The Implementation of Phasor Measurement Technology

WATT'S TECHNOLOGY JUN 2014

The modern era of power delivery is faced with many challenges due to decreased reserves, increased use of renewable
generation, Flexible Alternating Current Transmission System (FACTS) devices, and changes in customer/load behaviour.
In order to manage the risks introduced by these challenges, it is important to have near real time observability of the health
of the power system.
Phasor Measurement Unit (PMU) technology has come to the forefront in terms of smart measurement of the power system
in general because it provides accurate, high resolution monitoring of important analysis metrics. Eskom has identified the
need to enhance the situational awareness of operators at the National Control Centre by implementing synchronised
phasor measurements technology to improve the power system reliability and operational security during normal and highly
stressed operating conditions.
This paper discusses the implementation of Phasor Measurement Technology on a large scale in Eskom and summarises
the benefits of Wide Area Monitoring System (WAMS) applications that can contribute toward a smarter transmission grid.
INTRODUCTION
Eskom initiated a project for the installation of PMU modules and substation class Phasor Data Concentrator (PDC)
equipment at 15 key substations. This included: Installation of the Phasor Data Concentrator (PDC), synchronised
measurements historian, interface to Energy Management System (EMS), visualisation and stability applications at the
National Control Centre (NCC). The project covered the short term requirements but has been engineered to meet future
requirements as well. The project has completed the Factory Acceptance Test for the control centre WAMS application and
substation equipment. The installation of the PMUs is currently in progress. The Integrated Site Acceptance Test shall
commence after Eskom has completed PMU installations at 8 substations. The future WAMS scope includes an increase in
coverage of critical nodes in the network to install a total of 52 substations.
PMU MEASUREMENT TECHNOLOGY
One of the key benefits of PMU measurement technology is the use of highly accurate GPS timing to measure phase angle
which can be compared over a wide area. This is done in the following way: the GPS time is used to construct a time
reference (this is essentially the same over a wide area due to the high accuracy of GPS timing). This time reference is
transmitted to the PMU via a 1 Pulse per Second (PPS) signal which is used to define the start of the second. The phase
angle is then measured as the angle difference between the start of the second and the peak of the wave as shown in
Figure 1.
This time reference can then be replicated over each sample so that a relative angle can be calculated per frame. Therefore
if the voltage is at synchronous frequency (i.e. 50 Hz) and the frame rate is a factor of this frequency (i.e. 10, 25 or 50 Hz)
then the phasor would remain stationary.
However if the frequency of the voltage signal was off nominal, the phasor would then rotate at a speed of fm fn where fm
is the measured frequency and fn is nominal frequency.

This technique of phase angle measurement allows for the comparison of angles over a wide area. As per Figure 2, the
difference between the voltage angles of bus 1 and 2 can be determined simply because they share the same time
reference.
Combining this measurement with high resolution reporting (up to 50 frames a second), makes this data extremely useful.
As seen in figure 3, the Eskom WAMS consists of PMUs and Substation Phasor Data Concentrators (SPDC) deployed at
Eskom transmission stations and a Central Phasor Data Concentrator (CPDC) and WAMS Data Centre facility deployed at
the Eskom National Control Centre .
The PMUs transmit synchrophasor data using the IEEE C37.118 protocol. The data is time- stamped by the PMU using a
GPS clock and the data is then transmitted to an SPDC. All communications use the Eskom telecommunications WAN
routers and bandwidth management equipment.

SPDCs collect IEEE C37.118 data from up to 8 sources and combine the data into a single IEEE C37.118 PDC stream
transmitted to the Control Centre. The WAMS Control Centre consists of computer and data storage equipment running the
Psymetrix PhasorPoint software. Client workstations running the PhasorPoint workbench software can interface with the
central servers to provide phasor data visualization and analysis tools.
The CPDC and WAMS Data Centre consists of computer and data storage equipment running the Psymetrix PhasorPoint
software suite supporting data management applications, phasor measurement visualisation, and synchronised
measurement applications. The system inter-faces with Eskoms EMS (TEMSE) and Enterprise Information System (EIS). In
addition, client workstations running the PhasorPoint workbench software can interface with the central servers to provide
data visualization and analysis tools.
The CPDC and WAMS Data Centre consist of dual redundant mirrored servers, providing a highly available data storage
and processing facility.
SUBSTATION ARCHITECTURE
WAMS communications within the substation between the PMUs and the SPDC operates via a dedicated network switch as
shown in figure 4 above. The infrastructure supports up-to eight PMUs connected to an SPDC. PMUs and the SPDC are
allocated IP addresses and communications between devices is mediated by the IEEE C37.118 protocol over TCP/IP.
At the substation, SPDCs cache all received IEEE C37.118 stream data. This provides a short term data repository (seven
days) that permits data to be retrieved directly from the substation if required
Production Scheme Design
Each production scheme consists of two portions:
1. Common Services items such as the DC Secure Supply (chop over) system, sPDC, GPS, auxiliary relays, global MCBs
and the cubicle (panel) itself.
2. PMU Module items including the Alstom Grid MiCOM P847B, PK-2 test blocks, PMU-specific MCBs and a Voltage
Transformer Selection scheme.
Eskoms specification for the WAMS requires up to eight PMUs communicating with the WAMS Control Centre through one
sPDC. The design drawings, exploiting the modular concept, incorporate eight PMU modules in three panels as shown in
figure 5.
A 48U PMU panel accommodates a maximum of three PMU modules (capacity for 12 sets of currents, and three sets of
voltages).
The synchronizing unit Alstom Grid MiCOM P594 has one IRIG-B BNC output and 4 x 1PPS outputs and it is able to
synchronize up to 4 PMUs.
High Availability
A robust, highly available (HA), control centre architecture is provided. The core phasor processing, analysis, visualisation
and storage functions are duplicated across two PhasorPoint HA server stacks.
Each stack consists of two servers. In addition, two C37.118 proxy servers ensure a robust duplication of identical IEEE
C37.118 streams to each server stack.
The PhasorPoint server stacks operate as a mirrored redundant pair, both processing and analysing the identical IEEE
C37.118 data streams provided via the proxy servers.
Each consists of multiple servers operating together in a functionally integrated server stack. The two server stacks are
located on different floors of the NCC and operate together to maximise both the integrity and accessibility of the WAMS
data collection, analysis and event notifications. A single PhasorPoint server stack operates as the master server affording
workbench and CPDC configuration.

In the event of single server failure, no data is lost as both servers receive and store the stream data. In addition, workbench
clients connected to the failed server detect the service interruption and migrate to the available server. Once the failed
server is operational, recovery is automatically initiated, synchronising the configuration and restoring missing stream and
processed data from the operational server.
The two server stacks continuously monitor each other. If one server stack functionally becomes unavailable, an associated
event is created on the pair server.
Interfaces
The PhasorPoint servers provide the following remote service interfaces:
Access to HMI workbench applications which provide advanced visualisation,analysis and alarming.
PhasorPoint Structured Query Language (SQL) interface providing SQL access to all phasor and derived data via ODBC
or JDBC drivers.
IEC 60870-5-104 interface permitting transfer of phasor data and alarms to the EMS.
IEEE C37.118 interface allowing PDC streams to be transmitted to additional destinations.
Historical Phasor Data
Each PhasorPoint server stack maintains two logical data archives.
The Short Term Rolling Archive holds the data at the native (received) data rate for a fixed short duration. Data is kept in
the rolling archive for a user configurable amount of time, set by default to three months, before it is automatically deleted to
make way for new data.
The Long Term Rolling Archive holds the data for a significantly longer duration (duration depends on physical storage
space, selected data rate and quantity of high resolution snapshots).
In addition users can copy a time bounded selection (snapshots) of the WAMS data from the Short Term Rolling Archive to
the Long Term Archive. This copies all of the WAMS data (including phasor, analogue, and digital IEEE C37.118 data and
oscillatory stability analysis data) for the selected time to the Long Term archive. Snapshots can also be generated
automatically (triggered) upon events.
Development
Independent development servers are also provided which run a single central PDC PhasorPoint service to support
simulated data, testing, training and the development of applications which interact with PhasorPoint. The PhasorPoint
system on the development server provides the full range of functions (excluding HA related functions) as provided by the
primary PhasorPoint service. The development server has no direct access to the highly available IEEE C37.118 streams
provided by the C37.118 proxies. However, it is possible for the server to directly and independently connect to remote
SPDCs and PMUs, for example, for pre-commissioning testing. In addition, the operational PhasorPoint servers can be
configured to provide an IEEE C37.118 PDC output stream retransmitting user-selected live PMU data to the development
server. All data received from IEEE C37.118 is available to be sent via PDC output streams.
Security and Network Isolation
Functional isolation of the WAMS network from the EMS, EIS, and external networks is provided using physical firewall
devices. In addition, the firewalls mediate VPN access to the WAMS network.
The development environment is also isolated from the control centre WAMS network, utilising a different VLAN and firewall.
Workstations and GPS
Five (5) operator workstations are provided for access to the WAMS visualisation and application features.

Two (2) Meinberg M900 GPS Clocks provide Stratum 1 time sources. In addition, the clocks provide an NTP service for all
of the Control Centre servers and workstations and for the remote WAMS SPDCs.
NETWORK INFRASTRUCTURE
A redundant 10-100-1000 baseT standard Ethernet LAN underlies the WAMS networking infrastructure at the NCC for the
critical computer equipment. Dual redundant control centre network switch infrastructure is provided for communication
robustness.
Each of the servers in the server stacks implementing WAMS functions including the Development (PDS Environment)
Servers and the workstations in the NCC Control Room connects to two independent switches. The server stacks in
redundant configuration are integrated inside cubicles and are installed on the 1st and 2nd floors of the NCC building. The
general network architecture is as described in the Figure 8.
BENEFITS OF WAMS
Improved State Estimation
State estimation provides the initial condition for all online applications such as contingency and stability analysis. The high
accuracy of PMU data and the measurement of voltage angle difference allow the estimation algorithm to be vastly
improved. This will start as a hybrid system with the eventual goal being a linear state estimator, which uses non-iterative,
algebraic equations to find the system state.
Real time monitoring of importance power system metrics
WAMS is able to calculate and monitor voltage magnitude and phase, real and reactive power and frequency. These are
important metrics for understanding the stress in the network over a wide area at any given time. Voltage angle difference
between two points in the network shows a multidimensional metric for stress because this angle is dependent on the
voltage magnitudes, impedance and power flowing through the corridor. Alarming on all of these metrics allows the
controllers to easily diagnose system problems and react accordingly.
Real time monitoring of power system oscillations
All power systems are prone to power oscillations due to the nature of synchronous systems. Whilst these oscillations
cannot be eliminated they must be monitored and controlled to ensure that they dissipate quickly (i.e. good damping).
WAMS has real time estimation techniques that determine the oscillation mode frequencies, damping, amplitude and phase.
When these metrics are shown over a wide area, it is possible to determine how system is oscillating and where the mode
originates from.
Enhanced post disturbance event analysis
Current recording systems are prone to missing crucial data: disturbance recorders use high resolution sampling but record
for only short durations around fault conditions and SCADA recordings use a very low resolution (less than one sample
every four seconds). WAMS provides continuous, time- stamped, high-resolution data over a wide area, which has proven
critical for understanding system events and has aided root-cause analysis during investigations.
Validate our system models with real time recordings
Improves the simulation accuracy of offline studies therefore reducing risk and possibly increasing transfer capability.
Improved ability to re-synchronising islanded networks
Since all PMUs measure frequency these are useful during blackstart or islanded conditions to reconnect portions of the
power system. The phase angle and frequency differences between islands can be shown in real time and therefore aid the
controllers in moving the system to a state which will reduce the stress across synchronisation points and allow the sync
check relays to operate effectively.
Monitoring the overall power system stability

The high-resolution data is useful for monitoring stability in the network, i.e. frequency, small signal, transient and voltage
stability.
Provide the foundation for developing control staff Situational Awareness through new metrics and displays
The development environment allows for the facilitation of simulated exercises where large disturbances can be observed
and understood so that controllers and staff are better equipped to react to and minimise the effect of catastrophic events.
CHALLENGES
Cyber Security
Traditional Master to Remote communication over serial links, which implies a non-routable protocol for Cyber Security. The
allocation of IP addresses to conform to the Eskom IP allocation standard.
Outage Constraints
Delays in the implementation of the PMUs in the field due to the outage constraints.
Asset assignment
This device does not lend itself to any one of the traditional departments, i.e. Protection, Control, or Metering due to its
accuracy and application. This causes problems when defining work scopes and allocating resources since no department
feels it is directly within its responsibility.
Compliance to IEEE C37.118 standard
The newest revision of the standard was released at the end of 2011. Whilst vendors claimed to be compliant, little proof
was available at the time of tender evaluation. Due to the high expectation set by the standard, vendors have largely left the
issue of performance to the utilities. This means extensive tests must be done in order to verify and understand device
settings and performance.
Limited knowledge of the new technology
Many new technologies are met with reluctance to change, and PMUs are no different. Many training sessions have been
initiated in order to share the idea of the new technology and its benefits. In particular, the control shift staff needs to
understand how this technology is different to SCADA and how it can be used to their benefit.
FUTURE WAMS SCOPE
Rolling out this technology to more substations, more PMUs mean more visualization to support the real time grid
operators;
The ultimate end state is a Linear state estimator which uses only PMU data to provide full observability. For this, 33 % of
the transmission network will require PMUs. Future WAMS scope to increase coverage of critical nodes in the network to
install a total of 52 substations;
More Advanced situational awareness/stability monitoring applications;
Add new EMS applications to fully exploit PMU data;
Integrate the PMU functionality with the protection relay;
Synchrophasors and IEC 61850;
Grid Protection: develop protective control schemes that dynamically adapt to current power system conditions, to preserve
the integrity of the power system grid as an entity.

The implementation and acquisition of phasor measurements data is predicated on the following requirements as per Table
1 over the short, medium and long term.
CONCLUSION
After succesfull completion of a number of research projects, Eskom realised the benefits of Phasor Measurement
Technology and is currently in the process of installing 59 PMUs in 15 Transmission substations as part of Phase 1 of the
WAMS project. PMU technology is not a replacement to the traditional SCADA system but complements it.
The phasor measurements data with its advanced visualisation features and early detection of abnormal events, will be used
to enhance the control staff decision making allowing the staff to move from a purely reactive response mode to a predictive
and proactive process. The WAMS applications provide powerfull tools to improve power system security and ensure we
march towards a smarter grid.
References
[1] Power System Relaying Committee, IEEE Standard for Synchrophasor Measurements for Power Systems, IEEE Power
& Energy Society, IEEE Std C37.118.1TM-2011, Dec 2011.
[2] Performance & Standards Task Team (PSTT), PMU System Testing and Calibration Guide, North American
SynchroPhasor Initiative (NASPI), Dec 2007.
[3] Eskom WAMS System Design Report (ALSTOM GRID Inc. & Psymetrix Limited)
[4] Eskom WAMS Control Centre Detailed Design (ALSTOM GRID Inc. & Psymetrix Limited)
[5] Eskom WAMS Control Centre Functional Specificatiom (ALSTOM GRID Inc. & Psymetrix Limited)
[6] Eskom WAMS Substation Detailed Design (ALSTOM GRID Inc. & Psymetrix Limited)
able generation, Flexible Alternating Current Transmission System (FACTS) devices, and changes in customer/load
behaviour.
In order to manage the risks introduced by these challenges, it is important to have near real time observability of the health
of the power system.
Phasor Measurement Unit (PMU) technology has come to the forefront in terms of smart measurement of the power system
in general because it provides accurate, high resolution monitoring of important analysis metrics. Eskom has identified the
need to enhance the situational awareness of operators at the National Control Centre by implementing synchronised
phasor measurements technology to improve the power system reliability and operational security during normal and highly
stressed operating conditions.
This paper discusses the implementation of Phasor Measurement Technology on a large scale in Eskom and summarises
the benefits of Wide Area Monitoring System (WAMS) applications that can contribute toward a smarter transmission grid.