Académique Documents
Professionnel Documents
Culture Documents
Working Group
D2. 26
April 2011
Working Group
D2.26
April 2011
Members
Mehrdad MESBAH, Convenor (France), Robert EVANS (Australia) Dugald BELL (Australia),
Jan PIOTROWSKI (Poland) Jorge MENDES (Portugal) Pedro GAMA (Portugal),
Ion NEDELCU (Romania), Matjaz BLOKAR (Slovenia) Paul SCHWYTER (Switzerland),
Paul RENSHALL (United Kingdom) Claudio TRIGO de LOUREIRO (Brazil),
Elton BANDEIRA de MELO (Brazil) Suzana JAVORNIK VONCINA (Croatia),
Janine LEIFER (Israel) Masayuki YAMASAKI (Japan) Kazuto IWASAKI (Japan),
Eva LASSNER (Hungary) Lhoussain LHASSANI (Netherland) Olav STOKKE (Norway),
Jorge FONSECA (Portugal) Pedro MARQUES (Portugal) Danilo LALOVIC (Serbia),
Juan Antonio GARCIA LOPEZ (Spain)
Copyright 2011
Ownership of a CIGRE publication, whether in paper form or on electronic support only infers right
of use for personal purposes. Are prohibited, except if explicitly agreed by CIGRE, total or partial
reproduction of the publication for use other than personal and transfer to a third party; hence circulation on any intranet or other company network is forbidden.
Disclaimer notice
CIGRE gives no warranty or assurance about the contents of this publication, nor does it accept
any responsibility, as to the accuracy or exhaustiveness of the information. All implied warranties
and conditions are excluded to the maximum extent permitted by law.
ISBN: 978-2-85873-150-3
1
2
INTRODUCTION ................................................................................................................5
COMMUNICATION SERVICE IN THE POWER UTILITY ............................................8
2.1
Introduction...................................................................................................................8
2.2
EPU Communication Services......................................................................................9
OPERATIONAL APPLICATIONS ...................................................................................13
3.1
Protection Communication .........................................................................................13
3.1.1
State Comparison Protection Schemes (Command Schemes)............................15
3.1.2
Teleprotection Signalling Systems .....................................................................16
3.1.3
Analog Comparison Protection Schemes............................................................17
3.1.4
System-wide Protection Schemes .......................................................................17
3.2
Energy Management, SCADA and WAMS Communications ...................................18
3.2.1
SCADA Communications...................................................................................18
3.2.2
Inter-Control Centre Communications ...............................................................19
3.2.3
Remote Control Centre Operator Consoles ........................................................20
3.2.4
Generation Control Signaling .............................................................................20
3.2.5
Wide Area Monitoring System (PMU Communications) ..................................20
3.3
Remote Substation Control and Automation ..............................................................22
3.4
Operational Telephone System ...................................................................................22
3.5
Settlement and Revenue Metering and Customer Communications ..........................24
3.5.1
Energy Metering in the Deregulated Environment.............................................24
3.5.2
Customer Metering, Advanced Metering Infrastructure.....................................24
3.5.3
Advanced distribution applications and Smart Grid...........................................25
OPERATION SUPPORT APPLICATIONS......................................................................26
4.1
Collaborative Multi-media Communications .............................................................26
4.2
On-line Documentation...............................................................................................26
4.3
Substation Automation Platform Management...........................................................27
4.4
Condition and Quality Monitoring Communications .................................................27
4.5
Substation data Retrieval ............................................................................................27
4.6
Mobile Workforce Communications ..........................................................................28
SECURITY, SAFETY AND ENVIRONMENTAL MONITORING................................29
5.1
Security of Sites and Assets........................................................................................29
5.1.1
Video-surveillance of sites..................................................................................29
5.1.2
Site Access Control.............................................................................................29
5.1.3
Environmental Hazards Monitoring (Sites and Assets)......................................30
5.1.4
Intruder Detection ...............................................................................................30
5.2
Human Safety & Operational Security .......................................................................30
5.2.1
Earth Connection Monitoring .............................................................................30
5.2.2
Isolated Worker Safety Communications ...........................................................30
5.2.3
Public Warning Communications .......................................................................31
5.2.4
Hydraulic Structure Operation and Maintenance Applications ..........................32
5.3
Cyber-Security applications communication..............................................................34
OPERATIONAL CONSTRAINTS AND SERVICE LEVEL AGREEMENTS ...............35
6.1
Operational Coverage and Topology ..........................................................................35
6.2
Time Constraints.........................................................................................................36
6.3
Availability Constraints ..............................................................................................39
6.4
Service Survivability and Resilience ..........................................................................41
6.5
Service Security Constraints .......................................................................................43
6.6
Service Integrity..........................................................................................................44
6.7
Future Sustainability, Legacy Openness and Vendor Independence..........................46
6.8
Environmental Constraints..........................................................................................46
6.9
Defining Service Level Agreements ...........................................................................47
7 DISASTER RECOVERY AND SERVICE CONTINUITY..............................................54
7.1
Introduction.................................................................................................................54
7.2
Threats and Risk Management and Risk Assessment.................................................54
7.3
Business Continuity Plan and Disaster Recovery Plan...............................................56
7.4
Project Design Criteria................................................................................................57
7.4.1
Back-up Facilities ...............................................................................................57
7.4.2
Power Supply Independence...............................................................................58
7.4.3
Network Redundancy..........................................................................................58
7.4.4
Countermeasures against Natural Disasters........................................................59
7.5
Enhancing the emergency response capacity..............................................................60
7.6
Disaster Information Systems .....................................................................................62
8 TELECOM SERVICE DELIVERY MODELS..................................................................64
8.1
Introduction.................................................................................................................64
8.2
EPU Profiles - Telecom Service Users .......................................................................64
8.2.1
Coordinating or Operating Bodies without Network Assets ..............................68
8.2.2
Transmission System Operator (TSO) or Transmission Utility .........................69
8.2.3
Distribution Utility..............................................................................................70
8.2.4
Generation Utility ...............................................................................................71
8.3
Telecom Asset Ownership Profiles.............................................................................73
8.3.1
Introduction.........................................................................................................73
8.3.2
Physical layer assets............................................................................................74
8.3.3
Transport network assets.....................................................................................77
8.3.4
Application service networks and platforms.......................................................77
8.4
Telecom Service Provider Relationship to the User ................................................78
8.4.1
Integrated to the Operational User (Type A) ......................................................79
8.4.2
Sister Entity to the Operational User (Type B)...................................................80
8.4.3
Affiliated Service Company (Type C)................................................................80
8.4.4
Independent Service Contractor (Type D)..........................................................81
8.4.5
External Telecom Service Provider (Type E).....................................................81
9 FEDERATING OF SERVICES ON THE PRIVATE INFRASTRUCTURE....................84
9.1
Introduction.................................................................................................................84
9.2
Process and Organization Issues.................................................................................85
9.3
Technical Solutions.....................................................................................................87
9.3.1
Fibre Separation in Optical Cables .....................................................................87
9.3.2
Wavelength Separation through C- or D-WDM.................................................88
9.3.3
Bandwidth Separation through PDH/SDH .........................................................89
9.3.4
Virtual Network Separation (MPLS VPN or Ethernet VLAN) ..........................89
10
MANAGEMENT OF TELECOM SERVICE AND INFRASTRUCTURE ..................91
10.1 Introduction - Need for a Management System..........................................................91
10.2 Present State Assessment and Target Definition ........................................................93
10.2.1
Telecom Business Maturity Modeling................................................................93
10.2.2
Management Process Maturity ...........................................................................94
10.3 Management Frameworks & Best Practices ...............................................................96
10.3.1
Introduction.........................................................................................................96
10.3.2
ITIL Framework..................................................................................................96
10.3.3
NGOSS Frameworx.........................................................................................97
10.3.4
Business Process Framework eTOM ..................................................................98
10.3.5
Relating ITIL to eTOM Framework ...................................................................99
10.4 Towards a Utility Telecom Management Framework ..............................................101
10.4.1
Introduction.......................................................................................................101
10.4.2
Utility Telecom Management Operations Map (uTOM)..................................102
10.5 Upstream Management .............................................................................................105
10.5.1
Introduction.......................................................................................................105
10.5.2
Policy Definition & Business Planning ............................................................107
10.5.3
Strategic Deployment and Tactical Adjustments..............................................109
10.5.4
Business Development, Service Offer and Service Migrations........................112
10.6 Operational Management..........................................................................................113
10.6.1
Customer/User Relation Management ..............................................................114
10.6.2
Communication Service Management Process.................................................115
10.6.3
Network Resource & Infrastructure Management ............................................117
10.6.4
Provider/Contractor Relationship Management Process ..................................120
10.6.5
Enterprise Processes impacting Telecom Service Delivery..............................121
10.7 Management Tools and Information Systems ..........................................................123
10.7.1
Introduction.......................................................................................................123
10.7.2
Element & Network Management Systems ......................................................124
10.7.3
Operation Support Systems (OSS)....................................................................125
10.7.4
Inventory & Configuration Data Base ..............................................................127
11
COST CONSIDERATIONS.........................................................................................131
12
FURTHER ACROSS THE HORIZON ........................................................................135
12.1 Power System Evolution - Smart Grid .....................................................................135
12.2 EPU Organization and Environment ........................................................................137
12.3 Communication Service Provider Environment .......................................................138
12.4 Telecom Technology Evolutions ..............................................................................139
12.5 Information System Evolution - Cloud Computing..................................................141
APPENDICES ..........................................................................................................................144
A1.
IP Voice in Utility Telecoms ........................................................................................144
A2.
Sharing Mobile Emergency Service (TETRA).............................................................149
A3.
Satellite Communications in Power Utilities................................................................151
A4.
Disaster Counter-measures Learning from US 2005 Hurricanes ..............................154
A5.
Survey of Electric Power Dimensioning Practice in EPU Data Centres ......................158
A6.
Deploying a Management Framework Western Power .............................................161
A7.
ITIL Management Framework......................................................................................165
A8.
TM Forum NGOSS - Frameworx.................................................................................173
A9.
List of Acronyms ..........................................................................................................182
REFERENCES .........................................................................................................................185
1 INTRODUCTION
Some 15 years ago, a technical brochure prepared by CIGRE Study Committee 35 started with
the following lines, summarizing the situation and the opportunities facing the power utilities
in terms of telecommunications as seen in early 1990s:
Over the last few years an increasing number of utilities have seen their telecommunication
activities being influenced by technological and operational change. The choices that utilities
are faced with today may have a significant impact on their future development.
Traditionally, power utilities have used telecommunications networks primarily for control and
operation of the power system. Standardized telecommunications equipment could not
always be used in an operational environment. Leasing telecommunication services, on the
other hand, was unacceptable in many cases because of low availability figures and service
level from the monopolized public operator or simply because the public
telecommunications network did not have sufficient geographic coverage to reach all the
utilitys substations.
The use of telecommunication network for other purposes was generally prohibited by
legislation. As a consequence, the planning of future network developments was relatively
straightforward. The networks were planned and designed to meet the utilities particular
operational needs and this decided the extent of investment in new network infrastructure. The
type of network infrastructure provided was also influenced by operational needs with
specialized equipment such as Power Line Carrier (PLC) being used extensively. Although
these criteria still play an important role in many networks today, the focus has now shifted to
cover a broader range of issues:
The requirement for increased capacity, speed and response time of the operational and
administrative services
The need to improve the efficiency of network maintenance and quality of service
(Extracted from TB107 Power System Communications in the High Speed Environment, [1])
In the 15 years that separate us from these lines, if the essential issues have not changed, the
situation is radically different.
The creation of electricity market has created new participants and hence the need for new
communication services. A new group of services has been formed that can be called
Business and Market communications. The requirement to communicate with the
electricity customers and with individual producers is driving important metering
infrastructure projects which may represent an opportunity to implement other utility or
commercial services.
The great majority of electrical power utilities have implemented extensive optical fibre
networks with SDH as the core technology providing the required capacity, speed and
response time of their operational requirements.
Technological orientations for a packet data communication layer are no longer a subject
of discussion. The omni-presence, the ubiquity and the strong industry support for the fully
mature Ethernet and IP technologies make them the natural complementary layer for
providing new services in the electrical utility. The multi-service capability of the IP
technology as discussed extensively in CIGRE Technical Brochure TB249 [2], allows the
creation of a single Integrated Service Network (ISN) to cover different operational,
operation support and corporate IP requirements, or multiple networks, each dedicated to
one of these families of Utility communications. At present, IP connection to the electrical
substation is a strong requirement for many new applications and has been the subject of a
Technical Brochure TB321 Operational Service using IP Virtual Private Networks [3]
and of further ongoing work. In order to carry these IP connections and other more timecritical communications, Utilities implement new Ethernet transport over SDH, over fibre,
or over an MPLS core. Wide Area Ethernet transport has been the subject of a separate
publication in CIGRE [4].
Many new directions in the mode of service delivery, which were to be explored at the
time, are today sufficiently evaluated, to allow a new analysis in the light of more than a
decade of experience. Many pioneers of seeking commercial revenue from the operational
network infrastructure have evolved into standard telecom Service Providers, moving
away from their original goal. Those who have maintained their original scope have
survived due to particular legal, legislative or practical contexts which are interesting to
explore.
New problems and operational issues have appeared due to outsourcing or due to the
provision of commercial services in the liberalized telecommunication market. A previous
CIGRE publication TB108 Business Opportunities for Utilities in the Telecom Market
[5] published in 1997 needs to be reviewed in the light of these new issues and problems.
The effects of commercial service on the operational service provision are to be analyzed.
Finally, moving from the monopolized Public Telecom Operator to many concurrent
Telecom Service Providers have changed radically the cost/performance and quality of
service objectives for the procured services. There has been a clear change of orientation
from a uniform quality objective towards a competitive, avoid non-contractual
performance to reduce cost strategy. The relationship between the Service User and the
Service Provider and the principle of Service Level Agreement between the two is of great
interest and needs to be covered.
At present, Electrical Power Utilities (EPU) are increasingly dependent upon the existence of
fast, secure and reliable communications services. These services interconnect the participants,
platforms and devices constituting the technical, commercial, and corporate processes of the
Utility across the different sites. The communication services are provisioned, managed and
maintained in different ways depending upon different quality constraints, cost and regulatory
imperatives and company policy considerations. The services can be integrated together into a
Requirement for new services or change of scale in the existing services, incompatible
with the present provisioning model
Lack of satisfaction from the services obtained through the existing provisioning model,
Major capital investments and running costs required for refurbishment and extension of
existing facilities,
Lack of qualified staff and the ageing of the concerned technical work-force
EPU Telecom service providing entities who need to adapt to those same changing
requirements
Telecom Service Provider offspring of EPUs who have over time forgotten the service
imperatives of their EPU operational customers
Introduction
The term service is widely used often with a very loose definition and may lead to confusion
and misunderstanding. We shall therefore start with some definitions that shall be used
throughout the document. These definitions are illustrated in figure 2.1.
Service is the perception of a User from a process implemented by a Provider.
A Service Provider deploys a telecom infrastructure and corresponding management
processes in order to offer Communication Services satisfying the requirements of its user
community. The user perceives the service as a network cloud providing communication
connectivity for its user applications and processes. ITU-T E800 defines the service as a set of
functions offered to the user by an organization [6].
A communication service is delivered at a Service Access Point with a certain Quality of
Service (QoS) as stipulated through a Service Level Agreement (SLA) between the Service
Provider and the Service User. A service level agreement can be formally stipulated or implicit
between the provider and the user.
The process of assuring that the terms of the SLA are met is called Service Management. It
relies upon a proper Infrastructure Management and Maintenance performed by the Service
Provider.
Service
Access
Point
User
Application
Platform
Service
Management
Infrastructure
Management &
Maintenance
Dedicated
Telecom
Infrastructure
Application
Platform
User
Procured Service
Communication
Service
User Application
An Electrical Power Utility (EPU), on the other side, provisions telecom services essentially
for its own requirements. The provision process may be multi-layer, employing dedicated
infrastructures or procured services at different levels and presenting multiple User-Provider
relationships as illustrated in figure 1. In this case, the catalogue of services must be based on
detailed analysis and characterization of EPU applications in terms of communication
requirements.
Process User
Service
User
User
Provider
User
Provider
Customers
Telecom Connectivity
User
Service
Provider
Public Telecom
Transmission Medium
Operator
Fibre Connectivity
Public Telecom
User / Provider
Relationship
Provider
2.2
Communication services in the EPU can be identified according to the applications that they
address. In particular, wherever the Service User entities and the telecom service providing
entity are tightly related, there is a one-to-one correspondence between applications and
communication service, resulting in an application-oriented definition of communication
services (e.g. SCADA or Protection communication services mean communication services
respecting the requirements of SCADA or Protection applications). The communication
Service Provider is assumed to be sufficiently familiar with the applications to apply the
necessary precautions in the delivery of the required service (i.e. implicit SLA). Whenever a
new application is introduced or the requirements of an application change, the user and
provider must seek a new common understanding of the service requirements.
On the other hand, where communication service is provided by an external or formally
separate entity, then the service provision contract (explicit SLA) defines the service attributes
according to the providers service catalogue (e.g. Platinum, Gold, Silver, etc.). The Utility
user must then decide upon the suitable SLA for his applications.
In this report, we have identified communication services by the applications that they serve.
Consequently, we define communicating applications related to the operation of the power
system in chapters 3, 4 and 5, then the constraints and the required qualities in chapter 6,
before relating applications and constraints in section 6.9 (figures 6.6 and 6.7) allowing SLA
specification whichever provisioning scheme is adopted (as characterized in chapter 8).
The following application-oriented service categories can be identified:
1. Operational Services These communication services enable the coordination and
exchange of information between the staff, devices and platforms directly involved in
operational applications and processes used to operate, control and protect the power
system and its constituents. The processes are necessary for the proper accomplishment
of the Utilitys primary mission and therefore their communication services are referred
to as mission-critical.
2. Operation Support Services Closely related to the Power system Operation, there
exists increasingly a group of applications related to the maintenance and support of the
Power System infrastructure. This includes voice and data applications for the field
maintenance staff connecting them with central offices, servers and data sources
allowing them to perform their tasks, as well as remote monitoring and surveillance
applications. These are collectively referred to as Operation Support Applications and
their telecom requirements as Operation Support Communication Services.
3. Security, Safety and Environmental Services A whole group of new applications
related to the security of utility staff, public safety, Utility site security, and
environmental security is emerging in many countries due to growing security and
environmental concerns and consequent regulatory constraints. These applications
which were previously considered as part of the operation support are increasingly
considered as a distinct class of applications with extensive communication
requirements and constraints.
4. Corporate Communication Services These communication services are related to
the administrative applications of the Power Utility as a Corporate Enterprise, covering
the administration and corporate needs of the Utility organization and its employees
(including those located in Operational sites).
5. Business and Market Communication Services The Power Utility needs to
exchange information with its external Market partners and its power customers. This
includes communications between power generators, distribution companies, national
and other country TSO, trading platforms and energy consumers. The required
communication services are referred to as Business and Market Communication
Services. Smart metering and Demand-side Management communications are part of
this class of communication services.
6. Commercial or U-Telco Communication Services The Power Utility or an
affiliated entity may provide commercial communication services as a source of
revenue to other Utilities, to institutional customers (e.g. government or community
offices), to telecom Service Providers, or to multi-site companies (and in certain cases
to individual customers). The service may cover subscriber premises access (DSL),
core communications, or both. Providing U-Telco communication services can be
assimilated to the service provision model of a public telecom operator.
Criticality of communication services in the Power Utility can be assessed through the
consequences of service loss and degradation. It is clear that a high degree of criticality can be
10
attributed to the operational services. However, it should be noted that the operational
applications are not the only critical processes in the Power Utility. Security and human safety
related applications present also a high level of criticality. The same can be said about
communications related to Utility Business and Market activities where the financial
consequences of a loss of communication can be tremendous. Corporate communications may
be tolerant to longer periods of programmed unavailability, in particular for maintenance
purposes. The loss of commercial communications has the same degree of criticality as that of
other public telecom operators resulting immediately in a loss of revenues, the non-respect of
contractual obligations if prolonged, and in the long run in a loss of customers.
The performance objectives and the Quality of Service are also different among these different
service types. Many operational services, such as Protection Relay applications, have
extremely severe time delay and communication integrity constraints, whereas the other
communication service types are mainly transactional with less severe time sensitivity. On the
other hand, business and market communication services implicate access beyond the
perimeter of the power company and may raise more severe security issues.
The first three groups of services described above may collectively be called operationrelated services (serving operation-related EPU applications). The present brochure focuses
on these services even if other services are often mentioned, in particular when their provision
interferes with (or influences) the way in which operation-related services are provisioned.
Administrative/Corporate
Communication Services
Operational
Communication Services
Business / Market
Security transactions
internet-oriented
Communication Services
Operation Support
Communication Services
Security & Safety
Service-Provider
oriented
Commercial / U-Telco
Communication Services
Communication Services
Industrial
Communications
11
EPU applications to these categories can vary to some extent and may evolve with
organizational changes. Some of the factors that influence these allocations are as follows:
Security policy The definition of separate security domains in the EPU and the
consequent allocation of applications to these different security domains can result in the
communication service allocation. This means that the applications which are part of a
same security domain shall exclusively use a same group of communication services.
Organization The organizational entity in charge of a group of applications may require
exclusive usage of a service or a same group of communication services.
Company strategy Grouping of communication services may depend upon the
companys strategy, for example to merge corporate and operation-related IT and
telecoms, or to merge corporate and market related applications communications
provision, etc.
Regulatory issues Regulation authorities may prevent operational applications to share
communication services with non-operational, or may impose full separation of the UTelco activities.
The grouping of different applications communication services strongly impacts the service
integration strategy of the company as described in Chapter 9 on Federating of Services.
12
3 OPERATIONAL APPLICATIONS
3.1 Protection Communication
Power system faults disrupt normal power flow by diverting current through a short-circuited
connection and collapsing power system voltage. Power system fault clearing requirements are
very important design and operational criteria for power systems.
Faults can cause damage and breakdown to power apparatus such as circuit breakers,
transformers and cables. The repair work or full replacement in case of destruction is very
costly and may take considerable time.
Faults can also cause severe operational disturbances resulting in collapse of power
delivery and blackout for regions, and, in severe cases, even for several countries. Heavy
reliance of modern society on electric power consuming devices for business activities,
safety, lighting, heating, communication and many other conveniences make severe
disturbances and blackouts unacceptable.
Transients due to faults in the power system can also adversely affect sources of
generation and customer loads.
Consequently, faults must be detected and isolated very quickly. Electric power system
generators, transformers, Busbars, and power lines are monitored by Protective Relays
designed to detect faults and operate isolating devices designed to interrupt damaging fault
current.
Protection performance requirements specify the balance between the conflicting goals of
dependability and security:
Dependability goals require maximum sensitivity and fast response time to detect and
clear all faults quickly with very low probability of a failure to trip.
Security goals require maximum selectivity and slow response time to minimize the
probability of spurious operation leading to an unwanted trip on a faultless circuit. Security
is an issue during fault conditions as well as during normal, faultless conditions.
13
When protection schemes detect a fault on the equipment or line they protect, they signal (or
trip) isolating devices, called circuit breakers, to open, in order to isolate the faulty segment
of the system and restore normal voltage and current flow in the power system.
When the protection scheme and circuit breakers operate properly, the fault is isolated within
the required fault-clearing time. Protection applied on extremely high voltage systems, where
fault-clearing times are most critical, typically detect faults and operate in about one to two
cycles (or even less than one cycle in certain cases). Circuit breakers operate in one to three
cycles. The combination of high-speed protection schemes and fast circuit breakers can
interrupt a fault in about two cycles, although more common fault-clearing times range from
three to six cycles.
Many protection applications require the real time transfer of electrical measurements, signals
and commands between electrical substations to enhance or to enable the trip/operate decision.
Protection systems for substation units (generators, busbars, transformers, etc.) can
normally meet the fault clearing requirements without using telecommunication. Telecom
services may be needed in this case, only to command a circuit breaker at a remote end if a
local circuit breaker has been economized (Direct tripping) or exists but fails to interrupt
fault-currents (Breaker Failure).
Protection schemes for HV lines generally need to exchange information with the
protection device at the far end of the line to meet fault clearing requirements.
Communication between the protection devices may be the basis for fault detection as in
the case of a Current Differential Protection, or needed to ensure that time response and
selectivity requirements are met, as in Permissive Distance Protections.
The Teleprotection function is part of the Protection system that adapts the signals and
measurements from the Protection to the telecommunication channel. It may be integrated into
the protective device, or the telecommunication access equipment, or it may constitute a standalone device.
If telecommunication fails, backup protection schemes still ensure that power system faults
will be cleared, but they may not be cleared within specified time requirements. Then the
probability of uncontrollable power swings and partial or complete system blackout increases
significantly.
Protection communications between substations are at present carried through transparent
dedicated telecom circuits ranging from analogue (e.g. PLC), to a sub-E1 or E1 circuits
multiplexed into an SDH transmission system, a dedicated wavelength or a dedicated fibre.
The communication requirements of different protection schemes have been described in detail
in [7]. Their evolutions, in particular their interfacing and transport over an Ethernet
connection, are currently being assessed in CIGRE JWG D2/B5-30 and shall be the subject of
a separate Technical Brochure.
Building additional generating stations or transmission lines is generally the other alternative to
reduce the probability of fault-induced blackouts but is significantly more costly than
reinforced protection schemes with adequate telecommunication services. This is the reason
why Protection Relaying applications can, on their own, justify the implementation of
14
15
The terminology used to describe these state comparison protection schemes is basically
defined according to the impedance zone monitored by the protection relay as presented
below.
CIGRE Terminology
Alternate Name
Zone Acceleration
Transmission time is the maximum time (Tac ) for the delivery of the command at the
remote end, after which it is considered as having failed to be delivered. This is a
constraint to the time performance of the communication service, not only in terms of
nominal value but as a guaranteed limit.
Dependability is the ability to deliver all issued commands at all times without any
statistical considerations. It is expressed as the Probability Pmcof missing commands
(issued commands not arriving to the remote device, arriving too late or with a time
duration shorter than a specified limit). This sets a very severe constraint on the
availability and error performance of the communication service, challenging such telecom
service concepts as errored seconds and degraded minutes being counted in the
available time of a communication service.
16
Very short lines and cables where the low impedance makes the adjustment of settings
difficult for the use of Distance Relay
Multi-terminal lines where the intermediate in-feeds modify the impedance seen by the
Distance Relays, implicating that the observed impedance is not only dependent on the
distance to the fault, but also on the in-feed from the remote terminals, making impossible
an accurate measure of the impedance.
Situations where only current transformers are installed at each end of the line (no voltage
transformers)
EHV transmission lines where series capacitors may create protection issues.
The transfer of analog samples between the ends of the protected line can be performed in
several ways, the most common, at present, being the use of digital communications. The
instantaneous current values at each end of the power line are sampled, converted to digital
data and transmitted towards the other terminals with a sample rate ranging from 12 to 60
samples per cycle. Although the communication interface is generally a standard ITU-T (or
EIA) interface, it should be noted that the time, integrity and availability constraints for these
services are far from the standard telecommunication practice. Direct optical fibre connection
between protection terminals or wavelength multiplexing of the optical protection signal can
also be used with an enhanced reliability where dedicated fibre or wavelength is available.
Current differential Protections are particularly time-sensitive as their operation is based upon
the comparison of current samples collected from a remote point with those measured locally
at the same instant of time in order to detect a fault. An error in sample timing and the delay
compensation mechanism, results in a differential current that increases the risk of unwanted
tripping.
Modern systems provide a global time stamping of samples through GPS-synchronization.
However, the older generation relays, still largely deployed, use the total round-trip transfer
time to calibrate the time difference between local and remote current samples, assuming that
the go and return times are strictly equal. This creates a great sensitivity of the system to any
time difference and therefore implicates the same routing for the two senses of
communication.
This is a strong constraint on the operation of the communication network and the mode of
resilience employed for the communication channel.
18
Power transmission and distribution networks SCADA generally differ in their requirements,
cost objectives and hence suitable communication solutions. The number of outstations and
their corresponding size, cost, volume of traffic, and geographical dispersion are very different
in the national transmission grid and in regional distribution networks. The time constraints
and the required level of availability, fault tolerance and data integrity are also different. As a
consequence, transmission grid SCADA communication is often implemented through a
broadband private network with point-to-point or multi-point with a small number of RTUs
(typically 2 or 3) per circuit, while in distribution networks (in particular for the MV level)
lower capacity solutions such as UHF Multiple Address Radio systems (MARS), spreadspectrum licence-free radio systems or procured services (GPRS, VSAT, etc.) prevail.
Still today, the widest employed communication mode for the substation RTU remains the
Asynchronous Serial link through an RS232 interface, polled by the central control platform.
The communication protocol associated to this mode has been standardized as IEC 60870-5101 (IEC101), although many other protocols are still in use in legacy systems. The great
advantage of Serial link SCADA is its conceptual simplicity. The major drawback to serial
communication for SCADA is indeed its lack of flexibility (e.g. for back-up control centre
connection) and cumbersome installation in particular at the Control Centre.
Packet switching has been applied to SCADA services since the late 80s, essentially to save
leased aggregated bandwidth on Control Centre links and to enhance system flexibility and
resilience. SCADA networks have been implemented over X25 packet switching, Frame Relay
systems and ATM in certain countries. However, worldwide popularity of SCADA
communications over packet networks has been due to IP communications.
SCADA RTU communications are migrating to TCP/IP based protocol IEC 60870-5-104,
generally called IEC104. The RTU communicates through an Ethernet LAN access interface at
10 or 100Mbps, although the bandwidth allocated to each RTU communication remains often
around 10kbps. Legacy RTU may be connected through a Terminal Server encapsulating Serial
data.
The use of TCP/IP enhances considerably the flexibility of the SCADA communication
system, facilitating the relocation of an RTU or the switch-over of RTU communications to a
back-up facility.
The migration process for large legacy SCADA networks from existing serial communications
to TCP/IP is a major concern in many Utilities. This process may be extended over many
years, and does not necessarily cover at the same time the replacement of the RTU, its
communication interface, the telecommunication network and the Control Centre Front-end
facilities. Moreover, new RTUs dispersed across the network may be TCP/IP while the existing
may remain serial linked, up to their programmed end-of-life.
19
These communications are assured through the Inter-Control Centre Protocol (ICCP)
standardized as IEC 60870-6 and Telecontrol Application Service Element (TASE-2) protocol,
although earlier protocols may still be in use in certain older systems.
ICCP uses an underlying transport-service, normally TCP/IP over Ethernet. The required
bandwidth for an ICCP link is generally around 2Mbps (E1) which can be provisioned over an
SDH network, although lower capacity links (64-128 kbps or even lower) have been used
where no fibre and SDH capacity is available.
The time constraint for an ICCP connection is of the order of hundreds of milliseconds which
rarely constitute a constraint in an IP/Ethernet network over a digital communication
infrastructure.
Security is the fundamental issue in implementing ICCP connections. An inadequately
protected ICCP connection may form an open door to the control of the nation-wide energy
network.
Post-incident analysis and static modelling applications are offline systems where
collected data is used to analyze the cause of an event or to adjust the behaviour model for
a system. Data can be collected continuously, daily or only on request. The communication
service can be a TCP/IP file transfer service with no real time constraint.
20
Visualization and Situational Awareness applications collect data from sites and display
them for human operator observation. These applications which constitute the great
majority of present day systems have time requirements which are those of a human
operator and must additionally present a level of sample loss unperceivable by the human
operator. In terms of communication service a non-acknowledge UDP/IP is an adequate
solution in this case whether through a dedicated network or a public provider.
Monitoring & Decision Support systems use collected data to produce analytical
information helping operators respond to grid events and to position the grid for improved
security and resilience. Stability diagrams and corresponding voltage collapse margins, as
well as different monitoring applications (Voltage & Frequency stability, Power
Oscillations, Line Temperature, etc.) are among these applications. Monitoring and
decision support applications have time constraints which are similar to power system
SCADA. This is achievable through UDP over a private IP network or a Service Provider
VPN through a carefully specified SLA.
Closed Loop Applications are those which incorporate collecting of data from the grid,
processing, automatic recognition of a pattern, and remedial action upon the grid. The
systems are used for emergency situation control and special protection applications as
described earlier in section 3.1.4. Closed loop synchrophasor applications are not yet
widely implemented and their critical real-time nature necessitates particular attention on
time control. Furthermore the decision to act automatically upon the network in real-time
means that the data set (from different locations and sample stack from each point) must
be complete, that is to say almost lossless. Providing lossless data across a telecom
network generally implies error recovery which is constrained by time limitations.
PMU operation is specified by IEEE C37.118 which defines phasor construction using the
GPS-satellite timing signal, as well as the phasors data format. The exact data volume
associated with the transmission of a data packet from a PMU varies depending on the
incorporated parameters and the way each of them is coded (i.e. floating point or not, etc.) but
can be assumed to be around 80 100 octets. This data volume is to be transferred across the
network at a rate which is governed by the sampling frequency of the PMU. The sampling
frequency is expressed as a number of (or a fraction of) AC cycles. It is often 25 (or 30)
samples per second corresponding to one sample every two cycles to 100-120 samples per
second corresponding to two samples every cycle (Nyquist Rate). This latter rate allows the
processing of the signal corresponding to the AC fundamental wave.
The required communication throughput is then somewhere in the range of 16 100 kbps for a
50Hz power system although PDC links may require few hundred kbps upto 1Mbps or more.
21
Operation
time
Latency
Data
Availability
Telecom
Service
Human
operator
Minutes
seconds
Sample loss
not
perceivable
UDP/IP on
dedicated
network or
public provider
seconds
Sample
loss
tolerated
UDP over
private IP
network or
VPN with
well specified
SLA
Off-line
minutes
N/A
Non-critical
TCP/IP File
Transfer
seconds
or below
10-100ms
Very
Critical
Ethernet
VLAN with
fast recovery
Direct (hotline) telephone lines from the Control Centre to all major operational sites
control rooms
Switched telephone service through PBX and a closed numbering scheme
Additional redundancy and operation in situations of site isolation.
Interconnection with the public telephone network.
Voice and data traffic.
Mobile radio voice facilities for access to operational staff who visit facilities. (Mobile
workforce communications is treated in a separate section).
22
The Operational telephone service is today evolving into IP telephony and becomes
increasingly an Ethernet transported data service with particular time and bandwidth
requirements.
Some of the specific features of operational voice service are as follows:
Access restriction Use of the operational voice service is confined to operational staff
and not accessible to unauthorized users.
High availability Voice service access for the operational staff and in particular the
access of the Control Centre to the network substations and generating plants is essential
and must present a very high availability through adequate route resilience and equipment
duplication.
Resilience/fault tolerance The voice service must remain available even in the event of
network faults, node failure and route unavailability. In particular, a star-structured
network in which the failure of a single node may jeopardize the system is not acceptable.
Multiple homing (at least dual homing) of secondary sites and a mesh interconnection of
the main nodes is generally required to achieve the required level of fault tolerance.
Transfer to Backup Control Centre In emergency situations leading to the migration of
power system control to a Back-up Control Centre, the telephone network must rapidly
adapt in order to transfer the telephone calls for the Control staff to the Back-up facility.
This transfer must be possible even if the communication equipment in the main control
centre is no longer operational (e.g. fire, flood or power breakdown).
Very rapid call connection The call establishment time must be in line with the
operational emergency situations in which the voice communication may become
necessary. In particular, the structure of the telephone network (number of cascaded
transits) and the employed signaling scheme may greatly influence the call connection
speed.
Priority functions These functions allow critical communications to be established even
when all voice network resources are occupied. This can be performed through Forced
Releasing of facilities which are used by less critical communications, or by reserving the
usage of certain facilities (e.g. communication channels) for priority calls only. Similarly,
critical calls can Beak-in into an established communication of a busy called party.
Priority status can be attributed permanently to a given user line (i.e. Control Operator), or
obtained dynamically through a code for a given communication.
Caller identification and Call Queuing Control centre operators need to identify
automatically the source of incoming calls and to establish queues of in progress
communications in order to interact with many sites, in particular at times of power system
emergencies. In progress and queuing calls must be accessible and transferable between
different Control centre operator positions.
Mobile voice Control centre facilities and large power plants telephone systems must
have the capability to connect mobile voice terminals to fixed telephone extensions.
Depending on the implemented mobile radio network, these connections may require the
existence of PTT (Push to Talk) facilities and associated conversion of Half-Duplex to
Full-Duplex voice communications.
Ability to pre-select conference calls The voice system must present the capability to
establish pre-configured conference calls, in particular between operational staff in the
control centre, in multiple substations and maintenance staff.
23
Call Recording Control centre voice facilities include voice recorders which constantly
record all communications of the operators which will be archived periodically. These call
recordings are essential in order to establish the sequence of events and instructions given
by the Control Operators in emergency situations.
Appendix 1 presents some examples of IP telephony usage in Utility Operational Voice
systems.
24
enable, in certain cases, other operational and monitoring applications in particular in the
distribution network where the number of communication nodes is extremely high and the cost
that can be attributed to the access for each node is very small. Distribution SCADA for the
secondary network (e.g. 33/11 kV) and monitoring of MV transformers are typical examples of
operational applications profiting from the deployment of advanced metering infrastructures.
Advanced Metering Infrastructure (AMI) which covers the overall system composed of
consumer data acquisition and collection as well as bidirectional communication with the
electricity provider, is a further step from simple remote reading of customer meters. Several
AMI projects in different countries are assessing new technologies for dedicated network
coverage beyond the EPU sites perimeter (MV/LV PLC, meshed networking of packet radio
systems, etc.) and some telecom/internet operators are working towards new service offers to
occupy this promising market segment.
25
Networked office applications (e.g. mail and calendar systems, file transfer, remote
database access, intranet),
Work-order and ERP solutions (e.g. project control and time registration),
Video conferencing facilities and electronic white-board in branch offices and control
centres and PC-based video-streaming in the dispersed sites.
These applications require the secure extension of the corporate enterprise applications from
the branch office to the operational sites, while remaining fully isolated from the operational
applications. IT-support may be effectively administered from a corporate central site.
26
Protect substation assets and premises (temperature monitoring, substation fire detection,
etc.)
Protect the environment from industrial risks and hazards related to the substation assets,
e.g. chemical pollution detection, etc. These are part of the security and safety applications
as described in section 5.1.3.
Energy Quality Monitoring applications transfer data related to electrical power parameters at
commercial interfaces or boundaries where energy is transferred between different power
actors and may be sanctioned through financial penalties according to contractual quality of
service criteria.
27
29
allows differentiated accessibility in time and across locations for different classes of staff
(operational staff, service contractors, maintenance, etc.). These applications require fast and
reliable data communications for authentication and access registration.
30
1. Fixed Telephone service - The simplest communication service for isolated workers is
indeed the accessibility of a telephone at site, provided that the person is able to reach
the telephone in an emergency. This service must be available not only to the
operational staff but also to external parties contracted for specific tasks in the Utility
premises. The telephone system must provide an emergency number accessible to all
categories of users.
2. Radio communications Different categories of two-way radio systems are in use in
different Utilities for traveling staff and for employees working in large sites such as
power plants. Mobile workforce communications are covered in section 4.6 and may
be used also for assuring the safety during the trip to site depending upon coverage
constraints. Usage of private mobile radio as a means of assuring workers safety needs
careful location of base stations and identification of shadow areas as well as an
adequate procedure to assure the supply of charged batteries. Public mobile phone is
used in some Utilities as a cost-saving alternative, but may present serious drawbacks
with service coverage and service continuity particularly when the Power Utility
employees are intervening to re-establish electrical power in a region due to the
extremely short power autonomy of the public base stations.
3. Satellite Communication systems Satellite phones overcome the problems of public
mobile phone in poorly covered areas and the dependence on local power supply of the
base stations. Satellite systems also allow implementing Location Beacon Systems
determining the location of the employee through GPS and signaling this location to an
operational base permanently. Care should be taken however as their operation is
affected by damage to aerials, failure of vehicle power supplies, or vehicle damage.
4. Personal security systems These portable wireless transmitters are permanently in
communication with a central receiver and may include a non-movement sensor that
will automatically activate an alarm transmission if the transceiver has not moved
within a certain time
5. Emergency location beacons When working in particularly remote areas, emergency
location beacon systems which are automatically activated in emergency situations
may be used. These systems do not depend upon the vehicle power supply and do not
risk damage as satellite communication systems.
31
operates the Discharge Alert devices. Sirens and information display boards notify the public
that water will be discharged, preventing water-related accidents.
Kyushu Electric Power also informs its customers through a specific website about the location
and time of lightning strikes in Kyushu region. The lightning strikes forecast is a result of
collected field values measured by sensors across the operating region, this information is than
sent and processed centrally, and is used as the criteria of judgment for the generation of
lightning, thunder, and launching, making possible protection against lightning and safety
control.
Figure 5.1 Kyushu Electric Power Company Dam Management System [14]
32
they are picked up by conventional SCADA alarms. (e.g. amount of debris building up at trash
racks or in the vicinity of critical flow monitoring weirs). IP video cameras are also used as
confirmation of SCADA alarms in flood control situations and hence provide extra information
for difficult operational decisions or where associated events such as fallen trees, rain or snow
impedes access to the site.
Field staff can also use such a network to access water level and gate position data to enable
onsite calibration of transducers without the need to tie up office staff in relaying values back
to the field staff. They can also send photographs, etc directly back to specialist engineering
staff from remote locations thereby sometimes saving travel time to remote locations for
specialist staff.
In all of the above, security policies and technologies have to be rigorously applied due to the
openness of Wi-Fi networks. Typically technologies such as WPA2, strong authentication,
IPSec tunnels, firewalls, etc are mandatory because any Wi-Fi network has to be regarded as
completely untrusted in effect the same as an internet connection.
Many Hydro Generators also have UHF/VHF radio networks that support their operations by
transporting data regarding stream flows as well as snow and meteorological parameters. This
data is used in applications to optimize the Hydro Generators operations and also to account
for the use of water as part of their water license conditions.
Lastly Hydro Generators also have protection systems which are used for the reduction in
impact caused by the failure of a hydraulic structure, particularly for the detection of a burst or
significantly leaking penstock. These systems are usually based on differential flow
measurement (e.g. at either end of the penstock) with automatic control back to guard gates.
Similar to electrical protection systems, reliability is critical for this application and hence
duplication of systems and communications facilities is inevitably employed. However a much
longer transmission delay (e.g. seconds) is acceptable for these protection systems due to the
relatively slow speed of operation of hydraulic gates.
Hydro Generator hydraulic assets are often located in remote areas outside the coverage of
telecommunications providers. However in some cases the main hydraulic structure may be
able to be connected to a telecommunications Service Provider and used as a local hub to
aggregate the communications required to support the above applications.
33
34
Power Plant
Control
Centre
Provider
Nodal
Switch
substation
Service Provider
substation
400 kV Line
400 kV Line
substation
Dedicated Network
Fig 6.1 Topology of Telco and dedicated network and their correlation to the Power system
35
Communication channels used for protection relay applications may need absolute time
latency as low as 5 to 10 ms depending upon the protection scheme and the power
systems fault clearance time (around 80 - 100ms depending on the voltage level).
SCADA system overall performance can be degraded by a high time latency or even
made completely inoperable through RTU communication time-outs.
Voice communication can be seriously degraded by high time latency (more than
150ms) through echo.
Absolute time latency problems may be avoided through an appropriate design based on
Time Division Multiplexed circuits (e.g. SDH), constrained usage of switching and routing
and lack of traffic queuing for critical applications. For delay critical services delivered
over IP, pre-established static routes may be employed to ensure guaranteed end-to-end
performance.
It should be noted that the real-time requirements of SCADA RTU communications are
generally in the range of seconds, as compared to order of magnitude smaller transmission
times across a thoroughly designed SCADA Ethernet/IP infrastructure. The main issue here
is therefore the number of intermediate nodes in the routing of SCADA information as well
as the time for any encapsulation and concatenation.
b. Time Predictability (delay variations) and Timing Jitter
Time predictability determines the delay variation of a communication channel. It defines
the capability to predict the time delay of the communication network, independently from
the traffic load from other services being carried across the network, and whatever being
36
the networks state. Time predictability assigns a probability distribution around a nominal
time delay value and therefore a maximum acceptable delay.
Protection applications (in particular current differential schemes) and voice services are
particularly sensitive to delay variations.
Time predictability is achieved by avoiding traffic queues which generate variable service
times and by imposing constrained routing (e.g. maximum number of hops, pre-determined
back-up route, or no route resilience). In case of an SDH system, ring protection must be
avoided or carefully analyzed.
c.
d. Restoration time
Transporting operational traffic imposes a time limit on service restoration following a
network fault condition. Some operational services require Hitless Switching (no loss of
service during the switchover from a normal configuration to a backup state).
Restoration time depends upon the employed communication technology and the
topological complexity of the network:
End-to-end alternate route switch-over for each critical service is very fast but does not
assure a high level of resilience,
Ring protection across an SDH network (e.g. SNCP protection) can restore in less than
50ms.
The restoration mechanism of Ethernet, the Spanning Tree Protocol (STP) has a
convergence time which depends upon the complexity of the Ethernet mesh. The
restoration time may be too long for SCADA communications. More elaborate options
such as Rapid Spanning Tree (RSTP) reduce this time.
37
Multi-Protocol Label Switching (MPLS) enables the network to restore service using
pre-established alternate routes for each Virtual Private Network (VPN) and can in this
way reduce considerably the high restoration time.
The time behavior of the communication network is determined by the following aspects:
1. TDM versus Packet network The migration from conventional TDM (Time
Division Multiplexing) networks towards Ethernet and IP increases considerably the
networks bandwidth efficiency (avoid idle bandwidth), flexibility (network interfaces
and routing) and resilience. However, it is also a major source of concern for the
control of time behavior. Assembling data packets before transmission and store-andforward of the packet at each intermediate node causes additional buffering delay
which increases with the packet size and with the number of transit nodes. Furthermore,
dynamic data routing needed for network resilience, gives rise to delay variation and
lack of time predictability. Moreover, the fully deterministic behaviour of the TDM is
being replaced by the statistical behaviour of the packet-switched multi-service network
where queuing and traffic profiles determine the overall delay.
2. Multi-service Integration Bandwidth efficiency in packet networks is achieved
through integrating multiple traffic streams into a same packet network. Priority
queuing mechanisms are employed to assure better time performance for more critical
services.
This implicates that one or multiple traffic queues are established at each
routing/switching node. A queue-based store and forward communication system
provides essentially a Best Effort service with statistical time characteristics. This
issue is often masked through over-dimensioning of the network and order of
magnitude smaller time requirements of the applications.
3. Network Resilience versus Fixed Routing Network resilience ensures the continuity
of service in presence of network faults, but at the same time renders indeterminate the
routing of communications. Time predictability is generally sacrificed by improved
resilience. Traffic streams which are sensitive to delay variations must generally be
treated separately without service restoration mechanisms described previously under
restoration time.
4. L1/L2/L3 Partitioning and Topological Structuring In order to provide adequate
time performance to critical services while maintaining bandwidth efficiency,
flexibility, cost and resilience, it is necessary to design the network with an adequate
level of information forwarding at physical, link and network layers.
Direct or TDM connections for best time performance but at low bandwidth
efficiency, flexibility and resilience.
Ethernet Switching with Virtual Networking (VLAN) and priority assignment for
fast transfer of information packets (frames)
IP Routing for maximum resilience and multi-servicing
Different network topologies can in this way be obtained for different services over a
same telecom infrastructure leading to different numbers of intermediate nodes at each
layer.
38
Mean.Time.To. Re store.Service
Mean.Time.Between.ServiceOutages
Service Downtime
5.25 min/year
(~5 Hours/57 years)
52.5 min/year
(~5 Hours/ 5.7 years)
525 min/year
(~5 Hours/ 0.57 years)
Example of Service
Protection Communication
SCADA, Operational Voice
Data Service
39
Reduce the occurrence of network infrastructure faults. This can be achieved through more
reliable and fault tolerant network devices and infrastructure resources.
Reduce channel and timing impairments in the network. This can be done through suitable
transmission link design, synchronization planning, traffic engineering, performance
planning, and installation practice.
Reduce the impact of network component faults and channel impairments on the delivered
service. This can be achieved through a more resilient network design and duplicated
access.
Reduce the duration of service down-time following a network fault. This can be achieved
through faster detection and localization of network faults, faster identification of impacted
services and faster restoration of service through network maintenance. Reducing downtime is therefore dependant upon the service and infrastructure management and
maintenance processes and organization, well-trained staff, well-dimensioned stocks of
spares and adapted monitoring tools discussed in more detail in later sections.
40
Service Restoration Time This point has already been discussed under Time
Constraints section. Different levels of network resilience operate at different time scales
which may be compatible or not with different applications maximum acceptable outage
duration. Typically, what is commonly called Hitless Switching signifies that an
established connection for an application remains unaffected by the disruption time due to
the switching of the communication path or resource.
Routing Control Designing resilience into the network generally signifies injecting a
degree of uncertainty into the routing of information and the resources which are used for
delivering the service. This in turn, impacts the absolute time latency of the network and
generates delay variation. In the most time-sensitive applications (e.g. Protection Relay
communications), resilience is restricted to the highest level, that is to say to the
application itself (main/backup end-to-end routes). The two communication channels for
these applications must employ no common path, no common node and no common
equipment, so that no single fault disrupts the application. This requirement generally
requires a total control of the channels routing.
Underlying infrastructure Where the telecom Service Provider employs lower layer
connectivity service delivered by another infrastructure provider to build the network, it
becomes difficult or impossible to guarantee the independence of main and backup routes.
41
In particular, when a public telecom operator and managed services are involved, tracking
the full routing of connections can become extremely difficult.
Indeterminate Routing
Controlled Routing
?
N1
N1
N1
B1
B1
B1
N1
B1
B2
B2
B2
N2
N2
N2
B2
N2
Fig 6.2 Double routing and crossed routing for SCADA and Protection communications [16]
42
User
Access
Protected HMI,
Physical Protection
Network
User
Access
Network Node
Telecom
Management
System (TMS)
Protected Access,
User Authentication
Access Logging
Figure 6.3 Security risk mitigation measures for the telecom Service Provider
43
44
observed during ten consecutive seconds. These latter seconds count as available time. The
counting of degraded minutes is carried out only when the connection is available (excludes
unavailable periods).
Despite important technological changes in telecommunications, the error performance
objective and its related definitions are still widely used in power utility networks, in particular
for planning and testing 2Mbps connectivity through SDH infrastructure and the primary
access multiplexing systems. However, it can be fully inadequate in some situations.
For critical applications such as Protection Relay communication, available time with
degraded performance is not a reasonable definition. A system which presents one SES every
ten seconds (or even 1 SES every two seconds!) cannot be considered as available with
degraded performance.
Total Time
Unavailable Time
Available Time
ES
SES
DM
Data Integrity
Performance
Degraded
Performance
Availability
Performance
45
46
It should be noted that adopting managed services through a public provider does not remove
the expenditure and effort associated to these aspects because access equipment must be
installed at Utility sites.
Lastly, the impact of Earth Potential Rise (EPR) during a station earth fault needs to be taken
into account in connecting a new telecommunications service to a HV substation or power
station. As an example, an insulation breakdown of a 330kV asset to earth with a typical fault
current of 20kA may cause the station earth mat to rise up to 8kV or more above remote earth
potential. The exact figure depends on many factors including actual fault current, soil
resistivity, earth grid impedance, etc. The way the earth mat, fences and external connections
were initially designed and interconnected would have ensured safety of people on site and
remote to site. However subsequently adding a new telecommunications physical connection
without proper understanding of EPR could cause a very dangerous situation to occur to staff at
the station or remote from the station due to the difference in earth potential that exists during
the earth fault. This is easily solved by avoiding connecting copper communications cables to
HV stations by using optical fibre or radio solutions. If there is no other economical solution
other than connection of a HV zone to a telecommunications Service Provider by a copper
cable, then it is essential for safety reasons that appropriate isolation devices are used.
47
fulfillment of operational constraints as described previously and the Service Provider may not
be prepared to review his entire networks operation mode and operational process to meet one
customers SLA requirements. In this case, assessing the most appropriate SLA of the provider
against the operational constraints of the EPU applications allows the estimation of the gap and
the risk analysis associated to the potential impact of this gap. The following checklist has been
prepared to serve utilities for specifying or assessing SLAs in the EPU operational context.
Figure 6.5 - SLA checklist for EPU procuring telecom connectivity services
SLA Parameter
Description / Comments
Interface type
As required by the application, (e.g. Optical Ethernet, G703, RS232).
Choosing a physical interface such as Ethernet that can be scaled
remotely results in easier expansion of services as the need grows.
Bandwidth and
Guaranteed minimum and Peak bandwidth available to the service and
throughput
degree of flexibility
% number of
It is important to set a policy at the edge of the EPU network to avoid
packets allowed
exceeding the allowable limits, otherwise the policy on entry to the
per service for
Service Providers network with either drop the packets or remark them
each procured
to the least priority service, leading to poor service performance due to
Quality of Service oversubscription of the service by the EPU itself. Conversely you want
(QOS) level.
to see the Service Provider to apply limiting policies on entry to their
network in order to protect the EPU service from contention due to
oversubscribed services from other Service Provider customers.
Time Latency
For packet based services, these need to be defined for each class of
(end-to-end delay) service. Voice services for example will be processed via separate low
latency queues.
Delay Variation
For packet based services these need to be defined for each QOS level.
(Jitter)
As is the case with most data service parameters these are usually
expressed by the Service Provider as monthly averages. Consider how
to manage the situation of high peaks that dont cause the monthly
averages to exceed the Service Provider specifications. (High peak
jitters can cause voice degradation or network convergence problems
and still not hit the monthly average parameters.)
Go-Return delay
For certain protection relay communications. Asymmetrical delay will
difference
cause certain protection schemes to fail.
Service Restore
The time required for automatic reconfiguration mechanisms to act upon
Time on network
the network and hence to restore service (e.g. Spanning Tree Protocol,
change
SDH Ring Protection restore time, etc.)
Availability
Distribution, frequency, duration, and timing of service failures.
Integrity and
Specified for each procured class of service.
Packet Loss
Power Faults
Critical services not impacted by power system disturbances.
Correlation
Precautions for not losing service during disturbance.
48
Figure 6.5 (continued) - SLA Checklist for EPU Procuring Telecom Connectivity Services
SLA Parameter
Description / Comments
Resilience and
Control of the provider on the routes taken by services in normal time
Routing Control
and on anomalies (determines the capability of establishing duplicated
communications without common point of failure).
The Service Provider and Service User need to agree on the routing
protocol between their networks, and to set various metrics that impact
on the resilience of the interconnected networks.
Power Autonomy
The time duration for which the service can be delivered in case of A.C.
power outage
Maximum Time to Service Providers ability to respond to service failures and carry out the
Restore Service
necessary repairs within the maximum specified time. Different times
will be defined for urban, regional, rural and remote locations
depending on the location of Service Provider maintenance staff.
Dual Route
Ability to guarantee that specified connections between two points
Independence
never use a same equipment, cable segment, power supply, node or
cable conduit.
Physical
Specify the level of redundancy required for example at network level,
redundancy check equipment level, or at specific locations.
Service Isolation & Isolation between internal and external traffic, as well as between
Security
different internal services. Measures deployed by the provider to protect
against the risks of interfering third parties (confidentiality, denial of
service, integrity of information). An EPU will usually have to regard a
Service Provider as untrusted and employ security techniques such as
encryption.
Access
Most EPUs have special rules for site access for security and safety
Arrangements
reasons. These need to be communicated to the Service Provider and
factored into his support of the service.
Qualified/
Ensure that the Service Provider has sufficient depth in its workforce
Certified/ Insured
with the right number of personnel in the right locations to ensure that
Workforce
response time guarantees are realistic. Ensure appropriate insurances are
in place to cover accidents by the Service Provider workforce when
attending an EPU site.
Performance
Meaningful and comprehensible information to be provided in a timely
Reports / Fault
fashion. An EPU should consider implementing their own monitoring
Notification
tools to ensure the performance of the Services is appropriate. This is
especially important for packet based services using different QOS
levels.
49
Figure 6.5 (continued) - SLA Checklist for EPU Procuring Telecom Connectivity Services
SLA Parameter
Description / Comments
Penalties and
While penalties may not compensate for loss of critical services, they do
Liability
focus a Service Providers attention on the need to accurately monitor
the SLA guarantees. Usually a Service Provider will exclude
responsibility for contingent liabilities and cap their overall liability to a
percentage rebate of fees paid. It is worth considering inserting a
termination clause in the SLA that allows termination of the service for
a sustained poor performance. At least this enables an EPU to engage a
new Service Provider and potentially fix the problem using a different
service if the current Service Provider continues not to remedy the
problem.
Other Legal
Depending on the structure of the contracts (e.g. if there is a separate
Conditions
service provision contract or not) there may be other legal conditions
that may need to be covered off in the SLA including details for;
confidentiality between the parties, intellectual property, compliance
with all applicable laws (and governing law where the service is
provided cross jurisdiction), acceptance and payment, Force Majeure
and Termination of contract provisions to name the most common ones.
Figure 6.7 titled Typical Communication Service Requirements for EPU Applications
provides a cross reference of typical service requirements for the EPU applications discussed in
Sections 3, 4 and 5. The reader should use Figure 6.6 to provide the meaning behind the
numbers 1, 2, 3 and 4 in Figure 6.7.
50
Time Latency
1 5 sec
Human operator
0.1 1 sec
Few cycles
(20 - 100 msec)
4
Highest Severity
Beyond the grid,
(Energy farms,
customer sites, etc.)
Fraction of a cycle
( 5 20msec)
Time Predictability,
Delay Variation
Seconds
0.1 1 sec
10 100 msec
1 10 msec
May be through
different telecom
media
Few Hours
Uncontrolled over
the same telecom
system
Few Minutes
Controlled Routing
Identical path,
200s
Few Seconds
99%
Service may be lost
in the event of
anomalies
Public
Lost data recovered
(Acknowledge &
Retransmission)
99.9%
Survives one
module or one link
failure
Un-trusted
Not so sensitive to
recurrent data error
& loss
99.99%
In Confidence
99.999%
Survives major
system faults &
disasters
Protected
Sustainability,
Life-cycle Mgt.
Continuous upgrade
(type IT)
Yearly upgrade
Multi-annual upgrade
(Planned migration)
Constant over
application asset
lifetime
Environmental Class
Customer Premises
Admin Building
Control Centre
Power plant /
Substation (Control
& Relay Rooms)
Differential Delay
(go-return path)
Restoration Time
Availability
Service Survivability
& Resilience
Security Domain
Service Integrity
3
High Severity
51
Grid corridors
Switch-yard
Hydraulic Structure
Time Latency
Delay Variation
Differential Delay
Restoration Time
Availability
Survivability
Security Domain
Service Integrity
Life-cycle Mgt.
Environment Class
3-4
2-3
2-3
Operational Telephony
SCADA RTU
Generation Control Signaling
2
2
2
2
2
2
2
2
1
2
2
1
3
3
3
2
2
2
3
2
2
4
4
4
2
3
4
3
3
2
2
Remote Operator
Synchrophasor visualization &
monitoring (WAMS)
Settlement and Reconciliation metering
Smart Metering
2
4
1
1
1
1
1
1
2
1
1
1
1
1
3
3
3
1
3
3
2
1
Operational Applications
Applications
Protection Communications
Current Differential
Protection Communications
State Comparison (command)
Requirements
Coverage
52
Time Latency
Delay Variation
Differential Delay
Restoration Time
Availability
Survivability
Security Domain
Service Integrity
Life-cycle Mgt.
Environment class
1-2
1-2
2-3
1-2
1-2
2
2
1
1
1
1
2
1
2
1
1
1
2
1
4
4
3
1
3-4
1-2
2
2
On-line Documentation
3-4
Condition Monitoring
Video-surveillance of sites
2
2
1
1
1
1
1
1
1
1
1
1
1
3
3
3-4
1-2
2
3-4
3
2
4
2
2
2
3
4
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
3-4
3-4
3-4
3-4
3-4
3-4
4
1
1
1
1
1
1
1
3
3
2-3
3
3
3
1-2
2
2
3-4
3-4
4
4
2
Applications
Mobile Workforce Communications
Requirements
Coverage
Operation Support
Figure 6.7 (continued) - Typical Communication Service Requirements for EPU Applications
53
54
The decision making process for the risks treatment relies on information about the threats and
vulnerabilities that contribute to the likelihood of the risk occurring and the impact of its
occurrence, compared with the cost of mitigating the risk and the risk appetite of the EPU. A
considerable part of threats and vulnerabilities faced by EPUs are disasters, some common
include the following:
Making the risk assessment, you will find out that telecommunications assets will have
different risks associated with them, and a correlation analysis of different risks is needed.
Some risks will impact many of the assets of a company, such as the risk of a massive fire
destroying a main building and everything on it, or an earthquake destroying a large amount of
power lines, while in other cases, like a flood in a data center, will only affect a group of assets.
Based on the generic risk management process model, a EPUs risk management framework
was developed as shown in Figure 7.1 [22].
Figure 7.1 EPU Risk Management and Risk Assessment Model [22]
RA: Risk Analysis, RM: Risk Management
55
The four layers in Figure 7.1 illustrate the various hierarchical levels of the operations part of a
typical EPU which operates power plants and/or electricity networks. Typically each level will
have authority over one or more entities in the level below. Therefore, the single corporate
entity will have authority for most business units, and each business unit will have
responsibility for one or more power plant and/or electricity network, etc.
As each level has authority over the entities in the level below, it will set certain business
objectives for them, and monitor the level of achievement of these business objectives. We can
use this same concept to elaborate how risk is assessed and managed within the corporate
structure of EPU operations.
Each level of the organization has a different set of objectives, and is therefore exposed to
different sets of risks. However, as there are dependencies between each level in terms of
objectives, there are also dependencies between each level in terms of risk. In order to take a
holistic approach to risk management, it is important that these dependencies are recognized,
and linked unambiguously through the use of a common framework or language for the
identification and management of risks.
At each level within the organization, Risk Assessment activities should take place, and this
should result in risks being quantified and Risk Treatment actions being taken to bring certain
identified risks down to acceptable levels (where they are not already at or below acceptable
levels).
Probability of disasters;
Business reliance on telecommunications;
Growing corporate and social responsibility;
Standardization movements.
As a part of the BCP, a good DRP for telecommunications is needed to ensure an effective
response to a disaster that affects telecommunications services and minimize the effect on the
business. The major goals of the DRP are:
56
To achieve these goals, the most important factors that it shall take into account are:
Communication
o Personnel - Notify all key personnel for a certain problem and assign them tasks
focused toward the recovery plan;
o Customers - Notifying clients about problems minimizes panic.
Tools: Be sure that the plan includes the identification and access to all the tools needed
for the recovery, such as manuals, procedures, applications, devices, privileges, etc.
Backups: Backups should be stored in separate locations. If backup resources are taken
offsite, these need to be recalled. If you are using remote backup services, a network
connection to the remote backup location (or the Internet) will be required;
Facilities: Having backup sites (hot or cold) and mobile recovery facilities are also good
options;
Prepare your employees: during a disaster, employees are required to work longer, more
stressful hours, and a support system should be in place to alleviate some of the stress.
Prepare them ahead of time to ensure that work runs smoothly.
Testing the plan: provisions, directions, frequency for testing the plan should be stipulated
After identifying the potential impacts of disaster and to understand the risks and construct the
BCP plan itself, in order to realize business continuity, the BCP must be not only established
but also continually updated and maintained in a Plan, Do, Check and Act basis, to
ensure that it remains appropriate to the needs of the EPU in terms of covering the measures
and action plans to meet the Recovery Time Objective.
Select a safety location specially in terms of natural disasters risk that can be easily
accessed;
Be sure that all needed data will be present at the backup facilities in case of disaster,
through storage/database synchronous replication for the most critical information or
through tape restore for the less critical;
57
Include in the backup facilities all the tools and applications that allow visualization
capabilities that ensure that operating personnel have situational awareness of the BES,
and all those needed for the minimum business/corporate activities;
Assure all data and voice communications needed for the regular service operation,
including voice and control communications to the critical substations and power plants,
and to communicate outside the organization including Internet access;
Include reliable power sources such access to redundant distribution power lines, diesel
generators, diesel refill contracts, etc.;
Be sure that all the physical and cyber security requirements applied to the main facilities
and control centers are also guaranteed;
Implement an operating process for keeping the backup functionality consistent with the
primary control center.
Do not forget to assure food and medical first aid;
As all of the components of the BCP, the backup facility availability and functionality must be
tested in a regular basis. However, if cost effective, his usage as a hot site is recommended,
for example in a distributed processing architecture or under hot/standby or metro-cluster
architectures. Using backup facilities as a hot site can be very attractive in terms of
permanent updating and training, and providing easy redundancy even in non disaster events.
58
Some of the most common redundancy measures adopted and combined building high-reliable
and high-available telecommunications networks are:
Duplicated equipments for the same purpose, sometimes form different vendors;
Equipments with redundant architectures in terms of power supply, CPU, service cards,
access ports, etc.
Physical independent communication media (e.g. optical fiber and radio);
Different communication technologies (e.g. SDH and PLC);
Alternative/Mesh network routes;
Distributed processing systems;
Third party telecommunication services;
Out of band management systems.
The entities at various levels shall formulate contingent plans for different disasters to ensure
the avoidance of interruption of power lines or fast recovery of the interrupted power lines and
guarantee the safety and stability of power network.
b. Prioritize
Knowing which services are more important for the business continuity, for example
dispatching telephone or SCADA, and designing the network and recovery plans based on such
priorities, will help to get a fast and more successful recovering process.
c. Strengthen the integration, storage and sharing of emergency response resources
Emergency response resources include internal resources of the entities, power industry, public
communication network, various domestic trade organizations, public security services, and
relevant international organizations.
Integrate all resources available for use and improve the efficiency of interlinking between
public network and other special communication systems. Given public network and special
power communication networks are highly complementary in terms of the laying mode,
vigorous efforts should be made to cooperate with public network operators and other special
network operators to study contingent plans and sign relevant agreements with them to make
public network serve as one of ways for power communication emergency response system.
Energetically endeavour, in conjunction with public network operators, back upping, protecting
each other, supplying power lines each other to establish one independently functioning
hardware platform with sufficient competition and effective integration in case of emergency
as well as intercommunication and interlinking and set up emergency response joint-action
mechanism for resource sharing.
Intensify efforts in the work on the establishment and storage of emergency response materials
and devices to deal with disasters. List important materials, equipment name, model, storage
places and telephone are for contact necessary for emergency response in contingent plan and
relevant management rules.
d. Emergency response talent team
Optimize maintenance system and set up maintenance talent team for fast response to reduce
business interruption time.
Foster technically sophisticated maintenance team and launch enhanced training on technology
and process among maintenance personnel to mitigate operation accidents and reduce fault
location time.
60
Enhance OPGW or ADSS cables capacity of resisting disasters in terms of design standard,
laying mode and coverage density on the basis of rational technical and economic
comparisons.
Multiple communication modes are available. Besides the existing OPGW, ADSS and carrier
wave, underground optical cable, microwave and satellite communication are also added to
enable three-dimensional and diversified communication network.
Conduct scientific evaluation and rationally select location, improve the quality of construction
of communication facilities and tighten its standard and enhance the capacity of resisting
natural disasters; arrange equipment room, base station, transmission route in a unified and
rational manner to ensure multi-route and multi-mode feature of communication network and
avoid damage.
Reduce the length of transmission path and the number of transmission nodes and replace
back-to-back switching with high-capacity equipment.
f.
Enhance networks protection capacity to mitigate the impact from interruption of optical fiber
or node failure, and replace large-loop network deployment by small-loop-plus-small-loop
network deployment mode to mitigate the impact from interruption of optical fiber.
g. Adopt technologies with less risk of failure
The most unpredictable disasters and sometimes the most devastators are the natural ones.
Adopting also technologies that, at least in theory, are less dependent on physical
infrastructures damages, can be an effective way to improve virtual communications loss
immunity and faster recovery periods. Examples of such technologies are mobile radio and
satellite applications.
h. Invest on disaster recovery support systems to improve time response
Develop and implement disaster recovery support systems based on the concept that his
introduction will enable rapid understanding of damage status, online information sharing
through centralized management of the statuses of damage, planning and mobilization, and
providing information to customers, relevant government offices and other entities outside the
company. Having this kind of recovery support systems will end in the acceleration of recovery
plan.
61
Figure 7.2 Kyushu Electric Power Company Disaster Information System [14]
Having disaster recovery support systems using mobile devices to receive inspection
instructions, enables the recovery staff to have access to an online view of damage status and
facilitates formulation of recovery plans.
62
In Poland, a system called PERUN is in use at the PSE-Operator (Polish Power GridOperator). It collects and displays the information about lightning which is sent from the
Weather Institute (Meteorology and Water Economy Institute IMiGW). The information is
refreshed every 10 minutes using FTP protocol. Communication is through fixed, point-topoint links and is protected against intruders.
PERUN operates with two METEO-FTP servers in cluster. The main one is in the Control
Centre and the second in the Back-up Control Centre. The software program RAPOK enables
to view the lightning data on the country map using different scales.
Up to 5 local computer stations can be connected to the system which runs separately from the
real time systems and no data from PERUN is transmitted to other systems.
To summarize the storm information service is a stand alone supportive service for dispatchers,
which provides aggregated 10 min. snapshots about the lightning fronts.
The usage of disaster information systems enables the processing of a high volume of fault
information, and prompt collection of service-interruption information, making possible to
accelerate recovery in areas hit by power outages during emergencies or disasters, and quickly
provide service-interruption information to customers, government bodies and other relevant
entities outside the Utility in large-scale disasters.
63
In Europe, the necessity to create a single electricity market determined new rules in the
organization of electrical business, the most important being by far the implementation of the
transmission system operator entity (TSO) in charge with the operation of transmission grids,
acting completely separated from generation and supply companies.
In case where the company owning a transmission system is part of a vertically integrated
group, there are two options: ownership unbundling or, only in exceptional cases subject to
the status of the company at a fixed deadline (September 3rd 2009) [26, 27], a right to set
up the system operator independent from supply and generation interests and strictly
monitored by the national regulator. The ownership unbundling model is that the
electricity transmission network is operated and owned by one independent from supply
and generation interests company, which reveals and undertakes the incentives,
responsibilities and liabilities for the network.
Independent system operator model, where the system operator does not own the
transmission assets but is ownership unbundled from the rest of the system, as e.g. in US.
Independent transmission system operator, fully unbundled from the rest of the system
which owns and operates transmission assets. This is the case of the majority of European
countries
Legally unbundled transmission system operator, unbundled from the rest of system which
owns and operates transmission assets. This model meets the EU requirements and can
involve effective separation of transmission operation from the rest of the sector while
transmission assets remain under the same ownership as generation or retail. e.g. France
A hybrid model where both the independent system operator and the transmission
operation are ownership unbundled from the rest of the system. The independent system
operator is asset-light, while the transmission operation has no system operation function.
This is the case in electricity market in Chile and Argentina.
A system and/or transmission entity embedded in the vertically integrated utility, e.g.
traditional utilities in Europe. This is the model that Europe has sought to move away from
in successive directives; however it is still in de facto operation in some European
electricity markets.
As it can be seen, the most sensitive case in the new market organization, still in discussion in
many countries, especially in countries with large and dominant vertically integrated
65
companies, is the setup of this monopoly manager that means a TSO with/without assets, but
unbundled from generation and supply.
This is why all the efforts of EU Community are still concentrated to find optimal compromise
to convince even not yet convinced giants to accept a kind of legally unbundling (see third
bullet above).
A detailed description of ownership unbundling issues and the underlying reasons and
regulations is beyond the scope of the present document and can be found for example in [28].
Similarly, in the Distribution domain, unbundling is in progress resulting in the separation of
operation from supply (in practice this is monitored by national regulators). In the European
case, following EU Directives [26], the independent DSO ought to be established by July 2007.
Connecting these afore mentioned forms of EPU organizations with Telecom Service Provider
profiles described in section 8.4, one can conclude that:
Type B could be common for TSOs (independent and legal unbundled models), as well as
for vertically integrated companies
Consequently, the term Electrical Power Utility covers at present a wide range of organizations
whose telecom service requirements and dispersion of sites lead to different service
provisioning models: Transmission System Operators (TSO), Regional or National System
Operators, Energy Market Operators, Transmission Companies, Distribution Companies,
Generation Companies, Regulators and Service Contracting Companies, etc.
Further consolidation within the industry has seen acquisitions and mergers take place, creating
large national or multi-national utilities which operate various business activities within a
complex regulated and un-regulated environment and in this respect assuming several utility
roles. The communication requirements and attributes of the company shall therefore be the
sum of those for each utility mission.
Something that the resulting EPUs have in common is that they have to operate as an enterprise
organization, and as such are accountable to a variety of stakeholders including:
Parent Company and Investors
Customers
Regulators
Partners and providers.
This accountability aspect is of great importance when analyzing telecommunication
management issues, and in particular the upstream tasks of policy definition and business
planning (refer to section 10.4.2).
Different Utility roles in terms of their respective attributes and specificities as telecom Service
Users are presented in this section and summarized in Figure 8.2.
66
Figure 8.2 Utility Roles as perceived from the communications point of view
Role of Utility
Sites
Communication Services &
Applications
National or
Connect to Control EMS/SCADA, WAMS and Voice
Regional
Facilities (tens)
facilities around the National Load
Coordinating or dispersed across a
Dispatch Centre and connection to
Operating Body country or a region
Control Centres and Back-up
with little or no
facilities for Transmission Grid
power network
companies, large generators
assets (e.g. USand distribution companies,
type ISO)
Administrative facilities
Market participants and Energy
Trading platforms
Transmission
HV substations and Protection relaying
Grids, TSO and
transmission lines
Security monitoring of
other entities
10s to few 100s of
installations/assets
operating
sites dispersed
Condition monitoring of assets
transmission
across the country
HV metering for Settlement
network assets
or a large region
EMS/SCADA, WAMS and Voice
facilities around the Control Centre
and connection to
Grid substations,
Other CCs, Back-up facilities,
Power Plants,
Operation Support sites
Administrative facilities
Market participants and energy
trading platform (TSO)
Distribution
Distribution
Distribution Automation
Grid / Energy
Automation:
Demand Side Management
Retail Supply /
1000s of sites
Market communications
Dispersed
HV metering for Settlement
Generation
Demand Side
Asset monitoring & supervision
Management /
(transformers, overhead and cabinSmart Metering :
based switches, MV lines, etc.)
100s of thousands
Demand Side Management /
Smart Metering
Connection to Administrative
facilities
Security monitoring of installations
and assets
Generation
Small number of
Intra-plant & Inter-plant
Utility
sites
Between the plant and the Control
(HPP, TPP,
Centre
Energy Farms,
Communications of the associated
etc.)
HV substation
Market communications
67
Specificities
Extensively based on
dedicated telecom networks
using mainly optical fibres
68
The relatively large capacity of the installed fiber has lead to the integration of EPU
Corporate communication services. This integration has often transformed the operational
telecom entity into a separate Telecom Service Provider entity, serving both operationrelated and corporate services. The separation of the telecom entity from the power system
operations has often resulted in some boundary uncertainties, migration (and/or loss) of
technical skills and misunderstandings on requirements (e.g. in the domain of protection
communications). The previously informal relationship between telecom and electrical
engineering in the substation environment generally evolves into in a more formal user to
provider relationship based on explicit specifications and SLAs. On the other hand, the
ICT staff moving from a department inside the EPU in charge of a support activity to a
separate telecom entity (i.e. a core activity), get higher responsibility, more value and
more incentives for their quality of accomplished work.
The extensive Right-of-way and/or extra fibre capacity in the cables, in conjunction with
telecom deregulation, lead many Transmission Utilities into the provision of commercial
telecommunication services to varying extents. This evolution has had further important
impacts on the provision of operation-related services as discussed in a separate section.
69
70
Large Wind Farms (e.g. greater than 30 MW in Australia), like other large generation
plants, must provide SCADA data to the System Operator who may also require to
perform set-point Generation Control, that is to say, to reduce the generation when
required. As mentioned previously large wind farms, either land-based or off-shore, are
often located outside the coverage of public telecom Service Providers. The transmission
utilitys telecom network must therefore extend for an access to these generation plants.
The cost of communication access is part of the connection cost of the transmission lines
71
to the Wind Farm. Communications from the remotely located control and surveillance
platform is performed via an often unmanned communications hut connected locally to
each wind generator unit. Wind farm communications include local automation, SCADA,
metering, different monitoring and surveillance applications, voice access in off-shore
wind turbines and in collecting and transforming substations as well as wireless facilities
for operation & maintenance staff. Optical fibres are often used in off-shore systems with
microwave radio as a backup between the off-shore and on-shore facilities.
Small green power generators and producing consumers on the other hand, generally
inject power into the rural MV distribution network. Remote communication to these
installations is generally covered by the Smart Metering system which in particular
includes remote disconnection functionality (e.g. for distribution network maintenance),
status monitoring and frequently refreshed metering data. Many different service delivery
modes have been used across distribution utilities ranging from procured services (e.g.
cellular data services, wired internet or switched telephone network) to light dedicated
communication links (e.g. UHF radio, VSAT and broadband wireless data). Further
communications may be required in the future perspective of Microgrids. The impact of
these generators on the telecommunication delivery pattern is to be considered as a further
attribute for Distribution Utilities.
Typical telecom service provision model for generation plants can therefore be as follows:
External communications services through the transmission grid network (using the power
plants grid substation), or an access link to the transmission grid network, either dedicated
or procured from telecom operator. Power Generation Utilities constitute natural
customers for U-Telco communication services.
72
Physical layer assets Optical fibre, OHL right-of-ways, cable trays, RF towers,
frequencies
Transport network assets All electronic equipment used for the core transport of
information. We purposely separate the assets for the bulk transport of information from
those used for multiplexing and interfacing of individual applications which constitutes an
edge or a distribution layer, even if in technological reality these two layers can at times
be merged together and therefore render difficult the separation of their assets ownership.
Application service network and platform assets all specific systems delivering
particular communication services through the core transport capacity (e.g. low capacity
access multiplexing, Voice network, Teleprotection signalling, SCADA communication
network, etc.)
Assets owned by the Service User (the operational entity in the EPU)
Assets owned by the Service Provider (whatever be its relationship with the user)
Assets owned by another party (e.g. state-owned fibre, fibre leased from another provider,
fibre, equipment or bandwidth belonging to another utility, etc.)
This section analyses some specificities in each case. Ownership criteria and issues for some
common telecom assets as described in the section are summarized below:
Asset Layer
Physical Layer Assets
Asset Types
Optical Fibre
Conduits, Rights-of-way
RF Towers, Repeater Housing
Radio Spectrum and Licenses
Long Term Contracts
Bulk Data Transfer Connections
Core Network Infrastructure
Narrowband Telecom Links (PLC,
Radio, etc.)
Service Multiplexing
Teleprotection Signalling
Voice and Data Servers
LAN/WAN Assets
Surveillance Systems, Platforms
73
The most determining physical asset is the optical fibre. It can by its own lead an EPU to a
particular mode of service provision or prevent an EPU from adopting a particular mode.
Installing optical fibre cables between communication sites of the EPU necessitates Right-ofway, that is to say underground or overhead corridors where the cable can be laid. This is a
very precious asset that Transmission Utilities own due to their HV transmission lines.
Optical fibre infrastructure can be provisioned by the EPU through one of the following
manners:
1. Procure and install fibre cables through the EPUs right-of-way corridors (overhead
power lines, underground power cables, etc.). This is by far the most used scheme in
Transmission Utilities, and in Distribution Utilities owning HV lines. Spare capacity can
be used for corporate and other communication services and spare fibres can be leased to
external users for covering costs or for extra revenue. Possession of extra fibres may lead
the EPU into building a U-Telco activity.
2. Jointly financed procure and install This scheme is typically employed at the
interconnection between two EPUs, e.g. transmission line interconnecting two
transmission utilities
3. Fibre (or service) in exchange of right-of-way The EPU fibre requirement being far
lower than the capacity of an OPGW, it can grant a telecom carrier the right to draw
multi-fibre OPGW cables in exchange of its required fibres in those cables. However, this
scheme presents many issues concerning the maintenance of the OPGW which is
intimately related to the maintenance of the transmission line. Even if often envisaged
(e.g. for immediate availability of financing when a sizable fibre infrastructure is needed),
it often evolves into case 1 with leasing of extra capacity. However, where the telecom
entity of the EPU moves away into commercial service and becomes a distinct company,
it may inherit the fibres and consequently the EPUs right-of-way through a long term
leasing contract, in exchange of fibres or services left to the EPU.
4. Swap with other fibre asset owner This is typically used for providing route
redundancy where the networks topology does not provide the required resilience. The
other asset owner can be another utility, a telecom carrier, etc. Access from the fibre asset
owner to the EPU site may be an important issue. It should be noted that these swapping
schemes may raise regulatory issues regarding the non-payment of taxes and duties.
5. Lease fibres in another EPUs cables This scheme is often employed at sites where a
smaller footprint EPU connects to a more extended EPU. Some typical examples are:
o generation plant using transmission utility fibres at the transmission grid
substation,
o distribution utility access to a national facility using transmission utility fibres
74
It can also be used to close a partially open telecommunication ring using assets belonging
to a regional footprint utility. Dark fibres (rather than transmission links) are leased when
the distances are sufficiently short to avoid intermediate regeneration and when high
capacity is required (e.g. Giga Ethernet). Fibre leasing from another EPU is generally
performed at co-located sites and therefore avoids the last mile issue encountered in
other leasing schemes.
6. Lease from a fibre asset owner This is the typical situation for EPUs that require a
high degree of control over their telecommunication network but do not have the
necessary right-of-way (or the initial capital investment) for installing their own fibres.
Optical cables over transmission lines may also be in the state public property as part of the
line conductor in case of OPGW, but conceded to the TSO or to the Telecom Service Provider
entity of the power utility for its internal usage. This type of long term concession in general
does not authorize the entity to which the cable is conceded to lease dark fibres.
Using leased fibre from an asset owner other than another EPU, raises several important issues
that need to be considered:
a) The topology of the resulting physical network depends upon the fibre owners network
structures leading to far longer than necessary links and often far from optimal overall
network.
b) Last Mile issue the distance from the fibre asset owner access point to the EPU
premises, even if relatively short, needs right-of-way and non-negligible civil works
inducing important cost and delay consequences.
c) Physical routing issue The design of a fault tolerant transmission network is based on
the knowledge of physical medium routing which is controlled by the fibre asset owner. In
particular, where fault tolerance is concerned, the two routes must not pass into any
common nodes, cables or conduits. In the event of incorrect routing information or
cable route changes, the fault tolerance of the whole system may be unacceptably
compromised. The EPU has no other way than the providers word to keep track of
changes or of the correctness of the routing information. Moreover, it is particularly hard
to obtain two independent cable routes from the EPU premises to the cable providers
meshed network (and not the providers access point).
d) Maintenance works EPU need to have control of maintenance schedules which is not
the same thing as being informed of the date of maintenance works. A multi-customer
fiber provider cannot program his works according to EPU requirements. In case of
interruption, of its fibres, the EPU requires immediate repair which may lead to unscheduled interruption of other fibre users without prior notice. However, in case of other
users fibre interruption, the EPU cannot accept non-anticipated maintenance works. A
very non-symmetrical contract in general unacceptable to the provider is needed.
e) Cable reliability The majority of fibre providers have underground cable infrastructures,
particularly subject to cable cuts due to civil works, while overhead OPGW normally used
by the EPU is almost invulnerable. The extremely high levels of service availability
required by EPU operation-related applications are very difficult to meet with the
probability of cable cut that can be obtained from cable providers. (when using OPGW,
75
Another important category of physical layer assets in the EPU are those related to
implementing radio transmission networks and links. We add here repeater housing even if this
can also apply to optical regenerator housing:
RF Towers in HV substation premises and power plant sites, including tower lighting and its
associated power supply, are generally the property of the EPU. These towers as well as other
EPU structures which can serve as antenna support (e.g. electrical poles, power plant tower
structures, etc.) may also be used by other radio network infrastructure owners such as cellular
radio operators as a source of extra revenue for the EPU. Antenna support outside EPU
premises (e.g. microwave repeater or radio base stations) can be through co-location on towers
belonging to other radio networks. In particular when a wide zone coverage is required (e.g.
UHF data systems, mobile trunk systems, etc.), the optimal location of radio relays for
covering a given zone is often the same for all radio infrastructures, facilitating co-location.
Repeater housing, including air-conditioning facilities and repeater power supply can be the
property of the EPU, its telecom Service Provider, or leased from an external party. Microwave
link repeaters are often located on EPU premises in which case, they are generally EPU assets.
UHF and other zone coverage base stations on the other hand, are often on high sites, and may
be in shared housing leased from another asset owner or telecom Service Provider. The
maintenance of the facilities, in this case, is generally provided by the owner as an external
service to the EPU. When using externally provided power supply for radio relays, the
autonomy of the power supply and the dimensioning of batteries is an important issue for
operation during power outages.
Frequency licences with narrow directivity (e.g. microwave radio links) and narrow bandwidth
(UHF radio for SCADA or few channel mobile systems) are generally applied for under the
name of the EPU user and are therefore part of its assets. Licensed broadband spectrum with
wide coverage, on the other hand, cannot be allocated in many countries for the exclusive
usage of the EPU internal communications. It is, in this case, common to obtain shared usage
with other Utilities (e.g. gas, water, other EPU) or other critical services. This is normally
performed through procurement of services from a specialized operator, or setting up a service
that can be procured by other users. This latter case generally results in the separation of the
Service Provider entity from the EPU operational entity.
Power Line Carrier Narrow-band PLC whether HV, MV or LV is generally a dedicated asset
for a particular Utility application (Protection, voice and SCADA in HV, device monitoring
and metering for MV/LV). The physical coupling assets consisting of line traps, coupling
capacitors, Line Matching Units and the frequency spectrum assets, are indeed the property of
the EPU operational user. Particular attention has to be paid to the maintenance of these assets,
which usually fall under the responsibility of HV line teams, not belonging to the EPUs
telecom department (or a fully separate telecom entity). Broadband PLC, when used for
76
multiple purposes and in particular when commercial services such as customer internet access
are involved, the situation may be more complicated. A separate telecom service providing
entity shall be necessary to deal with this situation.
Ownership of underlying physical layer assets When the physical layer assets are not
under EPUs control, it is easier to admit lack of control over the transport layer (e.g.
leasing STM1, E1/T1 or Giga Ethernet connection rather than leasing dark fiber and
repeater housing, power and maintenance). This may lead to more straightforward contract
and SLA management and less interactions.
Required communication bandwidth Narrowband information transport on owned
physical assets, e.g. HV PLC access to substations or wireless SCADA systems, is always
performed with EPU-owned transport network assets. On the other hand, when the
communication requirement are small compared to the capabilities of the available or
suitable communication technology, bandwidth sharing with other users is necessary,
either to justify cost, or to overcome regulatory constraints. (e.g. Broadband wireless data
services, Satellite Communication Hub).
EPUs required level of control over the service The more a communication service is
critical in the EPUs process, the more it is inclined to keep full control of the associated
transport layer assets (e.g. communication services for Protection Relaying applications).
Total Cost of Ownership, Asset Life Cycle and Return on Investment (ROI) The cost of
implementing and maintaining a particular type of transport asset may lead the EPU to
renounce to its ownership. This indeed is to be traded off with the requirement to keep full
control. It should be added that unlike physical assets, transport network layer assets have
much more limited life cycle, meaning that the ROI must be possible in shorter time.
Required skills for managing and maintaining the transport assets The EPU may simply
not have the necessary skills, tools and organization to run a particular type of transport
network, or the organizational capability to keep it up-to-date. Large core data networks
and Network Operation Centre facilities to run bulk information transport are typical
examples of hard-to-maintain assets.
Here again, the short lifecycle of the assets (e.g. IT platforms) and the total cost of ownership
being mainly driven by the cost of upgrading and maintaining, the EPU is highly inclined to
procure services rather than assets. This is therefore a typical area where the Service Provider
is in a better position to invest and to obtain ROI.
Power Corporation
EPU
Telecom
Service
Operations Corporate
Provider
Activities
EPU
Operations
Corporate
Activities
Telecom
B
Operations
EPU
Telecom
Services
EPU
Corporate
Activities
Corporate
Activities
Operations
Telecom
Service
Contractor
Telecom
Assets
EPU
Operations
Corporate
Activities
Telecom
Service
Provider
A: Telecom is part of the operational activity. Corporate entity provisions telecom services separately.
B: Common Telecom (& IT) Services for both Corporate and Operational Applications.
C: TSP is a sister company to the EPU, providing services exclusively (or in priority) for the Power System
D: EPU procures its telecom assets but operates them using an external Service Contractor
E: Telecom services are procured under SLA by a TSP providing services to many customers.
78
79
An integrated telecom service provision scheme can scale up to cover corporate or other
communications inside and outside the EPU, but in this case, the evolution to a type B situation
is almost automatic in order to cover assets and running costs for the corporate
communications.
80
The liberty of the company in terms of development strategy, assets and human resources and
extra income from sharing the infrastructure with other users (or providing services to external
customers) normally results in a more cost-effective telecom service provision and should lead
to lower service costs for the EPU. On the other hand, the telecom Service Provider must
assume the responsibility for network planning, development and refurbishment of
communication network and service platforms in order to maintain the quality of the delivered
service (e.g. mitigate asset aging) and to ensure that the infrastructure is capable of responding
to new requirements (new services, increased bandwidth requirement, and service migration)
provided that the EPU ensure the financing. This requires periodic assessment of EPU
migration plans at the time of revision of the service catalogue and pricing.
However, delivering U-Telco services can also lead to telecom regulatory issues and in
particular fair trade regulations loosening the preferential links with the EPU. Depending on
the proportions that external service provision may take in comparison to the EPU service, the
danger is that in time, the affiliated telecom company may become simply a normal
commercial service supplier resisting the specificities of the EPUs operational services as
further discussed in chapter 10.
Service Management
Telecom Infrastructure Management
Field maintenance
The contractor provides organization, process and skills, even the absorption of EPUs
telecom staff and can often better maintain the skilled workforce through more competitive
salary policy than the EPU itself. On the other hand, the EPU shall lose technical knowhow in
medium/long term and consequently the control of its network and of its contractor.
The contractor is engaged with a Service Level Agreement governing its interventions and
services but is not responsible for the failure of ageing assets or their lack of performance
whose renewal policy remains with the EPU employer even if the contractor conserves an
advisory role in this respect. Typically, the service contractor must prepare a yearly
development and refurbishment plan of communication network and service platforms based
upon the EPU plan for application changes and the contractors survey of aging assets. The
contractor can only assume the responsibility of maintaining the quality of the delivered
service if the EPU accepts the refurbishment and new developments ensuring that the
infrastructure is capable of delivering the service.
81
Procuring telecom services liberates the EPU from procuring assets, deploying infrastructures,
employing skilled workforce, building processes and deploying tools for its management and
maintenance. However, as it will be seen in chapter 10, the EPU shall still need to manage the
external Service Provider with adequate processes (and tools) and adapt the procured
communication resources to the requirements of its internal users.
The infrastructure is extended, diversified, upgraded and renewed without any involvement
from the EPU. However, extensions, new services and service migrations need to be planned
long in advance to ensure that the provider shall have the capability of delivering the new
services (e.g. covering new sites, increasing capacity in remote areas, etc.). This will be
included in the yearly renewal or revision of service contracts.
However, this mode of service provisioning presents many drawbacks which are symmetrically
opposite to the advantages given in section 8.4.1 above. The EPU will have, in particular, to
provide considerable effort in the following domains:
1. Formally and precisely specify service requirements and constraints. It should be noted
that the terms and vocabulary do not have the same significance in public telecom and
in the operational EPU context (e.g. availability) and may lead to misunderstandings
with great consequences. Time behavior and predictability of the connections may be
an important point to consider.
2. Establish Service Level Agreements (SLA) and Sanctions for not respecting them It
should be noted that non-respect of SLA in the world of telecom is sanctioned by
financial compensation with no proportionality to the EPU risks due to lack of service.
3. Carry out Performance Measurement and SLA Monitoring with appropriate tools
4. Provide considerable effort in contract and conflict management,
5. Implement application interfacing and service multiplexing in operational sites where
the service operator cannot access,
6. Coordinate Security Policy and Disaster Recovery/Business Continuity Plan of the
Service Provider with those of the EPU. Perform audits to assure that they are not
compromised. In particular, power autonomy, or the capability of the telecom service
to be delivered in the event of a power outage through adequately dimensioned
batteries is of great importance for Disaster Recovery.
7. Schedule long in advance any extensions, changes and upgrades and negotiate in good
time with the provider.
8. Avoid monopolies and dominant positions for any single telecom provider which may
increase its prices and decrease the quality of service.
9. Service life expectancy has to be carefully analyzed before using extensively a
standard service delivered by a provider. Many cases can be enumerated where a
standard telecom service used by an EPU is abandoned or replaced by another service
not equivalent for EPU usage (e.g. leased digital circuits used for protection relay
communications).
10. Safety certified field maintenance workforce or safe location for providers assets.
82
As it was stated at the beginning of this chapter, different telecom service provisioning modes
often co-exist in the same EPU depending on the nature of services.
When operation-related telecom services are provisioned through an integrated entity (type
A), then corporate communications are generally through procured service (type E).
When operational and corporate services are integrated into the same provisioning model
and organization (type B, C or D), then Protection communications are often separated
from this integrated approach and performed directly through separate fibres (or
wavelengths).
Figure 8.5 summarizes main service delivery modes in different types of EPU.
Telecom Provider Profile
Integrated with
the Operational
User
Sister Entity to
Operational
User
Affiliated Service
Company
(to Service User
or to Holding)
Independent
System Operator
TSO /
Generation
Company
Distribution
Network Operator
Public Telecom
Operator
(Procured
Service)
Communicates
essentially to
other EPUs.
Has no telecom
infrastructure
National / Regional
System Operator
Transmission Grid
Company
Independent
Service
Contractor
Operationrelated service
with limited
resources
PTO cannot
provide
coverage for
HPP or off-shore
wind farm
Wireless SCADA
Mobile
workforce
Small number
of high traffic
urban sites for
TPP
Use of
backbone for
Metering,
Commercial
offices & IT
Generally using
customer access
for providing
other U-Telco
services
Large number
of sites with
little traffic,
Metering,
Mobile comms
Figure 8.5 Examples of use for different telecom service provision schemes in the EPU
83
84
Service Management
Organization & Formal Processes
Management of operational telecom
services is performed by Power
System Operation entity (Control
Centre and Operational Areas) or
by a separate Service Provider
managing operational and nonoperational communications?
85
Provider-client Service
Management Platform
86
Using
Separate Fibres
Commercial
Telecom Service
Most commonly
used
Corporate Data
Network
Corporate ICT
managed by fully
distinct entity
Protection
Applications
Used for
Protection comms
between
substations
terminating a
dedicated OPGW
Separate
Wavelength
Used to
economize fibre
or optical device
for some (long)
links
Same as Separate
Fibre but in a
fibre-constrained
context (requires
common optical
network
management)
Used to separate
Protection from a
multi-service IP
/MPLS network in
a fibreconstrained
context
Separate Phys.
Bandwidth
Not used except
for some special
access services
and when one
EPU sells
bandwidth to
another EPU or
another Utility
Only specific
corporate links
implemented
through dedicated
telecom
Separate
Virtual Network
Multi-service
IP/MPLS
network
integrating
operational,
corporate and/or
commercial
services with
VPN separation
IP access to HV
substations
Used when
Protection is
integrated in a
SDH/PDH
dedicated network
None at present
Figure 9.1 Commonly used Service Separation schemes employed in Electrical Power
Utilities
87
For commercial and/or corporate services mainly used when corporate and commercial
service is provided by separate entity or when an EPU has to account for regulated and
unregulated (commercial) activities using segregated accounting practices.
For Protection mainly used when telecom service providing entity does not provide
protection communication services
For Core IP/MPLS data network mainly used when the telecom entity operates an SDH
network and an IP/MPLS core data network.
Assure the coexistence of SDH multiplexing networks with packet-oriented IP/MPLS core
data networks for which traditional SDH cannot deliver the required scalability. CWDM in
this case is used as an overlay onto existing legacy PDH/SDH optical systems, thereby
enabling additional services to be delivered while preserving the assets for current legacy
systems.
The use of wavelength separation rather than fibre separation is also a particularly appropriate
solution where extra fibre is not available, cannot be provisioned economically or existing
extra fibres can be used in a more attractive manner (e.g. profitably leased). A typical example
is to break into a prior existing trunk cable to enable delivery of more localized services in a
region.
The significant cost reduction and reliability improvement in recent years has rendered the
WDM technologies an accessible and economical solution in existing private optical networks
for providing secure transparent connectivity services.
CWDM is the low cost technology for multiplexing up to 16 transmission channels over a
wavelength domain from 1300 nm to 1610 nm with channel spacing of 20 nm (ITU-T G.694.2
and G.695).
Its main market is cost effective transportation in metro networks and can be used in different
types of network configurations. The transmission range is often limited (around 40 km when
using all 16 channels, and around 100 km when utilizing only 8 channels).
DWDM on the other hand, is a technology used for long distance and/or very high capacity
links. It can typically carry from 32 to several hundred channels over a wavelength domain
from 1530 nm to 1624 nm (ITU-T G.694.1) with a channel spacing less than 1 nm. This results
in the requirement for precise wavelength control of the laser light source and temperature
stabilization in the multiplexing units. The transmission range is around 80-100 km.
The WDM scheme adopted by utilities depends on the planned expansion of activities. In most
cases, where the network is essentially dedicated to internal requirements (operational and nonoperational) with limited external Service Provider capability, CWDM is the cost-effective
88
solution, DWDM being reserved for larger capacity requirements and long segments requiring
amplification. High power WDM systems have been reported allowing long spans without
costly wavelength control [29].
89
Providing IP connectivity for operational and corporate (and perhaps U-Telco) services, over
the same physical network bandwidth remains controversial and subject to discussion in many
EPUs. While dedicated channel-based technologies such as SDH or WDM, rarely raise the
question of a service putting another at risk, there is a natural distrust for service provisioning
models based upon shared resource environments such as MPLS, and in fact defined security
policies need to be enacted to ensure that the separate VPNs are in fact securely delivered [30].
There are some basic rules that must be addressed in order to permit coexistence of internal
and external customer services in the same MPLS network:
MPLS being is a carrier oriented backbone technology, it is highly recommended not
to mix the carrier network with customer networks. Adopt the MPLS recommended
topology consisting of Provider, Provider Edge and Customer Edge routers.
Avoid connecting layer 2 customer equipments directly to the MPLS core network;
Before configuring an MPLS service one must ensure that all customer service
requirements (availability, bandwidth, delay, maximum re-convergence time, etc.) are
known and can be guaranteed. For some specific services MPLS may not be the best
solution;
Each customer service must have its own separate MPLS VPN service;
Customer service level requirements must be guaranteed through implementing
Quality of Service for service prioritization and data flow performance;
MPLS Traffic Engineering must be implemented to have a more efficient network
usage, to deal with network congestions and strong traffic pattern changes;
The network must be protected from the security threats that may originate from
customers. Configure and operate the network with a secure network management
system, secure protocols and secure procedures;
The network must be monitored in terms of equipment interconnection load, packet
loss, resilient connections availability and security events;
It may be envisaged to implement independent data networks for internal services
(MPLS or simply IP) and external commercial services (MPLS). Such architectures
get the best of both worlds at somehow higher cost, providing complete segregation
between internal and commercial services while taking advantage of multi-service and
multi-customer capabilities of MPLS.
To end this section, it should be noted that MPLS is now an aging technology first developed
a decade ago. It remains complex and does not fully respond to all requirements in particular
for delay- and jitter-intolerant services. A number of other Scalable Ethernet transport (or
Carrier Ethernet) technologies have been specified and developed, fulfilling exclusively the
transport objectives of MPLS:
PBB-TE (Provider Backbone Bridge Traffic Eng.) specified by IEEE (802.1Qay2009) delivering WAN Ethernet services without the protocol complexity of MPLS.
An example of such implementation at experimental stage has been demonstrated in
Argentina [31], elaborating a two layer architecture associating CWDM and PBB-TE
to deliver new and legacy operational services.
An extremely light organization with minimal processes and elementary tools may be
sufficient for dealing with a small network dedicated to a very limited scope of services.
However, if the scope of delivery grows in scale, then the light scheme will become
completely ineffective.
A large and complex formal process can overkill an activity if it is imposed onto a small
organization with a limited scope of service delivery. Large management tools in small
organizations may fully jeopardize the workforce, becoming occupied with the tools
rather than with delivering the communication service.
Management
Tools
Management
Scope
Management
Management
Organization
Process
Figure 10.1 Organization, Process and Tools adequacy to the Scope of delivery
91
In the case of EPU telecommunications, the management scheme has been trivial and implicit
in the past due to the small scale, simple and time-invariant nature of requirements, informal
user relationships, and non-recovery (sometimes non-assessment) of costs. The typical service
delivery model in this case has been the Type A or Integrated model described in section 8.4.
However, it is often observed that many EPUs undertake major changes in terms of scale and
scope of telecom service delivery focalizing on the infrastructure capacity and technology
without paying sufficient attention to the associated management aspects. The organization
moves from type A to types B or C (section 8.4) assuming that that same informal management
process may remain sufficient. The EPU telecom provisioning moves to external contractors
services for running and maintaining the telecom network or even to provide managed
telecommunication services (types D and E) hoping that the contractor or provider would find
himself bound by implicit constraints of the proper operation of EPUs critical applications,
resulting in non-satisfaction and conflicts.
Indeed, the investment in a standards-based, well-defined management framework is far from
trivial endeavor. In particular, managing the network cannot be reduced to purchasing high cost
sophisticated network management tools.
Some of the reasons justifying a redesign of management system and processes in the EPU
Telecom Service Delivery are given below. The triggering condition is generally the intention
to change the scope of service delivery (e.g. customer base), the scale of services, the
organization, or the evolution of the service and/or infrastructure beyond its initial deployment
objectives:
Complexity of the network and its underlying infrastructure necessitating the cooperation
of several complementary skills.
Reduced workforce and less interchangeability of staff implying that roles and
responsibilities must be more clearly defined
The present chapter aims to produce guidelines for the definition of management processes
governing the provision and delivery of telecom services in the EPU. It aims to establish a
reduced framework, applicable to the domain of Utility telecommunications, comprising a
well-defined set of processes to formalize the roles and the interactions. The framework is
based upon existing Best Practices and existing Management Frameworks which, as they are,
may appear too complex to be relevant in this context. It should still be noted that the specified
framework remains a pick and choose base to be used by each Utility according to its
specificities and its positioning across the business maturity timeline as described in the next
section.
92
93
Operational
Services
only
External
Leasing of
Facilities
External
Managed
Services
External
Non-IP
Wholesale
External
IP Service
Wholesale
External
Retail
Services
Supplier
Helpdesks &
Warranties only
Field
Maintenance
Contracts
Full Service
Delivery
Contracts
Service costs
estimated but
not recovered
Service costs
recovered at
no profit
Service price
established as
Service Catalog
Relationship
can get formal if
problems arise
Formal SLA
No systematic
measurement
Liable to SLA
Must constantly
produce formal
proof
Procure
Telecom
Services
Resource
repartition per
contributing
entity
Required QoS
is defined &
agreed upon
10.2.2
The Business Maturity model discussed in the previous section locates the EPU telecom
activity and involvement at present and in the future, but does not situate it in terms of present
and future management practice.
Moreover, transforming the work process of an organization cannot be performed overnight,
moving from chaotic, informal ad hoc relationships of a very small delivery unit to a fully
mature value-oriented scalable organization with tools, processes and work practice that
continuously align to business requirements and opportunities. This has lead to the definition
of Capability Maturity Models composed of different maturity levels. This stepped model
can be used to assess the capabilities of the management processes in the EPU Telecom
scheme as illustrated in figure 10.3.
A maturity level is a defined evolutionary plateau for organizational process improvement.
Each maturity level covers an important subset of the organizations processes, preparing it to
move to the next maturity level. The maturity levels are measured by the achievement of the
specific and generic goals associated with each predefined set of process areas. A capability
level for a process area is on the other hand, achieved when all of the generic goals are satisfied
up to that level [33].
It should however be noted that the target state is not necessarily the highest
maturity/capability levels for all management processes. Many processes may be fully
acceptable at a reactive stage (repeatable) considering the required cost and effort for moving
to a proactive stage. Moreover, no matter how proactive you are, there will always be
unexpected problems that call for an immediate reaction. It is much more important to check if
94
the documented processes are effectively being followed and not constantly worked around
due to being not adapted, too costly or requiring more effort that the organization can
afford.
Capability Maturity Levels
95
TMN Logical Layered Architecture (LLA) defined by ITU-T M3010 (Elements, Network,
Service, and Business Management) has, on the other hand, been devised for extending the
visibility of the management system through different layers of abstraction. It extends the
NMS vision beyond the individual network elements into the vision of the network
composed of many network elements, and then covering multiple networks to manage a
whole service, and different telecom services constituting a telecom business [37].
For analyzing the overall management process for telecom services in the EPU however, one
needs at the same time a functional view (Fault, Performance, etc.), a lifecycle view (long-term
decision making, deployment and upgrades, day-to-day operation) and a logical layer view
(Element, Network, Service and Business).
FCAPS and LLA have been widely treated in previous CIGRE Brochures (e.g. [38]). In the
present document we focus on Management Frameworks defining the business processes that
must be deployed and their interactions and in further sections on applications, tools and
information systems.
Service Support
Service Delivery
96
The replacement version V3 (since 2009) globally covers the same scope but organized in a
service life-cycle-oriented manner (i.e. Service Strategy, Service Design, Service Transition,
Service Operation, and Continual Service Improvement).
Through a scalable and flexible adopt and adapt approach, ITIL is applicable to all IT
organizations irrespective of their size or the technology in use.
Many EPUs are at present using the ITIL framework (or standards based upon it) for the
management and governance of their IT infrastructure and services, implying familiarity,
practice and trained staff. This covers the corporate and operational information platforms in
terms of IT service delivery, IT service support and other associated processes. Applying ITIL
may therefore be an obligation through corporate policy, on the IT part of the EPU telecom
provider activities and on common processes such as Service Support (e.g. Service Desk) if
Telecom and IT organizations are merged.
ITIL is further described in the Appendix 7 and in [39].
NGOSS-Frameworx and its constituents are described in more detail in Appendix 8 and in
[40].
97
Market, Product and Customer transversal This includes all upstream and operations
processes which are directly in contact with the customer/user. It is the customer view of
the Telecom Management. In this context, product means a productized service
proposed by the provider to his customers (e.g. a communication service with a preestablished level of quality and price a standard SLA)
Service This covers all processes required for the delivery and support of
communication services to its customers using owned resources and/or other suppliers
and partners procured services.
At first glance, eTOM Business Process framework looks terribly complex and mainly relevant
to competitive customer-driven Telecom Service Provider business with millions of customers,
very large number of assets, numerous support service suppliers and a large and continually
renewed service portfolio. Modeling the relatively modest scope of EPU telecom service
provision/delivery through the strength of a Telecom Operator Business Process Modeling tool
is not immediately obvious.
However, it should be noted that eTOM framework is structured for pick and choose of
relevant sections according to the scale and size of the activity and may be adapted to
requirements. Process elements representing the activity of a whole department in the case of a
Public Telecom Operator may be part of the mission of one person or entity in the EPU
telecom organization. eTOM identifies in a detailed manner the required interactions and
allows the detection of potential areas of automation and IT support.
98
99
100
Upstream Management comprises all the processes necessary for building the
telecommunication policy and the provisioning model in conjunction with the power
utility actors and the regulating authorities, setting up the business plan for the adopted
service delivery organization, and deploying the necessary infrastructure and/or contracts.
The outcome is an organization with staff, processes, assets and service contracts which
allow through day-to-day operation to deliver the necessary telecom services within the
EPU. The deployment process also includes the migration planning from the existing
scheme (or network) into the target service provisioning/delivery model and underlying
infrastructure. This aspect is discussed in section 10.5.
Current Operations consist in running a network and providing services to different users.
It comprises different work tasks in relation to external Service Providers, in relation to
EPU service customers, and in relation to the network services and infrastructures as well
as associated tools.
Upstream Management
Organization,
Process,
Network Assets,
Service Contracts,
Service Catalog
Policy Definition
Business Planning
Deployment
Upgrade & Migration
Business Development
Operational Data,
Assessments,
Audit Reports,
Asset Requests
Current Operations
Customer Relations
Service Management
Infrastructure Management
Provider Management
101
Customer / User
Provider / Contractor
Figure 10.6 Basic model demonstrating Service Provisioning from (external) Providers and
Service Delivery to (internal) Users
102
3. The management model for Utility Telecom services must cover the whole life-time of the
service provision cycle:
Building a strategy for the provisioning & delivery of telecom services
Building the capability to materialize the provisioning & delivery scheme
Defining the perimeter of services to be proposed to the customers
Operating the service
Upstream Management
Strategy
Operations
Capability
Building
Building
Build a strategy for the
provisioning & delivery
of telecom services
Define Policy
Identify Opportunities
Set-up Business Plan
Convince stake-holders
Set-up Strategic Partnerships
Solve Regulatory Issues
Business
Operating
Development
Scheme
This division corresponds to the eTOM verticals called S (Strategy & Commit), I
(Infrastructure), and P (Product) adapted to the Utility Telecom context.
4. In the great majority of Electrical Power Utilities, telecom services are delivered through
an owned telecom infrastructure requiring extensive network infrastructure management.
The telecom management framework must therefore include a Network Resource
management process layer. Adopting eTOM structures, this also includes the management
of all IT tools and platforms associated to the telecom network and service management.
5. Operations management processes generally follow a same pattern which is illustrated in
figure 10.8 as Initialize, Deliver, Provide Support and Get Paid. This corresponds to the
eTOM verticals called Fulfilment, Assurance, Support and Accounting (or Billing).
A user request (e.g. a new SCADA connectivity) is initialized on the basis of a
previously defined SLA (type of interface, capacity, QoS, etc.).
The service is delivered to the user from the moment it is registered, provisioned over
the network and established. Support is given to the user at different levels for this new
connectivity.
The telecom provider entity may measure the usage of the network and invoice the
SCADA entity for the communication service.
103
Deploy &
Adjust
Business Development
Service Portfolio Evolution
Service Migration Planning
Customer / User
Operations
Upstream
Management
Build Capability
Build Strategy
Strategy &
Planning
Provider / Contractor
Enterprise
Processes
Initialize
Deliver
Get Paid
Provide Support
104
The EPU is a commercial organization whose goal may not only be delivering reliable and
secure electricity but generating profit through diversified activities and pay dividends.
The stake holders are no longer the state representatives but also the Parent Company, the
Investors, the Customers and the Regulation authorities.
In such a context, the tasks of Upstream Management for Telecom Services widen up
considerably. These cover all the processes resulting in the adoption of a particular telecom
service delivery mode and the preparation of financing, facilities, organization, resources and
partnerships for the activity to be sustainable, economically viable and approved by the
different stakeholders.
Figure 10.9 presents the main constituents of Upstream Management in the Utility Telecom
context and figure 10.10 a typical process flowchart and organization for these tasks.
Deploy &
Adjust
Business Development
Service Portfolio Evolution
Service Migration Planning
Build Capability
Upstream
Management
Build Strategy
Strategy &
Planning
Operations
Enterprise Processes
105
Reliable and secure accomplishment of the EPU primary mission (e.g. power generation,
coordination, and delivery) thus reducing the risks of human accidents, losing customers,
sanctions and penalties, and costly damage to power system assets,
Shareholders satisfaction (usually long term growth) and good mid term returns
(dividends) at the EPU level,
Regulators approval (essentially delivering value for money to the end customer).
The strategic policy on telecom services must allows the EPU to define the focus and strategic
direction of the business over the next five year period:
Determine how the organization will deliver the services based on its existing resources
(both capital and people),
Define any requirement for the cessation or diversification of existing services, identify
opportunities or requirements for the creation of new services,
The strategic policy for telecom services depends upon a number of factors and information
from a variety of sources. Some important factors and issues are enumerated hereafter:
Parent Company Policy (and Business Plan) This will define the general framework and the
orientations to be taken towards telecom services (e.g. diversify activities to other profitable
businesses)
Regulatory Context Telecom activities of the EPU must respond to the requirements of any
Regulatory and Legislative frameworks to which it is bound. In the case of U-Telco business
operations this includes the Telecommunications Regulator as well as the Energy Regulator.
Regulatory constraints may have significant impact on the availability of funds to develop the
telecom facilities, may create new security or safety obligations leading to extended telecom
services or refurbishments and renewals, may favour certain delivery modes to others through
CAPEX/OPEX distinctions, and may present an obstacle to the integration of corporate
enterprise services with the operational services.
Organizational Opportunities ICT Convergence, the technological trend removing the
boundaries between information processing/storage and exchange and the similarity in the
required skills is driving utilities to consider the merger of these organizations bearing in mind
that the cost and effort saving is to be balanced against potential issues regarding certain
processes and work constraints. Such a merger between IT and telecom activities can greatly
impact the strategic orientations for telecom service delivery.
107
Similarly, the cost and effort for maintaining two independent telecom organizations within the
utility perimeter to fulfil operational and corporate enterprise communications is driving many
utilities to consider their merger.
Market Opportunities Revenue Generating Commercial telecom involvement through UTelco business are often envisaged by EPUs every time that major enhancements are planned
in the telecom infrastructure. Setting up a U-Telco business however necessitates the
preparation of a distinct business case based on detailed potential market surveys and
opportunity identification which is beyond the scope of the present document. Combining the
essentially non-profit mission of operational telecom service provision with the profitoriented commercial service in a same organization may rise a number of organizational and
regulatory issues that need to be carefully analyzed (e.g. business development in a competitive
environment without compromising the critical service constraints of the power system, or
access to financing for operational service development while maintaining competitive
commercial activity).
Strategic Partnership Opportunities Many utility services require coverage, workforce or
investment which is economically unfeasible for the EPU on its own. Resource sharing (cable
infrastructure, radio coverage infrastructure, emergency mobile network, or maintenance team)
between multiple utilities or multiple critical Service Users can be a way around this
economical obstacle.
Investment Context Financing the deployment of a large telecom infrastructure can be
performed through the companys own funds, or through international loans. This latter type of
financing cannot be applied to an essentially OPEX mode of telecom service delivery (e.g.
procuring telecom services). Similarly, obtaining financial support for developing the network
for competitive commercial services requires demonstrated Return on Investment, which
may not be required for financing operational service development. Combining the two may
cause difficulties for obtaining either type of financing, unless clearly separated.
Resource Context As previously stated, asset ownership, in particular optical fibres across
transmission lines, is a determining factor on the selection of service provisioning modes and
on cost assessment. It is also a determining factor for the feasibility and interest of U-Telco
involvement and hence of the envisaged strategic orientations.
Another highly decisive resource in the EPU telecom environment is the RF Spectrum.
Assuring that adequate RF spectrum is allocated (or maintained) for the operational usage of
Utilities is a major field of national and international action in many parts of the world in
particular in the US and Australia. Gaining (or keeping) access to specific parts of the
spectrum is a political issue necessitating power utilities to voice their concerns collectively to
legislators and regulators (e.g. UTC). Access (or lack of access) to adequate RF spectrum
highly impacts the mode of delivery for mobile services and power distribution network
communications.
Skills and Human Resource Context Unavailability of sufficient workforce and/or skills
for scaling up the telecom activity may lead to the merger of IT and corporate enterprise
services with the operational telecommunications, or to outsourcing/ contracting.
108
109
of existing equipment: compare the costs of replacement of old equipment with the cost of
upgrade and maintain. Timely replacement of old telecom assets may allow substantial
economy on operating expenditure (spare parts, obsolescence management, site interventions
and repairs, etc.) in addition to enhanced functionalities and performance.
A major issue in large-scale telecom network rehabilitation projects is the migration of
operational services from the existing network to the target network with minimal and
scheduled service interruptions.
Building Service Contracts Contracting of services for the deployment of network
infrastructure and management tools, and/or operation, maintenance and support necessitates
precise specification. If turn-key implementation projects are often well specified, precise and
covering responsibilities and liabilities of each party, service contracts are often poorly
defined. Service contracts are often based upon Service Level Agreements (SLA) describing
the engagements of the contractor to intervene upon encountered anomalies in constrained
time: this contractual time defines the grade of the SLA (e.g. gold, premium, diamond, etc.).
The contractor must provide people, organization, knowhow, tools and process. However, the
responsibility of the contractor, who has not designed the system, has not chosen and procured
the equipment, and has not decided to maintain rather than to replace old assets, cannot be
extended to the proper operation of the system with an acceptably low down-time. He can just
guarantee to intervene in contracted time limit with skilled staff. Coordinating the contracts for
equipment procurement, network implementation and operation & maintenance can prove to
be extremely hard, leaving gaps in the overall responsibility.
The liability of the contractor is another important issue. In general, the level of sanction
cannot cover the potential loss: risk sharing cannot be back-to-back.
Building organizational capability may also result in the need to outplace existing staff
previously involved with that particular service into the contractors workforce. This delicate
human resource issue needs to be taken into account and treated at the upstream management
level prior to the change. Moreover, it should be noted that the use of external resources in
whichever scheme is indeed a way to acquire rapidly and flexibly the required skills for
introducing new technologies, but the obtained skill and experience is not sustained in the
company. Outplacement of skills and experience into external contractor companies is often an
irreversible process and may represent a loss of control and capability in telecommunications,
rendering the EPU dependent of its contractor.
Deploying Management Tools & Processes Building organizational capability requires also
the definition and application of new operational processes which will be appropriate for the
new scope and scale of activity. These are described in a further section on operational
management.
Deploying management tools, beyond the vendor-specific equipment configuration and
management platforms, represents an investment that many utilities are at present finding
necessary. These tools consist of on-line and off-line information systems allowing interactions
with the network infrastructure, with Service Users, with Service Providers and across the
management organization. They include service desks, network and service configuration
management data base, alarm and event management platforms, performance monitoring
systems, incident management and work order dispatch systems, dynamic route tracking
110
systems, security management and intrusion detection systems, etc. These IT platforms and
applications are described in section 10.7 further in the document.
Management tools employed in small and simple telecom networks are often trivial, homemade and may have marginal cost of maintenance. However, scaling up the network and
formalizing the processes necessitates more elaborate, complex tools which represent cost and
effort to deploy and to maintain.
Deployment Phasing and Migration plan An important task in transforming business plans
and policies into deployed solutions is the elaboration of a phased deployment plan describing
the step-by-step migration from the existing situation to the target solution. The deployment
plan may in this way be extended over some years depending upon a number of parameters:
Business plan requirement
User requirements anticipated in the Service Catalogue/ Roadmap (see next section),
Lifecycle of existing telecom assets on the network,
Investment plan and availability of funds,
Deployment capability and skills
Power delivery constraints and minimal disturbance planning
Validation and Feedback The process of deploying a telecom solution allows the practical
validation of the long term policy and its related strategic decisions. Parameters and factors
which have not been taken into account and hypotheses which turn out to be invalid are in this
manner identified. The feedback can be used to adjust the business plan and the strategic
orientations. The deployment process must use previously defined metrics (KPI) and devise
measurement capability to validate strategic decisions.
On the other hand, operational management teams can provide, through their processes
described further, valuable feedback used for identifying:
Tactical Adjustment EPU communication service requirements are not static and evolve in
time. Anticipating new service requirements is described in the next section on Service Offer,
however adjusting the capabilities of the network (or provisioning contract) in terms of
coverage, bandwidth and network in order to meet encountered requirements is to be treated
with a higher reactivity. System and process upgrades and optimizations are performed within
the perimeter of the approved yearly budget. This is generally carried out through setting
projects and following their deployment using the operational management or contractor
workforce. The scope may cover optimization, corrective action, upgrade/renewal of
equipment and firmware upgrade.
111
As for the restaurant menu, the communication service catalogue needs to be updated
according to encountered and anticipated changes in the EPUs applications. Typically,
SCADA applications are in most EPUs migrating from serial point-to-point communication
channels to Ethernet-based TCP/IP networking. This change, together with other applications
migration to Ethernet, is reducing considerably the requirement for serial point-to-point
communications, but increasing sharply the requirement for time-controlled wide area
Ethernet. The service catalogue must therefore be adjusted correspondingly, leading to the
deployment of adequate capabilities in terms of network, management tools and skills.
Service migration planning is often to be phased according to a number of different factors:
112
Upstream
Management
Customer / User
Enterprise Processes
Provider / Contractor
Initialize
Deliver
Get Paid
Provide Support
Assurance (As) Running the schemes set up previously in order to deliver the
communication services.
Support (Su) Providing a response to different enquiries and requests for service
including the maintenance services
Accounting (Ac) Determination and recovery of costs related to the delivery of telecom
services to different Utility-internal or external Service Users according to rules which are
set by the Upstream Management processes, monitoring the usage of communication
resources according to contracted conditions and recovering of network revenues to settle
the operational expenditure.
113
114
connectivity to modify its throughput, quality, interface type or routing. The change may be
feasible from the telecom providers Operation Support System or may require field
intervention. In both cases, user change management process must include time coordination
for performing the requested modifications, and in the latter case, also site access coordination
for the field intervention staff.
10.6.1.4 User Problem Handling (As)
The telecom provider must comprehend the user problem expressed through the service desk
generally by telephone, create a trouble ticket and manage it up to the resolution of the problem
and closing of the ticket. Proper management of user problems and an adequate method for
their tracking and reporting is essential in a formal user/provider relationship.
10.6.1.5 User QoS/SLA Management (As)
The telecom provider must monitor the QoS for the communication service and elaborate a
User Dashboard which allows quality reporting to the User/Customer. The measured quality is
compared to the contracted SLA and violations are managed. User SLA monitoring allows the
notification of users and fast resolution of quality degradations. Critical services such as
protection relay communications can in this way avoid major application-level incidents and
consequent contingencies.
10.6.1.6 Customer/User Entity Invoicing & Settlement (Ac)
Customer relationship management includes also the creation of invoices for each user entity
(e.g. SCADA, Protection, Surveillance, etc.) based on the usage data collected at the Service
Management level and on pre-established service pricing as defined in the Upstream
management process. Managing customer entity invoice enquiries is also part of this process.
117
118
10.6.3.11
Estimate cost of running the network infrastructure (Ac)
This operations management of the Utility telecom service provision must keep track of its
running costs. This will in turn allow the upstream management to define the service prices for
the service catalogue or to define the basis for cost repartition between internal user entities.
The cost estimation must include the running costs for the associated management IT tools.
120
121
Fulfillment
(Setting up)
User/Customer
Communication
Service
Assurance
(Delivery) Running
Support
(Enquiry Supervision
& Maintenance)
Service Enquiry Desk
(User Technical Support)
Incident Management
Network
Infrastructure &
Resources
External Providers/
Contractors
Disaster Management
Fig 10.12 - Current Operations Process for Utility Telecom Service Delivery and Provision
122
Accounting
(Metering, Cost
Repartition, Invoicing
& Settlement)
Customer/User Entity
Invoicing & Settlement
Service Policing
& Usage Metering
Introduction
123
10.7.2
Although the present chapter's focus is on tools covering the overall infrastructure, the service
and the user/provider interactions, and therefore beyond the vendor sub-system management, it
should be noted that the basic level of management tools presently employed in the majority of
EPUs remains the vendor-specific, dedicated Element and Network Management Systems.
These systems cover Fault, Configuration and Performance Management for a single-vendor
telecom subsystem in an in-depth and optimized manner. In particular, these systems allow an
adequately trained maintenance engineer to perform preventive and corrective maintenance of
fibre cores as well as board-level fault diagnostics, read and modify equipment configurations
and program end-to-end connections across the uniform sub-system with the required
protection mechanisms. Dedicated Element and Network Management Systems are widely
documented by telecom equipment suppliers. In this section we rather focus on tools beyond
the vendor sub-system management, covering the overall infrastructure, the service and the
user/provider interactions.
Fibre maintenance systems are specific SCADA systems based on OTDR measurements
performed on dedicated pilot fibres or in-traffic fibres in order to detect change of
characteristics or to localize damage. These systems allow a decrease of the fibre down-time
and hence improve the overall availability. They improve also the efficiency of the
maintenance process and work scheduling.
A particular category of network management systems which are increasingly employed in the
EPU operation-related environment is those which manage IP/MPLS and Ethernet packet
networks. These tools may be vendor-specific or more general monitoring platforms and allow
the following tasks in addition to managing elements fault and configuration:
A number of Japanese EPUs have made particular developments for their IP Network
Management [44]. The objective of these developments has been the achievement of smoother
operation and maintenance through an adapted graphical display of network states (VPNs and
Label Switched Paths, LSPs) enabling a visual perception of route information. Additionally,
the network management system can be operated with graphical operation windows and the
O&M staff can set up routers and L2 switches through wizards without wrong settings and
operations. This function requires only simple operations for end-to-end settings, and facilitates
the system expansion.
124
10.7.3
The Operation Support System (OSS) consists of on-line and off-line information systems and
corresponding applications supporting the operational management organization in its
interactions and in its visualization of the network infrastructure, current problem solving,
incident management and service monitoring tasks. Some particularly important functional
requirements for a telecom OSS in the EPU environment are listed below [43], [45]:
Management Function
Fault Supervision and
Problem Management
(Root Cause Analysis)
Scope
Service Monitoring
125
Figure 10.14 illustrates comprehensive management platform architecture for delivering these
OSS functions as envisaged by Japanese Utilities [43].
..
Line Configuration
Management
Configuration
Mgt. Support
Equipment Config.
Management
Line Design
Line Provisioning
Management
Line Planning
Task Support
New Task
Support
.. ..
..
New Functional
Component
Fault Statistics
Line Registration
Equipment
Registration
Line Planning
Registration
Contractor Control
Server Log
Collection & Analysis
VPN Route
Verification
Equipment Status
Monitoring & Analysis
Trouble-shooting
Procedure Recording
Scheduled Outage
Registration
Information Platform
Integrated
Fault
NMS
Determination
I/F
Element
Management
Systems
..
Thin Client
Workflow Engine
Service Quality
Monitoring
System Down
Determination
Task Functional
Component Group
Network
Management
Function
Online Help
Work Management
Support
Work Control
..
Electronic
Notice Board
Work Plan
..
Maintenance Task
Support
Trouble-shooting
Knowledge
Reliability Status
Management
Equipment Status
Management
Task Support
Function Group
Supervision Task Support
Service Status
Management
User
Authentication
Remote Switching
Control
GUI
I/F
EMS
Contacts
Service Impact
Determination
I/F
I/F
EMS
Proprietary
Alarm
Masking
Performance
Monitoring
Network
Management
Data Base
I/F
I/F
I/F
EMS
I/F
EMS
SNMP
Syslog
I/F
EMS
New
I/F
Fig 10.14 Outline of functional configuration of new Network Management System [43]
126
Information Platform
Functional components operate based on information from the integrated NMS and
network management DB. An information platform plays the role to accommodate this
information and functional components as if they are on a common bus, to provide a
mechanism for linking functional components through loose clustering, in which
functional components are independent of each other and separated for each coarse grain.
10.7.4
User, Service and Network Inventory are basic information system requirements governing the
EPU telecom facility and business. In service- and resource-related areas, the inventory
information is composed of the configuration data encompassing existing assets, allocation,
implementation, installation, configuration, activation and testing of specific resources and
services. The information may be employed to meet the requirements from different processes
to alleviate specific service or resource capacity shortfalls, availability concerns or failure
conditions.
The relative importance and necessity of different inventory related activities depend indeed
upon the scale and scope of the EPU telecom delivery model. Regardless of these delivery
model considerations however, the efficiency of the inventory system (and hence the speed and
efficiency of the enterprise process) lies primarily in the accessibility of data across different
sources and different data domains and the association of these inventory information domains.
Inventory and configuration information systems are examined here in more detail in support
of efficient implementation of EPU operational processes [46].
Beside interrelationship, another important aspect of inventory information is the quality of
data. Since inventory related activities are a combination of manual and (possibly) automated
tasks, the only way to maintain high level of data quality is to incorporate respective attention
into the enterprise processes. Data quality becomes extremely important when some process
127
automation is implemented at the operational level (e.g. automated service and resource
provisioning, alarm correlation, impact and root cause analysis, etc.).
Functionality of an inventory system typically includes the following features:
In practical implementations, these basic inventory system features are always part of a broader
context which needs to be defined before the implementation. Broader context is defined by the
business requirements at initial step of an inventory project lifecycle. Business requirements
are high level use cases which originate from enterprise processes and their corresponding
expected business results.
The importance and influence of the inventory system is illustrated below:
1. Automatic generation of User Notifications The EPU telecom service provider may
be required to generate and transmit automatically advanced e-mail notifications to
customers (or end users) of a possible service outage in case of planned maintenance
actions as part of its SLA. User notification is based on service impact analysis which
can only be performed if the service to network mapping is available in the inventory
information system and its precision depends on the degree of resolution available in
the inventory system.
2. Circuit provisioning across a multivendor network The EPU telecom service provider
needs to document the creation of new circuits across a multivendor network and to
issue the corresponding work orders. A path finder application containing decision
rules needs access to inventory information in order to perform its task.
3. Root Cause Analysis Problem management in the EPU requires the identification of
the root cause of an avalanche of alarms received from many network assets at different
levels of the infrastructure. Performing this task requires the knowledge of
configurations and the relationship between the alarm generating assets, which is
recorded in the inventory information system.
A comprehensive multilevel and multi domain inventory and configuration data base fulfils
these requirements. Data models differ for each implementation due to the types of needed
entities as well as the level of detail needed for each entity in the data model.
128
between service instances and resource instances, created as a result of the Resource
Configuration and Provisioning Management processes. Configuration and topology
information of each resource domain in the Inventory management System shall be kept
coherent with that in the resource types database (e.g. Element Manager Database). When
possible this shall be done by resource discovery applications or uploaded from dedicated
Element/Network Management Systems through update or reconciliation.
Resource Inventory is the basis for a number of important applications including spare modules
and warehouse management, capacity planning and change management and network
migration planning. These latter applications are used to discover and manage underutilized or
stranded resources including cable pairs, wiring and distribution panels, and other passive
resources.
Resource Inventory requirements also include those related to application software and
firmware releases and licenses related to telecom management or related to telecom service
delivery platforms which may (or not) be associated to more general IT asset management
information systems of the EPU.
Supplier/Partner Inventory
The supplier/partner inventory maintains records of all commercial arrangements with
supplier/partners, and any modifications to them. It also records all details of contacts with
suppliers/partners as well as commercial information, including details of supplier/partner
products and services, required to support Supplier/Partner Relationship Management. The
supplier/partner inventory is also responsible for maintaining the association between product
instances, service instances and resource instances. This inventory is particularly important for
EPUs utilizing connectivity provided by public telecom service providers where an association
between rented facilities to the EPU resource and configuration database is highly important.
130
11 COST CONSIDERATIONS
Cost assessment for different modes of telecom service provisioning can only be performed in
a specific manner for a particular EPU depending upon its existing resources in terms of
infrastructure, skills, service procurement opportunities, regulatory issues and service
requirements. Similarly, as it has been stipulated in the previous sections, many hybrid
provisioning schemes combining service procurement and in-house provisioning may be
adopted to fit optimally in any particular case:
In the present section, we only enumerate some important cost parameters that need to be taken
into account when performing such an assessment considering three principal modes of service
provisioning which are:
Build and operate
Build and contract services
Procure service
Asset ownership investment issues on the regulatory side
The costs associated with the provision of telecom services can widely be classified into
Capital Expenditure (CAPEX) and Operation Expenditure (OPEX). The former comprise the
costs associated to the implementation of a telecom network infrastructure, and the latter to the
running and maintaining of a telecom network or to the procurement of services. Regulatory
constraints in the deregulated environment generally render high CAPEX and low OPEX more
attractive because the expenditure on infrastructure may be declared to the regulator as an
improvement of the reliability and security of the power system and hence incorporated into
the pricing of the service. It should however be borne in mind that the same favourable
regulatory environment can end up to be an obstacle to the integration of non-operational
services (and/or commercial revenue generation through external customers) across the same
infrastructure.
Similarly, the rehabilitation of the operational telecommunication facilities can be eligible for
financing through international funds which indeed may not be used for procuring externally
provided services.
Perimeter of Communication Services and Requirements
When comparing costs and solutions, defining the perimeter of services and requirements is
often a tricky issue:
When provisioning through procured services is envisaged, utilities tend to limit their
requirements to the minimal operational services currently in use. However, when planning
to implement a dedicated infrastructure, utilities dimension the system not only for present
services but also for estimated new site extensions, new applications and estimated rise in
application bandwidth usage during the system lifetime.
131
Non-operational services are generally not included in procured service cost estimations
although these services result in some cost to the utility, usually for another department
such as Deregulated Business department, or alternatively these are absorbed by a
project or maintenance group so that the true costs are not known. These are most often
planned into the dimensioning of dedicated networks.
There are often auxiliary applications and services that would be implemented if
communications were available but rarely deployed when supplementary channels need to
be procured.
When assessing procured service costs, it is usually the commercially available service
nearest to the required quality which is taken into consideration. There is often an
important gap with respect to the actual requirement meaning that the utility accepts to
lower its operational expectations. The high cost of actually meeting the operational
constraints for atypical services (if possible at all) is rarely taken into account. In the
build scenarios, however, it is considered that the operational expectation cannot be
lowered as technical solutions for meeting the constraints generally exist.
It should also be noted that quality constraints may be related to the proper operation of a
system (e.g. time latency in Protection Relays), related to the respect of certain national
regulations or international recommendations (e.g. fault tolerance), or purely cultural,
related to the behaviour of the previously employed systems or how we thought they were
behaving (e.g. invulnerable)!
Capital Expenditure
The cost assessment for the telecom infrastructure needs to be complemented with information
concerning the expected useful lifetime of the system components. Optical cable can be
expected to have a lifetime of 15-25 years, an RF tower even longer, but modern
communication equipment may need to be replaced due to unavailability of spare parts or
incompatibility with (or performance limitation of) more recent system components. The
addition of varying life factors complicates the cost assessment even more when we consider a
procured service counterpart: how to estimate the cost of a currently available commercial
service in 15-25 years time or the cost of adapting the operational applications to whatever
telecom service is available in future?
Moreover, it should be noted that the cost of optical cable, RF towers and electronic
communication equipment does not cover all the CAPEX of a Build project. Some other
important expenditure items often under-estimated are as follows:
Installation Wiring, Cable Trays, Distribution Panels,
Licenses and Right of ways
Power Supply, DC Distribution, Batteries, Back-up Generators, Fuel Tanks
Cubicles and Environment Conditioning, EMC Protection,
Management Facilities
Application interfaces and converters
The opportunity cost of outage times where critical assets such as transmission lines
need to be taken out of service for the installation of OPGW for example
Spares
Many of these items still need to be provided in a Procured Service provisioning scheme:
Customer Premise Equipment in many public telecom operators do not fulfil power system
interface requirements and is not conditioned for installation in the substation environment,
132
necessitating cost provision for interface converters, multiplexers, routers and switches as well
as cubicles, DC distribution (or AC UPS for Customer Premises Equipment) and installation
items in an assessment. Moreover, if the providers scope ends outside the perimeter of the
electrical substation (e.g. for safety and security reasons), one may also need to add shelters
and links to the substation interfacing points. Alternatively the Service Provider will need to
include the cost of gaining and maintaining site access accreditation where he does need to
access electrical substations.
Operation Expenditure, Management and Operation Support
133
A cost item which is so often neglected when performing cost assessments is the cost of
maintaining in-house skills and expertise. A first-level superficial analysis may consider that
contracting of services (management, maintenance, full procurement of bulk capacity or
process connectivity) leads to important cost saving in that the utility no longer needs to
maintain in-house skills and skilled workforce. However, experience shows that losing skills in
the utility generally leads to the build-up of a provision monopoly for one contractor (e.g.
Telecom Service Provider), rise of service cost and degradation of service, due to the inability
of the Utility to change its Service Provider (all the knowledge of the service residing with the
external contractor to be replaced).
In some other circumstances, outsourcing of services is not for saving on skilled workforce, but
a consequence of lack of skilled workforce or the inability to capture or to maintain skilled
workforce (e.g. due to non-attractive salaries).
Finally, it should also be noted that maintaining in-house telecom skills is costly and may even
be more costly if the maintained skills are only to be used for specifying and supervising
purposes. An adequate provision of operation cost is to be allocated for training, transfer of
knowledge between generations and acquisition of new skills facing technological migration.
Risk related costs
Cost assessment of telecom services often lacks consideration of the Risk parameter even if
this is precisely the reason to envisage anything else than the normal operator service as
described throughout this document.
A Cost versus Risk assessment must examine the Liability chain from the end-user to the
end supplier and contractor. The level of liability that any external telecom supplier is willing
to accept is completely out-of-proportion with the risk that the utility may assume if the
telecom service were not to be operational at critical times. It is often recommended to set the
liability of the telecom operator as high as possible (despite the impact on the service price),
not to cover the consequences, but to assure responsiveness.
The cost of non-performance of the telecom service comprises the cost of possible damage to
EPU assets, cost related to sanctions and penalties that the EPU may have to face, as well as
the loss of image for lack of performance and its associated costs in the competitive market
environment.
Cost Comparison and Selection of the solution
In order to properly analyse different solutions which have different CAPEX and OPEX cost
considerations, it is useful to use appropriate economic models that enable valid cost
comparisons to be made. One such standard economic technique is known as Net Present
Value Analysis. In principle this technique converts a stream of future payments (OPEX costs
such as continuing telecom service fees, maintenance expenses, ongoing management
expenses, etc.) into the present (taking into account the time value of money caused by
inflation and opportunity costs) so that they can be added to CAPEX costs. This results in the
ability to directly compare different solutions with different mixes of CAPEX and OPEX costs.
It is also possible to do a sensitivity analysis for different scenarios, such as changes to the
inflation rate during the lifetime of the telecom service delivery. A simple google will find
ample references to this technique which can be easily implemented using a spreadsheet
approach, or by using Discount Cashflow tables.
134
Some of the main attributes of this Smart Grid are given below. Their fulfilment relies
heavily upon the existence of secure, reliable and fast communications with wide coverage of
the concerned sites and devices:
a. Self Healing - The prompt reaction of the power network to changes, through a highly
coordinated automation system, and network-aware protection schemes for rapid
detection of faults and power restoration. These communication-intensive applications
and their service requirements have already been described in the section 3.1.
135
b. Enhanced Visibility and Control - Increased security of the power delivery system
through better visibility of power flow and the network state across the interconnected,
multi-actor, competitive market. These applications covering the visibility of the
system from the remote control platforms together with new ways to organize,
synthesize, dispatch and display large amounts of collected information are already
illustrated in section 3.2.
c. Empower Consumers - Empowering the consumer means incorporating consumer
equipment and behaviour into the design and operation of the grid. This implicates
reinforced Demand Response and Load Control through information exchange
with the energy consumer and potentially the electrical appliances in consumer
premises, allowing to regulate consumption and to absorb peaks. The Advanced
Metering Infrastructure (AMI) and Smart Metering programs constitute the basis for
this reinforced relationship requiring bi-directional real-time communication access
from the Service Provider to the individual domestic and industrial consumers.
d. Tolerant to physical and cyber-security attacks New utility security monitoring
solutions and their communication requirement have been discussed in chapter 5.
Security also becomes an important attribute of the communication service for utility
applications as described in section 6.6. Moreover, disaster recovery solutions such as
geographically distinct back-up control platforms, fast deployment and mobile facilities
described in chapter 7, are often part of the mitigation plan of utilities towards physical
attacks raising the issue of routing of communications to facilities, sites and staff across
the network.
e. Accommodate Green Power - The secure integration of the considerably increased
dispersed power generation, mainly large wind farms, but also energy producing
consumers (solar, wind, etc.) implicates incorporation of statistical data and
meteorological forecasts into the power dispatch program, but also reliable two-way
communications with dispersed generators constituting virtual power plants. In case of
a wind or solar farm, the local communications of the power plant are extended
through a large perimeter, sometimes hardly accessible, covering an array of tens of
individual generators.
f. Optimize Assets through Monitoring Asset and environment monitoring for an
increased lifecycle of assets without compromising system reliability has been
discussed in chapter 4.
As it can be observed, the different components enabling the smartness of the grid have
already been covered in terms of communication services in the present document. What still
remains unclear at the time of preparation of this document is the extent of deployment,
coverage and associated investment of power utilities for each of these attributes. This may
have considerable impact on the overall provisioning scheme of telecom services in the EPU.
One such scale-setting application is the Advanced Metering Infrastructure for access to
consumers. Depending on the adopted applications and their communication requirements, the
distribution utility may be driven into a procured telecom service or into building of an
extensive dedicated telecom network.
The concept of microgrids, autonomous islands comprising generation and consumption, will
redefine the Energy Management architecture and its consequent communication service
requirements, pushing the concepts of deregulated energy market down to the community
level.
136
Finally the management of devices may also be considerably modified with the change of scale
in the number of smart metering devices to be monitored, evolving from a power utility type
monitoring system into concepts and platforms that mainstream telecom operators have long
employed for payphones, ADSL and cable modems.
In a great number of countries the power sector has already undergone or is undergoing
important changes moving from a state-owned vertically integrated utility with a public
mission to a competitive market situation with investor-owned companies. A natural trend has
therefore been a great focus on Return on Investment with consequences on all activities which
do not contribute directly and immediately on the creation of added value. This unique
direction of attention is gradually being counter-balanced to some extent by costly incidents,
regulatory bodies directives, and liabilities: Return on Investment is being complemented by
the Risk of non-Investment.
Competitive companies will have to invest extensively in such domains as cyber-security,
surveillance and physical protection of sites and installations, primary asset condition
monitoring etc. in order to meet regulatory requirements (e.g. NERC). The safety of the
maintenance workforce is also getting renewed attention through innovative information tools
as already described in order to avoid accidents.
Continuity of supply is increasingly a requirement that is sanctioned through important
financial penalties encouraging the power company to invest in its energy management, grid
management and asset management facilities and the efficiency of its maintenance. Reliable
telecommunications in this context shall benefit as an enabler infrastructure, requiring
important investments, not necessarily to create a Return on Investment but to avoid important
consequential losses.
On the other hand, regulatory bodies encourage investment in operation-related
communications by accepting the expenditure as necessary for the security of power system
(and hence reflected into the total cost of power). This approach may impact considerably the
telecom strategies of the EPU.
Workforce and Expertise issues
Power Utility workforce is changing in distribution, skills and behaviour requiring revised
ways of operating.
Ageing Workforce Many European countries are facing workforce aging due to
demographic reasons with waves of retirement creating a loss of legacy expertise while
legacy systems and technologies still remain in operation
Shrinking Workforce Competitive strategies in companies have led to reduced workforce
and to externalization reducing their ability to conserve skills and expertise on a wide scale.
Accelerated Turnover Younger generation technical staff, increasingly do not remain a
long time in one activity, making it difficult to develop a wide knowledge and experience,
in particular where multiple disciplines are concerned (e.g. Telecoms and Protection
Relaying).
137
Based on these statements two important directions of focus and investment can be identified
for power utilities in the coming years to mitigate workforce and expertise issues:
Training and simulation tools for technical workforce
Embedded intelligence in operation & maintenance tools This includes enhanced network
management and situation awareness systems and data-rich maintenance tools. Remote
assistance of field workers through enriched information exchange terminals is described in
4.6 and [12]). Some utilities are investigating into Expanded Reality for their maintenance
workforce.
138
f. Mobile and fixed services of premium quality/reliability may be created to serve specific
blue-light applications. This can be performed through sharing of costs and
infrastructures (frequency band, cable, workforce, etc.) among different users creating
specific operators as an alternative to dedicated infrastructures. The case study in Portugal
described in Appendix A2 is an example of this trend.
g. Satellite communications for specific applications may lead to the creation of specific
service contracts and Service Providers as illustrated in the Brazilian case study in
Appendix A3.
Optical technology progress in recent years has rendered feasible and cost-effective in the EPU
context, a number of solutions which were previously too complex, too costly or not
sufficiently reliable. Wavelength Division Multiplexing once reserved for large mainstream
networks has made its way into dedicated EPU networks. In particular, Coarse Wavelength
Division Multiplexing (CWDM) is now a very accessible technology for federating different
139
services (or groups of services) and communication modes (refer to section 9.3.2). Wavelength
multiplexing allows a new dimension in service integration, keeping legacy TDM systems and
Protection circuits on separate wavelengths from new core data networks. Double transport
planes (packet and circuit) available in many equipment, complement the wavelength
multiplex for building a compact hybrid transport for dedicated networks.
Optical networking which brought the SDH/Sonet in the late 80s and early 90s is now moving
into the Optical Transport Network (ITU-T OTN) standardized by ITU-T (G.709 revised Dec
2009). Operators are deploying OTN equipment at backbone, metro and access levels to
enhance packet networks. Inclusion of WDM and enhanced OAM gives the OTN significant
advantages over SDH and a combination of OTN and Ethernet (carrying IP) is proving to be a
very attractive industry solution for converged networks [49].
ITU-T OTN provides considerable flexibility through implementing an Optical Control Plane
that automates the allocation of wavelengths to communications. Reconfigurable Optical Add
Drop Multiplexer (ROADM) allocates routes across a number of different wavelengths
automatically at the time of route configuration. This constitutes the basis of Optical Transport
Networks delivering packets as well as TDM circuits (Constant Bit Rate traffic CBR) in a
scalable manner combining cost-effective and flexible bandwidth usage with low deterministic
delay and restoration times available through SDH [50].
On the other hand, EPUs need to consider the presently available experience data on the
lifecycle issues of different fibre cable technologies which were previously only theoretical or
based on little experience. If optical cable lifecycle may prove to be less optimistic than that
predicted, then substantial extra investment may need to be made available in the years to
come. Moreover, the traffic expectation continuously rising, the replacement cycle of the
cables is shortened, requiring supplementary disturbance and down-time for the associated
power transmission lines.
Carrier Ethernet Technologies
For a number of years, mainstream telecom industry has introduced IP/MPLS as the scalable
solution for the backbone core of all IP or Ethernet-based networks. Many Utilities, in
particular those carrying a great amount of administrative (corporate) traffic or commercial UTelco services across their network have deployed an MPLS core composed of a few nodes
interconnected through high bandwidth Ethernet links. However, it is argued at present that
MPLS presents too much complexity for a pure carrier function and does not provide necessary
guarantees for time-sensitive traffic. What is required in EPU networks is to deliver Carrier
Ethernet services (E-line, E-LAN) in a scalable manner without too much complexity while
keeping SDH-like QoS. This will allow simple migration from the present stage Ethernet
VLANs over SDH into a more scalable solution if and when necessary.
Provider Backbone Bridge (PBB-TE) defined in IEEE 802.1Qay and Transmission Profile
MPLS (MPLS-TP) being defined by ITU-T and IETF follow this same objective of providing a
simplified and time-constrained solution for scalable carrier Ethernet provision. This is a
technological step which will certainly interest EPU telecom networks in future [51].
140
Another important technology change which will impact EPU networks is the one related to
the management of network infrastructures. Management platforms are growing from Element
and Network Management to Service and Business Management.
Moreover, from a previous situation of monolithic management platforms at high cost and
maintenance effort which were often out of reach for EPU networks, we are getting to
interconnected multi-vendor modular management platforms, allowing the association of event
management, incident management, performance monitoring, network inventory, etc. from
different vendors.
TM Forum is leading to common information models and standard application interfaces
allowing gradual deployment of management infrastructures using Service Oriented
Architectures (SOA). This will allow the EPU to tailor and to build up its management system
according to its needs over many years.
The EPU environment, with its specific security and reliability requirements governed by
specific regulations, is unlikely to adopt public cloud services in an extensive manner, at least
in the coming five to ten years.
However, driven by standardization and integration of IT and industrial processes, not
forgetting the green IT demand, the adoption of private clouds in EPUs, is already in place
and will grow gradually. This will concern both IaaS (e.g. servers and storage) and PaaS (e.g.
databases and web-service platforms).
Assuming that cloud computing in EPUs will be based on a private cloud principle, the impact
in telecom services will be more on the way they are built, managed and used (i.e. migration to
the cloud) than the way the services are accessed. In this way, the direct impact that can be
identified at this stage would be extra bandwidth and QoS requirements resulting from the
transfer of servers and storage from distributed technical rooms to the main datacenters where
the cloud will be housed.
An important potential application and an issue to be considered is disaster recovery. At
present, typical disaster recovery implementations require often a lot of manual procedures,
141
activation time, frequent tests and revisions for business continuity plans, and low efficiency of
assets usage due to cold or hot standby solutions. The adoption of cloud architectures can
change the way we look at disaster recovery solutions, enabling the adoption of semi
distributed IT/Telco architectures where the cloud and the processing capabilities are spread
along the main and the backup datacenters and used in an active/active basis. This can be
achieved through long distance storage interconnection, metro-clusters, real time long distance
database synchronization, balanced Internet access, server farms, etc. This possible evolution
may render disaster recovery simpler, more reliable, more efficient and (theoretically) cheaper,
but will also need communications in line with those requirements which will necessitate
investment in (or usage of) more resources (e.g. fibers, wavelengths, MPLS, etc.).
At the time of preparation of this document, to our knowledge, EPUs are not using cloud
computing but investigating and in some cases only starting to plan to implement for corporate
organizational information systems.
On the migration into a Public Cloud, there are still many concerns and questions to be
answered relating mainly to reliability and security. Many of the questions resemble those
already discussed for procured telecom services:
Service availability and capacity How can it be guaranteed that the Service
Providers will meet not only the present needs but also the future ones as they
appear? How can the EPU internal capacity management processes be matched with
those of the providers? And what about customization, will the providers be as
flexible to the EPU as the EPU has to be for its internal customers?
Failure remediation and disaster recovery Are providers ready to offer failure
remediation schemes and guarantees in line with those already existing in house?
Is the EPU current typical SLA model applicable and enough? Are the providers
able to offer disaster recovery with the same SLA commitments as those existing at
present for internal customers? At what price?
Transparency Once we ask for a certain service in the cloud to a certain provider
we are getting that service from the cloud. What is the cloud? Who is behind the
cloud? Are there subcontractors? Who really will operate our services? How do we
meet our requirements along the chain?
Ownership and security: EPU infra-structures/applications/data will go out of its
physical and logical perimeter. Who really owns the information, EPU or its
Service Provider(s)? Can we guarantee information privacy and confidentiality?
Will other providers/partners/customers with whom an EPU has confidential
agreements accept such a model? How do we guarantee the same security
requirements and commitments that we have in place today? How can we define an
effective security policy if we lose/reduce the capability of applying security
policies enforcement and auditing? Will we be able to audit our cloud providers?
Isnt that a huge and perhaps ineffective task? In case of security breach how do we
track it and who is responsible? How will we put in place risk management
processes in this new reality?
Maturity: Nowadays even corporate services like email, portals or file transfer
can play a significant role in the critical processes of EPUs. Are providers mature
enough to support critical services or will the support be on some kind of best
effort basis?
142
Regulatory issues: How will the cloud model meet the regulatory requirements of
EPUs? What are the legal issues if we are under a public concession contract?
ROI: The model is quite complex and does not give readily a real idea of costs and
benefits. Little information is at present available on the return on investment. EPU
stilll have to demonstrate the interest of the financial model for their organizations.
Regarding the communication services and network, the first issue will be Internet
access that must be similar to the actual datacenter bandwidth access (Gigabit?) and
implicitly communications quality of service and security requirements. Are ISPs
able to provide service levels as we have in house today? The access to the cloud
should be given everywhere (fixed or mobile access)? Are ISPs prepared to commit
the necessary SLAs?
Further information on cloud computing as well as available platforms, providers and different
services may be found in [52] and [53].
143
APPENDICES
A1.
Hungary
The Hungarian TSO, MAVIR has a digital Operational Telephone Network since 1992,
connecting the telephone exchanges of NDC, reserve NDC and high-voltage substations (the
present number 2+26). Although this network is independent of the other used telephone
networks- as the nationwide administrative telephone network of the EPU-s and the public
telephone network-, these exchanges provide the concentrator task (collecting the different type
of lines to the dispatching desk) also.
At first the connections among the exchanges operated through PLC and microwave
connections, later on optical network. Since 1997 the network used a double tree structure (6
substation exchanges on higher level than the other), the connections were 2 Mbps trunk lines
and/or 4 wire E&M).
Since 2007 this telephone network operates as one of the districts of the nationwide
administrative telephone network of the EPUs.
In 2009 began the IP migration of this operational telephone network. In case of NDC, back-up
NDC, and 4 selected substations LAN connections were developed instead of the earlier
mentioned double tree structure.
These LAN connections are operating with double routes (through NgSDH and SDH network
on 2 Mbps and 10 Mbps lines.
The hardware changes are: routers, Main Control Server/Enhanced Survivor Server at
NDC/reserve NDC (between these MPLS connection), at the 4 substations gate-ways Local
Survivor Processors and port cards to analogue trunks and subscribers.
The IP telephones at the substations operate from the Main Server or, in case of necessity,
could work from the Local Server also. This started change will continue at the other
substations.
Reasons for IP migration and expected results:
144
Portugal
The Portuguese TSO, REN has a national wide TDM Operational Telephone Network that
provides voice and data services to Electrical and Gas dispatch centres (main and redundant),
electric and gas substations and corporative building. Additionally this telephone network is
interconnected with other telephone networks from electrical companies and with the public
telephone network.
In 2006 it was decided to start the process of evolving the telephone network into IP telephony
since it was implemented a national wide MPLS/IP network that could provide IP VPNs for
interconnection of substations and dispatch centres into the same WAN/LAN.
This project had the following guidelines and objectives:
Analyze the differences between different VoIP implementations from different vendors. In
this case it has tested 3 different vendors.
Compare the implementation of VoIP systems in office and substation sites in terms of
usability, redundancy and operation.
Implementation of a call centre in order to support the Telecom and IT helpdesk activities.
The following sites were selected:
2 Office building (400 users).
6 Substations with 50 user extensions medium each.
IP Call Center implemented in the central site with 20 IP users distributed across 4 sites
10 remote wind generation sites (8 extensions each, 6 analog modems and 2 IP phones).
For installation it was adopted the following requisites and project general options:
Each IP gateway is registered in central redundant system.
Each site has a survival mode that allows internal services to be provided even without the
central management VoIP system.
For mobile service inside the substations the choice was DECT technology since WiFi is
more expensive and not adapted for industrial environment use in the substations.
During the implementation several difficulties were found, in resume the following aspects can
synthesize the problems and difficulties detected during the implementation and testing of the
VoIP systems:
As several models of modem and fax equipment are used across the network, in some cases
problems were detected in communication (pass trough/relay configuration).
The IT network is composed by several elements, namely firewalls, DHCP servers, DNS
servers, routers, switches, etc., when there is interdependency between the VoIP network
and all these components, as the complexity increases.
Mobile communication inside the substation with some problems, namely for hand-over
of calls between cells (1 vendor).
Difficulty and in some cases unavailability to interconnect several services: BRI
connections, analog connections (FXS, FXO and E&M).
It was detected that certain configurations implemented in the central management system
were not reflected in the network equipment, differences between local and central
management system (e.g. gateways registered though MGCP into the VoIP management
system).
145
Netherland
The Dutch distribution network company, Stedin, managing gas and electricity networks
composed of around 200 communication nodes (10-150kV electricity and gas stations) and a
total of 650 substations to cover, is deploying an IP telephony network for operational and
operation-support voice communications.
The company uses public mobile phone (GSM) services which are however unavailable most
of the time in case of emergency. Pagers (better availability than GSM) as well as TETRA
mobile system of the Public Safety services (Police, Fire brigade, Ambulance Services etc.) are
also used.
In case of electricity failure and lack of public telephone services, the technical staff must
travel to the nearest station and use the companys dedicated telephone facilities.
The dedicated telephone system which is migrating to IP can operate without connection to the
public network, hence allowing communication between HV stations and Control Centres. The
network uses redundant communication servers placed in the two geographically distinct
control centres.
A number of process efficiency features are sought from the new IP telephone network,
including:
-
A new numbering plan based upon geographical zones, electricity or gas, High Voltage
or Medium Voltage, functional departments, etc.
An outstanding issue to be solved at present is the connection of legacy voice modems to the IP
telephony network.
146
Australia
Snowy Hydro Limited is a leading provider of peak, renewable electricity to the National
Electricity Market in Australia. It owns and operates the 3800 megawatt (MW) Snowy
Mountains Scheme, an integrated water and hydro-electric power system located in Australia's
Southern Alps comprising sixteen major dams, seven major power stations, 145km of
interconnected tunnels and 80km of aqueducts, as well as two gas-fired power stations totalling
620MW, both located in the southern Australian state of Victoria. There are also support
offices including in Sydney and in Cooma, resulting in the need to provide voice services to
locations spread over a distance of 1000km.
In 2004 and 2005 a new voice system was implemented which carried both operational and
corporate voice services and covered all sites belonging to Snowy Hydro Limited. The system
chosen comprised a combination of IP enabled hybrid telephone exchanges interconnected
with server based soft switches. This gained the benefits of IP based telephony while still
preserving the resilience required for operational and safety telephony. Each site is
interconnected by multiple routes comprising both IP (as a separate VRF on a QOS enabled
MPLS network) and TDM (where available) or multiple IP routes where TDM is unavailable.
The use of the hybrid approach enabled the condensation of two separate telephone networks
(operational and corporate) into one system while not compromising resilience, but gaining the
advantages of the IP voice services for certain applications. Office workers were provided with
IP phones giving them mobility to work at different desk locations (eliminating costly moves
and changes) or at different geographic locations using the same contact phone number. The
main and backup control centres were set up with multi-appearance IP phones. This has
enabled the use of different control centres without the need to notify other parties to use
different contact numbers. TDM phones were also provided to cater for a failure of the IP
network. Safety phones spread throughout the power stations were kept as analogue extensions
but their resilience against link failures was enhanced by the provision of IP routes between
locations that were additional to the TDM routes. The provision of IP routes also provided a
means for the carriage of voice for free on data networks to achieve substantial cost savings
through toll bypass.
The main issues (additional to change control issues) that had to be solved during the
implementation of the system were:
large range of ports on the firewalls. E.g. too tight a policy can lead to one way voice
calls.
Echo issues often related to duplex mismatches in the data network or were associated
with the use of specialised phones such as conference phones, cordless phones and
headsets. Even room layouts can affect echo issues. Careful attention to duplexing and
echo cancellation settings was required.
Implementing voice quality monitoring systems was important so that the diagnostic
information was available to resolve internal network issues as well as to substantiate
issues with Service Providers.
Over the last five years of operation, the system has proved to be extremely resilient. The main
issues that have occurred at times have been caused on routes running through Service
Provider networks as follows:
148
A2.
EDP (Portugal) employs private mobile radio systems for voice and MV SCADA applications.
These radio systems comprise:
Analog VHF networks used for voice and data
Analog mobile voice used only when public cellular (GSM) is down
Analog data networks with better availability than GPRS (99% compared to 95% for
GPRS)
However, EDP needs to find alternatives mainly due to the pressure from the regulator to
abandon the VHF band and the desire to migrate to standard TCP/IP based SCADA protocols
such as IEC 60870-5-104.
In this context EDP signed a protocol with the Portuguese Government for setting up a pilot
project using SIRESP, the governments TETRA Network for Emergency and Security Forces.
The objectives and goals of EDP for the pilot project were to test the networks capability to
support TCP/IP SCADA (IEC 60870-5-104 light version), coverage and ease of use of voice
services and network availability.
SIRESP (Sistema Integrado de Redes de Emergncia e Segurana de Portugal) is the
Portuguese Government TETRA Network for Emergency and Security Forces serving 55 000
users
It provides a very dense coverage of the national territory through:
532 BTS + 2 mobile BTS
4 switches (+ 2 in the islands + 1 for tests)
66 dispatch centers
Avaliability (%)
Operacional
> 99.9
Coverage(%)
> 95
Rural
> 90
> 80
Buldings - Rural
> 50
Same cell
< 500
Switching delay
< 100
Rejected calls
< 2.5
Dropped calls
< 0.5
Capacity (%)
A 2 Mbit/s Ethernet line was installed between EDP SCADA SYSTEM and SIRESP to
guarantee connectivity between the two networks (for voice and data)
The SIRESP network provides a fixed IP address to each unit and the scada system command
center.
Voice Service:
Good coverage (even in places where no GSM operator exists);
Fast call setup (virtually instantaneous);
Good audio quality;
Users prefer handheld to vehicular mobile terminals (they prefer to carry communication
terminal with them where they go, e.g. up the poles, lines, etc) ;
Good network availability During a tropical windstorm (up to 220 km/h) in the west
region of Portugal all GSM operators went down (in the region). The SIRESP network
provided EDP the only means of communication in the region (in addition to EDP VHF
network);
SCADA (data) Service:
Reduced availability of incumbent 2 Mbit/s circuit between SIRESP switch and EDP
SCADA system. We were not authorized to use our fibre optic network;
High latency packet transmission time (from 700 ms to 1400 ms). We had to adjust
protocols timers to cope the high latency. Also, SIRESP network changed from 1 fixed
timeslot to 4 dynamic timeslots assigned to data, per site;
Long drops without recovery - Radio attempts to re-establish the connection are
unsuccessful (SN_ACTIVATE_PDP_CONTEXT_DEMAND ) -> local reset is needed.
This is the main reason for the ongoing pilot (only ends when this problem is solved);
Local Site Mode -> Useful for voice but not for data. Radios are locked to a single BTS
(there is no alternative if it losses connection with switch).
Constraints
SIRESP does not support terminal maintenance;
SIRESP helpdesk is not familiar with data issues;
CAPEX: Tetra terminals are expensive (vs GSM);
OPEX: Monthly fee is 10 times higher than GSM (is the premium service worth it?)
150
A3.
Satellite communications constitute a valuable option for specialist utility applications where
other services, either through deployed facilities or procured from telecom operators, may not
be accessible, economically feasible or sufficiently disaster-proof. Some typical situations are:
Black start-up of power installations When power supply fails, then telecom facilities
are lost after a time duration which depends upon the battery and standby generator
autonomy of the telecom infrastructure. Moreover, in disaster situations, telecom facilities
can suffer destruction in the same way as the power system facilities.
Access to remote facilities Power generation installations (hydro plants, off-shore wind
farms, etc.) may not be accessible through other deployed solutions in a cost-effective
manner or through procured communications.
Fast deployment facilities Satellite communication terminals can be integrated into fast
deployment mobile units and transported when necessary to temporary work sites as
illustrated in the Brazilian example presented in this section [m].
The usage of satellite communications for SCADA and monitoring applications has so far been
limited primarily because VSAT services, while being autonomous and reliable, are expensive
in comparison to other technology options, and scaling a solution to support the SCADA
platform for large utilities is cost prohibitive.
Consumer-based Broadband Service
VSAT
Consumer
based satellite
systems
GPRS
Traffic
These new satellite IP services may support a wide range of monitoring and command
applications where latency issues are not critical. They allow the benefits of satellite autonomy
and ease of installation to be extended to a wider range of applications, and enable a much
wider monitoring base, something which will be particularly beneficial in supporting the
requirements of future smart grid applications. In addition, the satellite platform also offers the
151
potential to support services and applications which require higher data rates such as video
applications and CCTV monitoring; services which on a widespread basis may not have been
cost effective in the past.
Consumer broadband services offered by satellite Service Providers are generally asymmetric,
with the downlink being five or six times larger than the uplink although other up/down link
combinations should be feasible according to application requirements.
An example of the standard hardware for this type of system includes a satellite modem, the
iLNB and an 80cm satellite antenna. The equipment is generally compact, easy to interface
and, because the consumer equipment is designed for self install, not costly to deploy. EMC
hardening may however be necessary for installation in severe EPU environment.
Services of this type require a high speed link between the hub station at the satellite operators
location and the IP backbone. This link may be provisioned through an ISP. Alternatively, a
second satellite link back to the EPU control platform can be provided, again using the low
cost service. Such configurations depend on user-specific requirements for latency, network
independence and cost.
Fast Deployment Mobile Units Brazil
Tower collapse incidents are particularly harsh situations requiring complex logistics and
intensive voice and data communications for moving workforce, materials and machines to
site, deploying a support base, and performing works to restore the line. The unavailability of
transmission services is heavily penalized by the power system operating organization and so
the company is strongly motivated to reduce the average line restoration time through
improved communications. In many cases, public cellular phone service is not available or
proves to be very costly as the usage is predominantly long-distance
Furnas Centrais Eletricas, a major Brazilian EPU, has developed container-based Telecom
Mobile Units (TMU) for facing transmission line tower collapse scenarios [54]. The TMU
provides local wireless voice and data and uses a satellite service to connect these remote
facilities to the companys core infrastructure. The container-based unit is of compact design
allowing easy transport and positioning, even when installed on irregular ground without very
rigorous levelling.
Furnas has negotiated an annual satellite service contract covering 30 days of usage per year
for a fixed lump sum. Each time that service is required, the Service Provider must make the
link available within few hours of the service request. This delay is compatible with the time
required to expedite the TMU to the incident site. Usage exceeding 30 days per year is billed at
a contractual daily cost.
The cost of the satellite service (including satellite equipment rental) in the present contract for
a 256kbps link is reported to be approximately 15.000 per year, with a cost of 350 per extra
day of use and the quality of service reported to be satisfactory.
152
153
A4.
During the months of August, September and October of 2005, the Gulf Coast of the United
States and both coasts of Florida were belted by a series of three devastating hurricanes:
Katrina, Rita and Wilma. Katrina put much of New Orleans under water; trees were flattened
as well as homes, public buildings and communications towers by Category 4 winds. Loss of
life from this storm was tragic, and unusually high. Rescue and restoration effort were severely
hampered by the loss of public fix switched network (PSTN) and wireless carrier
communications.
Still recovering, Gulf States were again pelted by Hurricane Rita just weeks later and then by
Wilma in October. While these storms were not quite as severe as Katrina, the rapid succession
tested the limits of overstressed public safety and utility repair personnel. It became clear as
response efforts continued that the recovery of nearly all other infrastructures was dependent
on electric power restoration. Moreover, all people have a vital need for safe drinking water
and public health demands reliable wastewater facilities. Within these industries, as always,
coordination of repair crews, rapid restoration and interaction with public safety personnel
depends upon working communications channels.
Communities of the Gulf Coast and Florida are served by a wide variety of electric, gas and
water utilities, including many small regional companies and municipal power/water
authorities whose service territories lie within those of larger entities. These include large,
investor-owned electric power companies, electric generating companies and natural gas
utilities.
2. Survey Questions and Results
The Utilities Telecom Council posed five basic questions to utilities in hard-hit areas to
determine how their various communications systems behaved during and after the hurricanes,
and analyzed the resulting data. Learnings and adequate counter-measures which could be
proposed to EPUs world are presented below.
154
Most utilities, regardless of service territory size or proximity to the centers of the storms,
reported that their communications systems behaved well during the hurricanes. This is in stark
contrast to the public switched network (PSTN) in the region and wireless carriers, who
suffered extensive loss of service and slow recovery time. The comparison points to the fact
that communications systems, if built extremely well, can withstand the intense wind and/or
flooding associated with these events; however, unlike public networks, EPU systems
redundancies and robustness can be limited in size and scope, since they are designed and
constructed to meet the specialized needs of a single entity or group of companies. Such
construction would be cost-prohibitive for a commercial system designed to serve millions of
the general public. Thus, in spite of the growth of various commercial communications
networks, there will be a continued need for EPUs to maintain their own private
communications networks for mission-critical functions, including backbone networks.
Most companies that suffered damage to their equipment and networks fixed it very quickly,
generally within 24-48 hours following the passing of the hurricane. Such repairs were carried
out according to detailed emergency plans, since communications networks are vital to power
restoration and infrastructure recovery efforts being undertaken area-wide.
Utilities reported that cell and fix networks were down, leased lines and satellite
communications suffered damages. One small water district authority depending entirely on
cellular communications, suffered serious communications difficulties.
The utilities that had the greatest communications difficulties were the smaller cooperatives
and municipal authorities.
Lack of interoperability among utilities and between utilities and public safety remains a
serious problem that hampers restoration efforts.
Hurricane rescue and restoration efforts provide an example of the benefit that could be reaped
from allocating a small amount of dedicated spectrum for EPUs, with systems to be built using
an open architecture and made available to all emergency participants as needed for disaster
recovery.
2.2 Private Mobile Radio, Backbone Survivor
The performances of private mobile radio (PMR) systems were very high during and after the
storms. Unlike most commercial wireless systems, these networks are built specifically to
weather such disasters and to continue to operate in extended power outages, in support of
restoration crews carrying out extremely hazardous duties. However, the superior performance
of these individual systems offset by the lack of interoperability between systems and the lack
of dedicated spectrum to share with other utilities and public safety.
PMR is at this time the most critical tool of critical infrastructure communications in
emergency situations. It provides for necessary mobility and quality of service as crews travel
throughout the damaged service territory. In many cases during events such as Katrina, it
provides the only means of wireless communications during the first critical days after storm
impact.
The overall performance of private utility communications systems during the catastrophic
storms, in comparison with consumer networks, reinforces the industry position that private
systems must be maintained, and encouraged for emergency response.
155
Due to the emphasis upon reliability for utility operations, it is clear for the foreseeable future
that critical infrastructure entities not depend upon commercial systems for core
communications.
2.3 Utility fiber networks benefited from pre-planning
The performance of private and commercial fiber networks during the hurricanes demonstrates
that utilities build more reliability and redundancy into fiber networks than their commercial
communications networks counterparts. The technology itself offers all system operators
multiple opportunities to secure communications through the ability to deploy features such as
intelligent, self-healing rings. The different performance levels of private and some
commercial fiber networks throughout the Gulf Coast and Florida point to the fact that the
entities have entirely different objectives in mind: utilities build for reliability because
communications is critical to the functioning of the core business (electric, natural gas or water
delivery), while commercial communications companies build to provide consumer services
for the general population.
An important conclusion is that EPUs private fiber networks survived the storms due to preplanning for worst case scenarios. They behaved very well to the storms and where there were
problems, they were overcome by inherent features of the technology. These features were preplanned and built into the networks. By contrast, commercial companies suffered more
difficulties with broken cables than private, internal networks and had more difficulty restoring
service. As well, smaller utilities relying upon commercial fiber-based networks spent weeks
for service to be restored.
2.4 Microwave systems survived to the storms
Like fiber networks, microwave systems survived very well to the storms, even the intense
winds accompanied these events. Although some damage to microwave towers and
attachments was reported, this damage was not as extensive as it was to other types of
communications towers such as generally taller broadcast towers. Additionally, utilities
reported that they employed detailed backup plans, including employing mobile towers and
safeguarding communications through redundant links. Any needed repairs, such as refocusing
dishes to restore links, were accomplished in the days immediately following the storms.
Katrina event reinforced the conclusion among public safety, federal, state and local officials,
that EPUs entities, along with other emergency responders, have to be included in disaster
planning.
2.5 Utilities and public safety need better coordination
The actions deployed for disaster recovery demonstrated that while utility communications
systems behaved well during the hurricanes, there was little or no coordination with state or
local public safety organizations aside from some informal sharing of resources. Usually, the
utilities played the role of assisting public safety, and not the other way around.
Beside the clear needs of dedicated spectrum for EPUs and clear advancements in
communications interoperability, EPUs entities responding to disasters should be included in
any State or federally developed coordination process.
EPUs have to be deeply involved with public safety and homeland security organizations,
especially in the areas of critical infrastructure protection and cyber-security and
156
interoperability effort, because of large enterprise ITC networks and critical control systems
such as Supervisory Control and Data Acquisition (SCADA).
Equipment that is not dependent on frequency assignments is extremely important to both
EPUs and public safety, while again, unlike traditional public safety, EPUSs have no dedicated
spectrum on which to operate a next- generation communications system. This definitely is a
long-term effort and one that will require massive investment by all parties concerned in new
infrastructure and new user equipment.
157
A5.
Introduction
A survey was performed with 9 responding Electrical Utilities from different countries across
the world and covering different sectors of activity (Transmission, Generation and
Distribution) in order to collect EPUs current practices regarding the dimensioning of electric
power infra-structures in datacenters.
In this document Data Center should be understood as all kinds of rooms used to house servers,
storage, critical workstations or telecom assets, and system signifies all kinds of IT or telecom
equipments housed in datacenters.
Survey Results
Although it is hard to generalize, based on the answers provided by the questioned companies,
it was possible to identify the following current practices and evolution opportunities:
All systems housed inside the datacenter are considered critical and therefore supplied
through uninterrupted electric power sources;
Beyond critical systems, some other equipment like essential lights, access control
devices and management/monitoring systems are also fed by the datacenter UPS. In
some other cases this equipments have their own internal battery incorporated;
UPSs are centralized but and constituted by different power modules in N+1
redundancy scheme to avoid partial hardware failure and to permit partial maintenance.
Centralized redundant UPS provide better management and maintenance capabilities
than distributed and is a most cost effective approach;
Every datacenter has standby emergency power generators, mainly diesel engines but in
some cases gas is also used. In case of power failure some other essential services on
the facility should be fed by the generator like air conditioned and lights. The generator
starts automatically in less than one minute and runs without refill (considering full
tank capacity) for:
o Control centers: From 1 to 7 days;
o Telecom rooms: From 3 to 7 days;
o IT Datacenters: From 1 to 4 days.
158
The generator fuel tank is typically refilled on demand by preselected providers but
without SLAs defined;
The general redundancy model wide adopted is N+1 but in a few cases 2N model is
also used;
Different datacenters may coexist in the same room. This leads to different power
supply systems rather than an integrated approach. This is partly due to different asset
owners operating without an integrated approach.
The typical EPUs datacenter electrical distribution concept is illustrated in the following
diagram:
Based on the datacenter availability tier classification defined by the Uptime Institute, the
typical electrical distribution topology implemented on EPUs datacenters can be located
between Tier II and Tier III levels, combining significant redundant components and partial
concurrently maintenance capabilities.
Evolution opportunities:
The EPUs essential role in the present day society renders the availability of power supply to
its IT and telecom infrastructure part of the critical components to guarantee the demanded
service level. To achieve such level, enabling maintenance activity without disrupting the
power supply and minimizing the impact of hardware failure or human error, a set of evolution
opportunities must be considered:
Adopt a 2N redundancy scheme, at least in the more important facilities like control
centres;
To avoid high temperature and humidity lack of control in case of main power failure, it is
recommended to use UPS to feed a part of the air conditioned equipment;
Deploy all electric systems, including distribution switchboards outside the perimeter of
the IT/Telco technical room;
Battery autonomy shall de dimensioned taking into account the probability of failure of the
medium voltage distribution network. Having really redundant medium voltage power
lines in the facility if a effective way to increase availability;
Introduce automatic selective shutdown mechanisms to bring significant improvement of
power autonomy for the most critical systems;
Establish contracts with fuel suppliers with SLA;
Adopting this evolution opportunities in some cases may not be cost effective, so EPUs must
adapt them according to each datacenters importance and service requirements.
o
o
o
o
o
o
o
o
o
o
CEZ (Romania),
EDP (Portugal),
EPS (Serbia),
FURNAS Centrais Eletricas (Brazil),
IEC (Israel),
KEPCO (Japan),
REN (Portugal),
Snowy Hydro (Australia),
Statnett (Norway),
Western Power (Australia)
160
A6.
1. Introduction
Western Power is responsible for the transmission and distribution of electricity in the south
west of Western Australia, otherwise known as the South West Interconnected System (SWIS).
Western Power operates as an independent business that transmits power from power stations
to residential, business and industrial customers. It also ensures equitable access to its
electricity network for any new generators or retailers seeking to compete in the electricity
market. As Western Australia's leading electricity transporter, Western Power employs over
2,500 staff and 700 contractors, supplies power to more than 930,000 customers and has over
$3.5 billion in assets. Western Power owns a network that covers over 322,000 sq km with
over 140 transmission substations; 58,000 transformers; 88,000km of power lines and 721,000
power poles. The SWIS can be seen in as the highlighted area below.
161
Western Power owns and operates an operational communications network with over
AU$300M in assets, over 250 dedicated communications sites, over 100 shared
communication sites (within substations) and in excess of 2000 network elements. An
overview map of WPs communications network can be seen in the above figure showing fibre
(shown as orange) and microwave (shown as green). The operational telecommunications
network primarily exists to support the electricity transmission and distribution business and to
meet other technical regulatory requirements placed upon Western Power (for example
telecommunications circuits for electricity generators and retailers). In addition to these core
requirements, the operational communications network provides a number of services for other
corporate users, such as corporate LAN / WAN systems, and non-operational asset monitoring
systems.
The operational services carried upon Western Powers telecommunications network require
high availability, in order to meet their regulatory requirements, and the majority of the
network is designed and built with this high availability in mind. As a result, the operational
telecommunications network is mostly duplicated to meet these availability requirements.
Western Powers operational communications systems provide as services: teleprotection
signalling; SCADA communications; telephony; voice mobile radio (VHF); Ethernet VLAN
services; and distribution automation communications (UHF radio). WP utilises a number of
communications bearers optical fibre, microwave, pilot cable, power line carrier and third
party carriers. The predominant technologies on these bearers are PDH, SDH and Ethernet.
Development of the communications network is primarily driven by transmission network
augmentation; however the communications network is heavily influenced by rapid changes in
communications technology and the relatively short asset life of communications equipment.
4. Before the installation of a new tool-set
4.1 Telecommunications Service Delivery model and upstream management
Until 2006, Western Power was a government owned vertically integrated utility comprising
generation, transmission, distribution and retail. The company was disaggregated in 2006, and
the new Western Power became a regulated monopoly transmission and distribution utility,
continuing as state government owned. This case study only considers Western Power as it has
been structured since 2006.
Western Power can be most closely modelled as a telecommunications Service Provider of
type A as defined in figure 8.4. The description of this type of Service Provider (as discussed in
section 8.4.1) is an accurate fit to Western Power.
The telecommunications services are owned, strategically managed, operationally managed,
and delivered in line with an overall corporate model that is summarised in the diagram
hereafter.
Corporate management of the telecommunications asset and systems is undertaken using
identical business processes to those utilised for e.g. substation or new transmission line build.
162
Western Power implemented, over the course of thirty years, an organically grown set of
processes that evolved to directly meet the requirements of the (internal, operational) customer.
There was never any conscious or planned attempt to align or streamline processes to meet any
standard (such as the ITU models).
These organically grown processes and management have generally followed, or been forced
to follow, the models implemented for the operation of the electrical network, and have utilised
similar tool-sets and systems. In the defence of Western Power, the processes are mature and
well-documented; however these processes are extremely specialised and completely
inflexible.
However, these organically grown processes led to organically grown tools to manage the
telecommunications network. These management systems were deployed in an ad-hoc and
reactive manner, driven by the needs of a new technology or to address capacity requirements.
Western Power could be described as (generally) being at maturity level 3 defined as per
the COBIT models.
4.3 Management Tools and Information Systems
To monitor, control and document the telecommunications network, Western Power has
operated electronic / database systems since the early 1980s and paper / offline electronic
systems before then. Western Power operated a number of standalone, un-connected
163
Due to extensive electrical system growth from 2004 to present (itself due to considerable
economic growth in Western Australia), WPs telecommunications network has also grown
considerably. During 2007, a forecast of telecommunications network growth identified that
WP would not be able to maintain a resilient network (to meet operational regulatory
requirements) from 2009 onwards.
Western Power embarked on a program to replace the existing tool-set of network management
systems. The tripartite management scope as outlined in figure 10.1 was not fully considered as
part of the initial scope the focus was strongly on changing the tools in reaction to the
business driver.
6. Following the implementation of a new tool-set
6.1 Telecommunications Service Delivery model and upstream management
The upstream management requirements and the way in which Western Power delivers
telecommunications services has experienced any underlying change (no perceived issues
driver - although it could be argued there may have been a missed opportunity driver).
During the implementation, and from a comparison to the models presented in this brochure, it
is apparent that Western Power delivers a standard suite of telecommunications services, when
compared against other electrical utilities.
The purpose of the initial implementation of the tool-set was to accommodate growth and to
asset replace an ageing management platform. Its cost was justified through OPEX offset
(people, maintenance of existing tools, and operational improvement through efficiencies in the
toolset).
6.2 Business processes
The work undertaken to replace our current toolset with an integrated NMS has not
dramatically changed or matured our business processes, except where there was a clash
between the toolset and our ways of working. However, by the very nature of implementing an
operational support system with modules and capabilities that can enable implementation of
TMForum Frameworx, Western Power has been forced into some models (for example naming
conventions). This has opened our eyes to the full possibilities of Frameworx, and the
associated cost benefits.
7. Future planning
Western Power, in line with every other EPU, is facing the upcoming challenge and
opportunity presented by the Smart Grid. Western Powers smart grid roadmap has clearly
identified the need for considerable communications infrastructure to enable two-way
communications between devices such as meters, electric vehicles, etc. and Western Powers
systems. The need for management tools to manage the explosive increase in communications
devices is evident, and the response required in terms of business processes and structure will
be considerable.
It is believed that the management tools, and the ongoing alignment of business processes to
the eTOM framework, will stand Western Power in good stead for the future.
164
A7.
Service Strategy,
Service Design,
Service Transition,
Service Operation
Continual Service Improvement
However, ITIL V2 (2001), in particular its first two components (service support and delivery)
are the most commonly used. The V2 grouped process guidelines according to different aspects
of IT management, applications and services into 8 logical sets:
Service Support
Service Delivery
ICT Infrastructure Management
Security Management
Business Perspective
Application Management
Software Asset Management
The structures of V2 and V3 are given in figures A7.1 and A7.2 hereafter.
The international standard ISO/IEC20000 (based on the British Standard BS15000) describes
an integrated set of management processes aligned with ITIL (but with reduced scope).
165
Service Support
166
Service Support
(Contd)
Configuration Management
(CM)
167
Service Delivery
IT Service Continuity
Management (ITSC)
Availability Management
(AM)
168
ICT Infrastructure
Management
Security
Management
Business Perspective
Application
Management
Software Asset Mgt.
(SAM)
169
Service
Strategy
Demand Management
IT Financial Management
Service Catalogue Management
Service Level Management
Risk Management
Service Design
Capacity Management
Availability Management
IT Service Continuity Mgt
Information Security Mgt
Compliance Management
IT Architecture Management
Supplier Management
Service Asset & Configuration
Management
Service Validation & Testing
Transition Planning & Support
Service
Transition
Evaluation
Service
Operation
Incident Management
Problem Management
Request Fulfillment
Access Management
Continual
Service
Improvement
(CSI)
170
171
172
A8.
The Tele-Management Forum (TMF), initially the Network Management Forum (NMF) is a
leading industry association of telecom operators and related industry actors focused on IT
solutions for communications Service Providers. TMF published the first release of the "New
Generation Operational Support Systems" (NGOSS) in 2001 and a Telecom Operations Map"
which evolved into eTOM in 2003. Enhanced Telecom Operations Map (eTOM) is at present
published by ITU-T as a series of Recommendations M.3050.
NGOSS is a Solution Framework for the enhancement of telecom providers business
operations, support processes and systems. It applies specifically to telecom operators or
similar Service Providers (internet, mobile services, etc.) but may be applied to any telecom
service provisioning such as those related to EPUs. It delivers a framework for producing new
generation OSS/BSS solutions, and a repository of documentation, models, and guidelines to
support these developments. The goal is to facilitate the rapid and less costly development of
flexible OSS/BSS solutions to meet the business needs of todays competitive and rapidly
evolving telecom environment.
TM Forum Frameworx Integrated Business Architecture was introduced in 2010 and provides
a master plan for the implementation of a Service Oriented Enterprise. Frameworx is designed
using service-oriented principles and supports major software standards such as ITIL (refer to
Appendix 7). It builds on the NGOSS Solution Framework releases with two additional major
documents: the Frameworx Statement of Direction (TM Forum TR155) and the Frameworx
Implementation Methodology (TM Forum GB945-M).
Frameworx Architecture
173
create components, and Business Services (through APIs) for interfacing those
elements to each other across a technology-neutral architecture (it does not define
how to implement the architecture, but the principles that must be applied to be
Solution Frameworks compliant). In addition, the Integration Framework includes the
TM Forums library of Interfaces for integration of applications, and as the basis for the
Business Services Repository.
The Application Framework (TAM) defines the role and the functionality of the
various applications that deliver management capability. It can be used by Service
Providers to compare vendors solutions.
The Enhanced Telecom Operations Map (eTOM) developed by TM Forum has been published
by ITU-T as Recommendations M.3050.x .eTOM is a business process model or framework
for use by Service Providers and their suppliers and partners within the telecommunication
industry. It describes all the enterprise processes required by a Service Provider and analyses
them to different levels of detail according to their significance and priority for the business.
It is used to analyze existing business processes, to identify redundancy or gaps in the current
strategies, and to re-engineer processes correcting deficiencies and adding automation. We
shall refer to this component of the NGOSS solution framework in the ongoing analysis.
Frameworx Implementation Methodology
175
level, the framework has three major process areas, reflecting major focuses within typical
enterprises:
The Business Process Framework has multiple groupings for the processes that it contains:
Vertical process groupings: Focus on end-to-end activities (for example, Assurance).
Each vertical group links together the customer, supporting services, resources, and
supplier/partners. Taken together, these vertical groupings represent a lifecycle view
moving left to right across the Framework from the initial strategy for the products and
their components, through development and delivery, and on into operations and
billing.
Horizontal process groupings: Focus on functionally related areas, like Customer
Relationship Management. These groupings can be visualized as a layered view of
the enterprises processes, moving from top to bottom, with the customers and products
supported by the underlying services, resources, and (where relevant) interaction with
suppliers and partners.
Where a vertical process grouping and a horizontal process grouping intersect across
the map, further process detail can be applied in either that horizontal or vertical
context, according to the users needs.
The process structure in the Framework uses hierarchical decomposition, so that the business
processes of the enterprise are successively decomposed in a series of levels that expose
increasing detail. As an example, below is the process decomposition for Customer
Relationship Management. Detailed description of this example is beyond the scope of the
present document.
176
Service Providers are increasingly incorporating IT-based services and therefore need to bring
their IT and Business aspects closer together. To address this need, TM Forum together with
itSMF community (whose members develop ITIL, originally the IT Infrastructure Library)
have analyzed and defined integration of the two frameworks that leverages the best of both.
As a result, the Business Process Framework has embedded direct support for ITIL processes
by integrating these as Best Practices processes within the TM Forum Business Process
Framework.
One of the ways to use the Framework is to build process flows. While TM Forum does not
mandate how to build process flows it does provide recommendations. The figure below
illustrates a process flow fragment demonstrating how the Business Process Framework
integrates with ITIL Best Practices (in this case for ITIL Change Management applied to a
resource-oriented change) and includes low-level Business Process Framework elements
mapped against a background of ITIL process steps.
177
Using the Framework for drafting or evaluating process flows is not the only benefit. It can
serve as a master plan for process direction and provides a neutral reference point for internal
process reengineering, partnerships, alliances, and general working agreements with other
enterprises. It can be used as a standard structure, terminology, and classification scheme for
analyzing an organizations existing processes and for developing new processes.
TM Forum is developing Conformance Certification program against the Business Process
Framework (eTOM).
2. Information Framework
TM Forum's Information Framework (known as SID Shared Information and Data) provides
a common reference model and vocabulary for Enterprise information that is used to describe
management information. Information Framework provides concepts and principles needed to
define a shared information model, the entities of the model, as well as the business-oriented
UML class models (Unified Modeling Language), design-oriented UML class models, and
sequence diagrams that enable implementation of TM Forum Frameworx conformant service
oriented solutions.
The Information Framework scope covers all of the information required to implement
business processes in a Service Providers operations based on the Business Process
Framework (eTOM) processes. It focuses on what are called business entity definitions and
associated attribute definitions. A business entity is a thing of interest to the business, such as
customer, product, service, or network. Its attributes are facts that describe the entity. In short,
the Information Framework provides the model that represents business concepts and their
characteristics and relationships, described in an implementation independent manner.
178
As shown in figure xx, the framework is designed as a layered model, which partitions the
shared information and data into eight domains. At the top layer (Level 1), each of the eight
information domains is aligned with the Business Process Framework (eTOM). It enables
segmentation of the total business problem into manageable pieces and allows resources to be
focused on a particular area of interest. In other words, for a particular business process that is
to be automated it is possible to identify the information within the Framework that is needed
to support that process.
The Information Framework Product and Service domains have been adopted by the ITU and
are included in the ITU-T recommendation M.3190.
The Information Framework can be used as a standalone framework, or when used in
combination with the Business Process Framework (eTOM) it creates a bridge between the
business and the Information Technology (IT) groups by providing definitions that are
understandable by the business, but are also rigorous enough to be used for software
development. TM Forum is developing Conformance Certification program against the
Information Framework (SID) as well.
3. Application Framework
179
The high level view of the Application Framework is shown in figure A8.xx. Beside seven
horizontal layers (consistent with the Information Framework SID), it is also divided into four
vertical columns (consistent with the Business Process Framework eTOM). Each box on the
map represents a level 1 real Application such as Customer Order Management or Bill
Calculation.
The Applications Framework is further decomposed into lower levels of functionality which
is beyond the scope of the present document.
The complete listing of all available decompositions can be found in the Application
Framework document (Document Number: GB929). One of the prime benefits of using the
framework is the ability to identify and document the key attributes associated with each
application, which then enables Service Providers to understand fully the functionality they
already have within their organization. With this evaluation of applications, the Service
Provider has the information to make an informed choice.
The key uses of the Application Framework include:
Procurement: Service Providers can use it across the entire procurement processfrom
initial request for information, through systems comparison, to guidance for
implementation.
Product Positioning: It helps Suppliers position which systems they supply (TM Forum
maintains web based Product and Services Directory).
Streamlining IT/Operation Systems: It provides a map to rationalize and combine
application stovepipes across multiple technologies and services (e.g., mobile or fixed) .
180
4. Integration Framework
For todays Communications Service Provider, software rather than the network is the enabling
function. This is driving the rate of service innovation to new levels. To cope, Service
Providers are turning to software technologies, such as Service Oriented Architecture, and
industry standards to gain business agility and flexibility. These are provided by the TM Forum
Frameworx Integrated Business Architecture.
For further reference, the interested reader can refer to the following list of TM Forum
documents and packs:
181
A9.
Ac
ADSS
AGC
As
ATM
BCP
BSS
CAPEX
CBR
CCTA
CIP
CMDB
CMMI
COBIT
CRM
CWDM
DCS
DECT
DHCP
DM
DMS
DNS
DR
DRP
DSL
DWDM
EIA
EMC
EMS
EMS
EoPDH
EOS
EoSDH
EPR
EPU
ERP
ES
eTOM
EU
FCAPS
Fu
GIS
GPRS
GUI
HMI
IAAS
ICCP
ICT
IEC
IP
ISP
List of Acronyms
Accounting (& Billing) Process (Management)
All Dielectric Self Supporting Cable
Automatic Generation Control
Assurance Process (Management)
Asynchronous Transfer Mode
Business Continuity Plan
Business Support System
Capital Expenditure
Constant Bit Rate
Central Computer & Telecommunication Agency (UK)
Critical Infrastructure Protection (NERC)
Configuration Management Data Base
Capability Maturity Model Integration
Control Objectives for Information (and related) Technologies
Customer Relation Management
Course Wavelength Division Multiplexing
Digital (Substation) Control System
Digital Enhanced (previously European) Cordless Telephone
Dynamic Host Configuration Protocol
Degraded Minutes
Distribution Management System
Domain Name System
Disaster Recovery
Disaster Recovery Plan
Digital Subscriber Loop
Dense Wavelength Division Multiplexing
Electronic Industries Alliance (US)
Electromagnetic Compatibility
Energy Management System
Element Management System
Ethernet over PDH
see EoSDH
Ethernet over SDH
Earth Potential Rise
Electrical Power Utility
Enterprise Resource Planning
Errored Seconds
Enhanced Telecom Operations Map
European Union
Fault, Configuration, Accounting, Performance and Security Management
Fulfilment Process (Management)
Geographical Information System
General Packet Radio Service
Graphical User Interface
Human Machine Interface
Infrastructure as a Service
Inter-Control Centre Protocol
Information and Communication Technology
International Electrotechnical Commission
Internetwork Protocol
Internet Service Provider
182
ISO
ITIL
itSMF
ITU
KPI
LAN
LLA
LSP
MARS
MPLS
MPLS-TP
MSP
N-CMDB
NE
NERC
NGN
NGOSS
NMF
NMS
NOC
NSO/RSO
OAM
OGC
OHL
OLA
OPEX
OPGW
OSPF
OSS
OTDR
OTN
PAAS
PBB-TE
P/C
PDH
PLC
PMU
PSTN
PTT
QoS
RAS
RIP
ROADM
ROI
RSTP
RTU
SAAS
SAS
SCADA
S-CMDB
SDH
SES
SIPS
SIR
SLA
SNCP
183
SNMP
SOA
SOC
SPD
STP
Su
TASE
TCP
TDM
TETRA
TMF
TMN
TSO
UDP
UPS
UTC
U-Telco
uTOM
VIU
VLAN
VoIP
VPN
VSAT
WAMS
WAP&C
WDM
184
REFERENCES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
185
Ch 1
Ch 1
Ch 1
Ch 1, 3, 4
and 5
Ch 1 & 2
and 3.2
2.1
3.1
3.2.5
3.5.1
4.6
4.6, 5.5.2
8.2.3
4.6
5.2.2
5.2.3, 7.6
5.3
Ch 6
6.4
6.5
6.6
6.6
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
186
6.8
7.2
7.3
7.4.4
7.5
8.2
8.2
9.3.2
9.3.4
9.3.4
10.2.1
10.2.2
10.2.2
10.2.2
10.3.1
10.3.1
10.3.1
10.3.2, A7
10.3.3
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
187
10.3.4
10.3.5, A7
10.7.1,
10.7.3
10.7.2
10.7, A6
10.7.4
12.3
12.3
12.4
12.4
12.4
12.5
12.5
A3
A4