-4 Commitment investment smart grid technology by region (USD $billion)

Public access to comprehensive data from these


trials is inconsistent and when it is available, direct
comparison of trial results is generally difficult. In some
cases, it is especially challenging to compare the
outcomes of smart grid technology trials around the
world with Australian results. Where possible, Ausgrid
have discussed comparisons of some international
trial technology results in each of the Technical
Compendia. The Technical Compendia can be found
on the Smart Grid, Smart City Information Clearing
House at https://ich.smartgridsmartcity.com.au/
Globally, the drive for investment in smart grid
technologies has been evolving since the late 1980s.
Early investment in smart grid technologies was
sponsored by industry, with improving operational
efficiencies (reducing costs) the focus of the
majority of trials. Later trials from around 2008 have
predominantly been sponsored by governments and
have focussed on peak demand reduction, energy
efficiency, reduction in greenhouse gas intensity, and
improving grid reliability.
In some regions of the world there has also been a
focus on the potential to create jobs across the supply
chain from smart grid manufacturing through to
technology supply and highly skilled jobs within utilities
(networks)10,11,12,13,14.

3.1 Global smart grid


supporting policies and
programs
In total, international governments are investing
over (USD) $3.2 billion in smart grid development
programs from 2013 through to 2015 (this excludes
the Chinese government investment). This
investment demonstrates the relative importance that
governments are placing on the future implementation
of smart grid technologies.
There has been a number of smart grid roadmaps
published globally. The Australian Government as part
of its commitment within the Smart Grid, Smart City
Program funded the development of Australias first
Smart Grid Standards Roadmap which was released
in June 2012.
Additionally the Future Grid Forum which released
its report Change and Choice in December 2013
was funded by industry and the CSIRO. This report
proposed a number of possible scenarios for future
grid structures, regulation and business models.
It also identified a number of policy and energy
market challenges that may present barriers to the
deployment of technologies within the Australian
electricity sector.
A review of publicly available literature suggests that a
number of countries have explicit smart grid policies
that include:
Formal targets
Smart grid technology programs including research
and development grants

10 KEMA, The U.S. smart grid revolution, KEMAs perspectives


for job creation, prepared for the GridWise Alliance,
December 2008.
11 Climate Solutions, Powering up the smart grid, A Northwest
initiative for job creation, energy security and clean,
affordable energy, July 2005
12 Lowe M, Fan, H, Gereffi, G (on behalf of Environmental
Defence Fund), U.S. Smart Grid: finding new ways to curt
carbon and create jobs, April 2011.
13 Global Smart Grid Federation, Global Smart Grid Federation
Report 2012
14 Ministry of Knowledge Economy, Koreas smart grid
policy and deployment, presentation from the Conference
on Integration 2012, http://www.conference-onintegration-2012.com/fileadmin/user_upload_COI-2012/
RE_PDF/Jung_Yonghun.pdf

Regulations including customer protection (security


and privacy) and communication and integration
standards and frameworks (many of these are still
under development)
Technology deployment standards (such as
minimal functional specification)
Distributed generation feed-in-tariffs
A comprehensive summary of these policies and
programs is provided in Table 1-4.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 29

30 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Table 1-4 Summary of global smart grid policies, targets and regulatory frameworks

Jurisdiction

EU

Smart
Grid
Roadmap

Targets

Programs

Voluntary
DG

EV

DS

Mandatory R&D
Grants
SMI

Denmark

Finland

France

Germany

Regulations
Standards
DG FiTs

Opt-out

Privacy

Overall

SMI

Security

Data

Italy
Netherlands

Spain

Sweden

United
Kingdom

Rest of the
World
Australia

Canada

British
Columbia

Ontario

China

Japan

New Zealand

South Korea

US

Texas

California

DG= distributed generation, EV= electric vehicle, DS = distributed storage, SMI = smart meter infrastructure (smart meters), DG FiTs = distributed generation
feed-in-tariffs

Based on the information presented in Table 1-4, it is


clear that the most common tool to facilitate local trials
and investments in smart grid technologies globally
has been government sponsored research and
development grants.
In addition, it is clear that different parts of the world
have focussed on differing aspects of smart grids,
based on local needs and opportunities. These
aspects range from simply forming a position on the
outcomes that smart grids can provide, through to
targets and mandating deployment of technologies
(such as smart meters). California has the most
comprehensive of smart grid deployment mandates in
the world enacted by Senate Bill (SB) 17 (2009):
it is the policy of the state to modernize the states
electrical transmission and distribution system to
maintain safe, reliable, efficient and secure electrical
services, with infrastructure that can meet future
growth in demand and achieve various goals aimed
at cleaner energy future, energy efficiency and more
engaged customers.
California has mandated that utilities submit a
smart grid proposal to the California Public Utilities
Commission (CPUC) for approval. In addition,
California has one of the most ambitious energy
storage targets in the world - 1.3 GW of energy storage
on the grid by 2020.

To date, Australian research funding has been directed


towards examination of the cost effectiveness of smart
grid technologies (Smart Grid, Smart City Program).
Other funding programs have been designed to
incentivise distributed generation (both federal and
state-based schemes).
An alternative approach to driving smart grid and
other grid-focussed innovation was developed by
the United Kingdom (UK), which in 2005, introduced
a research and development incentive model that
led to a six-fold annual increase in expenditure in the
distribution sector. The Innovation Funding Incentive
(IFI) mechanism was designed to encourage gas
and electricity transmission and distribution network
operators to apply innovation in the way they pursued
the technical development of their networks, i.e.
the aspects of network design, operation and
maintenance. The principal objective of IFI was to
deliver benefits to consumers by enhancing efficiency
in network operating costs and capital expenditure.
The two main drivers for providing incentives were
the increasing need for investment in end of life
asset renewal and an increase in small embedded
(distributed) generation.
In addition to the IFI, the UK Government has
established a 500 million Low Carbon Networks
Fund for innovative distribution network trials over the
period 2010-2015. By July 2013, it had awarded 120
million to a range of projects including those aiming
to develop a greater understanding of smart grid
technology, electric vehicles, heat pumps, micro and
local generation, Demand Side Management (DSM)
and the deployment of smart meters.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 31

3.2 Smart Grid,


SmartCity work
streams and trial
technologies
There were a wide range of grid and customer-facing
technologies trialled as part of the Smart Grid, Smart
City Program and each of these have been described
in the following sections. There were four work streams
within the program:
Common platform and NBN interoperability
Grid application technologies
Distributed generation and distributed storage
Customer application technologies
Each of the four work streams, along with the
individual smart grid technologies and trial results are
described in the following sections of this report.

32 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

4 Common platform study


and NBN interoperability
4.1 What is a
commonplatform?
For a distribution network service provider
(DNSP), implementing a smart grid is a project of
unprecedented scope and complexity. Every part
of the providers business is impacted including all
parts of the electrical network (the poles and wires,
substations), the customers premises, IT and other
back-office systems, network design, planning and
operations, and business operations.
The smart grid requires the integration of the electrical
distribution network with a high-speed, reliable and
secure data communications network and associated
IT systems. An enormous amount of data, much
of it real-time data streams, must be managed,
stored, backed up, correlated and presented to each
business stakeholder in a way appropriate for them
to do their job effectively. All of this must happen in
an environment where the security of the network is
guaranteed, and individuals privacy is respected.
Given the electrical network is constantly evolving
to meet the changing needs of its customers, the
smart grid will never be static - new smart grid
applications will continue to emerge that will need to
be incorporated into the system.
A common communications, data acquisition and data
presentation infrastructure that can accommodate
all smart grid applications dramatically reduces the
incremental cost of each additional application,
simplifies the process of implementation and,
most importantly, allows for advanced automation
applications that have visibility of data gathered
from right across the network. A common platform
is also an enabler for an integrated smart grid
approach because it allows the integration of different
technologies to be assessed and verified by advanced
systems and analytics.

4.2 The role of a


common platform
inasmart grid
Traditionally the deployment of new technology to
support the electrical distribution network has been
treated as a stand-alone project with dedicated
sensors deployed in the field, communication
network built or communications service purchased,
and dedicated data aggregation, processing and
presentation systems established. Resources are
shared only if they are suitable, have the necessary
spare capacity, and their owners are willing to do so.
Issues such as security and privacy were handled on a
case-by-case basis within corporate guidelines.
A common communications, data acquisition and data
presentation infrastructure that can accommodate
all smart grid applications will dramatically reduce
the incremental cost of each additional application,
simplify the process of implementation and,
most importantly, allow for advanced automation
applications that have visibility of data gathered from
right across the network.
Figure 1-5 illustrates a conceptual approach that
divides the smart grid into three layers, each one
enabled by the systems in the layer below. Across
the bottom row are the common platforms that are
shared by all smart grid applications. Being the most
expensive elements of a smart grid, these platforms
should be shared by the entire organisation, and be
viewed as a key corporate asset.
The common platforms enable the middle layer, which
is a series of applications that provide visibility of the
state of the network and behaviour of customers in
real time. It also provides a capability to change the
state and behaviour of the network or load through
actuation devices distributed across the network and
in customers premises.
The final layer, an additional value-add of the smart
grid, is the advanced automation layer. This is a series
of applications that make use of the network visibility to
intelligently manage various aspects of the network.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 33

Figure 1-5 Proposed smart grid common platform layers

Advanced
Automation

Wide Area
Control

AVVC

FDIR

Dynamic
Ratings

Demand
Response

Pricing &
Feedback

Network
State Viability
& Actuation
Capability

Transmission
Monitoring

Substation
Monitoring

Distribution
Monitoring

Distribution
Control

Wind Area
Measurement

Smart Meter
Infrastructure

Common
Platform

IT
Infrastructure

Communications
Infrastructure

Security
Architecture

34 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Operational
Model

Design
Standards

4.3 The Common


Platform and NBN
Interoperability
Project and the Smart
Grid, Smart City trial
objectives and timelines
Unlike the rest of the Smart Grid, Smart City trials
which evaluated specific smart grid technologies,
the purpose of this work stream was to provide an
evaluation of how successfully these trials were able
to leverage the common platform developed as part
of the Smart Grid, Smart City Program. In addition this
work stream sought to identify any learnings that could
inform the design and implementation of a smart grid
common platform for other network businesses.
This was done by describing the elements of the
common platform developed for the Smart Grid,
Smart City Program, identifying the common platform
components used in each of the trials, and then
critically evaluating the suitability and performance
of those components to support the smart grid
application.

The project aimed to evaluate the requirements for a


smart grid common ICT platform in the following areas:
The effectiveness of current standards and the
degree to which they permit interoperability of the
wide variety of systems and devices making up the
smart grid
The data security and information privacy
implications of the common platform
The processing, management, correlation and
storage of large amounts of data available from the
smart grid
The data centres, disaster recovery, backup, and
monitoring systems necessary to support the
common platform
In addition, this study assessed the potential for the
National Broadband Network (NBN), which was being
rolled out in Australia at the time of the trial, to provide
communication services to field-based smart grid
infrastructure. The work stream included:
The evaluation of options for using the NBN for
smart grid applications and developed reference
architecture
Completion of detailed traffic modelling of the
reference architecture to evaluate its viability
and scalability
Development of an economic model to
demonstrate the economic feasibility of
usingtheNBN for smart grid field area network
(FAN)15 applications, and compare this to
several alternatives

15  A field area network (FAN) is a network enabling


communications for the distribution network
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 35

4.4 Smart Grid,


Smart City common
platform and NBN
interoperability trials
The Common Platform Study and NBN Interoperability
Technical Compendium provide an extensive
discussion on each of the areas that the study
investigated and the issues and opportunities identified
through the Smart Grid, Smart City Common Platform
study. This can be found on the Smart Grid, Smart
City Information Clearing House website which can be
accessed from https://ich.smartgridsmartcity.com.au/

4.4.1 Trial design and deployment


Common platform
The smart grid common platform developed for
the Smart Grid, Smart City Program consisted of
a communications network providing connectivity
between the various elements of the smart grid,
together with supporting IT infrastructure (as described
previously in Figure 1-5). A smart grid common
platform must meet the following requirements:
It must support a broad range of technology
options including legacy services and current smart
grid technologies, and (as far as possible) new
technologies that will emerge in coming years
All platform elements should be standards-based
and employ well-defined interfaces that support an
appropriate level of security
Hardware and software should be based on
proven, enterprise-grade architectures and
technologies
The architecture should be highly scalable, secure
and suitable for deployment lifetimes in the order
of 5 to 15 years, depending on their location within
the network

The common platform, like the electrical distribution


network, must be highly secure and reliable reliability requires the probability of intrinsic faults or
their negative impacts be minimised, whilst security
describes the ability of the system to resist unwanted
accidental or deliberate actions from external sources.
To be effective, security and reliability must be
designed into the common platform from the outset.
Further, the large amount of data contained in the
smart grid carries with it the potential to infringe on the
privacy of individuals and companies. The US National
Institute of Standards and Technology (NIST) report
7628 provides guidelines for smart grid security
and privacy16.
The smart grid common platform developed for
the Smart Grid, Smart City Program consisted of
a communications network providing connectivity
between the different elements of the smart grid,
together with supporting IT infrastructure.
The Smart Grid, Smart City common platform
communications network consisted of networks for
each of the smart grid domains:
The wide area network (WAN) provided the
communications backbone that connected the sub
networks with each other and with the control and
data centres
The field area network (FAN) provided
communications for the distribution network
beyond the zone substation. The measurement and
control points on feeders, distribution transformers
and low-voltage lines and customer meters were
widely dispersed and numerous
The substation local area network (LAN) provided
connectivity between protection, control and
monitoring devices within the substation, and
between the substation and the control centre
The home area network (HAN) provided
connectivity within the customers premises for
connection of smart grid devices located there
including distributed storage and generation, load
control devices and smart EV chargers
16 NIST, Guidelines for smart grid cyber security,
http://www.nist.gov

36 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

The common IT infrastructure Smart Grid, Smart City


common platform consisted of:
The enterprise services bus (ESB) managing the
interaction and communication between software
applications. Its role is analogous to a hardware
bus over which computer hardware elements
communicate
A suite of common applications used by a range of
individual smart grid applications, for example the
data historian and the network model
Server infrastructure including the common
management of servers, server security, data
centres, backup, redundancy and disaster recovery
Storage infrastructure for the storage and archiving
of large amounts of real-time data, produced and
used by smart grid applications

National Broadband Network


The Smart Grid, Smart City program evaluated the
potential of the NBN to play a role in the smart grid field
area network. Initial assessment of connecting smart
meters to the NBN on-premises network-terminating
device indicated that this was not a practical solution due
to both cost and technical issues.
Consequently an alternative approach was developed.
The alternative approach used the NBN for backhaul
and a wireless mesh for the last-mile communication.
This had several advantages:
A much lower cost because a single network
terminating device and associated wholesale data
service serves a large number of meters
The wireless mesh is more flexible and is
self-healing
The proposed architecture was suitable for either
fibre-to-the-home or fibre-to-the-node topologies

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 37

4.5 High level


lessonslearned
The following sections describe the high level results
from the Common Platform trials.

4.5.1 Common platform


lessonslearned
Standardisation and interoperability
The smart grid trials undertaken by Ausgrid and its
partners for the Smart Grid, Smart City Program
consistently demonstrated that technologies
supporting well-defined and mature standardsbased interfaces were significantly easier to integrate.
Standardisation in the smart grid context is still a work
in progress, with a large number of organisations
around the world developing smart grid standards.
The results of this trial suggest that smart grid
standards should:
Be open that is be developed through a
collaborative process open to all interested parties
and not dominated by a single organisation or
interest group
Be international developed to support smart grid
deployments around the world, not to support
regional or national interests
Leverage existing standards many mature,
international and open IT and communications
standards suitable for smart grid applications
already exist
Along with standards, a conformance-testing
regime is necessary to support interoperability
between vendors. The Smart Grid, Smart City trials
identified many instances where individual vendors
implementation of standards was sufficiently different
to prevent interoperability. Independent conformance
testing would have identified these issues at the time
the systems were designed, saving both the vendor
and the smart grid operator significant time
and money.

IT and SCADA convergence


At Ausgrid, the primary means of retrieving data from
the field has been its SCADA (supervisory control and
data acquisition) system. This has been developed
and optimised to provide data to the control room via
the distribution network management system (DNMS)
which is used by network operators to monitor and
control the electricity network.
During the Smart Grid, Smart City field trials a
significant quantity of data intended for other
audiences, or requiring considerable processing
before being actionable by the operations centre,
was collected.
Some of the information sources did not lend
themselves to communication via SCADA protocols
and required systems that supported other protocols
(some of which were proprietary to vendors) to be
designed and built. In others, the raw information had
to be correlated with other information and translated
into actionable data prior to presentation to system
operators. In each of these cases, the initial destination
of the data was a back end system other than the
DNMS, a situation not envisaged when the Ausgrid
SCADA system was designed.
One solution would be to extend the SCADA system
to provide an auxiliary SCADA function. This would
reduce the burden placed on the DNMS by less
critical monitoring systems and allow asset owners
and maintenance staff to directly control condition
monitoring equipment in a system independently from
the DNMS.
The convergence of enterprise IT and power systems
operations is a feature of the smart grid, and will
require adaptation of current practices on behalf of the
network business and an understanding of the unique
features of the electricity distribution industry on behalf
of IT systems and services vendors.

38 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Alignment of data sources


The smart grid brings both the opportunity and
the challenge of integrating a diverse range of data
from sources across the distribution network. One
challenge identified during the Smart Grid, Smart City
trials was that of aligning and correlating information
stored in a range of different Ausgrid systems. For
example, each asset has both a functional identifier
describing a role in the network and an asset identifier
describing the asset itself. If an asset is replaced,
the asset identifier will change but the functional
identifier will remain the same. Conversely, if an asset
is redeployed within the network, the asset identifier
moves with it, but the functional identifier associated
with it will change.
No system provided a single source of truth for the
correspondence between these and other identifiers
used in the Ausgrid systems. This caused considerable
challenges during the trials as several sources had
to be consulted to correctly map the correlations.
Any widespread smart grid deployment will therefore
require careful consideration of the data model as
one of the first priorities. Where multiple identifiers
are necessary, a single source of truth for their
correspondence should be identified. It is also likely
that a degree of data cleansing and rationalisation will
be necessary.

Legacy equipment and systems


Utilities do not have the luxury of building a smart grid
from the ground up there are decades of legacy
equipment and systems with years of useful life to be
considered. The challenges faced in rolling out the
Smart Grid, Smart City grid application trials were
focussed on issues such as these and are discussed
under the respective technology areas in this section
of the report.
For example, as part of Smart Grid, Smart City
Program a new substation LAN architecture was
developed. While this architecture was deployed
in full in two greenfield substation builds, individual
components were also deployed on an as-needed
basis to meet the needs of the various trials in sites
where there was existing plant. This pragmatic

approach ensured that services were provided


in a consistent fashion without provisioning more
infrastructure than required. Planning for, and
understanding the challenges of, dealing with aging
brownfield sites is critical to understanding the costs,
benefits, challenges and opportunities of smart grid
technology deployment.

Security
Although the smart grid introduces many potential
benefits in terms of reduced demand, asset
utilisation and power quality, the extensive use of
ICT infrastructure does make the electricity network
more exposed to cyber-attack. The large number of
intelligent devices and associated communication
networks distributed over a wide area represents a
significant opportunity for cyber-attack and a challenge
to protect.
At the worst extreme, cyber-attacks could potentially
render the grid inoperative, steal confidential
information or damage vital infrastructure.
Data protection requirements for each smart grid
domain and application need to be clearly defined so
that manufacturers, operators and other stakeholders
participating in the smart grid technology development
and implementation can establish the necessary
security controls and the appropriate technologies to
protect smart grid data.
Data flow encryption, tunnelling, authentication and
non-repudiation, digital certificates, firmware validation
and patch management are areas that should
be addressed.

Privacy
The Smart Grid, Smart City trials identified a
scenario in which multiple organisations including
DNSPs, energy retailers and third parties were able
to co-operate to deliver services to customers and
therefore share access to customer information
securely. This may be possible with a number of
feedback or demand management technologies for
example, where identifying information, tariff details or
consumption patterns may be exchanged.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 39

Where personal consumer information crosses


corporate boundaries, the Privacy Act requires
customer consent to specific information sharing
arrangements. The adoption of a more efficient
mechanism that balances the consumers right
to privacy with more innovative electricity retail
serviceofferings is made possible within a smart
gridenvironment.

The analysis also showed that the business case for


NBN could be significantly improved if NBNCo had
a specific machine-to-machine wholesale offering.
At the time of this analysis, the lowest bandwidth
offering is substantially above that required for smart
grid applications. Bandwidth requirements would
have to increase by an order of magnitude for the
NBN to become the most cost effective solution.

4.5.2 National Broadband


Network lessons learned

Analysis also determined that if the data throughput


requirements of a smart grid were to increase
substantially, the cost of the 3G/4G solution could also
escalate as these costs are based on data transferred
rather than bandwidth.

The University of Newcastle was commissioned to


develop a network model to validate the proposed
architecture. The study concluded that the basic NBN
best effort service offering comprising a 12 Mbps
peak downlink and 1 Mbps peak uplink is easily able
to accommodate regular 10500 meters read on a
4-hourly cycle.
Modelling of the 20-year costs of the NBN architecture
and several competing solutions showed that:
3G/4G wireless is currently the lowest cost
backhaul solution The 3G/4G option was
consistently the lowest cost option for all regions
and applications (if satellite customers are
excluded). However, the quality of service provided
by 3G/4G may not be comparable with that of
theNBN
NBN is the most cost effective solution for very
remote customers Where all customers have
a smart meter, there will be a number of remote
premises that are expensive to serve due to
the need for satellite communications. The
NBN solution is based on a single price for all
connections regardless of location, making it a
costeffective option for these customers
Grid applications add significantly to cost The
majority of grid applications require a direct
connection and therefore add significantly to costs
for all solutions. This impact was more significant
where there is a large monthly cost differential
between communication solutions

The consultants DM KEMA were engaged to assess the


economics of using the NBN for smart grid applications.
The project developed a model to quantify the costs
of the reference architecture and tested its economic
feasibility by comparing the cost and performance of
providing communication services using NBN against an
alternative communication technology that can provide a
similar service.
This comparison was applied separately for the urban,
major rural, minor rural and remote regions and for
the different traffic levels associated with smart grid
equipment that may require communications. This
included equipment from the customer applications,
grid applications and distributed generation and
distributed storage projects.
In conclusion, the use of the NBN is currently not a cost
effective solution for the transfer of data from smart
meter infrastructure, but could be significantly improved
if NBNCo developed a wholesale offering. In addition,
the current bandwidth requirements for smart meter
infrastructure are not sufficient to require a broadband
solution, with 3G/4G solutions adequate.

40 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

5 Grid applications
workstream an overview
The Grid Application work stream included a
combination of projects which trialled grid-facing smart
technologies within Ausgrids distribution network. The
intent of the Grid Application trials was to deploy and
integrate smart grid technologies (generally) with the
existing infrastructure at various points on the electrical
network.
The Grid Application technology trials aimed to assess:
Whether smart grid monitoring and control
technologies could reduce the cost of running an
electrical network

How smart technologies could improve the


reliability of the distribution network
How smart technologies could improve the quality
of service
Which technologies had the potential to deliver
the best results for different network densities and
characteristics
Table 1-5 provides an overview of the different smart
grid technologies trialled in the grid applications work
stream.

Whether integrating smart grid technologies and


techniques had the potential to enable the design
of a lower cost electrical network

Table 1-5 Grid Application Project smart grid technologies trialled

Smart grid technology project

Abbreviation

Description of smart grid technology

Active Volt-VAr Control

AVVC

Automated voltage regulating and reactive power


controls to measure and maintain acceptable voltages
and high power factor at all points in the distribution
network under varying load conditions.

Fault Detection, Isolation and


Restoration

FDIR

Automation technologies used to quickly and precisely


detect fault conditions, isolate faulty equipment and
restore power to customers by operating remotely
controlled switches.

Substation and Feeder


Monitoring

SFM

A collection of technologies which monitor the network


state (voltage, current, frequency) and condition of
assets within the electrical distribution network utilising a
common ICT platform.

Wide Area Measurement

WAM

Measurement devices capable of providing high speed,


time-synchronised samples of network data, including
voltage, current and frequency, called synchrophasors.
Once deployed at strategic points on the transmission
and distribution networks, incoming data from these
devices can be used to accurately and dynamically
measure the state of the power system from a widerarea perspective.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 41

Figure 1-6 shows the physical locations where the technologies were deployed.
Figure 1-6 Physical locations of operating technologies installed under the Grid Applications Project

SCONE

MUSWELLBROOK

SINGLETON
NELSON BAY

CESSNOCK

NEWCASTLE

WYONG

GOSFORD

HORNSBY

SYDNEY
OATLEY

Legend
Active Volt-Var Control

Wider Area Measurement

Substation Feeder Monitoring

LGA Boundary

Customer Applications

Fault Detection Isolation + Restoration

Distributed Generation + Distributed Storage

Main Roads
Towns

42 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

6 Active Volt-Var
Control(AVVC)
6.1 What are Active
Volt-Var Control
(AVVC) technologies?
Active Volt-VAr Control (AVVC) is a term for a collection
of technologies which use automated voltage
regulating and reactive power controls to measure and
maintain acceptable voltages and high power factor
atall points in the distribution network under varying
load conditions.
Network voltage, which moves up and down in
response to load and source fluctuations, needs to
becontrolled to ensure it remains within an acceptable
range specified by the Australian standards17. To
minimise electrical losses in the network, and ensure
that the network assets are used optimally, the power
factor should be maintained as close to unity
as possible.
Traditionally, distribution utilities have achieved a
degree of control over the network voltage and
reactive power via Volt-VAr control devices such
as switched capacitor banks and tap-changing
transformers. These rely on localised measurement of
voltage and current to determine the control actions
of the device. They have a simple control objective (for
example, keep local voltage within a pre-determined
range), and operate independently of other similar
devices deployed at other points in the network.
The smart grid provides an opportunity to continually
monitor the voltage and current in the grid at various
locations, and to use this information to control the
distributed voltage and reactive power control devices
using sophisticated control algorithms.
As these control algorithms, known as Integrated VoltVAr Control (IVVC), can access multiple measurement
and control points across the network, they have the
ability to implement very sophisticated

control functions (for example, to minimise electrical


losses, to manage demand and to reduce energy
consumption). In some cases IVVC can adapt to
changing feeder configurations, load and source
conditions and varying operational needs. This means
that IVVC control systems have the potential to better
manage and control the dynamic effects of increased
distributed generation as part of an optimised
AVVC strategy.
Deployment of AVVC is becoming increasingly
important as greater levels of distributed generation
(solar PV) are deployed within the distribution network.
Each of the specific AVVC technologies trialled has
been described further in the following sections.

6.2 AVVCs role in a


smart grid
The Smart Grid, Smart City trials sought to understand
the benefits that smart grid technologies could achieve
in improving electricity supply security, making the grid
more efficient, and in integrating renewable energy
sources in an Australian context. AVVC technologies
role in the smart grid includes the potential for:
Deferring capital expenditure through improving the
utilisation of existing assets by minimising reactive
power in the network and through peak shaving at
times of maximum load
Reducing system losses by minimising real and
reactive power flow in the network through voltage
control and the control of reactive sources
Improving power quality through improved voltage
profiles
Reducing energy consumption through
Conservation Voltage Reduction (CVR)
The AVVC project aimed to investigate, validate and,
where possible, quantify these benefits.

17 AS60038-2012 Standard voltages, 2nd Edition, Standards


Australia, 2012
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 43

6.3 AVVC Smart Grid,


Smart City Project
objectives and timelines
The Active Volt-VAr Control Project aimed to
characterise the benefit that could be obtained
by implementing an IVVC scheme in a distribution
network. The project sought to:
Evaluate the benefits of centralised and distributed
Volt-VAr control schemes compared to current
methods, including their impact on voltage
regulation, power factor, peak load, and
energy consumption
Examine the potential benefits of implementing
a CVR control algorithm for reducing energy
consumption in a distribution network
Develop modelling tools, resources and techniques
to assist in the process of planning and evaluation
of AVVC systems in the context of the smart grid,
including:
The development of high-fidelity time series
load models
The development of geo-spatial visualisation tools

The project consisted of advanced modelling


(PS+EDGE tool, discussed in the following sections
ofthe report) and field trial elements. The field
trial results were used to validate the PS+EDGE
models and gain insights into the commercial-scale
deployment of these technologies.
Before commencing the field trials and associated
data collection and analysis, the AVVC Project required
an extensive planning, design and installation phase
which occupied most of the first two years. The trial
site, at Nelson Bay in New South Wales was used for
both AVVC and FDIR (Fault Detection Isolation and
Restoration) trials, allowing one distribution monitoring
program and smart meter infrastructure to support
both trials. The Nelson Bay zone was selected
because it had:
A reasonable diversity of load types
Moderately voltage-constrained feeders
Satisfactory telecommunications coverage
High penetration of rooftop photovoltaic (PV)
generation
The AVVC Project spanned three years, from October
2010 to September 2013 (Figure1-7).

The modelling of Static VAr Compensation


(STATCOM) technology to assess the potential
benefits of its use within the distribution network

44 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Figure 1-7 AVVC project timeline

2009

2010

2011

2012

2013

Initiation & Scope Definition


Trial Design
Field Installation
IT Installation & Test
Field Trial
Simulation
Final Report
2
Milestones

 irst Capacitor Bank Online


1 F
30/04/2012

 VR Automation Trial Started


3 C
23/04/2013

 oltage Regulator Online


2 V
15/08/2011

 owerOn Fusion Commssioned


4 P
1/01/2013

3 1

5 IVVC Into Automatic Mode


29/05/2013

6.4 Where are AVVC


technologies currently
deployed?

6.5 Smart Grid,


SmartCity AVVC trials

AVVC technologies have been extensively tested in


international studies, with a range of benefits identified.
Generally the international trials found that AVVC
technologies did produce a measurable benefit, and
that the technology was sufficiently mature to be
deployed on a commercial scale without undue risk18.

Field trials for AVVC were undertaken in the Nelson


Bay area, north of Newcastle in NSW. At the
time of site selection, this area consisted of three
zone transformers with nine feeders supporting
around 17,000 customers across 330 distribution
transformers. The feeders were a mix of short rural
types with voltage constraints and urban types with
capacity constraints.

AVVC algorithms have a number of objective functions


including voltage control, reactive and real power
minimisation along with secondary objectives such as
the minimisation of tap changes. Feeders with voltage
constraint issues are a prime candidate for AVVC-style
technologies due to the relatively low capital cost.
This suggests that there is high potential for economic
benefit through selective deployment of these
technologies. It should be noted however, that the vast
majority of the studies in the literature were carried out
in North America and very little work specific to the
Australian context has been published.

6.5.1 AVVC trial design elements

During the trial, the zone underwent a major planned


change, with the introduction of a new zone substation
and subsequent reconfiguration of feeders. The final
network configuration, which was used in PS+EDGE
model consisted of two zone transformers with
eight feeders supporting around 10,000 customers
spread across 210 distribution transformers. The
final configuration of feeders still exhibited signs of
moderate voltage constraint.

18  The Smart Grid, Smart City AVVC Technical


Compendium provides a description of some of
the observations of AVVC technologies based on
published data from other trials. This document can
be located on the Information Clearing House at
https://ich.smartgridsmartcity.com.au/.
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 45

6.5.2 Technologies
Volt-VAr control is required to maintain an acceptable
voltage profile and power factor in the distribution
network despite the heterogeneous nature of network
feeders. Although a group of feeders may originate at
a common source, and have been designed to exhibit
similar characteristics, the evolving nature of the load
often leads to a situation where the characteristics of
one feeder constrain the optimisation of the system
as a whole.
AVVC technology relies on the measurement of
voltage and current at many points in the network
and the coordinated activation of multiple control
points to achieve a specific objective function. This
measurement and control cycle is repeated at
predetermined intervals, to maintain the objective
function under changing network state, load and
generation conditions.
An AVVC control algorithm can be configured to
optimise one or more objective functions, including:
Maintaining an acceptable voltage profile.
Maintaining the voltage within the mandated range
is a minimum requirement of any Volt-VAr control
system. In addition, flattening the voltage profile
of voltage-constrained feeders provides scope
to operate other feeders in the zone at closer to
nominal levels, reducing the power throughput and
maximising plant utilisation and asset life.

In addition to capacitors, power electronic devices


such as STATCOMs can be used to compensate for
the inductive reactance in the distribution network,
bringing the power factor closer to unity this is
known as conservation voltage reduction. It works by
reducing the overall system voltage below the nominal
level to marginally reduce the overall load on the
distribution system.
The technologies required to implement active Volt-VAr
control fall into four broad categories:
Actuation or control devices that change the
voltage or power factor within the network. These
typically include capacitor banks and voltage
regulating devices such as transformer tap
changers and power electronic devices such as
STATCOMs
Monitoring devices measure the voltage, current or
power factor at various points within the network.
Monitoring points may be standalone transducers
or be functions associated with network elements
such as zone circuit breakers, low-voltage RTUs,
telemetered switches or smart electricity meters
Volt-VAr controllers that accept inputs from the
monitoring devices and provide commands to the
actuation device according to a control algorithm
with a defined objective

Maintaining near-unity power factor. Power


factor is the ratio of real power delivered by the
network (available to do useful work at the load)
and the apparent power. The apparent power is
normally higher than the real power due to the
reactive nature of the load. The reactive power in
the network, although not available to do useful
work, contributes to losses and must be taken into
account in the dimensioning of the network. Since
the typical distribution load tends to be inductive,
it is possible to compensate for the inductive
reactance by introducing the appropriate capacitive
reactance to the distribution network, thus bringing
the power factor closer to unity.

46 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

6.5.3 Enabling ICT Infrastructure 6.5.5 National Feeder


TaxonomyStudy
In the Ausgrid network, Volt-VAr control elements
outside the zone substation operate under local control
and are therefore rarely provided with connectivity.
The implementation of an AVVC scheme requires
a communications platform to link the monitoring
devices, actuators and control systems. This trial
used 3G wireless technologies to communicate with
the feeder-based devices, and Wimax technology
to communicate with the smart meters. The existing
SCADA system was used for devices in the zone
substations. This platform was shared with many other
projects within the Smart Grid, Smart City Program.
For full details of this common platform please see
the Smart Grid, Smart City Common Platform Study
Technical Compendium.

6.5.4 Production System +


Extended Data Grid Environment
To complement and validate the field trials, a range
of advanced analytical modelling activities were also
undertaken including development of a bespoke
modelling tool, known as the Production System +
Extended Data Grid Environment (PS+EDGE) this
was a sophisticated real time modelling tool, which
was extensively tested and validated against actual trial
data. The trial area of Nelson Bay was reconstructed
in PS+EDGE to replicate trial conditions and to allow
for trial data to be fed into the model for further testing
and analysis. This model contained over 100,000
separate objects interacting across the network.
The high resolution PS+EDGE platform was developed
to investigate the effect of various AVVC control
algorithms on the network. High-resolution modelling
was required as the effects measured were in the
order of one or two per cent in some instances. The
PS+EDGE platform model was capable of resolving
network quantities to well under 1 per cent at intervals
of 10 minutes. A load model of similar sophistication
was also developed. The model ultimately allowed the
load and power factor to be described for a given hour,
day (weekday or weekend), season, and
ambient temperature.

A significant challenge for each of the Smart Grid,


Smart City technology trials was the need to
extrapolate the results from specific local trials (in the
Ausgrid network) to a national context. As part of the
AVVC trials, the National Feeder Taxonomy Study was
completed to address this challenge.
This project conducted by Ausgrid and CSIRO, and
worked cooperatively with Distribution Network Service
Providers (DNSPs) from across Australia, created a
database of feeders that captured the diversity of
Australias distribution networks. From this database
a small set of representative feeders that succinctly
described a typical Australian distribution network was
developed and used to model smart grid technologies
on a national basis as part of the national cost benefit
assessment (Part Two of this report).

6.5.6 Static VAr Compensator


(STATCOM) technology
The AVVC project also evaluated the effectiveness
of static VAr compensator (STATCOM) technology.
Astudy by the University of Newcastle evaluated the
benefits of STATCOMs for improving power quality
in the MV network in the Singleton area. Aseparate
STATCOM field trial by Energex, together with
simulations (validated against the Energex trial) using
the PS+EDGE model, examined the ability of low
voltage STATCOMs to regulate voltage under high
levels of distributed generation19. Preliminary results
from the Energex field trial indicate the ability of the
STATCOM to reduce the level of voltage fluctuations,
when operating in voltage mode. The Energex trial
also illustrated the impact of the operation of a
60kVA STATCOM operating in voltage mode, when
located at different positions along the feeder. The
results indicated that irrespective of the location of
the STATCOM along the feeder, that it had the ability
to reduce voltage fluctuations. Final results from the
Energex trial had not been reported at the time this
report was produced.
19 GA1275 STATCOM Study Trial Report, Energex, SGSC
supporting document, 2013

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 47

6.5.7 AVVC trial deployment


The planning and design stages of the AVVC project
involved the deployment of six pole-mounted capacitor
banks and one voltage regulator; 20 telemetered
switches; 96 distribution monitoring devices; and over
500 smart meters in a production network supporting
over 17,000 residential and business customers. In
addition, significant substation augmentation work
was carried out together with the installation and
commissioning of the supporting control systems,
information technology and communications
infrastructure. This included remediation to enable
remote operation of transformer tap changers and
capacitor banks within the Nelson Bay Zone.
In keeping with the desire to fully utilize existing
network assets where possible, some equipment was
redeployed for these trials, requiring decommissioning
and removal prior to reinstallation. In some cases, the
dismantling process caused damage to the equipment
resulting in unforeseen delays. Two capacitor banks
ultimately could not be commissioned in time for the
trial due to damaged components and long vendor
lead times for replacement parts.

As previously discussed, the AVVC project was initially


planning to run a STATCOM field trial, however, this
was not possible due to the long supply lead-times
for this equipment. The Queensland DNSP, Energex,
had ordered similar equipment for a trial they were
conducting, and made their trial data and report
available to the Smart Grid, Smart City Program
(noting that not all results from the Energex trial had
been finalised at the time this report was produced).
To complement the field trials, the project built
a measurement-validated time series model of
the Ausgrid Smart Grid, Smart City production
trial network segment. To ensure the model was
production equivalent, it was validated against real
loads, energy consumption and end-of-line voltage
using production data. This PS+EDGE model was
then used to complete further analytical modelling
of scenarios that could not be practically tested in
the field trial. This allowed the investigation of much
higher densities, combinations or locations of AVVC
devices. This model was also used for other elements
of the Smart Grid, Smart City Program including each
of the grid application technologies, Electric Vehicles
and Distributed Generation and Distribution Storage
(DGDS) programs.

48 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

6.6 High level


trialresults

6.6.2 Conservation voltage


reduction

6.6.1 Voltage regulation and


energy saving
In general, it was clear from the trials that AVVC
is effective in managing the voltage profile and
power factor in a distribution network. The results
demonstrate that significant improvement in voltage
regulation is possible, together with a four to five per
cent improvement in peak demand, and a similar
improvement in power throughput over the base case.
However, the introduction of AVVC will also result in
more frequent switching of actuators and may lead
to an increase in maintenance costs and potentially
reductions in the service life of these devices.
In the trial, centralised controls performed marginally
better than decentralised control, although the
differences were sufficiently small to suggest that
either approach is viable. The deciding factor
between the two approaches is likely to be the ease
of integration within a distribution networks existing
secondary systems.

The field trials evaluated the CVR factor using day-onday-off testing. The results showed a CVR factor of
0.65, consistent with trials described in the literature.
In the Australian context however, the benefits of CVR
need to be considered in the light of the transition of
the nominal domestic electricity supply voltage from
240 V to 230 V. In 2000 the nominal mains voltage
in Australia was reduced from 240 V to 230 V. The
upper tolerance on the nominal 230 V level was set
to include the previous 240 V range. Because of this,
many electricity providers have been slow to adjust
their voltage regulation equipment to accommodate
the 230V nominal voltage. As and when providers
adjust their voltage regulating plant to the new nominal
level, many of the benefits of CVR will be automatically
captured, without the need for a full-scale deployment
of AVVC technology.
Given the high level of monitoring on the low-voltage
network, for example through the deployment of smart
meters, the transition to 230 V could be managed with
a high degree of precision, maximising the CVR benefit
without significant additional capital investment.

Distribution network service providers with a highly


sophisticated substation automation infrastructure
would likely benefit from the distributed approach,
while those managing the network via a centralised
distribution management system would be more likely
to benefit from adding AVVC as an application at
thislevel.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 49

6.6.3 Impact on planning and design


Electricity distribution networks are designed to
operate within two key constraints; within voltage
limits (voltage constraints); and within capacity limits
(capacity constraints). Existing approaches to network
design do not to require Volt-VAr management to
operate within these constraints. For new networks
the potential of the technology is to offer a wider
range of design options such as longer feeders. For
existing network the benefits of the technology, be it
automated or not, are limited to those feeders where
the power factor at the time of system peak is low.
Only in these zones is there is significant capacity that
can be unlocked by increasing the power factor.
Short rural and long rural feeders, as defined in the
feeder taxonomy study, are those most likely to benefit
from AVVC. Unfortunately, these are also located in
areas where the necessary communications links are
less reliable and more expensive. The analysis found
that an estimated 1,500 suburban fringe and short
rural feeders would benefit from this technology only
about 15 per cent of the total number of feeders in
Australia. Further information in the Modelling Inputs
Compendium20 and in the results of the AVVC net
benefit analysis described in Part Two of this report.
The project confirms that AVVC can regulate the
voltage profile along feeders very effectively. This
capability could potentially be exploited by inverting the
current design paradigm. Feeders today are designed
to not be voltage constrained - this means that the
designs are frequently over-engineered to the extent
that the voltage profile under projected peak load will
remain within acceptable limits over the 20 or 30year life of the feeder. It would be possible to design
feeders much less conservatively such that they would
rely on the presence of AVVC to keep voltages within
acceptable limits. Individual feeders could be longer, or
conductor sizes could be reduced, reducing
capital expenditure.

Today the network is explicitly planned for peak load,


as this is when the network is under the greatest
stress. Most of the design processes in use today
assess the network only under those conditions.
However, deployment of a range of smart grid
technologies such as AVVC, operate most effectively
away from times of peak load. Therefore the
opportunity for greatest benefit is well away from the
time of peak load.
To extract the maximum value from smart grid
technologies such as AVVC, a continuous view and
optimisation design process for the network is required
rather than boundary condition analysis using a single
peak snapshot of the network.
The AVVC trial demonstrated that a design process
optimised only for peak conditions could result in suboptimal designs for other times. For example, modelling
showed that a capacitor on a feeder would improve
voltage regulation outside peak times, but because
that particular feeder showed no problem under peak
load, no capacitor was deemed necessary under the
traditional (current) design processes.
As electricity distribution networks evolve to include
smart grid technologies, simply designing the network
for peak loads will not be sufficient to ensure reliable
and efficient electricity supply. Instead, time series
power flow modelling will be essential. Part Three
of this report discusses the opportunities for the
deployment of AVVC technologies as part of the broad
recommendations.

20 The Modelling Inputs Compendium is available on the Smart


Grid, Smart City Information Clearing House at https://ich.
smartgridsmartcity.com.au/.

50 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

6.6.4 Applicability of Solutions


toAustralia
Australian distribution networks generally have far
fewer actuation points than their North American
equivalents, for which much of the technology
available today was developed. Networks in Australia
also typically have voltage regulation plant only on
particularly long or voltage constrained feeders.
Feeder capacitor banks are rare therefore any
AVVC deployment in Australia is likely to require the
deployment of a significant number of new activation
devices, and consequent additional cost, compared to
the North American experience.
The fundamental physical driver for the difference
between the North American experience and the
Australian experience is the different voltage levels.
North American network operate at 110V whereas
Australia uses 230V. This fundamental difference
drives the low voltage network design in terms of
feeder lengths and customers supplied from each
distribution centre. In particular the North American
experience is to have far shorter and more contained
LV networks - lower voltage means higher currents,
consequently shorter feeder lengths and fewer
customers per device meaning a greater number
of controllable devices exist on the North American
network compared to the Australian experience.
The best opportunity for AVVC in the Australian
context is therefore longer rural feeders, which tend to
have a large number of actuation devices. AVVC would
be an ideal technology to coordinate the operation
of these devices, although as mentioned above,
communications links may be a limiting factor.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 51

7 Fault Detection, Isolation


and Restoration (FDIR)
7.1 What are fault
detection, isolation and
restoration (FDIR)
technologies?
Fault Detection, Isolation and Restoration (FDIR) refers
to the ability of an electricity network to discover a fault,
determine the location of that fault, isolate the faulty
equipment and deploy resources in a manner that
will restore power to as much of the affected area as
effectively and efficiently as possible.
Currently grid operators like Ausgrid would
generallybe aware that a fault has occurred on its
highvoltage(HV) network through a feeder circuit
breaker tripping, sending an alarm to the control room.
Customer calls reporting damage in combination with
feeder patrols working their way down a feeder is then
used to narrow down the location of the fault. If no
fault is located through this approach, a process of
elimination is used to identify the faulty section by field
staff sectionalising and manual switching on the lines.

Earth fault indicators are also used where available,


particularly for underground network sections. Earth
fault indicators are generally set by an earth fault and
then manually reset by a site visit. These allow field
staff to locate the feeder section where a fault has
occurred. Once the faulty section has been identified,
a detailed visual inspection or pulse fault location is
carried out to verify the exact fault location. Manual
switching is then performed to restore power to
customers on the healthy part of the feeder.
Figure 1-8 shows an indicative timeline for fault
location and restoration within a typical distribution grid
(source: Ausgrid). These times can vary significantly
depending on: the type of fault; the assets affected;
whether the fault is in a rural or urban area; and
whether the feeder is underground or overhead.

Figure 1-8 Indicative fault location and restoration time

Power Restored to
Healty Part of Feeder
Fault
Occurs

Field Staff
On Site

Fault
Located
Investigation and
Feeder Patrol

20-40 min

20-40 min

Manual
Switching

10-20 min

50-100 min

52 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Repairs

Power Restored to
All Customers

7.2 FDIR role in a


smart grid
The smart grid enables network providers to
continually monitor equipment at various locations
on the grid to detect fault conditions and to use this
information to isolate the faulty equipment and restore
power by operating remotely controlled switches.
The main objective of FDIR is to reconnect as many
customers as possible, as quickly as possible after the
occurrence of a fault. Faults can occur on the network
in many different ways and forms. The most common
causes on Ausgrids 11 kV distribution network
include: nature (i.e. lightning, weather, vegetation and
animals) and equipment failure.
Back office IT systems use event distribution automation
algorithms which process information from field devices
and either automatically operate the switches or provide
advice to the control room operator on a recommended
action. With multiple monitoring points throughout the
network and switches located to enable sectionalising
of the grid and rerouting of power, these algorithms are
able to execute complex control functions to optimise
the number of customers able to be restored in
an outage.

FDIR solutions can either be centralised or distributed.


The centralised approach uses a control room-based
algorithm and requires communications between the
control room and field devices, while the distributed
approach involves automatically operated field devices
and requires peer-to-peer communication between the
field devices.
The main benefits of implementing FDIR as part of a
smart grid could be expected to include:
Reducing the number of customers affected by a
fault due to feeder sectionalising and switching in
other feeder paths
Reducing the reported frequency and duration of
faults
Operational improvements through more efficient
use of field crews
The above benefits would improve network reliability,
commonly measured in System Average Interruption
Duration Index (SAIDI), System Average Interruption
Frequency Index (SAIFI), Momentary Average
Interruption Duration Index (MAIDI), Momentary Average
Interruption Frequency Index (MAIFI) and Customer
Average Interruption Duration Index (CAIDI). The
definitions of these key indicators can be found in the
Smart Grid, Smart City FDIR Technical Compendium.
The reliability of the distribution network is of significant
importance to both DNSPs and their customers. The
potential for economic benefit from the deployment
of FDIR technologies is discussed extensively in Parts
Two and Three of this report.

Figure 1- 9 FDIRs role in a smart grid

Fault
Detection,
Isolation and
Restoration
(FDIR)
Bulk Supply Point

Subtransmission
Substation

Zone Substation

Distribution
Substation

Domestic, commercial &


industrial customers

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 53

7.3 FDIR and the


Smart Grid, Smart City
trial objectives and
timelines
The Smart Grid, Smart City FDIR Project aimed to
ascertain the benefits that could be realised by
implementing FDIR technologies on the distribution
network. The project sought to:
Evaluate the reliability improvement (SAIDI and
SAIFI) from deploying remotely controlled switches

The project consisted of modelling, simulation using


the PS+EDGE model and field trial elements. The field
trial results were used to validate the modelling and
gain insights into the commercial-scale deployment of
these technologies. The data collected from the trials
was analysed to determine how these technologies
operate as part of an integrated smart grid and to
quantify and explore the potential benefits obtainable
from deploying FDIR.
The FDIR project spanned three years, from October
2010 to September 2013 (Figure 1-10).

Evaluate the reliability improvement (SAIDI and


SAIFI) from applying real-time FDIR algorithms
Explore if deploying FDIR technologies could
improve operational efficiencies
Explore the potential for smart meters to improve
fault detection and restoration
Undertake a cost benefit assessment and net
benefit analysis of FDIR to determine broader
societal benefit of improvements in SAIDI and SAIFI

Figure 1-10 FDIR project timeline

2009

2010

2011

2012

2013

Initiation & Scope Definition


Trial Design
Installation & Commissioning
IT Build, Test & Implement
Trial and Data Collection
Simulation Activities
Final Report
1
Milestones

1 Two IntelliRuptors commissioned


 ne IntelliRuptor and
2 O
three ELBS commissioned
 our IntelliRuptor and
3 F
four ELBS commissioned

2 3 4 5

6 7

 ne IntelliRuptor and one


4 O
ELBS commissioned

7 Two ELBS commissioned

 hree ELBS commissioned


5 T
29/05/2013

 ne IntelliRuptor commissioned
9 O
(10072)

6 DMS FDIR Algorithm


commissioned

54 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

8 One ELBS commissioned

7.4 Where are FDIR


technologies currently
deployed?
7.4.1 Australia
A key FDIR initiative in Australia is SP AusNets
Distribution Feeder Automation (DFA) program,
launched in 2009. It involves a sophisticated
automation scheme with the stated objective of
restoring supply to as many customers as possible
within one minute.
As of August 2013, there were over 480 DFA schemes
in service covering over 80 per cent of the SP AusNet
electricity distribution network and about 90 per cent
of their customers. The DFA schemes are able to use
an installed base of more than 1 200 intelligent field
devices for the applications.
Since its introduction, there have been over 650
scheme operations and the automated restorations
achieved by the DFA schemes have resulted
in a cumulative unplanned SAIDI reduction of
over 32 minutes and significant unplanned SAIFI
reductions. In addition to the automated restoration
benefits, it is estimated that over 100 minutes of
accumulated unplanned SAIDI savings have been
made by accelerated manual response based on the
information provided by the DFA operations.

7.4.2 International
A number of international trials have been undertaken
to establish if there are measurable improvements in
responding to grid outages by deploying FDIR. The trials
to date have predominantly taken place in the US under
the DOE Smart Grid Investment Grant Program.

The reported benefits of FDIR in the US studies


include:
Improvements in SAIDI measurements in the range
of 10 50 per cent
Improvements in SAIFI measurements in the range
of 10 40 per cent
Operational savings through the elimination of
manual switching and improving the productivity of
field crews
On average (and using normalised results from four
trials in the United States), reductions in the order of
20per cent in SAIDI and SAIFI were demonstrated21.
In addition to trials, there are already a number
of large-scale implementations where FDIR
technologies are being deployed within a distribution
network. The largest and most comprehensive of
these rollouts are in the US and include:
EPB Smart Grid Project, where automated
switches have been deployed on all 12kV and
46kV feeders. In total, 1,300 automated switches
were installed in 2012 at a cost of US$50 million.
The anticipated benefits include a reduction in
outage times and a reduced number of call-outs
during storms (in the order of 250 per storm event).
Reductions in SAIDI are anticipated to be greater
than 40 per cent, which has been valued at
US$35 million per year to customers.
AEP Ohios gridSMART deployment, where
70feeders were commissioned and integrated
with distribution management system (DMS).
300 switches have been installed and the full
project required an investment of US$150 million.
It is anticipated that in the order of 2.5 million
customer-outage minutes will be saved, with more
than 45per cent of all outages expected to net
customer benefits. Crew savings of around two
hours per event are also anticipated.

21  U.S. Department of Energy, Reliability


Improvements form the Application of Distribution
Automation Technologies Initial Results, Smart
Grid Investment Program, December 2012
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 55

7.5 Smart Grid,


SmartCity FDIR trials
The FDIR project required an extensive planning,
design and installation phase occupying most of the
first two years, before field trials, data collection and
analysis could be completed. The trial site, at Nelson
Bay in New South Wales was used for the AVVC and
FDIR trials, allowing a single distribution monitoring
and smart meter infrastructure to support both trials.
The planning and design stages involved the
deployment of 23 telemetered switches, 96 distribution
monitoring devices and over 500 smart meters in a
production network supporting over 17,000 residential
and business customers. In addition, significant
substation remediation work was carried out together
with the installation and commissioning of the
supporting control systems, information technology
and communications infrastructure.
To complement the field trials, the project built a
measurement-validated time series model of the
Ausgrid Smart Grid, Smart City production trial area
of Nelson Bay. To ensure the modelling tool was
production equivalent, it was validated against real
loads, energy consumption and end-of-line voltage
from production data. Further details can be found
in the supporting documents for further details. This
model was then used to perform further analytical
modelling of scenarios that could not be practically
tested in the field trial. This allowed for the investigation
of much higher densities, as well as different
combinations and locations of FDIR devices.

7.5.1 FDIR Trial Design


Elements Technologies
The technologies required to implement FDIR fall into
four broad categories, described below:
Switches Switches are used to sectionalise
the grid and reroute power when a fault occurs.
Switchesrange from:
Circuit breakers which are automatically operated
electrical switches normally located at the head of the
feeder in the zone substation. They are designed to
protect network assets in the case of short circuits.
When a fault is detected, it opens to interrupt the
circuit and resets once the fault has cleared.
Reclosers are protective switches located part way
along the feeder. They are designed to open before
the circuit breaker in the zone substation to clear
downstream faults. After opening in the event of a
fault, they repeatedly attempt reclosing to test if the
fault remains. If the fault is transient, it recloses to
re-establish connection, but if the fault persists they
remain open until reset. Decentralised, manually reset
reclosers have been used by networks for decades,
but communications enabled reclosers (pulse
reclosers) have only become available more recently.
Distribution Monitoring Devices Distribution
monitoring and protection devices used for fault
detection. Distribution monitoring devices are
commonly deployed across distribution networks
on zone and distribution substations, transformers
and feeders. They normally measure current, voltage
and power factor, information which can be used by
the network to improve asset management and fault
detection.

56 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

FDIR controller An FDIR controller is a software


algorithm used to optimise the network. It utilises
information from switches and monitoring devices
to either automatically operate switches or provide
advice to the control room operator on a preferred
action to optimise the number of customers
restored. It accepts a range of fixed and real-time
measurement data, processes it, and provides
control instructions to the switches. FDIR controllers
differ in the range of input data they consider, and
the complexity and flexibility of the algorithms they
implement.
FDIR trial deployment The field trial was designed
to test advanced distribution automation applications
based on a modern DMS platform. It involved the
deployment and operation of distribution monitoring
devices, remote switches and smart meters. To
enhance the probability of the FDIR algorithm and
field devices having to operate, Nelson Bay was
chosen due to its relatively poor reliability and other
network characteristics.
Ausgrid proposed installing 10 switches mid-feeder
for sectionalising and 15 for back feeding, which
according to the report provided by the reliability
planning group would reduce SADI and SAIFI by
59per cent and 15percent respectively. Almost the
entire Nelson Bay area has poor soil resistivity, owing
to the fact it is a rocky, sandy peninsula, which made
earthing design more challenging. Ultimately, two of the
earthing designs were deemed too difficult to physically
construct by the installation crew, leading to only
23switches being installed. The location of distribution
monitoring devices was also constrained by the
physical requirements of installation (not all distribution
transformers can be retrofitted with monitoring
equipment) and by gaps in communications coverage.

A key challenge for the FDIR project was deploying a


new DMS within a compressed time frame, specifically
how to test it and secure business acceptance. The
sheer number of inputs that needs to be exhaustively
tested becomes a barrier to deploying a flexible,
algorithm-controlled smart grid. To address this, the
project team developed a real-time load flow engine
with a SCADA interface connected to the DMS to allow
for comprehensive testing of the system.
The DMS was subsequently deployed in parallel with
Ausgrids existing DMS system (DNMS) and the FDIR
algorithm was set to advisory mode. This meant that
the trial DMS had access to all required network
information, but would not attempt any automated
switching operations but rather provide advice to the
control room operators on a preferred course of action
in the event of a fault.
The ability for operators to see the results from the
new DMS system side-by-side with their trusted
DNMS system proved to be an unexpected benefit of
this approach. The knowledge that they could fall back
on the production DNMS if necessary also helped
with the acceptance of the new and different. The
deployment of a parallel DMS system in this manner
worked extremely well for the project. Ausgrid found it
to be a low risk option as rollback is simpler and less
disruptive than completely replacing the existing DMS.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 57

7.6 High level


trialresults

period to demonstrate operational efficiencies by not


dispatching field crews to manually operate switches in
the field.

There were a total of nine fault events registered in the


Nelson Bay area during the field trial between August
2012 and August 2013, and only one of these involved
switching activities where the deployed FDIR solution
could have had an impact. As part of the installations,
vegetation management was undertaken to clear
bushes and trees encroaching on feeders, which
might be part of the reason why the number of faults
was limited over the period.

The DMS required several reconfigurations during the


trial period, which highlights the inherent complexity
involved when integrating new capabilities into an
existing system, as well as shortcomings of the testing
process. The scripted SCADA sequences used for
testing were not sourced from actual event logs,
and therefore were subject to incorrect assumptions
regarding how the SCADA points would behave in the
event of a fault.

The one fault event involved back feeding to restore


power to parts of a damaged feeder via an adjacent
zone substation. This zone substation was not part of
the field trial scope and the FDIR solution was therefore
unable to assist in the restoration process, with the
remote switching activities performed by the control
room operator. As a result, there was no measurable
reliability improvements realised from deploying the
automated FDIR solution in the field trial.

The FDIR PS+EDGE simulations were designed to


test the DMS system deployed as well as to determine
the incremental reliability improvement from adding
switches and automation capabilities. To ensure that
the PS+EDGE model was production equivalent, it
was validated against real loads, energy consumption
and end-of-line voltage from production data. The
PS+EDGE simulation environment allowed the project
to explore scenarios, for example the impact of adding
and removing remotely controlled switches, which
would not have been possible in the field trial due to
logistical constraints and cost considerations. The
results of this PS+EDGE simulation are shown
in Figure 1-11.

However, remote switching by the control room


operator was successfully demonstrated by the field
trial. These capabilities were used for planned outages
on a number of occasions during the trial
Figure 1-11 Simulated SAIDI and SAIFI reductions
Sectionalising Switches

Manual Remote Switching

0%
-5%
-10%

% Change

-15%
-20%
-25%
-30%
-35%
-40%
-45%
-50%

SAIDI

SAIFI

58 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Automated FDIR

The PS+EDGE simulations demonstrated that utilising


an FDIR algorithm to automatically operate distributed
switches reduced SAIDI and SAIFI by over 60 per cent
with the trial area saturated with switches. This is, with
switches deployed to ensure that there would always
be a switching option available for the FDIR algorithm
to restore power to the healthy part of the feeder. This
meant simulating the deployment of two switches per
feeder, one for sectionalising and one for back-feeding.
With switches deployed as per the field trial, the SAIDI
and SAIFI reductions simulated were 40 per cent.
The greatest incremental SAIDI reduction was
demonstrated from deploying remotely controlled
switches for feeder sectionalising. Switching could
either be performed almost instantaneously through
automation, or with a slight delay if executed by
a control room operator. Simply enabling remote
controlled switching reduced the duration of faults for
customers on the healthy part of a feeder from hours
to minutes. Adding automation capabilities reduced
the duration with the difference between the two
response times, from minutes to seconds.
The results from the smart meter desktop study shows
that smart meters could potentially be used to more
accurately determine fault locations, and consequently
to more efficiently direct field staff to the location of
a fault to reduce outage times. However, the project
experienced quality issues with the smart meter data
throughout the trial. For smart meters to be effectively
utilised for fault detection, timely and accurate data
would need to be assured.

7.6.1 Key learnings


The smart grid is a significantly more complex entity
than the traditional electrical distribution network. It
includes an array of interconnected intelligent sensors
and actuators and multiple layers of software. A
structured and rigorous approach to testing and
configuration management is mandatory for the
safe and reliable introduction of new technologies,
algorithms and software upgrades.
Changes must be validated and tested in offline
environments that sufficiently represent the behaviour
of the production network to identify issues before
deployment. Different test environments may be
needed for distinct purposes including development
of algorithms and processes, functional testing to
validate new features and integration/regression testing
to ensure proper interworking with the production
systems. These environments must be sufficiently
flexible to allow production network changes to
be regularly back-ported to ensure that the testing
environments continue to be truly representative of the
production system.
The deployment of FDIR requires that a
comprehensive network model be prepared and
loaded into a DMS. During the trials considerable
time was spent preparing this model. An attempt
was made to predict the final state of the network at
time of commissioning, which proved problematic.
An improved approach would be to agree and build a
process with the participating vendor(s) that recognise
upfront the dynamic nature of the network and agree
on simple project BAU processes (not part of an
exception or scope change process) that incorporate
network changes as they are made.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 59

7.6.2 Applicability of solutions


toAustralia
The project has shown that deploying FDIR has the
potential to significantly improve reliability, which could
deliver benefits to customers, energy retailers and
network operators.
In Australia, network reliability targets are set under
the Service Target Performance Incentive Scheme
(STPIS), which rewards or penalises network providers
based on their performance relative to those targets.
This creates a clear business case for at least partial
deployment FDIR (up to the value stipulated under the
STIPS reward).
Investment by network businesses will also be
determined by differing reliability standards in each
state and the relative proportions of networks which
have low reliability indicators. FDIR is only as valuable
as the reliability improvement it delivers, which means
that a targeted deployment on a networks worst
performing feeders would be expected to deliver the
greatest benefits, with diminishing returns as reliability
improves. As a result, care should be taken when
applying the outcomes from this trial to the wider
network reliability indicators, given the relative reliability
improvements for FDIR will be area-specific.

60 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

8 Substation and feeder


monitoring (SFM) technologies
8.1 What are
substation and feeder
monitoring (SFM)
technologies?

SFM technology potentially offers various benefits


including providing information to help manage
outages and optimise maintenance. The systems help
avoid outages and lower maintenance and
inspection costs.

The purpose of substation and feeder monitoring


(SFM) technologies is to monitor the network
state and condition of assets within the electrical
distribution network. The Smart Grid, Smart City SFM
Project investigations study was divided into three
components focussed on transmission, substation and
distribution network components.
Currently due to the high cost of monitoring equipment
and the need for dedicated communications links,
on-line monitoring has traditionally been limited to a
small number of high-value network assets in the zone
substations. Monitoring of transmission lines, feeders,
distribution transformers and other assets outside
the substation has typically been carried out through
scheduled inspections and tests.

The Smart Grid, Smart City Project allowed Ausgrid to


investigate whether SFM smart grid technologies and
approaches could provide an opportunity for network
businesses to cost effectively monitoring a larger
number of lower-value assets.

8.2 SFM role in


asmart grid
The electrical distribution network is a complex system
that includes a variety of electrical plant distributed
over a wide area. These assets may be located in
substations or switching yards or in less accessible
locations underground or mounted high on poles.
Monitoring the condition of these assets has always
been an important activity for network operators, both
to maximise the assets economic life and to identify
and rectify potential sources of failure before they
impact supply.

Figure 1-12 SFMs role in a smart grid

Substation Feeder Monitoring (SFM)

Bulk Supply Point

Subtransmission
Substation

Zone Substation

Distribution
Substation

Domestic, commercial &


industrial customers

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 61

Due to the cost of monitoring equipment and the need


for dedicated communications links, on-line monitoring
has traditionally been limited to a small number of
high-value assets in the zone substations. Even then,
monitoring of much of the plant has been limited to the
relaying of alarms back to the control room. Monitoring
of transmission lines, feeders, distribution transformers
and other assets outside the substation has typically
been carried out through scheduled inspections and
tests. Replacement or refurbishment of high-value
assets is usually based on conservative estimates of
useful lifetime derived from historical values. Low-value
assets are usually run to failure.
The development of low-cost sensor and
microprocessor technology has reduced the cost
of monitoring equipment so that it is now becoming
realistic to consider monitoring the large number
of lower-value assets further down the distribution
hierarchy. With the addition of a network-wide common
communications platform, the need for dedicated
communications infrastructure for each technology
is eliminated and the incremental cost of monitoring
assets across the network reduces further.
The smart grid therefore provides an opportunity to
deploy a range of current and emerging technologies
to remotely monitor the state of the network and
condition of assets. The near-real time collection of
data from all levels of the distribution network has
been shown to significantly improve network reliability,
power quality, asset utilisation and network
operating costs.

8.3 SFM and Smart


Grid, Smart City trial
objectives and timelines
The SFM Project involved the trial of a number of
different technologies that could form a part of a
future smart grid SFM program. While the individual
technologies each offer certain specific benefits,
theprimary purpose of the project was to investigate,
validate, and where possible, quantify the broader
benefits that could accrue from real-time monitoring
including:
The deferral or avoidance of capital expenditure for
a network business by maximising asset utilisation
through real-time thermal rating. Real time rating
allows the asset to be used at its maximum
capacity at any point in time, rather than capacity
being limited by worst-case assumptions
The reduction in capital expenditure made possible
by the deployment of a shared low-cost scalable,
common communication platforms using
open standards
The minimisation of capital expenditure by using
real operational data to inform the planning and
design processes, reducing the number and level
of conservative assumptions incorporated into the
process
Reduction in operating costs and improvement in
system reliability (SAIDI) by identifying and rectifying
potential sources of faults before they develop into
service-affecting problems
Reduction in operating costs and improvements in
system reliability through improved awareness of
the network state
Reduction in operating costs through fewer routine
inspections and prioritisation of preventative
maintenance based on the condition of assets

62 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

The SFM Project was focussed on commercial-scale


field trials of these technologies, supported by desktop
studies and PS+EDGE simulations where relevant.
The project consisted of seventeen individual trials,
one of which (the Resonant Earthing Feasibility Study)
was limited to a desktop feasibility study. The trials
ran at more than 40 sites across the Ausgrid footprint
involving the transmission network, substations, feeders
and the distribution network. Where possible, the trials
were combined with planned capital work projects to
minimise costs and disruption.
The SFM project was a three-year project, spanning
the period from October 2010 to September 2013 as
shown in Figure 1-13.

Figure 1-13 Timelines for the SFM project

2009

2010

2011

2012

2013

Planning and Trial Design


Installation and Commissioning
Trial and Data Collection
Analysis and reporting
6
Milestones

1  Protection relay statistics Trial


Start

5 Distributed Temperature
Sensing Trial Start

2  Partial discharge analysis Trial


Start

Earth potential rise event


monitor Trial Start

11 8

3  Environmental monitoring at
distribution substations Trial Start

6 Soil thermal resistivity Trial Start

4  Initial Smart Meters deployed


in Nelson Bay

8  Transmission line monitor


TrialStart

3 2

1 5 7

10

11

9  Insulator leakage current monitor


Trial Start
10  Oil filled cable monitoring
TrialStart
11  Dissolved gas analysis Trial Start

7  Dynamic cable rating Trial Start

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 63

8.4 Where are SFM


technologies currently
deployed?
A review of literature found that international findings
were broadly consistent with the findings from
Smart Grid, Smart City SFM Project including
that substation and feeder monitoring technology is
generally considered to be a mature and cost effective
solution for voltage and current measurement and
that monitoring of thermal, gas and partial discharge
was considered to deliver net benefits only at the
subtransmission level.
It should be noted that the SFM category is very
broad (as can be seen by the number of technologies
trialled in the Smart Grid, Smart City SFM project),
which makes it challenging to accurately compare trial
outcomes within the literature. Further, the terminology
used to describe SFM technologies also varies due
to the lack of a standard global vocabulary and the
relatively fast evolution of technology in this area which
has led to multifunction devices that are difficult to
classify. The ability for new smart grid technologies to
multifunction can also lead to benefits being reported
under different categories, such as reliability benefits
which may be reported under the FDIR category
of benefits.
The SFM Technical Compendium (which can be
found on the Smart Grid, Smart City Information
Clearing House https://ich.smartgridsmartcity.com.
au/) provides additional information on a number of
international trials.

8.5 Smart Grid,


SmartCity SFM trials
8.5.1 SFM trial design elements
and technologies
As previously described, the SFM project consisted
of 17 different trials deployed in a variety of locations
across the Ausgrid distribution network. Trials
were conducted in all the network areas including
transmission, substation and distribution and are
described in the following sections.

8.5.2 Transmission and Feeder


Monitoring Trials
In the distribution network, electricity is transported
from the bulk supply points to subtransmission and
zone substations via overhead transmission lines or
underground cables at voltages up to 33 kV. The
electricity is then distributed from the zone substations
to the distribution transformers located near the
customers via a network of feeders, usually operating
at 11 kV. Transmission lines or feeders may also be in
the form of aerial conductors or underground cables.
The role of the transmission and feeder network is the
cost-effective, safe and reliable transport of energy
from one point to another. Achieving these goals
requires attention to be paid to both the conductors,
through which the electricity flows, and the insulation
that keeps the high voltages on the conductors
isolated from each other and the surrounding
environment. This sub-project examined several
technologies that have the potential to increase the
utilisation, improve the reliability or prolong the working
life of transmission lines and cables.

64 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Dynamic rating of underground cables

Overhead transmission line monitoring

The key limitation for the current rating of underground


cables is temperature. The maximum allowable cable
conductor temperature is limited by the properties
of the insulating material. The temperature rise of an
underground cable is influenced by the load on the
cable, the size and configuration of its conductors,
and the ability of the surrounding soil or backfill
material to conduct heat away from the cable. Cable
ratings are typically calculated using conservative
assumptions of the soils thermal resistivity since
this can vary considerably depending on the nature,
compaction and water content of the backfill material.
Consequently, there is potential for a cables true rating
to be significantly higher or lower than the design
value, and for this to vary over the cables length and
under different environmental conditions.

Overhead transmission lines consist of conductors


suspended from towers or poles via insulators. The
rating of transmission lines is limited by a maximum
conductor temperature that should not be exceeded.
The conductor has a maximum operating temperature
beyond which it may suffer permanent degradation
via plastic deformation. In many cases however, safety
clearances will be violated due to thermal expansion
(sag) of the conductors before this temperature has
been reached.

In the case that the design rating is conservative, the


cable may not be fully utilised, and in the worst case,
it may be prematurely replaced by a higher capacity
cable. Conversely, if the cables actual capacity is
lower than planned (for example, due to a localised
soil characteristic), a hotspot may develop leading to
premature failure22.
These investigations into the dynamic rating of
underground cables consisted of three parts including:
The first trial investigated distributed temperature
sensing technology capable of measuring
the temperature profile along the length of an
underground cable
The second trial evaluated an emerging technology
to allow the continuous monitoring of soil thermal
resistivity in the vicinity of buried cables
The final trial in this group examined technology
which utilised the cable temperature and soil
resistivity to calculate cable ratings in real time, and
to compare the results with ratings derived from the
static models currently in use by Ausgrid

22 As an example, a recent feeder failure in Ausgrids network,


was traced to the incorrect replacement of backfill during
the installation of a nearby oil pipeline more than 20 years
previously.

Maintaining the correct clearance between conductors


themselves, the ground, surrounding structures or
vegetation is particularly important in the Australian
context because of the high risk of bushfires. A range
of factors influence the temperature of a transmission
line including the load, the mechanical and thermal
properties of the conductors, and external variables
such as solar radiation ambient temperature, wind
speed and direction.
Distribution network service providers (DNSPs) design
transmission lines using thermal and mechanical
models that predict sag under various conditions, to
ensure that appropriate clearances will be maintained
throughout the lines operating life. A lines current
rating is therefore based on conservative assumptions
of temperature, wind velocity and the like. The true
value of these parameters can vary significantly from
the assumed values, and may not be identical at all
points along the line. In some cases therefore, the
design rating of the line may be underestimated, and
the asset may be underutilised.
As part of the Smart Grid, Smart City program, Ausgrid
completed a trial of technology to assist in the prediction
of temperature using real-time measurement of load and
local environmental conditions.
The objective of this trial was to evaluate the
effectiveness of continuous monitoring of transmission
line and environmental parameters to calculate
conductor temperature and tension to be used in
real-time rating of transmission lines. For this study,
technology to measure the environment, conductor
temperature and tension was deployed.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 65

Insulation breakdown monitoring


Insulation breakdowns are one of the most common
causes of failure in the transmission network. This
component of the project assessed different aspects
of insulation failure monitoring through three separate
trials including:
The evaluation of on-line partial discharge
monitoring in transmission cables
The performance of high-voltage insulators in the
presence of high levels of environmental pollution
Technology for monitoring the pressure of oil in oil
insulated transmission cables

8.5.3 Substation
MonitoringTrials
Electrical substations are important nodes in the
electricity supply network. Substations contain
transformers that convert the incoming high voltage
into a lower voltage for distribution further out into the
network. They also contain equipment that performs
switching, protection and control functions.
As well as the primary electrical system, substations
can contain one or more secondary systems. A
secondary system is defined as any system that
supports the safe, secure and efficient operation of the
primary system. Typical secondary system functions
include protection, monitoring and control. Tertiary
systems, which do not directly support the primary
electrical system, such as enterprise connectivity and
voice communications are also deployed
in substations.
The role of the protection systems is to detect fault
conditions, isolate faults to prevent damage to the
rest of the network, and to de-energise equipment
or conductors that may be in an unsafe condition.
The protection system will also attempt to maintain
electricity supply to as many customers as possible
during the fault. The reliable operation of the protection
systems is therefore critical for the safe operation of
the network.

Monitoring systems provide both real time and


historical information about the status and
performance of the network and the condition of
the network to assist the safe, reliable and efficient
supply of electricity. Control systems permit network
operators to remotely control the state of the network
in an efficient and timely manner such that the network
is kept in an optimal configuration for the
prevailing conditions.
Monitoring and control are increasing in importance
with the evolution of smart grids as improved
visibility of network state along with increased
actuation capabilities supports advanced distribution
automation technologies. This sub-project
investigates the evolution of substation-based
secondary systems in the smart grid environment,
and examines a small number of representative
technologies that this evolution enables. The
substation monitoring trials address:

Smart substation infrastructure


This project included the design of a local area
network (LAN) to support secondary and tertiary
systems within the smart grid substation. An
architecture based on the IEC 61850 protocol was
proposed. The trial assessed the first deployments of
this architecture and identified the lessons learned.

Representative smart substation


technologies
These trials examined three different monitoring
technologies that were representative of the types
of applications enabled by the smart substation
infrastructure. These included:
The on-line analysis of dissolved gas in substation
transformer oil
The use of video surveillance for substation security
The monitoring of earth potential rise at the
substation during fault events

66 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

8.5.4 Distribution
MonitoringTrials
Distribution networks carry electricity from
zone substations to customers. They consist of
underground and overhead feeders and lines that
carry the electricity, as well as distribution substations,
transformers and switching and monitoring equipment.
The distribution substation steps down the voltage
before the electricity reaches the customer. In
Australia, the majority of customers receive electricity
at around 230 volts.
The role of the distribution network is the costeffective, safe and reliable transportation of energy
to the end-customer. Achieving this goal requires
accurate information about the network state and the
health and utilisation of assets on the network.

Smart meter monitoring This trial examined the


potential to use smart meters as a tool to monitor the
low voltage network. The relative benefits of monitoring
via smart meters and via DM&C were compared.
Protection relay statistics This trial assessed
the benefits of using information available from
existing protection relays to predict failures. Statistical
techniques were used to evaluate historical data with a
view to identifying pre-fault signatures.
Resonant earthing feasibility study Resonant
earthing solutions can potentially improve safety and
reduce the threat of arcing that can cause bushfires,
under single-line-to-ground faults. This desktop study
evaluated the feasibility and cost-benefit of providing
such a system on an Ausgrid zone at high risk.

These trials examined several technologies that


have the potential to increase the utilisation, improve
reliability and asset maintenance, or prolong the
working life of distribution assets and included:
Distribution substation monitoring The
trials examined the potential benefits of extending
monitoring activities to distribution substations. This
monitoring offers the potential to improve asset
utilisation, reduce operational costs and provide early
warning of potential failures. The trial evaluated the
benefits of monitoring the distribution transformer itself,
as well as the environmental conditions at distribution
transformer sites. In addition a distribution monitoring
and control (DM&C) platform was deployed and its
performance evaluated.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 67

8.6 High level trial


results
Given the large number of trials completed as part
of the SFM Project, a high level account of each of
the previously described trial elements has been
provided in the following section. In-depth results
and discussion on the technical findings have been
included in the SFM Technical Compendium which can
be found on the Smart Grid, Smart City Information
Clearing House https://ich.smartgridsmartcity.com.au/.

8.6.1 Transmission and feeder


monitoring trial results
Dynamic rating of underground cables
trialresults
These trials demonstrated a number of aspects
including that:
Distributed temperature sensing is a very useful
component of a smart grid strategy. The sensing
technology is robust, and is able to provide real
insight into the thermal characteristics of buried
transmission feeders that are difficult or impossible
to obtain any other way. The technology is
inexpensive compared to the capital costs of the
infrastructure it monitors, with a single monitoring
node at the sub-transmission station potentially
capable of monitoring all the feeders radiating from
that point. The incremental cost of installing DTS
ready feeders is minimal.
The Soil Thermal Resistivity trial showed that the
implementation of a dynamic monitoring station
in a real world environment is possible, and that
continuous data collection potentially can aid the
dynamic rating of underground cables, as well as
suggestions to direct further research.
Real time thermal rating of underground
transmission cables showed significant potential
to increase asset utilisation without compromising
reliability. This study showed that up to 38 per cent
additional capacity was available on a five-hour
basis for one of the feeders studied, and that 114
per cent was available on another. The technology
used, while generally performing as expected,
was difficult to integrate with the DTS equipment,
and showed some anomalous behaviour at the
extremes of its measurement range.

68 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Overhead transmission line monitoring


trialresults
The trial demonstrated that accurate results for
conductor temperature could be calculated from the
conductor load and local weather conditions. The
results also highlighted the conservatism built into the
traditional static rating methods, which assume that
the wind speed is only 0.5 or 0.6 ms-1 at the time of
peak load.
In contrast the trial data has shown that the wind
speed can be substantially higher and thus the cable
may be underutilised. The use of localised real-time
weather data to calculate transmission line conductor
temperatures could therefore provide valuable input
for the real-time rating of transmission lines. Such
data could be used to avoid load shedding in peak
periods where the weather conditions permit higher
ratings than the static models permit. Further, long
term thermal data could be used to better estimate the
service life of transmission lines potentially delaying or
avoiding capital expenditure.

Insulation breakdown monitoring trial


results
The partial discharge monitoring trial was not able to
demonstrate the economic and operational benefit
of on-line partial discharge monitoring as no partial
discharges were detected, although a number of
valuable insights into this technology were obtained.
End-to-end monitoring of a sub-transmission cable
longer than 2 km would be possible only where
multiple monitoring points can be installed (presumably
at joint bays) together with the connecting optical fibre
communications infrastructure. Given the results, there
is unlikely to be an economic justification for this online
technology in the Ausgrid environment, as portable offline partial discharge analysis equipment meets cable
management needs.

Detailed analysis of the performance of insulators


in polluted environments suggested that there was
minimal risk of flashover of either the porcelain or
composite insulators. Monthly testing at Kooragang
and Anna Bay demonstrated the superior performance
of composite over ceramic insulators due to their
hydrophobic properties and significantly larger surface
creepage distances.
Results from the leakage current measurement system
demonstrated that the maximum leakage current
measured on both the porcelain post and composite
longrod insulators are well below industry recognised
warning threshold limits. Given the low levels of risk of
flashover in very polluted sites, limited benefit would be
gained by more widespread deployment of insulation
leakage current monitoring equipment.
The results of the trial of oil pressure monitoring
equipment showed that monitoring trends in oil
pressure in oil filled cables did provide useful insight
into the behaviour of the cables beyond that available
from simple low-pressure alarms. Anomalies in the
oil pressure could indicate slow leaks and equipment
failures allowing preventative maintenance to be
scheduled, and emergency call-outs to be avoided.
However, the system as trialled is not a suitable
replacement for the existing alarm system because
it takes almost two minutes to complete polling of
all the transducers. The current system can trigger
an alarm within one second of oil pressure deviating
from acceptable levels if a catastrophic leak occurs.
Any online monitoring system must be capable of fast
response to major deviations in pressure and provide
trend information in quasi real time.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 69

8.6.2 Substation Monitoring


TrialResults
The substation monitoring trial results were extensive
and complex and are discussed fully in the SFM
Technical Compendium and within the supporting
documentation. These can be found on the Smart
Grid, Smart City Information Clearing House
https://ich.smartgridsmartcity.com.au/.
The objectives of the substation monitoring trials
were to develop a suggested design approach for
a common communications platform for secondary
systems within the substation, and to evaluate three
representative technologies that such a platform could
enable. The high level results of the trials are presented
in the following section.

Smart substation infrastructure trial results


Within Ausgrid the functions of the protection
system and the monitoring and control system have
traditionally been engineered as discrete systems and
kept quite separate. However due to convergence of
technology the industry has seen this division become
progressively blurred. It no longer makes technical or
economic sense to consider the protection system
separate to the monitoring and control system. This
has also been enabled by the introduction of the IEC
61850 substation communications standard23.

This study showed that a LAN designed to support


the IEC 61850 protocol can effectively deliver
significant additional data from the substation than
current systems allow. With these additional data
comes a broadening of the stakeholders who have an
interest in the data, and so the substation LAN must
transition from being a tool for specific operations and
engineering activities to a corporate asset with many
competing requirements. This study demonstrated that
the barriers which need to be overcome in maximising
the benefit of this technology are organisational and
cultural as much as they are engineering.

Representative smart substation technology


trial results
Continuous monitoring of dissolved gas in subtransmission transformers was shown to help identify
potential points of failure, but the value that this brings
depended to a large extent on the operators asset
management strategy. Ausgrids current strategy for
managing transformers is one of run to end of life, and
the network is dimensioned to tolerate the in-service
failure of a transformer (as per legislative requirements)
and a comprehensive spares strategy ensures timely
re-establishment of network security.
The current processes for monitoring transformers for
dissolved gasses are therefore adequate for Ausgrids
purposes, and investment in the widespread rollout of
dissolved gas monitoring equipment is unlikely to yield
an economic return. If a different asset management
strategy were employed, on-line monitoring of
transformer oil dissolved gas may be an appropriate
technology, especially as the technology matures and
costs come down.
Video monitoring of substations was shown to be
capable of increasing security however the trial found
that video inspection did not replace any significant
asset management processes that would be conducted
as part of Ausgrids existing maintenance program.

23  IEC 61850 Power utility automation, International


Electrotechnical Commission, 2013

70 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

During the field trial period, the earth potential rise


event monitor proved to be robust and reliable system
to measure earth fault currents in earthing conductors
and soil voltages. The system was also successful
in monitoring the impedance of the earth grid, an
electrode, and both overhead earth wires, under both
transient and sustained earth fault events.
The number of earth faults at Nelson Bay zone
substation that resulted in potential touch voltages
hazards was found to be significantly fewer than
estimated. Monitoring found that of the 47 interruptions
recorded during the trial period, 28 had an earth fault
component, of which only four events (7.6 per cent)
resulted in a voltage exceeding 50 V. A fully integrated,
controllable and configurable production quality device
was deployed for the trial, expanding on previous
Ausgrid trials undertaken with a less sophisticated
prototype. The findings align with results from the
prototype trials.

8.6.3 Distribution monitoring trials


Distribution substation monitoring
trialresults
While continuous monitoring of distribution transformer
voltage, load and oil temperature could help identify
potential failures before they occur, the benefits
depend to a large extent on the operators asset
management strategy and the cost of the devices.
The network is dimensioned to tolerate the in-service
failure of a transformer (as per legislative requirements)
and a comprehensive spares strategy ensures timely
re-establishment of network security. Under these
circumstances, the economic value of this technology
is limited. However, the data gathered throughout the
trial may be used to improve the understanding of
asset lifecycles that could be useful for replacement
prioritisation.

The environmental monitoring trials found that


various environmental measurements are technically
feasible however are of varying benefit. Ambient
temperature was not found to be strongly correlated
with transformer oil temperature, but helped to
analyse the premature failure of field-based RTUs, and
identified issues with other temperature monitoring in
the field. Humidity monitoring was less valuable than
temperature as it is currently not an input for asset
management models at Ausgrid. A flood sensor was
installed at a distribution substation that was prone
to flooding, and four flood events were detected over
the trial period. Airflow sensors were not found to be
justifiable on Ausgrids network due to the very limited
number of suitable distribution substations.

Smart meter monitoring trial results


The trials showed that smart meters could be used
for low voltage distribution monitoring purposes.
Since Ausgrid already has an extensive DM&C
program underway, deploying smart meters solely
for distribution monitoring purposes is not justifiable.
However, smart meters do provide other potential
benefits such as support for remote meter reads, timeof-use tariffs and other feedback technologies. These
benefits are discussed at length in the Customer
Applications Technical Compendium and in Parts Two
and Three of this report.
In summary, the smart meter monitoring trial results
showed that smart meters could potentially be used to
determine the location of faults which could be used
to more efficiently direct field staff and reduce fault
restoration times.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 71

Protection relay statistics trial results


The trials showed that protection relay statistics were
useful as a forensic tool to analyse pre-fault conditions
on assets to determine the cause of the fault. However,
the results from the trial shows that the pre-fault
conditions were not picked up early enough for the
operator to take any action to avert the fault.
More evaluation of this approach over a longer time
period and across a wider range of devices is required
to determine the likelihood of solid faults actually being
averted through the use of the technology.
The ability to log protection relay statistics is a free
and common feature of modern relays, meaning that it
would be preferable to turn on this feature to determine
if there is a class of pre-faults that can be prevented.
Importantly, this is achievable at a low cost once the
smart grid common platform enables the central
retrieval of the full range of data, including
waveform files.

Resonant earthing feasibility study results


Trial results showed that resonant earthing was a
prohibitively expensive technology to deploy at existing
sites. In a new substation build, a resonant earthing
solution would be more cost-effective given that the
ancillary works that dominate the installation costs are
carried out as part of the new build.

8.6.4 Summary of the substation


monitoring trials
The overall findings for the substation monitoring trials
showed that given the reduction in costs of sensors
and electronics, combined with a deployment of a
common communications platform, the potential to
cost effectively monitor the small number of highvalue assets at a zone substation is improving. The
substation and feeder monitoring projects have shown
that while most of the technologies examined bring
value, the business case for any one of them at this
stage will be difficult to justify on a stand-alone basis.
The true benefit of the smart grid can be realised only
when a whole-of-business view is embraced.

8.6.5 Applicability of solutions


toAustralia
A consistent result from the SFM trials is that these
smart grid technologies allow for optimisation of a
number of variables and grid performance over a
longer period of time, whereas existing processes are
focussed on boundary condition analysis focussed on
network peak events. In order to extract the full value
from the smart grid technologies, a continuous view
and optimisation design process for the network is
required, rather than boundary condition analysis using
a single peak snapshot of the network.
Smart grid technology vendors should also be
aware that their technology will be used as part
of a multi-vendor smart grid implementation.
Scalability, interoperability and fully conforming and
validated standards-based interfaces are mandatory
requirements.
Given these factors impacting the integration of
SFM technologies, Utilities must not only ensure
that equipment supports appropriate protocols but
should also seek assurances from vendors of the
conformance of protocol implementations. Utilities
must also be capable of and prepared to assist
vendors with testing equipment protocol conformance.
This may include providing a vendor with an
appropriate level of access to test environments.
Many of the SFM technologies trialled as part of the
Smart Grid, Smart City Program required alternative
approaches to analysis, data and communication
between organisational structures in order to be
successfully deployed.
Another important finding that network businesses
will need to recognise is the need for effective change
management within an organisation. This will require a
range of activities and commitments at all levels within
the business. A number of these issues are discussed
in Part Three of this report.

72 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

9 Wide Area Measurement


(WAM)
9.1 What are Wide
Area Measurement
(WAM) technologies?

This data is then used to provide a global view of the


health of the transmission network including stability,
power flow direction and congestion.

The deployment of Wide Area Measurement (WAM)


technologies resulted from a series of record-breaking
blackouts that occurred in both North America
and Europe in 2003. The US-Canadian blackout
investigation report that followed determined that many
major blackouts were caused by inadequate situational
awareness for grid operators, and recommended
the use of WAM technology to provide real-time,
wide-area grid visibility as a means to help eliminate
future blackouts24. These events and subsequent
investigations had an impact globally, as the electric
power industry recognised more urgently the potential
consequences of the current limitations in wider-area
network awareness.
The concept of WAM involves the installation of
monitoring devices, called phasor measurement
units (PMUs), at a number of key points across the
transmission network. These devices are required to
provide a high speed sample of data (voltage, current
and frequency) at strategic points on the transmission
network. Referred to as synchrophasor, it is measured
with a highly accurate synchronised time signal.

The ability to dynamically monitor and analyse the grid


is becoming increasingly important as greater levels
of power electronics and variable energy sources are
deployed within the distribution network. WAMs role in
a smart grid is described in Figure1-14.

9.2 WAMs role in a


smart grid
At the highest tier of power supply, transmission
networks usually stand as the first line of defence
against impending wide-scale outages, making
effective protection and safe operation of primary
transmission assets the main responsibility of network
operators. This is a key area of improvement in the
smart grid environment, as protection equipment can
begin to take into account a wider view of the network
state, and increased visibility of network conditions
can improve safety outcomes.
Protection equipment on a transmission network is
traditionally configured to provide localised protection
of the asset it is installed on. Hence, this protection
scheme is set to view a localised portion of the
transmission network with no particular awareness of
the state of the transmission network as a whole.

Figure 1-14 WAMs role in a smart grid

Substation Feeder Monitoring (SFM)

Bulk Supply Point

Subtransmission
Substation

Zone Substation

Distribution
Substation

Domestic, commercial &


industrial customers

24  U.S.-Canada Power System Outage Task Force (2014)


Final Report on the August 14, 2003 Blackout in the
United States and Canada, April 2014
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 73

This localised view of the network gives rise to the


possibility of a wide area outage through cascading
protection operations across the network, which may
be initiated by a fault or switching error at a particular
site that is not dealt with quickly or effectively.

However, it is now widely accepted that SCADA


and EMS can be significantly enhanced with the
complementary use of WAM technologies because
this new, high frequency data clearly gives operators a
more real-time view of network events.

WAM systems enable highly precise visibility of


electricity system conditions over multiple levels of
the electric grid. Continuous system measurement
is not a foreign concept for utilities conventional
Supervisory Control and Data Acquisition (SCADA) and
Energy Management Systems (EMS) have traditionally
provided this capability for protection and control
scheme purposes.

Figure 1-15 provides a high level description of the


progressive improvements that have occurred, and
are possible because of smart measurement systems
such as WAM.
As Australias current grid evolves, many more devices
(in particular, distributed devices) will be connected to
electricity networks and the capability to monitor key
network indicators with greater accuracy is becoming
critical for grid operators.

Figure 1-15 Evolution of grid modelling with synchrophasor data

Past

Current

Opportunity

Future

Peak
Modelling

Months

Static Loads
in Real-Time

Minutes

Dynamic
Modelling of
Interacting
Components

Seconds

Transient
Analysis

Cycles

74 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

9.3 WAM Smart


Grid, Smart City trial
objectives and timelines
The overarching goal of the Smart Grid, Smart City
WAM Project was to explore and analyse the use of
WAM technology in the Australian context. The project
sought to inform the business case for deploying
synchrophasor measurement technologies across a
portion of the transmission and distribution networks.
Potential project benefits were investigated by:
Consolidating and disseminating synchrophasor
data to appropriate stakeholders
Using synchrophasor data to analyse network
events and monitor key network health parameters
Using synchrophasor data for the comparison and
validation of models and key measurements
Conducting a literature review and developing
PS+EDGE simulations to understand WAC
applications in the changing grid environment
The WAM Project consisted of a field trial element
involving several PMUs along transmission and
subtransmission high voltage corridors, along with
additional studies and PS+EDGE simulations to
investigate WAC feasibility for Australian distributors.

The WAM Project spanned three years, from October


2010 to September 2013 (see Figure 1-16).
The first two years of the project consisted of
design and installation phases. The WAM system
was operational by the third year, and analysis and
reporting was carried out during the final phase.
PS+EDGE simulation work was also carried out during
the final phases of the WAM project.
WAM technologies were chosen as part of the Smart
Grid, Smart City Program because it was expected
that the findings would contribute to its broader
objectives by:
Improving reliability from increased post-event
data and information to facilitate more effective
responses and minimise the impact of future events
Eliminating poorly informed investment decisions
by incorporating highly accurate data into current
modelling and analysis practices
Deferring capital expenditure through improving the
utilisation of existing network assets
Full details on the WAM project deployments and
results can be found in the Ausgrid WAM Technical
Compendium and on the Smart Grid, Smart City
Information Clearing House which can be found at
https://ich.smartgridsmartcity.com.au/.

Figure 1-16 Timelines for the WAM project

2009

2010

2011

2012

2013

Planning & Trial Design


Procurement &
Compliance Testing
Installation & Commissioning
Trial and Data Collection
Analysis & report
1
Milestones

1 Beresfield STS 14/06/2012

4 Peakhurst STS 14/03/2013

2 Singleton STS 14/08/2012

5 Mason Park STSS 28/03/2013

3 Ourimbah STS 20/06/2013

6 Newcastle BSP 29/04/2013

3 4 56 7

7 Muswellbrook BSP 29/04/2013


Beaconsfield West BSP 29/04/2013
Sydney South BSP 29/04/2013
Sydney North BSP 29/04/2013

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 75

9.4 Where are WAM


technologies currently
deployed?
Although synchrophasor technology has been utilised
over the past few decades, major benefits have not
been realised, typically because measurement points
have been deployed in isolation.
Whilst unproven, there is a widely held hypothesis
that high-speed analysis of these data across a
broad transmission grid can be used to provide
early detection of transmission network failure.
This hypothesis is supported by current literature,
which proposes control strategies that have been
successfully implemented on simulated networks.

9.4.1 Australian deployments


Primary WAM deployments in Australia have been
installed on behalf of the Australian Energy Market
Operator (AEMO), who is responsible for enforcing the
National Electricity Market (NEM) Frequency Operating
Standards. AEMO currently models and monitors three
known mainland inter-area oscillatory modes25:

These range of different oscillation frequencies


represent the swinging of many machines against
others26.
At the present, AEMO also receives data from five
PMUs:
New South Wales from TransGrid Sydney West
Substation
Victoria from Rowville Substation
South Australia from Para Substation
Queensland from South Pine Substation
Queensland from Greenbank Substation
PMUs have been mainly beneficial to AEMO in
identifying oscillatory stability issues. Synchrophasor
data is also known to be used in disturbance
monitoring, stability and risk assessment, control
design and separation (islanding) management. AEMO
does not, however, use PMUs in wide area control
(WAC) applications at present27.

QNI mode: Has a frequency in the 1.6 to 2.2 rad/s


(0.25 to 0.35 Hz) range. Queensland machines
oscillate against the machines in the other three
mainland states at this mode frequency
I25 mode: Has a frequency in the 2.4 to 3.1 rad/s
(0.38 to 0.49 Hz) range. South Australia and
Queensland machines oscillate against the New
South Wales machines at this mode frequency
I35 mode: Has a frequency in the 3.2 to 3.8 rad/s
(0.51 to 0.6 Hz) range. South Australia and New
South Wales machines oscillate against Victoria
machines at this mode frequency

25  Electro-mechanical oscillations between


interconnected synchronous generators

26 Prasertwong et al, Understanding low-frequency


oscillations in power systems, International Journal of
Electricity Engineering Education, July 2010 
27  Ausgrid, GA1239 WAM Benefits Study, SGSC
Supporting Document, 2013

76 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

9.4.2 International deployments


International WAM activity has been led by PMU
deployments in China, North America, and Europe,
with an estimated installed base of 3,000-4,000 PMUs.
Public data shows that North American networks
have installed the majority of PMUs; however, some
sources suggest that China is actually leading in PMU
deployment. Utilities within these regions are exploring
a range of core WAM applications, as well as more
advanced wide area monitoring, protection and control
(WAMPAC) systems28.
In North America, the American Recovery and
Reinvestment Act (ARRA) of 2009 has been a key
catalyst for WAM system development across North
American transmission systems. Twelve funding
recipients, including both transmission and distribution
networks, are spending $400 million to deploy
synchrophasor technologies and are exploring a
variety of applications, including wide area monitoring
and visualisation; voltage stability monitoring; islanding
and restoration; post-event analysis; and
model validation29.

Major WAM technology deployment in Europe has


been largely enabled by public-private partnerships
between utilities and nation-states as well as funding
from the European Commissions Framework
Programmes. EU nations are exploring WAM
technologies to provide more operational synergy
across the multitude of transmission system operators
(TSOs), where there is a recognised need to enhance
network visibility and improve monitoring across
interstate transmission corridors and interconnections.
WAM systems have been deployed noticeably over the
past decade as European TSOs aim to achieve
these objectives.
China is leading PMU deployment in Asia, whose
networks notably procure all equipment from domestic
manufacturers and use their own specifications and
standards. Chinese PMUs cover all 500kV substations
and generators over 100MW, and networks are
exploring WAM applications such as:
Real-time dynamic status monitoring and alarm
processing
Low frequency oscillation identification and control
Model validation and parameter identification
Wide area protection against cascading trips
State estimation using PMU measurements30
There is also a considerable amount of technical
research underway which is testing the feasibility
and architecture of WAM systems within distribution
networks. Details on some of these studies are
provided in the Ausgrid WAM Technical Compendium.

28  Kruimer B., Wide Area Measurement integration into


grid operations, MD Quanta Technology Europe, 6
December 2011
29  U.S. Department of Energy, Synchrophasor
technology and their deployment in the Recovery Act
Smart Grid Programs, August 2013

30  Bi T., The dynamic behaviour of PMU and the latest


development in China, North China Electric Power
University, 8-9 May 2013

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 77

9.5 Smart Grid,


SmartCity WAM trials
9.5.1 Trial design and deployment
Within the Smart Grid, Smart City Program, Phasor
Measurement Unit (PMU) devices were deployed
over approximately one year at select sites in the
Hunter and Sydney regions of NSW. Devices were
installed at both transmission and subtransmission
substations, therefore involving Ausgrid as well as
TransGrid, the NSW transmission network operator.
Trial sites were selected based on three main criteria:
proximity to TransGrids network; direct connection
to subtransmission substations; and geographic
dispersion enabling the measurement of phasors
across a wide area forming the basis of the
WAM system.

A literature review was carried out as part of the Wide


Area Control (WAC) study component of the WAM trial.
Given WAC is a relatively new concept, findings from the
research were used to assist engineers and decisionmakers in understanding the challenges and benefits of
WAC schemes in the Australian context.
The WAC study also included PS+EDGE simulation
work, which used a dynamic model developed
as part of a STATCOM study undertaken at the
University of Newcastle. The goals of the PS+EDGE
simulations included: the accurate modelling of events
that are transient in nature within the networks; and
to demonstrate and understand the dynamic grid
impacts of future scenarios involving the addition of
dynamic generation sources at the distribution level.

To manage and consolidate the complex multiple


synchrophasor streams, a communications and
backend system was necessarily deployed. An
open platform was chosen due to its scalability
as a software solution and for ongoing software
development. A prototypical web dashboard was
alsodeveloped which was used to visualise and
analyse real-time WAM datasets and ultimately to
demonstrate the capability for real-time monitoring
inthe operational environment.

78 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

9.6 High level


trialfindings
9.6.1 Operational use of WAM data
Synchrophasor data captured several events over
the course of the WAM trial. Three of these events
were analysed in detail including a generator trip,
a flashover31 and the opening of a switch on a
subtransmission feeder. Processed WAM data
provided several benefits to engineers for post-event
analysis including:
For the generator trip event, the availability of
WAM data made the investigative analysis faster
than would have been otherwise possible, and
also demonstrated how a transmission incident
can be wide-reaching even though the overall
impact is low. From this event, it was found that
a WAM system could provide near-term benefits
in the form of supplementary data used to more
quickly identify the causes of similar events, either
due to rate of change of frequency or excursions
of system frequency, information of which is not
available from current SCADA systems.
WAM data analysed from the flashover event further
showed the limitations in using current SCADA
systems to effectively analyse short-lived but
damaging fault events. The data from WAM was
able to prove that the event started four seconds
before it was indicated by the SCADA system.
More useful was the ability to use the WAM data
to accurately view resulting fault currents at subsecond intervals to confirm the sequence of events.
Analysis of WAM data during the switching
event allowed for both the frequency of resulting
oscillations to be estimated and the dampening of
modes of the oscillation, a critical component of
stability analysis.
31  A flashover in terms of an electrical network occurs
where an electrical current travels through the air (or
other insulation) rather than the intended conductors.
Generally this would occur because the insulation
around the conductors is made conductive through
(for example) salty water, smoke, pollution, a tree
branch etc. These other materials offer an alternative
path for the electrical current other than its designed /
intended path. A flashover can potentially cause a fire
and / or also damage the equipment and / or pose a
safety risk.

Analysis of the trial concluded that minimal process


change would be required for the addition of WAM
in current operations. However, a commercialscale rollout would need to consider how phasor
measurement data would be retrieved, stored,
analysed and presented to operators in conjunction
with existing data retrieved via SCADA.
Understanding of communications (especially latency),
GPS clock signals for timing, and limitations of existing
transducers are required for implementation and
analysis at the operational level. Further information on
WAM technical functionality that is typically available in
current relay technology is provided in the Substation
Feeder Monitoring Technical Compendium32. However,
while that functionality exists, enablement of that
functionality also requires suitable relay configurations,
telecommunications infrastructure and back-end IT.

9.6.2 Planning benefits through


model and measurement validation
In addition to fault analysis, WAM data were used
to examine the accuracy of existing measurement
processes of a number of network systems including:
WAM data were used opportunistically
to investigate power and line impedance
measurements from incumbent systems by
taking snapshots of SCADA measurement data
from two feeders and comparing them to raw
WAM data over a period of one month. This
comparison showed that errors in SCADA power
measurements were as large as 164 per cent. This
may have been due to limited verification of power
measurement accuracy during the commissioning
process, however, considering SCADA
measurements are used in existing modelling and
design practices, there is an inherent risk if the
appropriate level of data on which to make network
investment decisions is not available. A consistently
high erroneous load measurement could result
in a feeder capacity upgrade or the installation of
capacitor banks earlier than is actually required.

32 Ausgrid, Smart Grid, Smart City Substation Feeder


Monitoring Technical Compendium

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 79

Conversely, a consistently low load measurement


could result in assets being overloaded without
operator knowledge, eventually leading to
premature asset failure.
These investigations found large discrepancies with
variances from 0-28 per cent. These discrepancies
can lead to errors in fault current calculations and
potential damages to primary network assets from
protection relay misconfigurations.
WAM data also provided an opportunity to validate
grid models, which are primary sources for capital
expenditure rationalisation. Engineers in protection
and planning groups in both the Sydney and
Hunter regions compared network WAM data to
modelling outputs. The Sydney subtransmission
planning group found that measured WAM
values and model estimates were within one per
cent. For the Hunter group, WAM data identified
misalignments between equipment in the model
and physical equipment in the field. This was able
to explain large power flows that were experienced
when an open point between two sites was closed
by operators.
It was determined that WAM data could present a
key opportunity for utilities to upgrade their current
measurement practices, verify modelling and design
approaches and ultimately make more efficient
investment decisions33.

9.6.3 Applicability of solutions


toAustralia
Australian networks are largely radial, as opposed
to the highly interconnected systems and multiple
network market participants that are seen in North
American and European networks, in which much
of the synchrophasor technology is applied. This
feature reduces the likelihood of wide-scale outages
on Australian networks. Therefore, the benefits of
eliminating such outages via the use of WAM
systems on its own may not justify the cost for
Australian networks.

However, the trend of an increasing number of market


participants and small distributed generators in
Australia (and in most parts of the world) may influence
decisions to deploy WAM in the future. The ability
of network operators to monitor and control these
distributors and other dynamic devices is low with
current SCADA systems. Not only is accurate and
high-speed data necessary to fully understand the
power quality impacts of distributed generators and
dynamic devices on the grid, but it could serve as
an enabler for the optimal placement, operation and
utilisation of dynamic generation sources and
storage devices.
In the near term, because the trial demonstrated that
WAM data could validate network models, Australian
utilities may see benefits if improved data is used to
inform investment decisions.
A major challenge for Ausgrid as well as other
networks in the near term will be to access sufficient
synchrophasor data at strategic sites throughout a
network. Strategic partnerships among transmission
and distribution networks, as well as coordinating
efforts from AEMO to improve grid stability, could
assist and set the stage for successful future
WAM deployments.
As is evident from the Smart Grid, Smart City WAM
trials, although the benefits from WAM deployment
could be significant, the inherent complexity of these
benefits and the lack of benchmarking data currently
available in the industry make it challenging to quantify.
This is a key learning from the Smart Grid, Smart City
Project and highlights the need for further studies
within the industry to generate these data and inform
the future business case.

33 Ausgrid, GA1179 WAM Data and Model Verification


Trial Report, SGSC Supporting Document, 2013

80 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

10 Electric Vehicles (EVs)


10.1 What are
electric vehicle (EV)
technologies?
There are a number of different types of conventional
internal combustion and electric vehicles which are
available both within Australia and globally including
the following categories:
Internal Combustion Engine Vehicle (ICE) Represents the majority of private vehicles,
powered by a standard internal combustion engine
using petrol, diesel or gas.
Hybrid Electric Vehicle (HEV) - Combines both an
ICE with an electric engine. The electrical energy
is stored in a battery with the battery charged by
the internal combustion engine, however, battery
capacity is generally limited. Vehicle propulsion
is a mix of the ICE and electric engine, but is
predominantly powered by the ICE. This vehicle
does not use energy from the electricity grid, that it,
it does not plug in.
Plug in hybrid electric vehicle (PHEV) - Combines
both an ICE with an electric engine. Electrical
energy is stored in batteries with the batteries
charged by plugging into the grid. Vehicle
propulsion is a mix of the ICE and electric engine,
but is predominantly powered by the electric
engine. The ICE component is used to extend
driving range beyond battery capacity for longer
distances and to recharge the battery itself.
Battery Electric Vehicle (BEV) - Powered only by
energy stored in batteries with batteries charged by
plugging into the grid.

Electric vehicle technology includes the vehicles,


charging infrastructure and the electricity grid itself.
The exchanges between these components,
both physical and financial, define the electric
vehicle ecosystem.
The market demand for fuel efficient vehicles has
increased sharply over the last decade due to the
increasing cost of fuel and, at least to some extent,
environmental concerns. Electric vehicles offer
considerable benefits over even the most efficient
conventional internal combustion engine (ICE) vehicles.
For electric vehicle owners, benefits include reduced
fuel costs and potential convenience where home
or workplace based charging is available. Broader
societal benefits include reduced local environmental
impacts and energy security by reducing dependence
on oil based fuels. In some countries, electric vehicles
offer carbon reduction benefits, depending on the
carbon emission intensity of the local electricity supply.
In recognition of these benefits, many governments
across the world now offer incentives for consumers to
purchase electric vehicles.
Even so, electric vehicles are currently cost prohibitive
for the majority of road users, and for the most part are
being purchased by a small group of early adopters.
The attractiveness of electric vehicle technology to
consumers is further limited by the capacity and
reliability of battery technology and the availability
of charging infrastructure. These factors, whether
perceived or real, continue to influence the technology
uptake rates.
Notwithstanding these factors, many mainstream
vehicle manufacturers are now introducing a range
of electric vehicles into their product range. This is
leading to improved volumes with consequential
impacts on manufacturing meaning that the end-price
of EVs is reducing and may become more financial
viable for a greater number of consumers in the
near future.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 81

10.2 The role of electric


vehicles in a smart grid
Wide scale adoption of electric vehicles has the
potential to transfer a large proportion of Australias
demand for transport energy from liquid fuel to the
stationary energy sector. This new load represents a
potentially disruptive impact for the electricity industry
and for electricity network operators in particular if a
significant number of vehicles are being charged during
the existing network peak periods.
Conversely, electric vehicle load also represents a
potential opportunity for network operators. Network
utilisation34 could be improved where electric vehicle
charging is managed through the use of pricing signals
or directly controlled to occur during non-peak times.
There is also the potential for electric vehicles to
operate as a battery and export electricity back into
the grid, known as Vehicle-to-Grid (V2G) technology.
This could further enhance the opportunities for
electric vehicles to contribute to load management
within the electricity system.

The challenge for the electricity industry is to


adequately plan for increased numbers of electric
vehicles to ensure that the potential impacts do not
impose unacceptable costs in terms of network
upgrades due to charging during peak periods and
that any potential opportunities are captured. This
issue was captured in the final advice to the Australian
Energy Market Commission in 2011 by AECOM in its
report on electric vehicles35 where it noted that:
to recharge an EV, it must be connected to,
and draw electricity from, a distribution network
(or embedded network). This connection may
cause both direct connection costs (such as
the cost of an extension to the consumers
premises) and shared augmentation costs
(thatis, costs to augment the shared segments
of adistribution network).
The smart grid has a significant role to play in
managing electric vehicle load though the provision of
data and information to allow controlled charging and
to encourage efficient customer charging behaviour.

Figure 1-17 An electric vehicle from the Smart Grid, Smart City trial charging

34  Network utilisation is the proportion of time, usually


expressed as a percentage, that an electricity asset is
being used. It is more economically efficient to build
an asset that will used for a high proportion of the
time, rather than invest in assets that will only be used
for a relatively small number of hours each day or
year.

35  AECOM final advice to AEMC, Electric vehicles


NEM arrangements to incentivise efficient charging
behaviour, 2011, page 17.

82 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

10.3 The electric


vehicle Smart Grid,
Smart City Trial
objectives and timelines

trials, uptake and behaviour modelling and grid impact


modelling, which are described further in the
following sections.
The timeline and high level schedule for the Smart
Grid, Smart City Electric Vehicle Project is described
in Figure 1-18.

The overall objective of the Smart Grid, Smart City


Electric Vehicle Project was to understand the potential
impact of wide-scale uptake of electric vehicles in
Australia on the electricity distribution network.

Twenty Mitsubishi iMiEV electric vehicles were


delivered to Ausgrid in December 2010, with the
vehicles being showcased and a data collection
system for the trial being developed in 2011. The
vehicles were assigned to departments throughout
Ausgrid and were driven for business duties until the
data collection system was field tested.

To date, the majority of studies in Australia have


focussed on travel behaviour, driver perceptions,
uptake rates and the impact of electric vehicles on
the electricity network at a whole-of-system level
(a number of these are described in the following
sections). Many of these studies identified that
the greatest potential impact is likely to be at the
distribution level, but to date investigations into the
magnitude of this impact tend to be generalised. In
addition, existing studies did not link spatial analysis
of electric vehicle uptake with localised feeder
impacts. This study represented the first investigation
in Australia which has investigated the sensitivity of
distribution network to spatial variation in electric
vehicle uptake.

The first Better Place charge point was installed in


June 2011. Prior to the charging points being available,
vehicles were charged with 15A wall outlets. In
February 2012 the vehicles were reallocated, some to
new owners in Ausgrid for the fleet trial, the remainder
to households as part of the home trial.
The first fast charger was commissioned in February
2012 in Homebush and in June 2012 the first public
charging point was commissioned at Newcastle
Airport. The road trials ran until May 2013, collecting
detailed vehicle usage data for the entire period.

The project was informed by four distinct trial


components: charging infrastructure deployment, road

The analysis, uptake and behaviour modelling, grid


impact modelling and reporting were performed
throughout 2013.

Figure 1-18 Electric vehicle trial timeline

2009

2010

2011

2012

2013

Data Collection System


Fleet Trial
Home Trial Round 1
Home Trial Round 2
Uptake and Behaviour Study
Simulations
Analysis & Reporting
1 2
Milestones

1 First Quick Charger Commissioned

3 Last Charging Point Commissioned

2 First Public Charge Point Commissioned

4 Last Quick Charger Commissioned

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 83

10.4 Where are electric


vehicles currently
deployed?
Since around 2009, there have been a number
of Australian and international studies which have
investigated a range of aspects relating to electric
vehicle operations, safety and user behaviour and
acceptance. These have been described briefly in the
following sections.

10.4.1 Australian Electric Vehicle


deployments and studies
Victorian Government Electric Vehicle Trial
The Victoria Government Electric Vehicle Trial,
scheduled to be completed in mid-2014, involved
trialling a range of electric vehicle types by 120
individual household participants and 70 corporate
fleet participants for three months or more. The
vehicles trialled were both Original Equipment
Manufacturer (OEM) products and post-sale electric
vehicle conversions.
The first round of the trial in 2010 to 2011 was open
to all Victorians. Applicants to the trial were found to
be environmentally conscious, with positive attitudes
towards electric vehicles and supported government
action to investigate and promote electric vehicles.

In addition, charging load management was also


successfully achieved using partial load reduction
strategies tailored to the vehicle characteristics. This
tailoring meant that the charging of vehicles can be
slowed instead of being stopped during periods of
higher demand.
All the vehicles in the Electric Vehicle Trial were
supplied using GreenPower from AGLs Bogong hydroelectric plant.
The Victorian Electric Vehicle Trial Mid-Term Report
discusses experiences and lessons learnt with
charging infrastructure rollout. These findings relate to
technology maturity, timelines to deployment, hurdles
and costs. Additionally, a driver survey was performed
as part of the Victorian EV trial.

Western Australian Electric Vehicle Trial


The Western Australian Electric Vehicle Trial (WA
EV Trial) was the first EV fleet trial in Australia and
commenced in November 2010. Eleven Ford Focus
sedans were converted to fully electric operation
because at the time there were no OEM EVs in the
Australian market. The vehicles were used in eleven
government and private sector fleets in the Perth
metropolitan area for a period of two years. The trial
also included a network of 12 public charging stations.

The more recent round was only offered to customers


within specific locations within the electricity network
in order to better understand the impact on the
distribution grid. The Victorian trial also incorporated
smart meter infrastructure to control electric vehicle
charging, trialling both demand response and direct
load control schemes.

84 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

AEMC - Energy market arrangements for


electric and natural gas vehicles

AECOM Impact of electric vehicles and


natural gas vehicles on the energy markets

Significant policy work has been completed by the


Australian Energy Market Commission (AEMC) to
develop a national electric vehicle charging policy.
Its December 2012 report into energy market
arrangements for electric vehicles (as well as natural
gas vehicles) identified areas for reform to promote
efficient EV charging behaviour and to enhance
consumer choice. These reform areas are principally
in relation to the role of pricing signals and metering
arrangements for electric vehicles.

AEMCs 2012 report was underpinned by technical


analysis undertaken by AECOM into the likely uptake
rates and associated grid impacts at the National
Electricity Market (NEM) level. The key conclusion from
AECOMs analysis was that if charging is unmanaged
(i.e. there are no signals to encourage electric vehicle
consumers to charge away from times of peak
demand), then this could result in significant additional
peak demand, imposing further costs to the electricity
system. Given these findings, it is important that
there are appropriate energy market arrangements in
place to manage the impact of electric vehicles on the
electricity system.

Key recommendations made by the AEMC included:


Electric vehicle charging should eventually all be
on a cost reflective tariff. Pricing signals should be
used to manage charging to encourage customers
to charge at times that lead to efficient
market outcomes
All electric vehicle chargers energy usage should
be measured with interval meters to enable time
varying tariffs
Arrangements should be made such that electric
vehicle owners can delegate to another party the
right to control how their electric vehicle charges
Metering arrangements should be made such
that all electric vehicle charging energy is metered
separately to general premise energy
Customers should have the ability to have a
different electricity retailer for electric vehicle energy
than for general premise energy
Electric vehicle charging should have the ability to
be sub-metered to reduce electrical and metering
installation costs for electric vehicle chargers

The study presented plausible uptake rate scenarios


and justifiable assumptions. The study identified that
there would be significant spatial differences in uptake
rates, however it did not present a method or perform
a spatially sensitive study.
The study considers the increase in peak demand on
the NEM under different charging schemes such as
uncontrolled, time-of-use and smart charging. The
study found that the NEM will be able to support the
additional load of electric vehicles at the projected
uptake rates with minimal required upgrades if drivers
are sufficiently motivated to charge at off-peak times
through the charging scheme. Without incentives there
would be a significant increase in peak demand in the
NEM, and billions of dollars of generation upgrades
would be required to support electric vehicles under
this scenario.
Additionally the study investigated the financial
implications specifically the capital expenditure
required to support electric vehicles under given
scenarios as well as the benefits gained through
increased network asset utilisation. The study found
significant reductions in the retail price for electricity
could result from increased network utilisation as a
result of electric vehicles charging off-peak.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 85

This study investigated the future electric vehicles


network impact challenges at a macro level (the
entire NEM), rather than at a more granular level that
concerns distribution networks. Notwithstanding,
the study did show that electric vehicles will have the
largest impact at the distribution level but that further
work is required to determine the extent of this impact.
The Smart Grid, Smart City Electric Vehicle Project
aimed to measure this impact.

CSIRO Electric Driveway Project


The CSIROs Electric Driveway Project investigated
methods to integrate electric vehicles into household
power systems as a distributed system for storing
off-peak power from the national grid or from on-site
renewable sources.
Field trials included the development and testing of
two special-purpose electric vehicles integrated with
a home energy management system, used to partially
power homes.
Four separate research reports were produced which
evaluated the impact of widespread electric vehicle
use in an Australian context, detailing the barriers
and benefits of adopting the technology, and
exploring potential future synergies between the
different components of Australias electricity and
transport sectors.
The Phase 1 report describes the initial findings the
project. A discussion on electric vehicle adoption in
Australia is presented along with detailed modelling
to be executed later in the project. Like the AECOM
study, this study undertakes an analysis of the impact
of EVs with respect to the available generation
capacity. However, the CSIRO analysis is relatively high
level and is only applicable to Victoria.
The CSIRO also investigated three charging scenarios
- valley filling smart charging; vehicle-to-grid; and
uncontrolled charging. The initial findings concur
with other studies, suggesting that valley filling
smart charging is required to leverage available
generating capacity. The study also found that there

is no business case for implementing vehicle-to-grid


charging from the customer perspective. The study
also presented the findings from a case study on two
distribution feeders on SPAusNets network in Victoria
for three charging scenarios.

AGL Energy - Electric Vehicles in the NEM:


Energy market and policy implications
This research paper by AGL examined the likely impact
of electric vehicle charging on the NEM. An analysis
of the market, costs, drivers and possible uptake
scenarios is presented. High, medium and low
uptake scenarios are presented based on modelling
and drawing from other studies, but are not
spatially sensitive.
Two scenarios are considered for impact on the NEM
generation capacity, convenience charging and off-peak
charging using a model of vehicle availability for charging
developed from household travel data. The study found
that even under the high uptake scenario, the impact
on the NEM would be minimal and is unlikely to trigger
investment in additional generation and transmission
network capacity. The study considers a shock uptake
scenario with 50 per cent penetration and concluded that
even if all these vehicles charged on the peak demand
day in 2009/2010, there would still be enough available
generation capacity in the NEM if drivers charged offpeak. The paper did not consider vehicle-to-grid enabled
charging as the technology was perceived to be too far
off at this point in time to warrant consideration.
The paper also recognised that the electricity grid was
most likely to experience capacity issues as a result of
electric vehicle charging at the distribution level, rather
than the generation sector.

Electricity Supply Association of Australia Sparking an electric vehicle debate


In November 2013 the Electricity Supply Association
released a discussion paper exploring the potential
of plug-in electric vehicles to contribute to the
transformation of transport vehicles in Australia. The
paper outlined the range of benefits of electric vehicles

86 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

including greenhouse gas emission reduction, local


environmental benefits, energy security, improved
vehicle performance and safety and most significantly
improved utilisation of electricity networks.
Rather than advocate any particular policy, the report
sought to encourage a discussion about the benefits,
and how to encourage EV uptake in order to realise
these benefits.

10.4.2 International Electric


Vehicle deployments and studies
Pacific Northwest National Laboratory
(PNNL) - Impact assessment of plugin hybrid vehicles on pacific northwest
distribution systems

The US EV Project
The US EV Project37 was the worlds largest
deployment of electric vehicle charging infrastructure
and is being used to support thousands of electric
vehicles over 21 major cities. This trial investigated the
use of time-of-use tariffs to shift charging demand
to off-peak.

Green eMotion
The European Green eMotion trial38 includes ten
demonstration regions where deployment of more
than 2,500 charging stations is underway. The trial
investigated usage patterns to suggest the likely
impact of electric vehicles on system peak.

Cenex - The Smart Move Trial

This study aimed to determine the extent to which


the existing US light duty vehicle fleet (cars, utes, SUVs
and vans) could be supported by the US distribution
networks if all these vehicles were plug in electric
hybrids. The unutilised capacity available in the US
power grids was examined and compared with the
required energy to power the US light duty vehicle
fleet. The study builds upon a previous PNNL study
into available generation capacity but focuses on the
distribution feeder and distribution substation level.

Cenex, a UK based consortium, undertook a road


trial39 where participants were given an electric vehicle
to drive for a period of time during which qualitative
and quantitative data was collected. There were
several other notable trials that followed including the
Cabled trial, run by Aston University in the UK, and the
BMW Mini-E trial, run by BMW in the UK and Germany.
The aim of these trials was to assess the real world
feasibility of using electric vehicles as everyday
passenger cars.

The study adopted a power flow model for two typical


distribution feeders in different climatic regions with an
electric vehicle load to investigate the extent to which a
typical feeder could support electric vehicles.

All three trials presented information on driver


behaviour and vehicle performance with metrics such
as: average trip duration, time of day vehicle used,
vehicle efficiency, trip distances, power required and
charging times. In addition, a qualitative analysis was
performed into drivers response to electric vehicles
including perception of performance, suitability of the
available range, drivers attitudes to electric vehicles
and charging preferences.

36

36 Pacific Northwest National Laboratory, Impacts


Assessment of Plug-In Hybrid Vehicles on Electrical
Utilities and Regional US Power Grids. Part 1
Technical Analysis, 2007 http://energyenvironment.
pnnl.gov/ei/pdf/PHEV_Feasibility_Analysis_Part1.pdf
- 2007

37  US Department of Energy, ECOtality, The EV Project,


Ongoing http://www.theevproject.com/
38  Green eMotion, Ongoing,
http://www.greenemotion-project.eu/
39 Cenex, Smart Moves, 2011,
http://www.cenex.co.uk/case-studies/smartmoves/

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 87

10.5 Smart Grid,


SmartCity Electric
Vehicle trials
10.5.1 Trial design and deployment
Electric vehicle technology is not, in itself, a smart grid
technology. Wide scale uptake of electric vehicles,
where charging goes unmitigated, has the potential
to result in significant impacts to the electricity grid in
terms of increased system wide and localised peak
demand and capacity constraints. Such impacts are in
direct conflict with the broad objectives of a smart grid.

Figure 1-19 Summary of electric vehicle trial outputs

Charging
Infrastructure
Deployment

Road Trails

Uptake and
Behaviour
Model

Individual
charging profiles
(by location)

Charging
Characteristics

Trial
Outputs

Charging
infrastructure
cost and
feasibility

Travel behaviour

Travel behaviour
(by location
Australia wide)

Operating costs
and GHG
emissions

Uptake rates
(by location
Australia wide)

Used to inform / validate

Grid
Simulations

Used as direct input

88 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Aggregated
charging profiles
(by network level
over time)

Grid impact
(by network level)

The aim of the Smart Grid, Smart City Electric Vehicle


Project was designed to understand the likely scale,
location and timing of these impacts and to inform the
appropriate industry response. The project involved
two key activity streams including:
Field trials
Road trials of 20 Mitsubishi i-MiEV electric
vehicles including 13 vehicles assigned to 24
residential customers and seven fleet vehicles
Deployment of six fast and 46 standard
charging points at locations within the
Ausgrid network
Trial participant and broader public
perception surveys
Modelling
Uptake and behaviour modelling of electric
vehicle uptake rates and travel behaviour across
Australia over the next three decades
Grid impact modelling of the potential impacts
of electric vehicles to electricity distribution
networks using the PS+EDGE advanced
modelling platform
Grid impact modelling and evaluating
techniques to mitigate the impact of electric
vehicles on electricity distribution networks
The overall design of the trial and its outputs is
described in Figure 1-19 and discussed further in the
following sections.

Field trial components


Charging infrastructure deployment The
charging network consisted of 46 standard charge
points and six fast charge points distributed over the
Sydney, Central Coast and Newcastle regions. The
charging network was deployed in conjunction with
the charge point provider, Better Place. There were
many difficulties overcome in this process due to
the inherent immaturity of electric vehicle technology
and the local industry in Australia. In general,
significant barriers to the installation of public charging
infrastructure were found - for standard charge points,
the majority of the barriers were non-technical and
related to the selling of the business case to site
owners, whilst for fast charge points, it was difficult to
find locations in the existing low voltage infrastructure
where equipment could be accommodated electrically.
Home charge points were also installed for the home
trial participants.
Road trials The road trials occurred between
August 2011 and May 2013 and involved the
deployment of 20 Mitsubishi i-MiEV electric vehicles
to private users in the Home Trial and business
users in the Fleet Trial over a period of 24 months.
Thirteen of the vehicles were assigned to Ausgrid
staff for personal use in the Home Trial and seven
were assigned to Ausgrid, Lake Macquarie Council,
Newcastle City Council, Sydney Water and Newcastle
Airport for use in the Fleet Trial.
A survey of Home Trial participants was undertaken
both before and after the trials as well as a broader
public perception survey to understand the likely
factors influencing electric vehicle uptake into the
future. The road trials provided data and information
on charging patterns as well as the advantages and
limitations of various charging regimes.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 89

Uptake and behaviour modelling The project was


informed by an electric vehicle uptake and behaviour
model prepared by AECOM40. The model provided
spatial analysis of predicted electric vehicle uptake
rates and travel behaviour under various scenarios
across different areas of Australia. The model
assessed the spatial and temporal characteristics of
electric vehicle ownership and travel patterns which
ultimately influenced charging behaviour.
The AECOM model provided highly granular data
on EV usage over a 25 year timeframe, from 2015
to 2040. Outputs were provided for each SA25 area
across Australia and included the following:
Number of vehicles per household by vehicle type
Travel behaviour in terms of arrival and departure
times and average trip distance
Electric vehicle uptake rates
Electric vehicle efficiency and battery capacity
overtime
Outputs were provided for weekday/weekend,
summer/winter, by vehicle type and by garage location,
under three macro-economic scenarios (high, central
and low). The data collected in the road trial was used
to verify and refine a number of assumptions in the
AECOM model, thereby increasing its robustness and
accuracy. In particular, the road trial data was used
to validate the assumption that electric vehicle driver
behaviour does not differ significantly from travel
behaviour in internal combustion engine vehicles within
the range constraint of the electric vehicle.

Grid impact modelling To complement and


validate the field trials, a range of advanced analytical
modelling activities were undertaken as part of Smart
Grid, Smart City Electric Vehicle Project. The modelling
tool developed and utilised for assessment of potential
impact of EVs is known as the Production System
+ Extended Data Grid Environment (PS+EDGE).
PS+EDGE was used to complement, extend and
validate the field trials through a range of advanced
analytical modelling activities which used actual results
from the field trials and performed further analytical
modelling of scenarios that could not be implemented in
the field trials.
The trial area was reconstructed in PS+EDGE to
replicate actual trial conditions and to allow for real
trial data to be fed into the model for further testing
and analysis. This model contains over 100,000
separate objects interacting across the network. The
high resolution PS+EDGE platform was developed
to assess the impact of electric vehicles on the
grid, using time-series load flow analysis on parts
of Ausgrids distribution network and parts of the
distribution network of other DNSPs collected through
the Smart Grid, Smart City feeder taxonomy. The
grid impact modelling allowed a detailed assessment
of network performance to be performed under
various scenarios for various locations. This enabled
an understanding of when and where a network was
over-stressed or where supply was likely to go out
of specification.
Outputs from the electric vehicle uptake and behaviour
modelling were used to develop estimates of individual
charging behaviour under various charging schemes,
including an unconstrained scheme, which either
incentivise or control electric vehicle charging. The
charging behaviour varied by individual customer
depending on their travel behaviour.

40 AECOM, EV0795 - Electric Vehicle Uptake and


Behaviour Modelling

90 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

A number of grid impact modelling scenarios were run,


varying parameters each time, to test different rates
of electric vehicle uptake, different charging schemes
and the impact on different distribution feeders within
the electrical network. To keep the number of grid
impact modelling scenarios to be run at a practical
number, three typical combinations of feeder were
selected. These three combinations were considered
either typical or likely to cover the significant impacts
of electric vehicles across the majority of feeders
in Australia.
Grid impact models were run for a number of different
scenarios to test the effectiveness of various charging
schemes on reducing the potential impact of peak
demand and mitigating potential overloading of the
distribution network. Full details of the modelling
results can be found in the Electric Vehicle Technical
Compendium on the Smart Grid, Smart City
Information Clearing House website which can be
accessed from https://ich.smartgridsmartcity.com.au/

10.6 High level trial


findings
Modelling showed that by around 2025, electric
vehicles were projected to make up approximately five
per cent of the Australian vehicle fleet. At these levels,
the electricity grid would begin to require significant
and costly capacity upgrades to support the additional
demand of electric vehicle charging if charging was left
unmanaged. Urban residential feeders were likely to
see the greatest impact.
However, if managed carefully (incentivising vehicle
charging to off-peak), electric vehicles could
be accommodated by the existing primary grid
infrastructure without need for major upgrades even
at levels of up to 40 per cent of the Australian vehicle
fleet.

10.6.1 Requirements for public


charging infrastructure
The lack of available public charging infrastructure
was perceived to be a barrier to electric vehicle
update notwithstanding the fact that part of the broad
appeal of electric vehicles is the convenience of home
charging. The road trial combined with the AECOM
electric vehicle uptake and travel behaviour modelling
suggested that the majority of charging was likely to
occur at home. It also indicated that public charging
was likely to be limited to charging for infrequent long
distance trips, for unexpected or emergency charging
requirements or for electric vehicle owners without
the ability to install charging infrastructure in their own
homes (noting that people unable to install a home
charger are unlikely to purchase an electric vehicle).

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 91

10.6.2 Charging rates at home


At the time of writing, the charge time for electric
vehicles at home charge points (typically over two
hours) represented a barrier to electric vehicle uptake
for customers accustomed to short refuelling times for
conventional ICE vehicles. In the road trials, customers
reported slow charge times to be a significant barrier
to electric vehicle uptake. Consumer demand for
shorter charge times was therefore likely to drive
technological development into the future.

10.6.3 Impact of lower EV numbers


on the distribution network
The grid impact modelling highlighted that the impact
of electric vehicle charging was likely to be more
extreme at lower levels in the network, i.e. distribution
substations. This was due to electric vehicle chargers
representing a potentially large point load and an
increased probability of coincident charging at points
in the network with fewer customers (and therefore
less diversity). As a result of this reduced diversity, the
grid impact modelling showed that the proportional
increase in peak demand at the distribution substation
level tended to be greater than at the feeder level.
Distribution substations, particularly those that are
heavily constrained, and the corresponding low
voltage networks were therefore more likely to require
capacity upgrades before the feeder itself as a result of
electric vehicle charging.

10.6.4 Range of intervention


responses
The primary aim of any intervention was to ensure
that electric vehicle charging load did not add to the
existing peak and did not create a new peak, thereby
reducing or eliminating network augmentation costs.
A number of different strategies were investigated to
mitigate the impact of electric vehicle charging and
which utilise the discretionary nature of the vehicle
charging load. There was a need to deliver a set
amount of energy to each vehicle between the time it
arrived home and needed to depart on its next journey.
However, there was flexibility in terms of when and
at what rate the energy was delivered. The nature of
this discretionary load gave rise to a range of potential
intervention responses.
Potential intervention responses included time-of-use
charging, controlled off-peak charging, constant power
charging and smart charging. Both the smart charging
and constant power charging proved to be the most
effective in reducing the impact on peak demand at
both the feeder head and distribution substations.
The constant power scheme was simple and
effective and could be encouraged by the use of a
capacity-based tariff.
While both schemes were observed to be effective
in reducing peak demand, the smart charging
scheme had a far greater benefit than the constant
power scheme in terms of improved load factor. The
controlled off peak scheme, using existing technology
for off peak hot water heaters, also held potential with
careful implementation.

92 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

10.7 Further
considerations for
Electric Vehicle
deployment for
Australia
10.7.1 Need for pricing signals
Should the uptake of electric vehicles occur in the
future as modelled by the Smart Grid, Smart City
Electric Vehicle Project, the electricity supply industry
will need to take action to manage the increased
load these vehicles will place on the grid. Failing to
adequately manage electric vehicle charging has the
potential to result in a number of negative impacts.
Proactive action in the form of pricing signals and
metering arrangements when electric vehicle numbers
begin to increase would reduce the likelihood of
requiring significant policy changes or costly grid
impacts in the future. Where appropriate charging
schemes are in place, electric vehicles also offer
potential benefits to in the form of improved network
utilisation and therefore the economic efficiency of
network operations. This need for action has also been
identified by AEMC in their December 2012 report
into energy market arrangements for electric vehicles,
principally in relation to the role of pricing signals and
metering arrangements.

10.7.2 Vehicle-to-grid technology


Part of the DGDS trial of the Smart Grid, Smart City
Program was to understand the potential benefits
associated with the use of plug-in electric vehicles as a
means of distributed storage, or vehicle-to-grid.
Similar to customer batteries, the main customer
value that may drive this concept is highly dependent
upon the retail tariff arrangement or other customer
incentives for battery discharge provided by retailers or
network operators. The strongest current price signal
is the price differential in ToU prices between off-peak
charging and consuming energy at peak times. For
example, where a partially or fully charged electric
vehicle is at home during peak times (2:00 pm to 8:00
pm on working weekdays), there is the potential to use
energy stored in the electric vehicle to offset electricity
purchases at the high peak rate of 52.547 c/kWh41.
Under the EV Project an advanced modelling
framework (PS+EDGE) was used to investigate the
impact on distribution networks of three uptake
models for electric vehicles. These uptake and travel
behaviour models were developed by AECOM, based
on modelling of Australian Bureau of Statistics (Census
2011) data. This allowed for the modelling of individual
electric vehicle trips associated with each vehicle,
which was then associated with each residential
property and thus the impact on the network from the
additional temporal load from charging was able to
be determined.
While the grid impact modelling showed that vehicleto-grid technology could be used to reduce peak
demand, there are technical and financial barriers to
its wide-scale deployment in the near term future.
The grid impact modelling for the Elermorevale feeder
demonstrated that charging and discharging of the
vehicles was able to enhance grid utilisation from
3pm 6pm. However vehicle-to-grid technology was
less effective at reducing the morning peak due to the
majority of vehicles disconnecting and leaving for work
at that time.

41 FY2012-13 NSW residential regulated retail prices for


Ausgrid network
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 93

10.7.3 Electric Vehicle


environmental credentials
While the local environmental benefits afforded by
electric vehicles, including improvements in local air
quality in urban areas and noise reduction are well
established, it is not clear whether electric vehicles in
the current Australian context have a greenhouse gas
reduction benefit.
Further, the grid impact modelling showed that
electric vehicle charging profiles were not generally
complementary to solar photovoltaic generation
profiles. Both the results of the road trials and the
AECOM EV Uptake and Travel Behaviour Model
suggested that the majority of private vehicle charging
was anticipated to occur at home during the evening
and overnight, whilst solar generation is available
during the day. This suggests that the concept of
large scale supply of solar energy for electric vehicle
charging either directly from a rooftop solar PV system
or via the grid was unlikely.

94 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

11 Smart Meter
Infrastructure (SMI)
11.1 What are smart
meters and smart meter
infrastructure (SMI)
technologies?

The purpose of trialling smart meter infrastructure in


the Smart Grid, Smart City Program context was to
evaluate the effectiveness of the most advanced smart
meter technology available at the time of deployment
to support sophisticated smart grid applications both
in customer residences and on the grid.

Smart meter infrastructure (SMI) is a complex and


integrated assembly of technologies that need to
work together seamlessly as part of an effective
smart grid. The elements include the smart meter,
communications network, meter management system,
meter firmware and associated back office systems.
Smart meters themselves are an integral and enabling
component of the broader smart meter infrastructure
and smart grids.

Importantly, the smart meters chosen for the Smart


Grid, Smart City trial had to work with the previously
chosen communications and meter management
system technologies, and meet the operational
requirements of the other Smart Grid, Smart City
trials. The meters also had to meet Ausgrids existing
requirements for customer metering and AEMOs
market rules, given they would provide billing
information to Ausgrids customers.

The Smart Grid, Smart City Program defined a smart


meter42 as an electronic device that measures and
records the consumption of electricity over time,
usually in time periods of one hour or less, and
that has bi-directional communications capability
permitting communication with the network service
provider and potentially to other devices.

11.2 Smart meter


infrastructures role
ina smart grid

Additional smart meter functionality also includes the


ability to undertake network power quality monitoring,
power outage notification, enables direct load control
of hot water systems and provides an interface for
other utility meters such as smart gas or
water meters.

Within the Smart Grid, Smart City Program, smart


meter infrastructure played an enabling role in the
customer applications, Active Volt-VAr control (AVVC),
electric vehicles (EV), distributed generation and
distributed storage (DGDS), fault detection isolation
and restoration (FDIR) and substation and feeder
monitoring (SFM) trials.

Like smart meters, interval (or time-of-use) meters have


the ability to monitor power consumption in specific
time periods, but lack the two-way communications
features that form part of a smart meter. Interval
meters enable the option of time-of-use tariffs for
customers, but do not provide the platform to deliver
the broad range of applications that smart meters
can enable.
42  Within the Smart Grid, Smart City Program, a smart
meter was defined as an interval meter (i.e. an
electronic meter that records electricity consumption
in intervals of an hour or less) with bidirectional
communications capabilities. As defined in the
National Electricity Rules (Version 62, 2014), the
smart meters trialled were a Type 4 meter system.
Interval meters without communications capabilities
are classified as Type 5 meters, while spinning disc
meters, which represent the majority of meters
currently in Australia, are classified as Type 6.
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 95

Real-time electricity usage data supplied by smart


meters enables the introduction of more sophisticated
network and retail tariffs, informs customers about
their pattern of consumption and provides valuable
data sources for grid side applications which can
inform network operators of outages and network
conditions. More specifically, during the Smart Grid,
Smart City trials smart meter infrastructure enabled:
Remote collection of customer electricity usage
data, the storage and analysis of these data, and
the delivery of these data to feedback devices in
the customers residence
Remote control of household appliances such as
the cycling of air-conditioning units for the purpose
of managing demand during peak events
Off-peak scheduling for hot water systems
(replacing the currently used ripple control
system)43
Data logging capability, capturing time-series
measurement data that contributed to the
understanding of the behaviour of the distribution
network at a very detailed level for the Distributed
Storage and Distributed Generation trials. This
included the provision of voltage and power quality
data to assess the effect of distributed generation
and distributed storage devices on the network
Real time data collection to assess the impact of
charging electric vehicles on the grid as well as
providing the consumption and billing information
for each vehicles charge station
Real time data collection that allowed grid
applications like AVVC, SFM, and FDIR to evaluate,
monitor and control technologies to help improve
the asset utilisation, power quality, reliability and
efficiency of the electricity network

The Smart Grid, Smart City Program also investigated


the possibility of multi-service smart meters which
could undertake remote readings for not only electricity,
but also water and gas. Details of the findings of
these investigations can be found in the Smart
Meter Infrastructure Technical Compendium on the
Smart Grid, Smart City Information Clearing House
website which can be accessed from https://ich.
smartgridsmartcity.com.au/

11.3 Smart meter


infrastructure and the
Smart Grid, Smart City
trial objectives and
timelines
The purpose of trialling smart meter infrastructure
in the Smart Grid, Smart City trials was to evaluate
the effectiveness of the most advanced smart meter
technology to support sophisticated smart grid
applications both in the residential locations and within
the grid. The aims of the Smart Grid, Smart City Smart
Meter Infrastructure Project were to:
Supply the enabling smart meter infrastructure
required by a range of Smart Grid, Smart City
trials and evaluate the effectiveness of the selected
enabling technology in a commercial-scale trial
Use the experiences of the Smart Grid, Smart
City smart meter infrastructure deployment to
add to the body of knowledge in Australia and
internationally, including the relative benefits of
retail and network driven rollouts (this is discussed
further in the following section)
Inform the integrated cost benefit analysis
(presented in Part Two of this report) to the
Australian Government
The timeline for the Smart Meter Infrastructure Project
is shown in Table 1-6.

43 Note that hot water load control scheduling using SMI was
not applied to Ausgrid customers during the Smart Grid,
Smart City Program due to software issues that occurred
which meant that updating of the meter schedule caused
the loss of billing data or the meter to stop responding.
These issues are discussed in the Smart Grid, Smart City
Customer Applications Technical Compendium

96 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Table 1-6 Smart Grid, Smart City Smart Meter Infrastructure Project timelines

Task

Start (dates inclusive)

End

Technology Selection Selection of communications


medium, meter hardware and
software and smart meters

Prior to Smart Grid,


Smart City Project
Commencement.

Mid 2011

Technology Testing

Testing Release Cycles

April 2011

June 2013

Technology
Procurement

Delivery of meter and


communications modules

Sept 2011

July 2012

Communications
Medium Rollout

Rollout of WiMAX Cellular 4G


network solution

Prior to Smart Grid,


Smart City Program
Commencement

Early 2013

Back Office IT
System Integration

Integration of the end-to-end


IT solution

2010

2013

Meter Rollout

Customer Applications
Network Trial rollout

Oct 2011

Aug 2012

Customer Applications Retail


Trial rollout

Nov 2012

July 2013

Grid Trial rollouts

Sept 2011

TBC

DGDS and EV Trail rollouts

April 2011

TBC

De-provisioning of
communications

March 2014

March 2014

De-provisioning
of SMI

Description

11.3.1 The importance of smart


meter infrastructure systems
integration
There was a significant degree of technical complexity
in selecting the smart meter technology to ensure that
the chosen system would integrate and communicate
with other common IT platforms; either operational
within Ausgrid at the time of the project or for those
selected for the Smart Grid, Smart City Program.
Some of these issues were related to the lack of
standardisation of different smart meter infrastructure
systems, and in some cases, the preference by
vendors to rely on proprietary interfaces to achieve
interoperability between different equipment
and platforms.

These issues presented a challenge for Ausgrids


existing systems and processes and took a
considerable amount of time to overcome. Some of
the challenges Ausgrid needed to overcome included:
Selection of an appropriate metering configuration
Interfacing the smart meters with the meter
management system (MMS) which provided the
interface between the smart meters in the field and
Ausgrids back office systems
The need for the smart meter infrastructure to work
with the Field Area Network, Home Area Network
and Meter Management System technology
configurations

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 97

Interfacing the smart meters with Ausgrids back


office systems - Ausgrids business systems did
not support all of the required functionality for
the smart meter infrastructure. Ausgrids existing
systems at the time were batch-oriented systems
better suited to traditional quarterly meter reading
than the real-time reading capabilities of smart
meters. As a consequence, additional back office
systems were developed to manage this capability
including:
Customer Information System (CIS) which
functioned as the master repository for Smart
Grid, Smart City customer information
Customer Relationship Management system,
located in the CIS, was developed to support
helpdesk operations and the execution of
marketing campaigns to trial participants
Customer acquisition application which required
an iPad application to be developed in-house to
support and manage the customer acquisition
campaign and provide the sales agents with
real time updates of the status of sales targets
Meter data repository which was a separate
application built in-house to store mater data
Asset configuration system which held details
of, and the relationships between, Smart Grid,
Smart City assets including meters, generators
and on-premise devices
Work management system (WMS) which
supported the provisioning workflow for
customers. The WMS system received work
requests from the CIS, tracked the progress
of the provisioning of each technology and
reported progress on workflow back to the CIS
The challenges and solutions that Ausgrid developed
are discussed in full in the Smart Meter Infrastructure
Technical Compendium on the Smart Grid, Smart
City Information Clearing House website (https://ich.
smartgridsmartcity.com.au/).

11.4 Where is smart


meter infrastructure
currently deployed?
11.4.1 Smart metering
infrastructure in Australia
Policy positions around the deployment of
smart meters in Australia
Discussion around the relative merits and costs of a
mandated national rollout of electricity smart meters
in Australia has been ongoing over the last decade.
In 2012 the Australian Energy Market Commission
(AEMC) recommended in its Power of Choice review
of demand side participation in the National Electricity
Market (NEM), expanding competition in metering
and related service provision to residential and
small business customers. The Council of Australian
Governments Energy Council44 (COAG Energy Council)
agreed to support expansion of competition in
metering and related service provision to all customers,
consistent with a business-led, optional approach
to adoption of more advanced metering in states
where a widespread roll-out is not underway. COAG
Energy Council also agreed to remove the provisions
in the National Electricity Law relating to mandated
smart meter deployments, given the perceived risk to
investment of leaving this option in place45.
The primary driver for many of these considerations
has been the agreed need to implement demand
management mechanisms to shift electricity demand
away from the peak period. Demand management
is intended to reduce the need to build additional
network infrastructure which is only used for a few
hours each year during periods of extreme demand
and which contributes to higher network (and retail)
electricity tariffs for consumers.

44 Formerly known as the Ministerial Council on Energy (MCE)


and most recently the Standing Council on Energy and
Resources (SCER) and includes all Australian Government,
state and territory government energy Ministers
45 National Smart Metering Program Business Requirements
Work Stream, 16 May 2011, Smart Metering Infrastructure
Minimum Functionality Specification

98 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

The Victorian Government mandated the deployment


of smart meters46 across almost three million
customers in the state, to be completed by 30 June
2014. As at the end of 2013 the deployment was
around 90 per cent complete.
At the time this report was written, no other state or
territory governments had mandated the deployment
of smart meters in Australia.

Standards and regulations around smart


meter infrastructure
At the commencement of the Smart Grid, Smart City
Program, there were no agreed standards or regulations
regarding smart meter infrastructure functionality in
Australia. Although minimum functionality requirements
were released in May 2011, they were published
for information purposes only and were not official
standards or regulations. At the time this report was
written, the Australian Energy Market Operator (AEMO)
had been tasked with establishing and maintaining a
smart meter functionality specification in the context of
market-led meter deployments (termed customer-led in
this report).

Smart meter infrastructure trials in


Australia
In 2013, the then Department of Resources, Energy
and Tourism commissioned DNV KEMA to provide a
comprehensive national picture on the status of smart
meters and interval meters throughout Australia47.
At the time of the DNV KEMA report publication, there
were over 1.5 million smart meters installed in Australia
(representing fewer than 15 per cent of all meters). Of
these, 96 per cent were deployed in Victoria, where
the network smart meter rollout is nearing completion.
The remaining 52,000 smart meters were installed in
the other states as part of trials similar the Smart Grid,
Smart City Program.

In addition, there are more than 1.5 million interval


meters (not two-way communications enabled)
installed in Australia, generally where new or
replacement meters were required, or for large
electricity consumers. Interval meters allow the
network operators to charge network time-of-use
tariffs and retailers to charge retail timeof-use tariffs
to customers.
The DNV KEMA report indicates that a number of
network operators, retailers and a small number of
independent service providers throughout Australia
were conducting, or had conducted, smart meter
based trials, driven by a range of objectives including:
Investigations of dynamic peak pricing
Different communication methods
Load and metering research including solar PV,
remote meter reading, direct load control
Customer service improvements including
outage detection, home area networks, remote
disconnection and reconnection
Grid improvements including neutral
integrity testing
Comparing all of these trials, the Smart Grid, Smart
City SMI trial had the widest range of objectives and
deployed, by far, the greatest number of meters. This
excludes the Victorian Government mandated rollout
of smart meters which is not considered a trial.
The DNV KEMA report suggested that drivers for
smart meter deployments included:
Government mandated rollout
Technical trials of meter functionality,
communications and pricing by network operators
Support of distributed generation (mostly solar) trials
Facilitation of demand management technology
Customer or retailer request
Introduction of time-of-use tariffs

46 Noting that as per the previous discussion, the smart meters


being deployed in Victoria lack the functionality of two way
communications and could be described as interval meters
rather than a smart meter
47 DNV KEMA, National Smart Meter Infrastructure Report,
Department of Resources Energy and Tourism, 2013
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 99

The report noted a number of key commercial risk


factors or barriers for smart meter infrastructure
deployments in Australia including:
Uncertainty around market rules and regulations
Complexity around customer communications and
product offerings
Inconclusive or negative business cases for smart
meters
Split incentives for smart meter deployment
including challenges between allocating the
benefits between network operators and retailers
The potential for stranded assets and cost recovery
challenges
Technology risk including lack of national smart
meter infrastructure standards
A number of these challenges were experienced and
wherever possible, overcome, within the Smart Grid,
Smart City Program. The resulting conclusions and
recommendations relating to these challenges are
detailed in the Smart Grid, Smart City Smart Meter
Infrastructure Technical Compendium (https://ich.
smartgridsmartcity.com.au/).

11.4.2 International deployments of


smart metering infrastructure
As previously shown in Table 1-3, most regions in the
world have programs which have seen the deployment
of varying numbers of smart meters, often as part of
broader smart grid deployment strategies. To date
investment in smart grid technologies has centred on
smart meter investment (given it is a critical enabling
smart grid technology). This investment has typically
been mandated by governments and includes large
scale deployments in California, Texas, Italy, Sweden
and Victoria. Smaller rollouts also occurred in New
Zealand and the Netherlands. Currently European
countries are the global leaders in deployment of smart
meter infrastructure, whilst countries like Australia,
Germany and the US (as a whole) are significantly
further behind.
Early deployments (first generation) smart meter
infrastructure occurred in Italy, Sweden and
Pennsylvania. Second generation metering technology
deployments occurred in California, Texas, Ontario and
Spain, whilst current rollouts (third generation (or the
newest technology)) are occurring in France, British
Columbia and the United Kingdom. As whole, it can
be seen that significant funding has been invested in
these deployments. A significant proportion of these
large-scale smart meter infrastructure deployments
were mandated either by industry or government. In
other cases they have been voluntary. Over the past
five years many smart meter infrastructure trials have
been sponsored by governments with drivers including
peak demand reductions, carbon reductions and job
creation (particularly in the US).

100 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

A review of globally significant deployments of smart


meter infrastructure conducted as part of the Smart
Grid, Smart City Program found that grid operators
were the most common entity responsible for the
deployment of smart meter installations. Only in two
instances (New Zealand and the United Kingdom)
have electricity retailers been made responsible for
meter deployment.

Network operator vs retailer smart meter


deployment
The DNV KEMA Australia report48 reviewed the
experience in two markets that are rolling out smart
meters using different regulatory mechanisms. In New
Zealand the major electricity retailers have demonstrated
a business case that justifies the deployment of smart
meters and therefore have implemented programs to
introduce smart meters for their customers. It is expected
that around 75 per cent of New Zealand customers will
have smart meters by 2015.
In contrast, smart meter infrastructure is being
deployed in Texas (in the United States) by its
transmission and distribution utilities. The four largest
utilities have commenced deployment plans with three
of these due for completion by 2013 and the smallest
of the four due for completion in 2016. The business
case for smart meter infrastructure predicted savings
and benefits for customers, distributors, retailers and
the transmission system operator. Retailers in Texas
are currently using the installed smart meters to market
new tariff products, with pre-payment a commonly
selected option.

48 DNV KEMA, Smart Meter Infrastructure Report, Department


of Resources, Energy and Tourism, 2013

United States Green Button initiative


One creative and innovative program which has
focused on addressing a number of the issues
identified by the DNV KEMA report is the US
Governments Green Button Initiative49.
Green Button is an industry-led effort that responds
to a request from the American Government to
provide electricity customers with easy access to
their electricity usage data in a consumer-friendly
and computer-friendly format via a Green Button
onelectric utilities websites.
Green Button is based on a common technical
standard developed in collaboration with a publicprivate partnership supported by the Commerce
Departments National Institute of Standards
and Technology (NIST) and the Smart Grid
Interoperability Panel (SGIP). Voluntary adoption
ofa consensus standard by utilities across the US
has allowed (and continues to encourage) software
developers and other entrepreneurs to leverage a
sufficiently large market to support the creation of
innovative applications that can help consumers make
the most of their energy usage information.

49 www.greenbuttondata.org

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 101

Initially launched in January 2012, utilities committed


to provide Green Button capability to nearly 12million
households in 2012. Two of the major utilities
Pacific Gas & Electric and San Diego Gas & Electric
implemented live functionality on their websites.
Following this adoption, 35 major utilities and electricity
suppliers signed on to the initiative, committing to
provide many more millions of households secure
access to their energy data. In total, these additional
utilities commitments will mean that the current
42million households with this functionality now will
increase to around 57 million by the end of 2014.
The focus of this initiative is to provide households
and businesses with consumer-friendly information in
order to assist them save money on their electricity
bills, determine the optimal size of roof top solar panels
and identify opportunities for energy efficiency. Energy
usage information includes the electricity or energy50
consumed every 15 minutes, every day or every month,
depending data availability.
Consumers are also able to use the Green Button
Connect My Data icon on the website which allows
them to authorize a third-party service provider to
receive direct access to their Green Button data.
These authorizations are valid for an agreed time and
can be revoked at any time by the consumer.

11.5 Smart Grid,


Smart City smart meter
infrastructure trials
11.5.1 Trial design and deployment
Smart meter infrastructure played a key enabling role
within many of the Smart Grid, Smart City trials and
18,229 meters were installed as part of the Smart
Meter Infrastructure Project.
These smart meters were allocated amongst the trials
as follows:
17,134 smart meters for the Customer Applications
trial
560 smart meters for the Grid Applications trial
460 smart meters for the Distributed Generation
and Distributed Storage trial
75 smart meters for the Electric Vehicle trial
The smart meter infrastructure was supported by a
meter management system and back-office systems
as previously described.

Given the standardisation of the functionality, a


number of companies are developing Web and
smartphone applications and services for businesses
and consumers.

50 Gas and water information may also be available in the


future for some consumers, dependent on the utilities joining
the program

102 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

11.6 High level


trialfindings
Ausgrid made 10 key high level observations and
conclusions based on deploying and trialling smart
meter infrastructure in the Smart Grid, Smart City
Program, summaries of each are described in the
table below. For a more detailed discussion of the trial
findings and observations, refer to the Smart Meter
Infrastructure Technical Compendium which can be
found at https://ich.smartgridsmartcity.com.au/.
Table 1-7 Smart Grid, Smart City Smart Meter Observations and conclusions

Observation

Discussion

Conclusion

Testing of
smart meter
infrastructure

Smart meter infrastructure testing occurred over a


period of 18 months, significantly longer than the
typical 6 month period to test meters in other trials.
The extended duration of meter testing was due to
a change in vendor mid-project and re-setting of
the testing program due to meter metrology update
and the communications module. The integration
of the meter management and back office
systems commenced in 2010 and continued to be
developed, refined and tested through to 2013 as
each part of the solution was deployed.

In planning the rollout of smart


meter infrastructure, sufficient
time should be factored into
the market research, planning,
testing and refinement of
smart meters, communications
modules and supporting IT
infrastructure.

The time required


to test smart meter
infrastructure
was longer than
expected.

Training and
equipping of staff

Appropriate planning, training and equipping of staff


is important in reducing issues associated with a
rollout and improving rollout efficiency. Training for
Appropriate
field staff should include those who will be installing
training and
new metering technology as well as office-based
equipping of staff is
staff who are responsible for customer support,
critical
data processing and management.
Training is required to deal with new
communications and advanced customer
application technologies.

Training and equipping of staff


in operational processes related
to the installation, testing and
maintenance of SMI is required
to ensure that issues associated
with a smart meter rollout can
be appropriately resolved and
the rollout can be carried out
efficiently.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 103

Observation

Discussion

Conclusion

Technology
selection and
maturity

The Smart Grid, Smart City Program evaluated a


wide range of SMI-enabled applications, which
required smart meters and the associated MMS
to have a comprehensive list of features not
previously demanded in the market. As different
vendors were at different stages in the technology
maturity spectrum, some vendors offered part of
the required functionality, however few were able to
offer all of the required features off the shelf.

The technology selection


process for the Smart Grid,
Smart City Program found
that smart meter technologies
were at different stages of
maturity and that some desired
functionalities were not available
at the time. Care should be
taken in specifying smart meter
requirements that arent yet
developed in the market and
should ensure that a platform
exists for value
added innovation.

Smart meter
infrastructure
technology was
found to be
relatively immature

It is recommended that businesses intending


to provide smart metering should apply an
evolutionary rather than revolutionary approach,
involving the establishment of a platform focused on
delivery of business as usual metering obligations,
with additional functionality added over time.
Deployment of
smart meter
infrastructure
A large proportion
of sites were found
to be unsuitable
for the deployment
of smart meter
infrastructure

In the Smart Grid, Smart City Program, almost 30


per cent of sites were found to be unsuitable for the
deployment of the smart meter infrastructure used in
the program, despite a program of pre-qualifications
and site visits. Multi-dwelling units were a particular
problem with an estimated 70-90 per cent
unsuitable for the smart meters contemplated by
this trial. The high level of unsuitable sites in the trial
was due to a range of factors, including installation
issues and communications coverage. Customer
dissent and was driven in part by the nature of
the deployment.

A market-led (or customer-led)


deployment of smart meters
will be limited by the number of
unsuitable sites where it is not
possible or it is uneconomic to
install smart meter infrastructure
due to physical, locational or
technological constraints.

Reducing the number of


unsuitable sites will therefore
require additional investment by
the business, with the amount of
Based on the findings of the smart meter
this investment expected to vary
infrastructure trials, reducing the fraction of unsuitable by site.
dwellings to a minimum would require greater
investment in detailed radio planning to minimise
black spots when FAN technology is used, the use
of both wireless and wired PAN technologies and
the establishment of processes and protocols for the
remediation or replacement of meter boards or other
significant installation works.

104 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Observation

Discussion

Conclusion

Technology risk
management

Implementing smart meter infrastructure using a


multi-vendor approach has risks associated with the
integration between different vendors systems
and technologies.

Risk management around the


selection of the smart meter
infrastructure technology
platform, including vendors,
integration and the development
of new functionalities, is
essential for enabling robust
business as usual metering and
billing operations

The ability of a
business to meet its
business as usual
metering and billing
requirements is
significantly reliant
on the selection
and integration
of its metering,
communications, IT
software and back
office systems.

Standardisation
of customer side
infrastructure
The size of the
smart meter board
was problematic

In circumstances where multiple vendors are


involved, it is important to set clear integration
responsibilities for all parties involved in the
end-to-end solution. Significant extension of
the functionality of the meter management
system was required to meet all of Ausgrids
requirements, including support of the WiMAX FAN
communications technology.
The establishment of relationships with a number
of suppliers of metering and communications
equipment can minimise delays in the case where
particular equipment is found to be faulty or needs
to be re-tested. At times, the Smart Grid, Smart City
Program was delayed whilst attempting to pin point
the exact cause and thus responsibility of a fault.
The size of the smart meter board created
installation challenges in many cases during the
trial. The larger footprint of the smart meters in
comparison with non-smart meters (and even other
smart meters) meant that, in many cases it was
unable to be installed due to insufficient space. No
current standards exist at the time of writing and the
NSMP minimum functionality specification contains
an empty placeholder with regard to meter size.

Policy makers should consider


the standardisation of meter
dimensions, mounting points,
terminals and space for antenna
to ensure that smart meter
equipment is able to be installed
and cost-effectively replaced
with metering equipment from
another vendor

A large number of sites, particularly on multidwelling units, were found to be unsuitable to


install communication antenna due to the lack of
suitable location. Policy makers should consider
the standardisation of meter dimensions, mounting
points, terminals and space for antenna to ensure
that smart meter equipment is able to be installed
and cost-effectively replaced with metering
equipment from another vendor.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 105

Observation

Discussion

Conclusion

Standardisation
and competition

The lack of competition appears, in some part, to be


driven by the prevalence of proprietary technologies
and interfaces. This is particularly problematic in
the applications-level interface between meters and
the meter management system and the physical
and electrical interface between meters and the
communication module. Competition is likely to
be enhanced as the level of standardisation, and
therefore interoperability between vendors, improves.
It is recommended that policy makers, in consultation
with industry, develop a metering functionality
specification that initially involves the provision of
a limited subset of smart metering functionality,
including the requirement that SMI technology is able
to support the introduction of additional features
over time.

Competition in the market for


smart meter infrastructure
technology is likely to be
enhanced through the
standardisation of interfaces
between the meter,
communications module and
meter management systems,
enabling interoperability
between technologies from
different vendors.

Minimum
functionality
standards

The lack of standardisation of smart meter


infrastructure in Australia provided the following
issues in the Smart Grid, Smart City Program:

The lack of
standardisation
created a range of
issues in the trial

The lack of an end-to-end standard between the


meter, communication module and MMS created
interoperability issues at the application layer. Policy
makers should set a common market protocol in
Australia, which is consistent with the AEMCs Draft
Framework for Open Access and Communications
Standards, however it is inconsistent with the
AEMCs Final Report on this framework (March 2014)
which allows alternative pathways of communication,
meaning that more than one market protocol can be
in use at any one time.

Policy makers should consider


implementing an application
protocol standard for the end-toend process between the meter,
communications device and the
MMS and a hardware standard
for the physical connection
between the meter and the
communications device

The level of
competition in the
market for smart
meters is relatively
low

The lack of an interoperability standard for the


physical connection between the meter and
communications device. The development of such a
standard would enable meters and communications
devices from different vendors to be interoperable
and allow communications devices to be repaired
replaced over time with new communications
technologies, avoiding the need to replace the
existing installed meter.

106 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Observation

Discussion

Conclusion

Separating the
smarts from the
meter

This was observed in the Smart Grid, Smart City


trial when fault issues were experienced with
communications modules, which then needed to be
removed and addressed by the manufacturer. As
the communications modules were separate from
the meter, they could be removed and fixed while
the meter remained in-situ and continued recording
customers energy consumption data.

The Smart Grid, Smart City


project found that separating the
smarts from the meter provided
potential benefits in the form of
greater operational flexibility, cost
savings and ability to upgrade the
communications capability as the
technology develops over time.

Installing separate
meter and
communications
devices provides
flexibility in the
operation of
smart meter
infrastructure

Installing separate meter and communications


devices can provide cost savings. If there is an issue
with the communications device it can be removed,
avoiding the cost of replacing the installed meter
and the cost of performing data substitution and
estimation analysis while the meter is being repaired.
In addition, as communications technology generally
improves over time, installing smart meters with
in-built communications modules risks locking
customers into a communications technology that
becomes outdated

Market driven
rollout of
smart meter
infrastructure

The Smart Grid, Smart City trial found that the costs
of installing meters and communications equipment
varied significantly for a variety of reasons, including
access issues, insufficient meter board space and
poor signal strength. The issue of how to equitably
distribute the cost of upgrading private shared
equipment is one of the key challenges for any
market driven rollout.
If the issues associated with a market-led (also
called customer-led in this report) rollout of SMI
arent able to be overcome then the benefits from
the products and applications that SMI enables will
not be realised.

The Smart Grid, Smart City


SMI trial found that the cost of
installation for smart metering
infrastructure varied significantly
between sites and concluded
that a market-driven rollout
of SMI would likely target
customers whose dwellings
enabled low cost installation.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 107

11.6.1 Applicability of solutions


toAustralia
Based on the findings of the Smart Grid, Smart City
SMI trials, businesses considering undertaking a full
deployment of smart meter infrastructure need to
consider the nature and estimated costs of:
Technologies (meters, communications modules
and MMS) that are available in the market
Back office systems that may be required to
be developed to support a rollout of smart
meter infrastructure
The Smart Grid, Smart City Program found that
the costs of installing meters and communications
equipment varied significantly from dwelling to dwelling
for a variety of reasons, including access issues,
insufficient space and signal strength. In an extreme
but not unlikely example, installing one smart meter
for a customer in a multiple dwelling unit may require
the complete replacement of the meter board for the
entire complex. This cost will not be justifiable on any
basis other than as an investment in a large-scale
smart meter infrastructure rollout. The issue of how to
equitably distribute the cost of upgrading private shared
equipment is one of the key challenges for any market
driven rollout. In some cases, the installation of smart
meter infrastructure will be limited or prohibitively costly
for some dwellings.

108 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

12 Customer Applications
12.1 What are
customer application
technologies?
The Smart Grid, Smart City Customer Applications
Program trialled different methods of interacting with
customers purchasing electricity as part of the broader
Smart Grid, Smart City Program. Historically, traditional
spinning disc electricity meters have been installed in
the majority of residential homes in Australia, which
are read manually on a quarterly basis. This meant
that network operators and retailers only had access
to quarterly information meaning that both network
operators and retailers were limited in their ability to offer
tariff incentives or use other tools that could encourage
behaviour change to improve the efficient use of the
grid to meet electricity demand and moderate growth
in peak demand. With the introduction of smart meter
infrastructure, more accurate, real time, information
is becoming available to all participants in the market
place, providing an opportunity for more innovative
tariffs, technologies and customer information products.
Three groups of products were trialled as part of
the Smart Grid, Smart City Customer Applications
Program:
1. A Network Trial measured the effectiveness of
smart meter based products without changing the
customers retailer or electricity retail tariffs.
The network trial tested feedback technologies,
financial incentives (rebates) and provided a lifestyle
audit. Eight products were offered to customers either
individually or as a bundle. The products consisting of
an online portal, an in home display, appliance control
and sub-metering devices, an interruptible load (air
conditioning) control rebate, a dynamic peak rebate
and a lifestyle audit. Further information on these
products is discussed in the following sections.

A Retail Trial measured the effectiveness of


alternative electricity tariffs either as standalone
products or bundled with feedback technologies.
The retail trial tested smart meter based tariffs,
feedback technologies and a rebate. In total, twelve
products were offered to customers with each product
including a tariff (dynamic peak pricing, seasonal timeof-use or top-up plan) or a rebate (interruptible load
(air conditioning)) and optionally one or more feedback
technologies (an online portal, an in home display or
appliance control and sub-metering devices). Further
information on these products is discussed in the
following sections.
2. A Smart Water Meter Trial which trialled the
effectiveness of integrating smart water meters with
smart electricity meters.
These products were supported by smart meter
infrastructure, described in detail in the previous
section of this report, along with other common
platform technologies discussed in Section 4.
As part of this study, the Customer Applications
Program also investigated the differences in responses
and behaviour change outcomes across different
demographic groups. Particular emphasis was
placed on assessing the effect of interventions on
vulnerable energy users, including the elderly, low
income households, households with small children
and customers on life support systems. These issues
are discussed further in the Customer Applications
Technical Compendium and Parts Two and Three of
this report.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 109

12.2 The role for


customer applications
in a smart grid
Customers play a major role in smart grids. However,
little is known about how they perceive or how they
might respond to the opportunities that smart grid
technologies offer.
The Customer Applications trials tested residential
customer behaviour in response to the provision of
information, differential pricing and technologies that
provided them with the ability to make informed decisions
about their electricity costs based on near real time
information. It was hypothesised that this would enable
customers to make informed choices and potentially
better control their consumption of electricity over the
long-term and/or during peak events.
The Customer Applications trial analysis and results
have also informed the broader Integrated Net Benefit
Analysis (Part Two of this report) in order to determine
the efficiency value for network businesses and
for electricity customers, both in terms of potential
reductions in peak demand and broader costs and
benefits. The Customer Applications trial analysis and
results will also inform regulations and standards around
security and privacy.

12.3 Customer
applications objectives
and timelines in the
Smart Grid, Smart City
Project
Electricity customers in most Australian markets are
currently provided with limited information, incentives
and tools to manage their own domestic electricity
use. In most cases, a quarterly electricity bill is the
main source of customer communication. The bill
shows them the total amount of energy used during
the previous three month block and provides limited
information on their electricity usage patterns and
therefore limited opportunity to systematically modify
their behaviour in order to save electricity and money.
The majority of residential customers in Australia have
no visible price signal on which to respond to the true
cost of energy supplied at times of peak demand.
Unlike other parts of Smart Grid, Smart City Program
which focused on specific in-grid or supplementary
generation and storage methods of managing peaks, the
Customer Applications Program was focused on how
behaviour changes by residential customers enabled by
smart technologies, applications and products, could
contribute to peak demand management and overall
reductions in electricity consumption. The Customer
Applications Program evaluated the following aspects
relating to smart meters:
Customer acceptance of a smart meter installation
The attractiveness of the selected smart meter
based products to customers
The effectiveness of the selected smart meter
based products in assisting customers to better
understand and manage their electricity usage or
shift electricity load from peak periods, which may
in turn help them to reduce electricity costs

110 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

The impact of the selected smart meter based


products on total electricity usage and/or peak
demand and the subsequent effects of these
changes on a distribution network service
providers operation and management of their
electricity network

For the secondary program purpose of investigating


the benefits of integrating water and gas metering with
electricity smart metering, a combined utilities online
portal was developed to examine the attractiveness of
such a proposition to customers. Gas was originally
included in the trial scope, but was removed during the
Program when it was found that there were no smart
gas meters commercially available on the market. Both
the network and retail trials tested the integration of the
water meter.

The Customer Applications Program was conducted


8th October 2010 and the 28th February 2014.
The extension to February 2014 was made in order
to capture data from an additional summer from
participants in the retail trial.

The Customer Applications Program elements are


described in Figure 1-20.

Figure 1-20 Timelines and processes for the Smart Grid, Smart City Customer Applications Program

2010

2011

2012

2013

2014

Enabling technologies
Sampling and trial design
Network Trial meter deployment
Network Trial customer acquisition
Network Trial
Network Trial Extension
Retail Trial customer acquisition
Retail Trial meter deployment
Retail Trial
Retail Trial Extension
Sydney Water Trial customer acquisition
Sydney Water Trial meter deployment
Sydney Water Trial
1
Milestones

3 4

5 6 7

 tart of SGSC project


1 S
08 June 2010

 irst DPP peak event called


4 F
26 February 2013

7 Initial customer survey


commenced 29 August 2013

 irst smart meter installed


2 F
07 October 2011

 irst water meter installed


5 F
01 July 2013

 nd of Customer App extension 28


8 E
February 2014

 irst DPR peak event called


3 F
17 January 2013

 ast smart meter installed


6 L
29 July 2013

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 111

12.4 Where are


customer application
technologies currently
deployed?
12.4.1 International trials
As part of the global interest in smart grids, there
is an increasing appreciation of the importance of
residential customers managing their overall electricity
consumption as well as peak demand. It is generally
agreed that customers react to different signals in
ways that reflect their own price sensitivity, concern
for environmental and social issues, and that if the
electricity market can appeal to these concerns,
customers will decide to proactively manage their
electricity consumption.
Unlike the Smart Grid, Smart City Program, most
international projects have typically focused on specific
elements of the smart grid (such as smart meter
infrastructure or different grid applications), or have
been limited in their scale. This has meant that the
majority of these projects did not integrate different
smart grid applications to understand the interactions
between each of the component parts, nor did they
implement the technologies at acommercial scale.
Due to the number of international trials involving
customer applications, only a small number are
summarised briefly below. Additional details can
be found in the Customer Applications Technical
Compendia (which can be found on the Smart Grid,
Smart City Information Clearing House https://ich.
smartgridsmartcity.com.au/).

The role of smart metering infrastructure


global studies
Smart meter infrastructure is one of the most common
smart grid technologies deployed globally. Smart
meters (and the associated back office and common
platform technologies) facilitates the introduction of
alternative tariffs and incentives for consumers to
change their electricity consumption behaviour.
There is a significant and increasing amount of
literature which examines customer experiences with
smart metering. The literature includes:
Kaufmanns Customer value of smart metering:
Explorative evidence from a choice-based conjoint
study in Switzerland51 modelled the importance of
the perceived customer value of smart metering
The US Department of Energys Voices of
Experience: Insights on Smart Grid Customer
Engagement52 compiles information about
successful smart grid technology and deployment
engagement approaches. The guide is the result of
a stakeholder engagement led by the Department
of Energy to capture lessons learned and insights
on smart grid customer engagement that will help
steer a stronger utility-customer relationship.
Studies in Hong Kong investigated attitudes by
customers to price-responsive tariffs Customer
perceptions of smart grid development: Results
of a Hong Kong survey and policy implications53
which discussed how customers could contribute
to energy saving, energy efficiency and peak load
shifts, facilitated by real-time electricity information
and smart meters that are linked with dynamic
pricing system.

51  S. Kaufmann, K. Knzel, and M. Loock, Customer


value of smart metering: Explorative evidence from a
choice-based conjoint study in Switzerland, Energy
Policy, vol. 53, pp. 229-239, 2/2013
52  U.S. Department of Energy, VOICES of Experience:
Insights on Smart Grid Customer Engagement
2013
53 D. N. Mah, J. M. van der Vleuten, P. Hills, and J. Tao,
Consumer perceptions of smart grid development:
Results of a Hong Kong survey and policy
implications, Energy Policy, vol. 49, pp. 204-216,
10/2012

112 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Smart meter devices and the effect of feedback


on residential electricity consumption: Evidence
from a natural experiment in Northern Ireland
estimated the effect of real-time usage information
on residential electricity consumption in Northern
Ireland54. Starting in April 2002, the utility replaced
prepayment meters with advanced meters that
allowed the customer to track usage in real-time.
The findings suggested that the provision of timely
information was associated with a decline in
electricity consumption

Dynamic pricing studies


Dynamic pricing refers to a mechanism where the
price of electricity changes over a period of time for
example, during a day or different seasons. It is often
considered an essential part of demand reduction
programs, particularly when considering the rate of
introduction of new customer facing technologies.
Several companies within the US are currently
trialling or rolling out customer application focussed
technologies and pricing, including:
Critical peak pricing, time-of-use and other
pricingsignals
Customer demand response during critical
network events both through customer-initiated
response and automated demand reductions (e.g.
direct control of air conditioning compressors and
programmable communicating thermostats)

Lessons learned from smart grid enabled pricing


programs55 presents case studies of several dynamic
pricing programs, including different proposed rates,
enabling technologies and incentives. Program
successes were evaluated based on:
Peak load reduction
Customer bill impact
Customer satisfaction

12.5 Smart Grid,


Smart City customer
application trials
The majority of the technologies used in the Smart Grid,
Smart City Customer Applications Program were not
only new to Ausgrid, but were also generally considered
immature in national and international markets.
Implementing these technologies to provide an
end-to-end complete solution in a commercial-scale
trial of customer focussed products within the trials
timeframes presented significant challenges these
are discussed at length in the Customer Applications
Technical Compendia (which can be found on the
Smart Grid, Smart City Information Clearing House
https://ich.smartgridsmartcity.com.au/).

A range of feedback technologies to provide


information to customers on their real-time electricity
usage for the purposes of peak load shifting

54  D. N. Mah, J. M. van der Vleuten, P. Hills, and J. Tao,


Consumer perceptions of smart grid development:
Results of a Hong Kong survey and policy
implications, Energy Policy, vol. 49, pp. 204-216,
10/2012

55  J. Wang, M. A. Biviji, and W. M. Wang, Lessons


learned from smart grid enabled pricing programs, in
Power and Energy Conference at Illinois (PECI), 2011
pp. 1-7.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 113

12.5.1 Smart meter based


technology applications
The Smart Grid, Smart City Customer Applications
Program utilised and trialled a series of smart meter
based technologies to meet the Programs objectives.
The technologies used in these trials were categorised,
as described in the following sections.

Enabling technologies
A set of technologies were deployed to enable
selected customers to trial the customer applications
products. These enabling technologies were specific
to the Smart Grid, Smart City Program and had to be
defined, selected, built, tested and deployed before
customers could commence trialling the products.
These enabling technologies included:
The Smart Grid, Smart City smart meter (17,134
smart meters were deployed as part of the
Customer Applications trial)
The various communications technologies required
to transport data to and from the smart meter, the
feedback technologies and the back office systems
The firmware that controlled how the smart meter
interacted with the meters communication module
The meter management system (MMS)
Back office operational systems at Ausgrid,
EnergyAustralia and Sydney Water
The customer acquisition application

114 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Feedback technologies
One of the objectives of the Smart Grid, Smart City
Customer Applications Program was to measure the
behavioural changes of electricity customers based on
their use of a variety of different feedback technologies.
Five individual feedback technologies were selected to
form part of the products that were trialled as part of the
Smart Grid, Smart City Program including:
1. Customer web portal including a mobile
device application the customer portal was
an online electricity management website giving
customers access to information about their
household electricity usage and costs to assist
them in finding ways of reducing electricity bills.
This information included the customers own
meter data (usage and cost) in near real time in
time as well as increments of a week, a month or a
year. It also allowed households to compare their
electricity usage with other households.

2. In home display the Smart Grid, Smart City


in home display (IHD) displayed a dwellings
electricity usage and real time pricing information.
Text messages could be sent by the network
operator/retailer to the IHD to inform the customer
of price changes or peak events. The current price
of electricity or a relevant tariff could always be
displayed on the IHD. Theprovision of electricity
usage and cost data to households through the
IHD had the potential to drive behaviour change
particularly during peak demand periods. The IHD
presented three datasets using a simple interface
including metering, price and messaging data. The
Smart Grid, Smart City IHD is shown in Figure 1-21.

Figure 1-21 Smart Grid, Smart City IHD

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 115

3. Appliance control and sub metering


Appliance control and sub metering provided
the ability for customers to monitor the electricity
usage of up to ten specific devices in their home
in near real-time using panel meters and smart
plugs to measure electricity usage. In addition to
viewing individual appliance usage, the smart plugs
could also be used to switch appliances on and
off through a portal control. These elements are
described in Figure 1-22.

Figure 1-22 Components of the appliance control and sub metering trial

116 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

4. Cognitive metering cognitive metering is


alow resolution load disaggregation algorithm.
The load disaggregation algorithm analysed the
instantaneous electricity demand profile of a
home, to determine the appliances that were being
operated in real time and the amount of electricity
they were consuming.
Cognitive metering was used to identify appliances
within the home that had the largest electricity loads,
the times they were in use, and the impact they were
having on a customers quarterly electricity bill.
5. Home energy management system the home
energy management system (HEMS) used a suite
of feedback technologies that was similar, from a
customers perspective, to the appliance control
and sub metering offering, described previously.
Both included smart plugs to sub-meter and
control individual appliances and a HEMS customer
portal for remote control of appliances.
Unlike the appliance control and sub metering product,
HEMS was a standalone product offering that did
not require support by smart metering infrastructure.
A ZigBee IP residential gateway device was installed
in each dwelling that connected to a customers own
modem (provided by their ISP) to access the cloud
based customer portal.

One of the primary objectives of this deployment


was to trial a HEMS that did not require smart meter
infrastructure. The application of HEMS in the Smart
Grid, Smart City Program provided an understanding
of the alternative technologies and resources available
for network management.
The HEMS customer portal was hosted by the
vendor (Jetlun) in Amazons cloud and accessible to
the customer and Ausgrid via a website or mobile
devices56. One of the most powerful features of
this product was the ability to provide users with a
customer choice rule book, whereby the customer
could create a set of rules to manage and control
particular appliances based upon a budget or price
based events. Additionally, the portal was able to notify
the customer of peak price events if the customer had
previously entered any applicable tariff structure to
the dwelling.
The HEMS customer portal was capable of displaying
electricity usage information for all sub metered
appliances in the home in near real time as well as on
a daily, weekly, and monthly basis. An example of the
HEMS customer portal is provided in Figure 1-23.

Figure 1-23 Jetlun HEMS customer portal and mobile device

56 both iOS and Android devices were supported


National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 117

Network operator controlled demand


response device
Demand response devices enable a network operator
to override a customers normal electricity usage
patterns at times when electricity availability could
be jeopardised. Two examples of network operator
controlled demand response technologies were
trialled as part of the Smart Grid, Smart City Customer
Applications Program: the ability to shift the heating
of hot water systems to off-peak periods (using smart
meter functionality instead of the traditional ripple
control systems used by some network businesses
in Australia), and the ability to interrupting an air
conditioners use of electricity during peak event
periods. Each of these elements isdescribed below.
Demand Response Enabling Device (DRED)
enables the remote control of residential appliances
with high electricity loads by electricity distribution
network operators and retailers via a smart meter.
Ultimately, DRED technology provides the ability to
control residential load on the network during peak
demand periods by remotely switching off certain
components of household devices for small amounts
of time.
For the Smart Grid, Smart City Program, the DRED
module for an air conditioner was selected for this
analysis Using DRED in pool pumps was investigated,
but not pursued due to the lack of compliant pool
pump controllers available at the time.

Applied to a customers air conditioner, the DRED


module allowed Ausgrid to remotely reduce the air
conditioner electricity consumption by reducing the
percentage of time that its compressor runs. This
action was intended to be performed during critical
peak periods that impact either the network operator
or the retailer. The duration that the compressor
would be turned off would be regulated to ensure
the customer was unlikely to be significantly
inconvenienced and the air conditioners fan would
continue to operate during this time.
Customers who agreed to participate in this process
were offered compensation for their willingness to
allow their air conditioner to be controlled. They were
also notified during events and could choose not to
participate on each occasion.
Hot water load control (HWLC) was the
mechanism used by Ausgrid to manage the time of
the day during which power was supplied to domestic
hot water systems. The electricity required to heat
water for residential use constitutes a significant
component of the total electricity grid load. Many
network operators in Australia use an existing ripple
control system in order to control the periods in which
hot water systems are heating the water (generally hot
water heating is not available during peak network
periods). The trial of HWLC technologies sought to
demonstrate that the existing hot water load control
functions could be replicated with smart meters
including the options of:
Using the inherent timing capability built into a
smart meter to switch the hot water load on and off
at pre-determined times
Emulating the more fine grained control achieved
with the ripple control system by having the ability
to vary the on/off times in response to network
loads or other considerations

118 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Education products
Because the Smart Grid, Smart City Customer
Applications Program was focussed on enabling
customers to make informed decisions about their
energy usage, in addition to technologies, the
Program also provided educational material. The
Smart Grid, Smart City Customer Applications
Program Home Energy Assessment included a free
assessment of a customer home electricity usage with
recommendations for reducing electricity costs along
with gas and water consumption. Recommendations
that were selected as part of this process were
intended to be implemented without adversely
affecting a customers lifestyle.
By providing the education assessment, the Customer
Applications Program sought to evaluate whether
customers would change their behaviour as a result of
directed education about what was driving electricity
usage and costs in their dwelling.

Network rebate product


The Home Energy Rebate was included in the Smart
Grid, Smart City Customer Applications Program
Network Trial to support customer behaviour change
and help determine whether customers would
reduce their total electricity usage in response to a
price incentive. The Home Energy Rebate provided
customers with a payment for reducing electricity
usage during peak events. Peak events occurred on
very hot or very cold weekdays between 2 pm and 8
pm for either two or four hours.
Participating customers receive a rebate ranging from
$0 to over $100, depending on how much electricity
was reduced compared to their baseline use at that
time of day, with the average rebate estimated to be
$20 per event. This rebate was designed to be available
for between 10 and 15 special rebate peak events
during winter and summer months. The peak events
were anticipated to occur when electricity intensive
appliances, such as air conditioners and heaters, were
most likely to be used. This would typically be between
2pm and 8pm on working weekdays for either a two or
four hour period.

Customers were notified of peak events with an SMS


notification and the rebate was paid to the customer if
that customers actual electricity usage was less than
their baseline usage during the peak event period.

Retail tariff and rebate pricing structures


The Smart Grid, Smart City Customer Applications
Program included a series of retail pricing structures,
(tariff, rebate or reward), that were developed to
address key project outcomes of increasing electricity
efficiency and reducing peak demand. Two base
pricing tariffs, one financial reward and one financial
rebate were developed and trialled as part of the
Customer Applications Retail Trial. These pricing
structures were then offered as standalone products
or bundled with feedback technologies to result in a
total of 12 products trialled with customers.
The pricing structures that were selected to form part
of the products that were trialled for Smart Grid, Smart
City Program were:
1. Dynamic peak pricing tariff marketed as
PriceSmart
2. Seasonal time-of-use tariff marketed as SeasonSmart
3. Top up plan reward marketed as BudgetSmart
4. Interruptible load for air conditioning rebate
marketed as FlowSmart
It is important to note that pricing for these new
products was developed on the principle that the
customers annual bill should be cost neutral if they did
not change behaviour. The pricing structure products
are described below.
PriceSmart was a dynamic peak pricing (DPP) tariff
designed to reduce peak demand and to test whether
customers would reduce their electricity usage upon
receipt of a price signal, which was intended to be
more cost reflective of the costs incurred at times of
peak demand.
The price signal was a substantial increase in rates
(six to seven times higher than the standard time-ofuse peak rate) during the dynamic peak events. The
dynamic peak price rate was $3.00/kWh ($3.30/kWh
including GST). Outside of the peak events, customers

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 119

were incentivised by being charged at either a


shoulder rate (7 am to 10 pm all days) or off-peak rate
(all other times).
During the dynamic peak events, it was anticipated
that customers would limit electricity usage and
therefore avoid higher costs of electricity during
these times.
Responsibility for calling events was equally split
between the retailer and network operator, with
each entitled to call seven events. The retailer had
final approval of all events and was responsible for
notifying customers of an impending event. Events
were designed to last between one and four hours
andwereonly to be called between 2 pm and 8 pm
onworking weekdays.
A bespoke time-of-use tariff was designed for
the Smart Grid, Smart City Program. Figure 1-24
describes the existing pricing structure compared
tothe PriceSmart pricing.

designed to reduce peak demand during summer and


winter by charging peak rates that were higher than
those in the standard ToU tariff, and to test whether
customers would reduce their electricity demand
during summer and winter peaks by shifting their
usage to alternative time periods.
During the lower demand spring and autumn months,
customers would enjoy a shoulder rate during the
standard tariff peak period. The tariff was designed
to better reflect higher wholesale electricity costs
during summer and winter. During summer and winter
periods, it was anticipated that customers would
limit electricity usage by shifting load to off-peak or
shoulder times or be more efficient.
Customers were billed on a three-part ToU pricing
structure (peak, shoulder and off-peak) during summer
and winter and a two-part ToU pricing structure
(shoulder and off-peak) during spring and autumn,
asoutlined in Figure 1-25.

SeasonSmart was a seasonal time-of-use (ToU) tariff

Figure 1-24 Dynamic Peak Pricing tariff used for Smart Grid, Smart City PriceSmart trial participants

Existing Pricing

New PriceSmart Pricing

With current price structures, customers on an All Time plan


pay the same rate all day, every day. For customers on a
Time of Use plan, the rate varies according to the time of day.

PriceSmart reduces the peak rate and offers consistent


savings from 2pm to 8pm. Thereare further opportunities to
save by reducing your usage during the peak price events.

Price Inc-GST

All time
26.84c/kWh

13.09c/kWh
24.53c/kWh
24.53c/kWh
330.00c/kWh
Saving

12am

7am

2pm

8pm

10pm 12am

8pm

10pm 12am

All day, every day

12am

13.09c/kWh

Price Inc-GST

Price Inc-GST

Time of Use
52.55c/kWh

21.34c/kWh

7am

2pm

Time (weekdays)

12am

7am

2pm

Time (weekdays)

120 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

8pm

10pm 12am

Figure 1-25 Seasonal ToU tariff used for Smart Grid, Smart City PriceSmart trial participants

Existing Pricing

New PriceSmart Pricing

With current price structures, customers on an All Time plan pay the
same rate all day, every day. For customers on a Time of Use plan,
the rate varies according to the time of day.

Season Smart - Spring / Autumn


26.84c/kWh

12am

Price Inc-GST

Price Inc-GST

All time

SeasonSmart reduces the peak rate in spring and austumn for


consistent savings. The higher the peak rate in summer and winter
means you can save money with smarter energy use.

7am

2pm

8pm 10pm 12am

7am

2pm

8pm 10pm 12am

Season Smart - Summer / Winter


52.55c/kWh

Price Inc-GST

Price Inc-GST

Saving

Time (Spring / Autumn weekdays)

Time of Use

12am

21.34c/kWh

21.34c/kWh

12am

All day, every day

13.09c/kWh

13.09c/kWh

21.34c/kWh

7am

2pm

8pm 10pm 12am

Time (weekdays)

BudgetSmart was a top up reward product that


provided customers with a greater understanding of
their electricity usage and costs through maintaining
an electricity account in credit. BudgetSmart was also
designed to test whether customers would reduce
their electricity demand in general, due to an increased
awareness of electricity usage behaviour.
By keeping their account in credit, customers received
a financial incentive (a 12.5 per cent discount off their
total bill). Customers were able to do so by making
regular top-up payments in response to account
balance prompts from their retailer. It was anticipated

12am

13.09c/kWh

74.47c/kWh
52.55c/kWh

21.34c/kWh

7am

2pm

8pm 10pm 12am

Time (Summer / Winter weekdays)

that by providing customers with prompts to top-up


as their account approached a debit balance, they
would have a better understanding of the impact
their electricity usage had in dollar terms and could
take steps to reduce or shift electricity usage to drive
downcost.
Two standard regulated tariffs were used for this
product: inclining block tariff (known as domestic
all time) and ToU (known as PowerSmartHome),
asdescribed in Table 1-8. Customer tariffs were
notchanged when they signed up to this product.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 121

Table 1-8 Seasonal ToU tariff used for Smart Grid, Smart City PriceSmart trial participants

First 10.9589 kWh per day

Next 10.9589 kWh per day

Balance

Inclining block tariff


Ex GST

24.90

26.38

28.48

Incl. GST

27.39

29.018

31.328

47.77

19.86

11.97

52.547

21.846

13.167

Time-of-use
Ex GST
Incl. GST

FlowSmart was an interruptible load rebate for


customers with air conditioning, where payment
was provided to customers in return for reducing air
conditioning usage during up to six pre-notified peak
events. FlowSmart was designed to test whether
customers would reduce their electricity usage during
periods of peak demand by reducing the electricity
required by high electricity usage appliances such as
air conditioners.

Utility interoperability

Using a demand response enabling device (DRED),


a customer air conditioning compressor (the cooling
component) was switched off for 15 minutes per hour
during the peak event to reduce the demand on the
grid. The air-conditioning fan continued to operate
during this period to provide some cooling benefit.

To investigate the benefits of integrating water metering


with electricity smart metering, a combined customer
portal was developed. More detailed information
about smart water meter solution can be found in the
supporting documents to the Customer Applications
Technical Compendium57.

This product did not proceed past testing stages and


therefore limited data was available for analysis.

To enable the measurement of the benefits of a


combined utility portal within a smart grid, an endto-end solution was required to be established. This
solution consisted of the smart meter infrastructure,
the feedback technologies provisioned at the
customer dwellings, the network tariffs provisioned
on the smart meter, the retail tariffs implemented in
the retailers billing and settlement systems and the
various IT systems used by Ausgrid (as the DNSP),
EnergyAustralia (as the electricity retailer) and Sydney
Water (as the water retailer).

An important part of the Smart Grid, Smart City


Customer Applications trials was to investigate
whether electricity, water and gas meter data could be
presented to customers through a combined utilities
portal which would influence their behaviour. Gas was
originally included in the trials scope, but was removed
when it was found that there were no smart gas meters
commercially available on the market.

57  At the time this report was prepared, Sydney Water


had not finalised its report on the smart water meter
trials

122 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

12.5.2 Customer application trial


design elements
The Customer Applications Program was implemented
in a commercial-scale environment in which the costs
and benefits of different smart meter based products
could be evaluated to meet the overall Program
objectives within the Programs scope and limitations.
Specifically, the design enabled the Program to:
Test whether customer applications within a smart
grid would have a measurable impact on reducing
total electricity usage, including peak demand, and
electricity cost in an Australian context; and
Investigate the costs and functional efficiencies
of, and understand customers attitudes and
responses to, integrating smart water and gas
metering with smart electricity metering
Meet the objectives of the Customer Application
Program, including determining the following trial
design factors:
1. Product selection the customer applications
(tariffs, feedback technologies, rebates, rewards
and educational products) required to be trialled to
meet the Programs objectives
2. Enabling technologies for the products selected,
the enabling technologies required to perform the
trials in the Ausgrid environment
3. Customer selection for the products selected,
the customer selection and eligibility requirements
to be able to supply the statistical data required for
the trials analysis

The products selected for the trials were evaluated


against the following metrics:
The attractiveness of the selected smart meter
based products (the tariffs, feedback technologies,
rebates and lifestyle audit) to customers (where
attractiveness refers to the benefits the customers
perceived they would receive from using the
product and includes measurements of uptake
rates, active and passive opt out rates and the
retention activities required to ensure customers
continued to use the product)
The effectiveness of the selected smart meter
based products in assisting customers to better
understand and manage their electricity usage and
shift electricity load from peak periods, which may
in turn have helped them to reduce electricity costs
The impact of the selected smart meter based
products on total electricity usage and/or peak
demand and the subsequent effects of these
changes on a distribution network service
providers operation and management of their
electricity network

12.5.3 Deployment processes and


customer selection of customer
application trial products
As previously described, the key trials completed
as part of the Smart Grid, Smart City Customer
Applications Program were:
A Network Trial which sought to measure the
attractiveness, impact and effectiveness of smart
meter based products without changing the
customers retailer or electricity retail tariffs
A
Retail Trial which sought to measure the
attractiveness, impact and effectiveness of
alternative tariffs either as standalone products or
bundled with feedback technologies

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 123

Both the Network and Retail Trial


supported the Smart Water Meter Trial
Given the complexity of conducting a commercial
scale customer-focussed trial, there were a number of
key processes that were designed prior to, during, and
after the rollout of the trial products. These processes
included the development, testing and provisioning of
the customer products and the selection, engagement,
management and ongoing support of the customers
trialling the products. These processes have been
discussed at a high level in the following sections,

with the detail of these processes provided in the


Customer Applications Technical Compendium (which
can be found on the Smart Grid, Smart City Information
Clearing House https://ich.smartgridsmartcity.com.au/).

Customer applications products


A summary of the products selected for each of the
trials is provided in Table 1-9. Each of these products
has been described previously in this section of
the report.

Table 1-9 Summary of the Customer Application Program trial products

Trial

Customer Applications Program products

Network Trial feedback


technology, rebates and
educational products

SGSC Home Energy Online Portal


SGSC Home Energy Monitor
SGSC Home Energy Assessment
SGSC Home Energy Rebate
SGSC Home Energy Network Control System with Home Energy Online Portal
SGSC Home Energy Rebate with Home Energy Monitor
SGSC Home Energy Online Portal with Home Energy Monitor
SGSC Home Energy Air (not deployed)

Retail Trial tariffs, and


feedback technology
products

BudgetSmart Plan
BudgetSmart Plan with PowerSmart Monitor
BudgetSmart Plan with PowerSmart Online Portal
BudgetSmart Plan with PowerSmart Online Portal and Home Control System
PriceSmart tariff
PriceSmart tariff with PowerSmart Monitor
PriceSmart tariff with PowerSmart Online Portal
PriceSmart tariff with PowerSmart Online Portal and Home Control System
SeasonSmart tariff
SeasonSmart tariff with PowerSmart Monitor
SeasonSmart tariff with PowerSmart Online Portal
SeasonSmart tariff with PowerSmart Online Portal and Home Control System
FlowSmart tariff

Smart Water Meter Trial

Water meter with Home Energy Online

124 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Community awareness and engagement

Customer acquisition and selection process

Engagement with the general community was


required to communicate the benefits of the Smart
Grid, Smart City Program and in particular the
benefits of the trial products to residential customers.
Community engagement involved raising awareness
and building knowledge of the Smart Grid, Smart City
Program. Raising awareness was required because
the Customer Applications Program trials were opt
in trials, i.e. customers had to volunteer to become
involved. Engagement activities were undertaken upon
commencement of the trials to encourage participation,
as well as throughout the trial implementation,
management and decommissioning phases, including
smart meter rollout and other acquisition phases of the
Customer Applications Trials.

Different approaches were used for the deployment


of smart meters for the Network and Retail Trials:

Community engagement within the Smart Grid, Smart


City Program faced inherent challenges:
Electricity is a product that typically exhibits low
levels of community engagement.
The trials were available to only a small percentage
of customers in the Ausgrid network therefore
the majority of the community were not able to
participate in the trials, regardless of their level of
engagement and awareness.
The benefits of the Smart Grid, Smart City
Program were tailored to individuals, retailers and
network service providers, rather than the broader
community.

Network Trial: Smart meters were installed at


selected customer dwellings prior to inviting
customers to participate in the Trial. This process
emulated a mandatory rollout (also termed a
fulldeployment in this report), however customers
had the opportunity to decline the installation of
the smart meter (referred to as dissenting to
theinstallation)
Retail Trial: The installation process required a
customer to initially agree to participate in the
trial, prior to the smart meter being installed. This
process emulated a market driven rollout (also
termed a customer-led deployment in this report)
One of the key principles of the deployment of smart
meters for the Smart Grid, Smart City Program was to
adopt, as closely as possible, Ausgrids business as
usual meter installation processes. This allowed any
Program-specific smart meter implementation issues
to be highlighted, as distinct from any BAU meter
installation issues. These Program specific issues
could then be investigated to assist in the development
of any future large scale rollout.
By virtue of this being a customer focussed trial, it was
necessary to acquire customers to trial the selected
products. These customers needed to be acquired in
such a way as to be able to supply the statistical data
required for the trials analysis.
For the Network Trial, the customer acquisition process
commenced after the smart meters were installed
and had passed all checks. For the Retail Trial, the
customer acquisition process was performed prior
to the installation of the smart meters. For the Smart
Water Meter Trial, the Smart Grid, Smart City Program
partnered with Sydney Water to deploy water meters to
customers who were already part of the electricity trials.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 125

Sampling and Trial Design Project This project ensured


alignment between customer selection for the trials and
meeting the objectives of the Smart Grid, Smart City
Customer Applications Program. The Sampling and Trial
Design project was relevant to both the Network and the
Retail Trials customer selection process and was aimed
at ensuring a representative sample of customers would
be testing each product to enable the relevant analysis.
The purpose of the Sampling and Trial Design
Project was to:
Define the most statistically appropriate sampling
variables for the Customer Applications Program
Determine the cut-off points to convert the
recommended sampling variables into categorical
variables appropriate for input to the Customer
Applications Program sample frame
Define the sampling approach for each of the
selected products in the Network and Retail trials
Develop a methodology to select customers from
the Customer Applications Program sample frame
Customer selection - The Sampling and Trial Design
Project produced a methodology for the selection of
customers to partake in the trials. Based on a series of
constraints, assumptions and estimates, an initial set of
potential customers were selected from the
Ausgrid database.
For the customer selection process, sampling
variables were first identified, based on domestic
electricity usage, electricity efficiency and electricity
conservation behaviour.
There was a requirement for a control group of 600 to
be selected. This control group needed a smart meter
to be installed, however the control group customers
received very limited information about the trial (to
avoid introducing bias).

Product provisioning and activation


The provisioning of trial products was carefully
managed to ensure optimal customer experience
and therefore results for each trial. The provisioning
process was similar for Retail and Network Trials,
however product content, information and user
guides were branded separately to appropriately
reflect the variation in products between the two trials.
Product documentation and guides were distributed
to customers via email, mail, or inserted into product
packaging prior to delivery.
All provisioning was directly managed by the Smart
Grid, Smart City provisioning team, who were
also responsible for clarifying the various aspects
of the provisioning process for all participants
including customers, sales representatives, internal
stakeholders, vendors and partners (e.g. Sydney
Water and EnergyAustralia). This process also included
smart meter infrastructure registration, installation and
provisioning by Ausgrid.

Trial management
The trial management process describes all aspects of
managing both the Network and Retail Trials, following
on from the provisioning and activation of the products.
Trial management activities undertaken throughout the
Network and Retail trials included the following:
Customer retention activities
Ongoing general customer support
Managing those customers wishing to opt out of
the trial
Ongoing product management

126 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

12.6 High level


trialresults
The Smart Grid, Smart City Customer Applications
Program trialled 22 products comprising pricing
structures, network controlled devices, feedback
technologies, emerging technologies and educational
products with over 7,000 residential customers.
A significant amount of data was collected and
analysed by Ausgrid, EnergyAustralia and third parties
which included Woolcott Research, the Institute of
Sustainable Futures, NERA Economic Consulting and
Frontier Economics. Analysis was performed to help
inform the final observations and conclusions over the
Customer Application Program, as outlined in
this section.

Sampling and trial design


The Sampling and Trial Design Project was intended to
ensure that customer selection for the trials included
a representative sample of customers and enabling
statistically relevant analysis of the results for
each product.
The population of prospective customers for the
trials (both the treatment and control groups) was
restricted to Ausgrid customers within the trial area,
filtered by the various Program constraints which
resulted in a reduced total population available for the
trials. Whilst this did not impact the ability to analyse
results, it was likely to have produced some biases
in the estimates. Although mitigation measures were
taken to correct this bias, these measures were only
partially successful.

The resulting sample (both treatment and control)


could not truly be classed as random from a
statistical analysis point of view because although
customers were selected randomly, the trial managers
placed an emphasis on attempting to ensure sufficient
representation for each product from each of the
statistical cells.
Self-selectivity bias was another limitation that could
not be mitigated against. Self-selectivity bias refers
to pre-disposition of participants to some aspect of a
trial. It is impossible to directly correct for this bias in
an opt in rollout, therefore the possibility of this bias
should be taken into account when analysing the
results of future like trials.
There is further discussion on the sampling and
trial design elements of the project in the Customer
Applications Technical Compendium.

Smart meter infrastructure


Smart meter infrastructure was an enabling platform
for many of the Smart Grid, Smart City trials, with the
largest deployment of meters required to support the
Customer Applications Program trials58. The overall
observations and conclusions of the smart meter
rollout as part of the overall Smart Grid, Smart City
Program is provided in the Smart Grid, Smart City
Smart Meter Infrastructure Technical Compendium.
The results below pertain specifically to the interactions
of smart meters with the Customer Applications
Program. It was found that in order to increase the
willingness of Customers to consent to an in home
installation of smart meters, Retailers and Network
operators need to increase the attractiveness of smart
meters. Critically, in order to fully realise the benefits of
smart meter infrastructure, the smart meter products
must be attractive to customers.

58  17,134 meters were installed at 15,966 NMIs (i.e.


customer dwellings) for the Customer Applications
Program trials by the end of the meter rollout in
August 2013.
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 127

Furthermore, the Customer Applications Program


found that the lack of customer acceptance of smart
meters themselves was a barrier to achieving the full
potential benefits of a smart grid. This issue is further
discussed in Part Three of this report.
This finding is supported by two separate sources.
In their 2013 national survey, Woolcott Research59
found that 21 per cent of customers were not willing
to have a smart meter installed and in the Smart Grid,
Smart City trial area, where extensive community
engagement activities had been undertaken in order
to educate the community on the benefits of the
Customer Applications products, this figure was
found to be even higher at 31 per cent. It is possible
that these figures underestimate the percentage of
customer opposition to smart meter installation, as
the term smart meter was not clearly defined to the
respondents and it is possible that some customers
understood the term to refer to an interval meter60.
If negative customer perception of smart meters
remains at this level and a market driven roll out is
adopted, it will not be possible to achieve the full
benefits of a smart grid. Therefore, if a market driven,
rather than mandatory smart meter rollout is assumed,
governments, network operators, retailers, industry
groups and other relevant stakeholders need to
work together in order to address negative customer
perceptions, misconceptions and concerns regarding
smart meters.

59  Woolcott, Research, Smart Grid, Smart City


Community Tracking Study, August 2013
60 This is evidenced through the 13 per cent of
customers who claimed that they already had a smart
meter installed

Pricing structures
The Smart Grid, Smart City Customer Applications
Program trialled a series of pricing structures (tariff,
rebate or reward) that were developed to address
the key Program objectives of testing the ability of
customers to make informed decisions on their overall
electricity usage and usage during peak
demand periods.
Frontier Economics were retained by the Smart Grid,
Smart City Program to perform econometric modelling
to estimate customer uptake of the products trialled
in the Customer Applications Program and customer
responses to the trialled products with respect to
electricity consumption patterns61.
The full results of all of the pricing structures is described
in the Customer Applications Technical Compendium,
and notable results are described below.

Dynamic Peak Pricing (DPP)


EnergyAustralia set an upper temperature limit for
DPP events, due to the perceived risk to vulnerable
customers whose health may be adversely affected by
their attempts to respond to pricing signals (e.g. elderly
customers turning off their air conditioners during peak
periods on very hot days). One attempt to call a DPP
event on an extremely hot day (which was forecast
to be over 40C, ultimately reaching 37.8C at Sydney
airport) was abandoned due to a trial rule in which
there was agreement that an event would be cancelled
if the forecast temperature was 40C or higher.

61  Frontier Economics, Smart Grid, Smart City Customer


Applications Trials: Data methodology report, 2014

128 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Some customers did provide negative feedback on the


proposed peak event62 after they were notified of the
event. This feedback included:
Hardly a fair test. Forecast today is 39C, my
husband is 75 and we are caring for our three
year old grandson. Not safe to switch off cooling
appliances! Is it possible to reconsider please?
Will be terminating my contract. Doing this on the
hottest day of the year and with one days notice is
price gouging.
Part Two of this report discusses trial participant
behaviour responses and interest in peak event
products and some of the challenges for network
operators and retailers associated with non-firm
demand response.

Top up reward (BudgetSmart)


Customers who trialled BudgetSmart in combination
with an online portal or a portal and HAN had access
to two estimation features through the online portal.
In addition to the estimation features, all BudgetSmart
customers could receive low or debit balance
notifications via SMS. Eighty per cent of BudgetSmart
customers who received an SMS prompt to top up
their account did so, a result which indicates that
the SMS prompt was effective in changing customer
behaviour. Eighty per cent of BudgetSmart customers
who received an SMS prompt to top up their account
did so, a result which indicated that the SMS prompt
was effective as a communications mechanism.
Analysis of customer survey data indicates that
whilst BudgetSmart did not perform as well as the
other pricing products across a range of indicators
including frequency of engagement, perception of bill
savings and impact on energy awareness, it exhibited
unexpectedly high product satisfaction and customers
were very likely to recommend BudgetSmart.
This result may indicate that a unique feature of
BudgetSmart the more deliberate and constant

engagement with actual accrued bill costs was


popular with customers.
A large number of BudgetSmart customers who
received SMS notifications and emails about their
credit status received either incorrect information
(i.e.messages suggesting they were in debit when
they were in fact in credit), were not receiving
information at all (and therefore had to guess how
much money they owed) or were receiving messages
when they claimed they knew nothing about the trial.
These issues were traced to a combination of poor
customer data (mainly sourced from Ausgrid records
as the network operator) or software errors.
Some customers also requested a more transparent
approach (i.e. the ability to view their real time
account online).

Dynamic peak rebate (Home Energy Rebate)


Fifteen dynamic peak rebate events were called during
the trial. Of these, three events were called in summer
of 2012/2013, four events were called in autumn or
spring of 2013; four events were called in winter 2013
and four events were called in summer of2013/2014.
The analysis performed on the dynamic peak rebate
events customers usage data indicated that:
On average across all events, customers who
trialled the Home Energy Rebate with no feedback
technology exhibited a greater reduction in total
usage than customers who trialled the Home
Energy Rebate along with an IHD. Customers who
trialled the Home Energy Rebate as a standalone
product reduced their total usage (in kWh) by
approximately 19 per cent, compared to a 14 per
cent reduction, on average, for customers who
trialled the Home Energy Rebate along with an
IHD. This indicated that an IHD was not effective
in helping customers trialling the Home Energy
Rebate reduce their usage

62 Attempts were made to contact these and other


customers who raised similar complaints and to
explain the purpose and effect of the product, i.e.
that customers are charged lower rates throughout
the year to compensate for the calling of DPP events
and that customers did not have to switch off their air
conditioning in order to respond to pricing signals
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 129

On average across all events, customers who


trialled the Home Energy rebate as a standalone
product and those who trialled the Rebate along
with an IHD, received a similar average rebate
($14and $13.85 respectively). That the rebate
received was similar despite differing usage
impacts, may have been attributable to Ausgrids
methodology for calculating baseline usage63.

Technologies
As part of the Smart Grid, Smart City Program
analysis, a customer survey was commissioned. The
full results of this survey are presented in the Institute
for Sustainable Futures (ISF) customer research report
which is available on the Information Clearing House
https://ich.smartgridsmartcity.com.au.
The IHD was considered by customers to have been
the easiest to set up (with 77 per cent of survey
participants finding it easy or very easy to set up)
and the IHD became easier to use over time. The HAN
was considered the least simple to set up and use,
however the difference in customer survey responses
was not statistically significant which suggests that
while the HAN was more complex to set up and
operate, this did not strongly impact on customers
ability to use the technology.
Empirical evidence showed that elderly and nonEnglish speaking customers had the most difficulty
in completing the set up process and effectively
operating the IHD.

With the exception of the ability to track greenhouse


gas emissions, all of the feedback technology features
were perceived to be useful by over 80 per cent of
customers. The most useful feature, as rated by
customers, was the ability to see near real time usage
in dollars (66 per cent), followed by near real time usage
in (kWh) (60 per cent). The ability to see near real time
usage was more important to customers than the ability
to see change in usage over time for all technologies
other than the HAN, which suggests that the HAN
encouraged longer term efforts to use appliances more
efficiently most effectively.
The ability to track greenhouse gas emissions was
by far the least valuable feature to customers, with
47 per cent describing it as not at all useful. The
presentation of greenhouse gas emissions, for
example to determine if the concepts were readily
understood by the average household user was
not examined.
Customers who trialled the IHD were the least likely
to rate the ability to see change in electricity usage
(in kWh and dollars) over time as useful, which
suggests that the IHD did not provide this information
as effectively as the other feedback technologies.
Customers also rated the ability to see near real time
usage as more useful than ability to see change in
usage over time. This may suggest that customers
are more interested in seeing the difference between
appliances than in using their appliances more
efficiently in the longer term.

63 Frontier Economics note that the correlation between


kWh savings and rebates was not strong. This was
due to the fact that Ausgrids baseline usage for
calculating rebates used a different methodology to
Frontiers methodology. NERA Consulting noted that
Ausgrids methodology led, at least for one event, to
the selection of an inappropriate baseline usage level.
Ausgrids methodology took no account of temperature
differences between event days and baseline days.
This could have been the reason for the occasional
selection of inappropriate baseline usage

130 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Customers who trialled the IHD showed a significantly


higher level of regular user engagement than both
customers who trialled the online portal and the online
portal in combination with the HAN, with almost
half (49 per cent) using their IHD at least weekly and
almost a third (33 per cent) using it daily. This was
presumably attributable to the lower entry barriers to
engagement with the IHD as the IHD was in open sight
of customers and did not require a dedicated login.
Thirty-six per cent of customers were found to have
never used their feedback technology, or only used
it only once or twice. Customers who did activate
and use their feedback technology did so mostly on
a weekly basis (40 per cent), with some using their
technology on a daily basis (22 per cent). However, half
(51per cent) of participants reported that their use of
the technology decreased over time.
Overall, the findings indicate that for the majority of
customers, the reason for reduced engagement was in
fact positive, as the feedback technologies contributed to
their awareness over their electricity usage.
When considering the impact of feedback
technologies on customer awareness of their electricity
usage, the IHD had the greatest impact which was
perhaps due to its ease of interaction. The portal
in combination with the HAN had much greater an
impact than the customer portal alone, presumably
due to the additional functionality of the HAN which
led to a greater level of customer understanding
and awareness. The online customer portal had the
lowest impact on improving customer awareness.
79 per cent of customers with an IHD reduced their
overall electricity usage, compared to 73 per cent for
the customer portal plus HAN and 71 per cent of the
online customer portal alone.

Demographic analysis
The following conclusions were able to be drawn from
an analysis of customer survey data in regards to the
responses of differing demographic groups to the
Customer Application Program trials:
Low energy users had a lower probability of
reporting that participation in the trials had
increased their ability to reduce bills. This could be
attributable to this demographic group having less
discretion over their usage
Financially vulnerable households were less likely
than other households to perceive that they could
shift a large portion of their usage to outside peak
times. This demographic group was also more
likely to recommend the products they trialled
Pensioners were generally more likely than other
customers to perceive that they could shift a
large portion of their usage to outside peak times;
however they were only slightly more able than
other demographic groups to actually be able to
shift usage to outside peak times
A substantially higher proportion of households
with children reported feeling unable to pay their
electricity bill, compared to households with no
children. However households with older children
were more likely to report perceived savings. This
indicates that households with young children may
have been more vulnerable to pricing regimes
For all demographic groups, the pricing structure
trialled was the primary driver of overall reduction
in usage and peak demand reduction. Feedback
technologies on their own had little effect

The portal in combination with the HAN was the most


effective of customer trialled feedback technologies
when considering how improved awareness went on
to empower customers to reduce their bills. The online
portal alone was the least beneficial in this respect
from a customer perspective.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 131

Product satisfaction
Overall, customers were positive about the products
they trialled. Most products were viewed very positively
by customers, in particular the DPR products.
Customers who participated in the Network Trial
with an IHD only reported the lowest average level of
satisfaction. Customers were found to have formed
their opinion on their product after a short period of
time and this opinion did not change much over the
course of the trial.
In terms of overall product satisfaction (not related to
any particular product), the following conclusions were
able to be made:
Customers with English as their first language were
more likely to be satisfied with their products
Product satisfaction was higher for customers who
stated they understood and were able to obtain the
benefits of their products. Whether this is a cause
or effect however is unclear, although customers
who understood the benefits of their product
were more likely to demonstrate active engagement
and behaviour change, which resulted in
financial savings

Overall, 69 per cent of customers were interested in


continuing use of their product if it were to be available
in the future. The products which recorded the highest
levels of customers interested in continuing use were
the Home Energy Rebate products, SeasonSmart and
BudgetSmart (both along with the online customer
portal). Products consisting only of the IHD had the
lowest levels of customers willing to continue use of their
product. These findings suggest that the presence of
tariffs increases customers desire to continue product
use. However it should be noted that participants
who opted out of the trials are not represented in
these results.
Bundled products (i.e. a combination of pricing
structure and feedback technology) were also the
most likely to be recommended by customers.
Thiswas particularly the case for the Home Energy
Rebate products.

Customers who trialled the Home Energy Rebate


reported the highest levels of satisfaction, followed
by PriceSmart, BudgetSmart and SeasonSmart
Customers who engaged with their product were
more likely to report product satisfaction
Feedback technology only products resulted in the
lowest levels of customer satisfaction

132 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Change in usage
An analysis of change in usage data indicated that
customers who trialled the following products achieved
the greatest reductions in overall electricity usage:
BudgetSmart (average overall reduction of 3.5 per
cent across all offered BudgetSmart products)
PriceSmart (average overall reduction of 2 per cent
across all offered PriceSmart products)
SeasonSmart (average overall reduction of 1.8 per
cent across all offered SeasonSmart products)
These results indicate that a top up reward pricing
structure was the most effective in reducing overall
customer electricity usage.
An analysis of change in usage data indicated that
customers who trialled the PriceSmart products
achieved the greatest peak demand reductions, with
an average peak reduction of 38.9 per cent across all
offered PriceSmart products. Whilst customers who
trialled pricing structures also achieved reductions in
peak demand, as these customers were not notified of
peak events, there could have been no relevant peak
demand analysis over their usage and reductions in
usage were able to be calculated on an overall basis
only. Analysis of trial results showed that the DPP
product was the most effective in reducing customer
peak demand. Despite this finding, there remains the
practical challenges of network operators and retailers
in managing issues associated with non-firm demand
response. This has been discussed in greater detail in
Part Two of this report.

Providing customers with a feedback technology had


varying impacts on their overall usage. Whilst the IHD
and customer portal (when trialled as a standalone
product) were effective in assisting customers to
reduce their overall usage, when a customer portal
was trialled in combination with appliance control and
sub metering, this combination actually increased
overall customer electricity usage.
Customers who trialled feedback technologies by
themselves (i.e. with no associated pricing structure
and therefore no change to their retailer tariff),
exhibited negligible change in overall usage. These
results indicate that the pricing structure trialled, rather
than the feedback technology, had a greater impact on
electricity usage.
Providing customers with a feedback technology
also had varying impacts on their peak demand
reduction. Customers who trialled one of the peak
event based products (either PriceSmart or the
Home Energy Rebate) in combination with an IHD
reduced their peak demand by a smaller amount than
customers who trialled the same products without a
feedback technology. However, customers who trialled
PriceSmart with a feedback technology other than
the IHD reduced their peak demand by 5 per cent
more than customers who trialled PriceSmart as a
standalone pricing structure.
These results indicate that the pricing structure trialled,
rather than the feedback technology, had a greater
impact on peak demand reduction.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 133

13 Distributed Generation and


Distributed Storage (DGDS)
13.1 What are
distribution generation
and distributed storage
technologies?
Distributed generation refers to generation devices
installed at the distribution network level, typically
solar PV panels, small-scale wind turbines and gas
fuel cells. Distributed storage refers to battery
technology which can be used to store electricity to
provide back-up supply to a customer premise for
reliability purposes or store surplus generation during
times of low demand which can then be exported

back to the grid, or used locally, during periods of


high demand. These devices may be installed at
a customer premises connected to the electricity
network, either directly, or through the customers
existing network connection.
The installation of distributed generation within
an electricity network represents a shift from the
traditional view of planning, design and operation
of electricity distribution networks which have been
predicated on the basic premise that energy only
flows from centralised generation sources, through the
transmission network, to the distribution network, and
onto the customer (described in Figure 1-26).

Figure 1-26 Traditional view of electricity flows

Bulk Supply Point

Subtransmission
Substation

Zone Substation

Distribution
Substation

Traditional views of energy flow

134 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Domestic, commercial &


industrial customers

The increasing adoption of new distributed generation


technologies by customers, particularly rooftop
solar PV generation, represents a paradigm shift
for network operators. The past decade has seen
a marked escalation in rooftop solar PV panels with
installed capacity rising from 23 MW in 2008 to around
1,450MW by the start of 201264.
As the number PV systems continues to grow and
potentially other technologies like distributed storage
are deployed, the role of networks is changing from
being solely focussed on delivery of electricity to
customers, to a wider set of services.
Importantly, electricity no longer flows in a single
direction and distribution network operators must
now plan, design and operate networks taking into
consideration generation at customer premises and
reverse energy flows (described in Figure 1-27).

Figure 1-27 Emerging view of electricity flows

Bulk Supply Point

Subtransmission
Substation

Zone Substation

Distribution
Substation

Domestic, commercial &


industrial customers

64 
AEMO Rooftop PV Information Paper, National
Electricity Forecasting 2012
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 135

Increasing levels of distributed generation and


distributed storage provide an opportunity to
improve the operation of the distribution network, by
increasing reliability, reducing costs and facilitating
increased levels of renewable generation. However,
from a network perspective, planning, designing and
operating networks in this environment will be more
challenging because there will be limited information
on consumer demand within the lower voltage network
and the large variance in network demand across
the day.

13.2 The role for


distribution generation
and distributed storage
in a smart grid
One of the key drivers of capital expenditure for the
electricity supply industry is peak demand where,
unlike many other industries, failure to supply peak
demand can result in disruption of supply to many or
all customers, not just the last customer to increase
their load.
This technical feature of electricity supply systems
and the very large economic cost of disruption or loss
of supply means that generation, transmission and
distribution systems have been designed to meet the
peak demand with a high degree of reliability. As a
consequence some network assets may only be fully
operational for a short period of time, i.e. utilisation
levels for those assets is low.
Distributed energy resources (distributed generation
and distributed storage) have the potential to defer
asset investment for assets that are only installed to
meet peak demand including:

Avoided or deferred capital expenditure on


large scale generation capacity by generating
electricity and/or heat near the point of use or by
using storage devices charged during off-peak
times, there is a potential to reduce expenditure on
large scale peak generators
Avoided or deferred transmission and distribution
capital expenditure distributed energy resources
have the potential to reduce some network capital
expenditure by reducing the peak load growth. In
this case network augmentation has the potential
to be deferred, delayed or downsized providing
savings for customers and higher asset utilisation
rates for network operators
Avoided costs of electricity production Higher
short run marginal costs associated with peaking
plant relative to base load plant, can lead to higher
prices in electricity markets during peak demand
conditions. If distributed generation and distributed
storage assets are operational at these times, they
could potentially lower wholesale electricity spot
prices during periods of peak demand (refer to
discussion in Appendix One on the operation of
Australias electricity markets)
The successful integration of additional distributed
generation resources into distribution networks will rely
on improvements in the level of detailed information on
the underlying demand for electricity combined with
an improved understanding of the low voltage network
impacts (for example, the impact on voltage and the
safe operation of the network).
It is reasonable to expect that the continued
deployment of distributed generation, and potentially,
distributed storage will occur in a generally random
manner. Accordingly an important aspect of the Smart
Grid, Smart City DGDS Project was to emulate this
approach to customer distributed energy resource
deployment. This included the deliberate sub-optimal
location of a number of these technologies in the
DGDS Project in order to examine to the potential
negative impacts on the distribution network.

136 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

13.3 Distribution
generation and
distributed storage
objectives and timelines
in the Smart Grid,
Smart City Project
The Smart Grid, Smart City DGDS Project was
initiated to provide critical data and information to help
understanding the implications of a commercial scale
rollout of distributed storage integrated with distributed
generation and other smart grid technologies.
The DGDS Project sought to:
Understand the maturity and suitability of
DGDS devices
Assess the impacts on the grid from increased
penetration of DGDS devices
Understand the value DGDS devices can potentially
deliver for network operators and customers

The program consisted of field trials, advanced


modelling and PS+EDGE simulation trial elements. The
field trial results were used to validate the models and
gain insights into the commercial-scale deployment of
these technologies. The advanced modelling and field
PS+EDGE simulations provided additional analytical
tools for different penetrations of DSDG devices which
are not commercially and physically possible in a
field trial.
The key technology components utilised in the DGDS
Project included solar PV, small wind turbines and fuel
cells, customer battery storage, grid battery storage
and diesel generation (to simulate high penetration
of PV generation). A range of measurement and data
monitoring technologies were also installed for data
acquisition and control purposes and to provide data
acquisition and centralised control of some devices
including smart meters, modems, data loggers, remote
terminal units and back office systems.
The trials were based in three locations (Figure 1-28)
in Newington (a suburban environment located in
Sydney), Newcastle (an urban environment) and Upper
Gundy (a rural environment located near Scone
in NSW).
The DGDS Project commenced in October 2010
running through a number of phases as shown
in Figure 1-29.
Throughout the entire DGDS Project, community
consultation and communication with participants
was undertaken.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 137

Figure 1-28 Distributed generation and distributed storage trial locations

Figure 1-26 Traditional view of electricity flows

These three areas were selected because they cover


the range of networks within Ausgrids network area,
which were intended to be used as proxies for the
majority of different types of networks in Australia.

138 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

13.4 Where are


distribution generation
and distributed storage
technologies are
currently deployed?
There is a wide range of literature and research
available pertaining to distributed generation and
storage technologies in different parts of the world.
This literature examines a range of technologies,
each at different levels of maturity, cost and scale.
Further, these studies also were undertaken in a range
of different electricity markets. Additional detail is
provided in the DGDS Technical Compendium.

13.5 Smart Grid,


Smart City distribution
generation and
distributed storage trials
As previously discussed, the DGDS Project was
conducted in three trial areas, Newington (Sydney),
Newcastle, and Upper Gundy (Upper Hunter Valley).
As part of this project 60 consumer batteries, 25 fuel
cells and 10 small wind turbines were installed. In
addition, a large number of existing PV installations in
the suburb of Newington was also utilised. A diesel
generator was also installed in Upper Gundy to
emulate high penetration of solar PV to understand the
impacts on a rural network.

Very few studies however, have examined the impact


of a commercial-scale deployment as was attempted
by the Smart Grid, Smart City DGDS Project.
Figure 1-29 Timelines for the DGDS Project

2009

2010

2011

2012

2013

Initiation & Trial Design


Trial Site & Participant Selection
Installation & Commissioning
Field Trial and Data Collection
Analysis and Reporting
1 2
Milestones

1 Community Consultation

3 4

5 6 7 8 9 10

2 Newington Network Reconfigured

7 Automatic battery Dispatch


Algorithm Activated

 etwork Monitorng
3 N
devicesinstalled

8 Vendor selected to supply and


install 10 wind turbines

4 Community Update

9 Community Update

5 All Fuel Cells Commissioned

10 All Batteries Commissioned

6 Community Update

11 Gundy 8 Turbines Commissioned

11

12

13 14

15

12 Community Update
 etwork Simulation Model
13 N
developed
 iesel Generator (Solar Farm
14 D
Simulation) Commissioned
 ewcastle 2 Turbines
15 N
Commissioned

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 139

Given that some of the technologies were new


technologies to Ausgrid, two aspects of the trial
presented challenges:
Initially, installation of the selected customer
batteries and fuel cells was slow due to there
being no previous relevant installation experience.
However installation times improved as experience
was gained and the process was refined
The small wind turbine installations were delayed
due to difficulties in obtaining Development
Application (DA) approval for suburban areas within
Newcastle. As a consequence the original program
was adjusted and eight small wind turbines were
installed in Upper Gundy (rural area) by August 2012
The suitability, maturity, impacts and value of a variety
of different technologies were trialled and modelled
using PS+EDGE across these three areas, which
represent a range of different types of networks
common across Australia. The trials investigated
the impact of DGDS on power quality parameters
including:
Voltage levels
Power factor (PF)
Voltage unbalance
Total harmonic distortion (THD)
Flicker (PST and PLT)
On selected days, controllable DGDS devices were
operated in a way to mimic rooftop solar PV systems
to simulate an increase in penetration of solar PV
systems in the network. In the Newcastle Area and the
Newington Area, power quality meters were placed at
the Distribution Centre (DC), a relatively robust part of
the network, and at the end of the LV feeder, a weaker
part of the network.

13.5.1 Distribution generation and


distributed storage trial design
elements
The three key trial locations for the DGDS Project are
described below.

Trial location Newington (high


penetration rooftop solar PV)
The Newington suburb housed the Athletes Village
for the Sydney 2000 Olympic Games. The housing
stock was built to high energy efficiency standards,
and one of the features of the suburb was the
installation of solar panels on almost every home.
Newington includes an estimated 1,104 rooftop solar
PV installations. These are mainly 500W and 1kW
systems on residential rooftops. A larger system is at
the Newington Public School, and another adjacent to
an industrial building, Newington Armoury (owned by
Sydney Olympic Park Authority).
Newington represented a saturated suburban
network and was selected to understand the low
voltage grid impacts of a growing number of PV
systems in Australia. The distribution network in
Newington is underground. In total, Newington has
approximately 1,800 customers, with most houses
having an existing 1kW or 0.5kW PV system. This
area therefore presented an opportunity to examine
the impact on the network of high penetration rooftop
solar PV systems. Examination of meter data at the
beginning of the trial showed that around 75 per cent
of the PV systems were functional.
The network was originally arranged with most of the
photovoltaic systems on ten distribution substations
supplied by five 11kV feeders from two zone
substations (Homebush Bay and Flemington). The first
stage of the trial was to rearrange the 11kV network to
increase the penetration of rooftop solar PV systems
on a single 11kV feeder. The final network state for the
trial consisted of 9 out of 10 distribution substations
being connected to the same 11kV feeder, supplied by
Homebush Bay zone substation.

140 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Trial location Newcastle


(Smart Future Urban Network)
Newcastle was chosen to trial a Smart Future urban
network, where the effects of a mixture of DGDS could
be examined on a typical urban network. The intention
of the Newcastle trial was to test the impacts and
benefits, on a predominately residential feeder, of the
types of DGDS technologies customers may install
in the future. The Newcastle trial area was a typical
suburban 11kV radial network, with some existing PV
systems. In addition 40 distributed storage devices (5
kW ZBM flow batteries) and 25 distributed generation
devices (1.5kW solid oxide gas fuel cell) were installed
and actively managed to demonstrate a potential
Smart Future network with a variety of distributed
generation and storage devices. The trial area included
a mixture of overhead and underground network.

The choice of Newcastle for the trial was based on its


ability to facilitate these proposed technologies. Such
considerations included:
The site had a reasonable number of existing
rooftop solar PV systems65. Of approximately 1,800
customers supplied by the feeder, 138 had PV
systems installed at the beginning of the trial, with
approximately 153 systems present during the key
part of the trial period from 1 July 2012 to
30 June 2013
Siting requirements included an existing gas supply to
fuel the solid oxide fuel cells, and/or space to be able
to house either a fuel cell, or a zinc bromine battery
The intention to trial the installation of small wind
turbines in a residential area meant there was also
a requirement for small parks, and green spaces
along the feeder length

65 Likely to have been installed due to incentives from


both the federal and state government programs
National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 141

Within the Newcastle trial area different devices


were clustered in locations supplied by the same
distribution transformer. This installation strategy
allowed for an increased penetration of distributed
generation and storage technologies on the low
voltage network thereby increasing the observed
effects of the technologies.

Grid battery trials

Trial location - Upper Gundy


(Thin Grid Rural Network)

It was hypothesised that in the future distribution


network operators would own, install and operate grid
batteries to reduce peak demand on critical network
elements, or improve power quality by offsetting the
transient impacts of non-dispatchable generation (such
as solar PV).

A rural location at Upper Gundy near Scone was


chosen to trial the impacts of DG and DS on a lightly
loaded, low capacity, high impedance, sparsely
populated overhead 11kV feeder with a small number
of customers (31) spread approximately 12km. Twenty
2.4kW small wind turbines were installed to test the
impact and value of DG and DS on a rural network.
Typically networks like the one at Upper Gundy have
lower reliability levels, and more power quality issues
than suburban networks. Thin grids are likely be more
susceptible to negative impacts from high penetration
distributed devices, and could demonstrate
significantly higher value in solving existing power
quality and reliability issues. Ausgrid also used this
opportunity to demonstrate the potential for creating a
microgrid, using smart grid technology and customer
located DG and DS devices so that this section of
the network could successfully and safely operate
independently or islanded from the grid.
The section of the network on which the trial occurred
was a spur, or tee, running north from Gundy into
the Upper Gundy area. An important aspect of
the selected network was that the spur had only
one source of supply, which means that any fault
disconnects all customers downstream of the fault,
until the fault is repaired.

A grid battery is a battery which is sized to provide a


significant amount of energy, approximately five per
cent of the load on an urban 11kV feeder, equivalent
to 100s of KVA. They are also physically much larger
than the batteries located at customers premises,
often being installed inside large shipping containers.

Ausgrid had intended to deploy grid batteries in


all three trial areas as part of the DGDS Project,
however, this project element produced significant and
unforeseen challenges.
During the trial design stage, Mitsubishi Japan
approached the Smart Grid, Smart City Program
to offer a 1MVA lithium ion battery. Mitsubishi had
developed a large grid battery and wanted to test it in
a real-world environment which Ausgrid intended to
deploy in Nelson Bay. However, the offer of the battery
was retracted after the earth quake and tsunami
occurred in Japan.
A grid battery was an essential aspect of the microgrid
design intended for the Upper Gundy trial because
it was to provide grid stabilisation so that distributed
generation and storage could remain connected.
When it was determined that the planned trial was
not possible without a grid battery, the trial design for
Upper Gundy was reviewed. Instead Ausgrid decided
to use a 200kW diesel generator as a cost effective
device to mimic the behaviour of a 30kW PV system
and grid battery combination.

142 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

The Newington trial was intended to show the impact


on an urban network of high penetration rooftop
solar PV systems. The role of the grid battery was to
simulate a generation profile similar to that of extremely
high rooftop solar PV penetration on a feeder. This
remained an essential component to the Suburban
Saturation trial design and May 2012. In June 2012
Zenergy was awarded the contract to supply a
grid battery.
One of the biggest challenges was finding a suitable
location for the grid battery. The size of the battery
(installed in a large shipping container), limited the
number of places it could be installed in a residential
area. Although lengthy negotiations were undertaken
with the strata management in a particular location,
Ausgrid were unable to secure approval. This resulted
in the Newington grid battery trial not proceeding66.
Instead a PS+EDGE simulation study was conducted
to analyse the impact of a Battery Energy Storage
System (BESS) connected at the end of a suburban
feeder in Newington. The Newington 11kV feeder
comprised nine distribution substations and was fitted
with online distribution monitoring. The BESS had a
capacity of 60kW for two hours, was able to charge /
discharge at various rates and was connected to the
network via a four quadrant inverter. The objectives of
the PS+EDGE simulations were to:
Determine the ability of the BESS to deliver
improvements in the voltage profile of the 11kV
feeder on the network
Ascertain what impact using the BESS as a reactive
load or source can have on the voltage deviation of
the network
Determine the impact of increased PV penetration
and the possibility of reverse power flow

66 In October 2013, the DGDS Project transferred


responsibility for the grid battery project to Ausgrids
Demand Management team to implement the Zenergy
80kVA lithium ion battery in the Ausgrids network.
A suitable site was found in Homebush, close to
Newington, to install the grid battery.

13.6 High level


trial results
Given the extensiveness of the trials, the high level
trial conclusions are presented here under a number
of categories including:
Technology (suitability, maturity etc.)
Network impacts
Customer impacts
Further detail on the results and conclusions from the
DGDS Project Technical Compendium can be found
on the Smart Grid, Smart City Information Clearing
House website https://ich.smartgridsmartcity.com.au/

13.6.1 Distributed generation and


distributed storage technology
results
Solar PV
Solar photovoltaic generation (solar PV) is considered a
mature technology. In the case of the DGDS trials, the
program utilised existing PV installations in Newington,
Sydney. Approximately 25 per cent of the existing PV
systems were not operational, mainly due to inverter
failures. Customers were encouraged to repair these
systems however only 11 per cent of those customers
contacted did so.
The trials demonstrated that the generation profile from
PV systems does not necessarily align with the local
network profile which therefore diminishes the benefits
that solar PV can offer in terms of reducing peak
demand on individual distribution network assets.
The trials also showed that the orientation of solar PV
panels towards the west may offer better alignment
of photovoltaic output with the late afternoon summer
peak demand periods. However, westerly facing
rooftop solar PV systems generate lower total energy
over the whole day (and year) therefore reducing
the financial value that customers may obtain from
a westerly facing PV system under current tariff
arrangements.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 143

Redflow battery
The RedFlow battery is a developmental technology
using a flow process based on zinc bromine. This
product was just emerging from the research and
development phase when the Smart Grid, Smart City
Program started. The relative immaturity of the product
meant that many of the challenges in trialling this
technology stemmed from the lack of demonstrated
practical applications and lack of experience
implementing this type of technology in the field.
The DGDS project trialled the RedFlow R510 zinc
bromine flow battery. The customer battery storage
was sized to store up to 10kWh of electricity, or about
half the daily requirements of a typical household.
The trial found that battery characteristics including
the discharge/charge profile, full cycle efficiency, and
ambient temperature performance all impacted the
battery output.
One key aspect of these batteries is they are net
consumers of energy. The customer based batteries
used in the trials consumed 300W in float mode,
and this loss of energy had an impact on overall
performance.
Another challenge was that the storage capacity
of the batteries was quite limited which meant that
correctly predicting when the network peak was
going to occur was important so that the batteries
were able to be used for peak demand management.
The trial found that if the customer based batteries
were dispatched too early, particularly on hot days
when high temperatures reduced the effective battery
capacity, the battery output dropped off before the
peak demand event occurred, or had ended. Similarly
if the batteries were dispatched too late the peak may
have already occurred.
The predictability of network peaks is generally better
for larger sections of the distribution network (i.e.
where network operators are predicting the behaviour
on average of a larger number of customers over a
greater geographical area), compared to load profiles
at individual feeder or customer level. This suggests
the scheduling of grid-embedded batteries for charge
and discharge connected higher up in the network

may be less problematic than for customer (home)


based batteries.
The value of this trial was in testing how this new
technology could be applied in practice and
identifying issues across technical, customer and grid
performance areas. It is anticipated that many of the
issues discovered through this process could be used
as lessons across the battery industry generally. - The
RedFlow customer batteries performed relatively well,
with battery export meeting device expectations, for
the first five months of the trial. After this period, the
performance started to decline prompting a range of
modifications. As a result of the discoveries in this trial,
RedFlow are currently redesigning aspects of
the battery.

Fuel cells
Although slightly more mature than the Redflow
batteries, the Solid Oxide (BlueGen) fuel cells were
a relatively new technology at the commencement
of the trial. The BlueGen solid oxide fuel cells
that were trialled in this project were at the precommercialisation stage of development.
One of the key constraints for installations of the
BlueGen fuel cell was the requirement for the cell to be
connected to all utilities: i.e. electricity, gas and water.
To incentivise customers to trial the product they were
provided with hot water during the trial at no charge.
The units provided continuous electrical output of
1.5kW and waste heat for hot water. A gas hot water
booster was also installed to ensure hot water supply
quantity and temperature satisfied the customers
requirements. Trial participants were very enthusiastic
about their involvement in the DGDS project and
willingness to trial the technology was high. Their
feedback at the end of trial was generally very
positive and many indicated disappointment when
the units were decommissioned (it is important to
note customers were not liable for any of the capital
or operating costs of the units during the trial, and did
receive the financial benefits associated with free hot
water produced by the fuel cells).

144 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

Fuel cells typically take a long time to start-up and


to shut down, and as a consequence they are most
efficient and economic when running at steady
output, typically near, or at the rated capacity. At
lower and higher output levels the efficiency of the
fuel cell is lower, and cycling the cells through wide
ranges of output can have implications for the life of
components. Given these factors, fuel cells are more
likely to be used to reduce network demand at all
times, rather than just during the peak.
Given the pre-commercial nature of the BlueGen fuel
cells, there were a number of major failures of the
systems which required repair during the trials. In
addition it was noted that there was the potential for
damage to components if not shut down correctly, for
example, during a network outage. The devices also
require constant water and power supply, along with
continuous monitoring via an ADSL internet connection
which means this device is unsuitable to areas which
have unreliable utility supplies.

Small wind turbines


Small wind turbines are considered a relatively
mature technology. Maintenance requirements for
wind turbines are also typically very low, i.e. a visual
inspection annually to see that the blades are clean
and operating correctly.
Existing regulations governing the installation of wind
turbines presented key lessons for the Project. Existing
legislation was prescriptive, (e.g. these devices could not
be installed within 100m of a residence) and planning
approval processes were time consuming. Planning
approval processes differed between different councils
and the Project found that the development approval
was easier in a rural environment as its requirements
were limited to a less rigid site selection process and a
feasibility assessment.

Grid batteries
As previously discussed, implementation issues
prevented the installation and trialling of a grid
battery during the Smart Grid, Smart City trial
period. However, advanced modelling and PS+EDGE
simulations were performed in order to investigate the
potential network value of grid batteries in reducing
peak demand, reducing network capital expenditure
and improving asset utilisation.
Advanced modelling using PS+EDGE model was
performed for the Newcastle trial area with a large
grid battery with a rated capacity of 1MVA and battery
storage capacity of 2MWh. In addition, a PS+EDGE
simulation of a smaller grid battery (60kVA/ 120kWh)
was performed for the Newington trial area by the
University of Newcastle.
The objectives of the grid battery PS+EDGE simulation
were to determine:
If the introduction of a BESS had the ability to
reduce the peak demand on the local network
(11kV feeder load)
What impact using the BESS as a reactive load
or source can have on the voltage deviation of
the network
The impact of increased rooftop solar PV
penetration and the possibility of reverse power flow
The models and calculations were based on field
measurement data from the Homebush Bay zone 11kV
feeder for the Newington trial area, collected at 10
minute intervals, for the period from December 2011 to
December 2012. A load forecast tool was developed in
MATLAB for predicting the next days half-hourly load
demand of the Newington feeder, based on historical
temperature and load data. This tool was then used to
determine when to discharge the battery to maximise
the potential reduction in feeder peak demand.

It was found that the majority of the small wind turbines


provided their peak power output in the early afternoon
between 1 pm and 3 pm. The customer experience
was generally positive, with several customers indicating
that they were confused by concerns expressed by
some people over wind turbine noise as they had not
experienced these issues.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 145

The findings from the PS+EDGE simulations


indicated that:
Real power discharged from a BESS can reduce
voltage deviation (<1V) in a distribution network.
This improvement was small because of the
relatively low impedance of the Newington feeder
and the limited capacity of the inverter
The reactive power charge from a BESS can
reduce the voltage deviation in a distribution on
a limited basis (0.85V). This voltage deviation
improvement is small because of the relatively
small impedance of the feeder and the size of the
inverter. The PS+EDGE simulation results indicate
that the addition of active inverters can increase
the effect of a BESS on the voltage profile in
comparison to real power discharge. PS+EDGE
simulations demonstrated that the effects of a
larger BESS device on higher impedance networks
will have a larger effect on the voltage profile.
Enhanced solar rooftop PV penetration (simulated
by operating a 60kW/120kWh battery) did not
cause reverse power flow and over-voltage
problems on the selected peak day (however
reverse power flow can occur on days when the
load is low)
Reverse power flow (simulated by operating an 180
kW/360 kWh battery) caused LV voltages to vary in
the range of 1 1.5V which correspond to 2.72 per
cent 4.08 per cent of the allow LV voltage range
Batteries, with their limited storage capacity, are
inherently sensitive to the energy required to reduce
the demand, therefore unexpectedly long peaks
can reduce the potential effectiveness of batteries in
reducing peak loads. This storage capacity limitation
has also highlighted the necessity for combining
accurate load forecasting with battery control
methodologies in order to optimise the potential value
of grid batteries to network operators.

13.6.2 Network technical impacts


No issues were found to arise with power quality from
the DGDS trials. It was determined that this result was
reflective of the strength of the Ausgrid network in
which these trials were conducted. Although a voltage
rise was anticipated as penetration of distributed
generation increases in the network (particularly
increased solar PV generation), the trials demonstrated
that there is no single threshold level of penetration at
which voltage problems arise (within the context of the
Trial configuration).
However it was concluded, that a range of different
factors specific to the local network characteristics
such as network strength and network setting influence
voltage levels. High penetration of solar PV systems may
cause no problems on one part of the network, whereas
the same level of penetration may cause voltage issues
on another part of the network (i.e. weak parts of the
network are more likely to start experiencing issues at
relatively modest penetration rates).
The increased role of automated monitoring and
management of distributed generation and storage
technologies will be critical in identifying opportunities
for optimisation of network operations. For example,
non-conforming voltage delivery to customers
was identified through enhanced monitoring which
otherwise may not have been discovered. More details
on monitoring within a distribution network can be
found in the Smart Grid, Smart City Substation Feeder
Monitoring Technical Compendium.
There were a number of issues associated with
islanding of a live network which were found in the
Gundy trial. A technical study was commissioned to
examining the potential engineering responses for
the detection, localisation and clearing of electrical
faults on islanded high voltage rural microgrids.
Additional discussion can be found in the DGDS
Technical Compendium.

146 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

13.6.3 Value of DGDS


Although it is likely that customers who install
technology such as photovoltaic generation or
small wind turbines are driven primarily by the value
proposition of the device, other factors may also drive
their decision. In total, these customer drivers
may not result in optimal electricity market or
network considerations.
Siting considerations can have a significant impact
on energy production due to topography, vegetation
and existing structures. Rooftop solar PV systems
may not always be installed on the north-facing roof,
and investment may not necessarily be confined to
locations that receive the most sunlight. Wind studies
may not always be conducted prior to installation of
small wind turbines due to costs and/or physical
siting constraints.
As a consequence the locations and installations of
these technologies can be sub-optimal, implying that
the energy production and actual impacts of these
distributed energy sources on the network will also be
sub-optimal.
Where rooftop solar PV systems are widely installed
across a network, they may contribute to a general
reduction in the network load, however whether this
impacts the network peak is highly dependent on the
specific load profile characteristics in the area (e.g. the
timing of the peak varies depending on whether
the load is mostly industrial, business, residential
or a mixture).

The non-dispatchable nature of wind turbines and


photovoltaic installations makes it difficult for network
operators to gain benefits in the operation of the
electricity network from these forms of distributed
energy. Wind turbine generation profiles vary in
accordance with local wind conditions and bear no
relationship to typical household energy usage and
rooftop solar PV installations normally generate the
majority of their electricity between 10 am and 4 pm
(assuming limited cloud cover). Residential peak load
typically occurs between 4 pm and 9 pm. This means
that whilst photovoltaic systems reduce the average
load on the network, their impact on peak load is far
less significant. The non-dispatchability of photovoltaic
and wind generation was demonstrated in trials and
PS+EDGE simulations in the Gundy area.
PS+EDGE trial modelling found that orientation of
rooftop solar PV panels can be varied (to face west)
to improve alignment with local residential summer
peaks and hence potentially be of benefit to network
operators, however this results in lower total solar PV
generation for customers. Based on existing tariffs and
incentives there are no customer benefits to reorienting
new or existing rooftop solar PV systems.

A common characteristic of renewable technologies


such as wind turbines and rooftop solar PV
installations is that any individual system is incapable
of being dispatched to the electricity market as the
availability of the underlying resource (wind or sunlight)
cannot be guaranteed at any given time.

National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP 147

148 National Cost Benefit Assessment: Part One Smart Grid, Smart City Trials ARUP

You might also like