Application of Synchronized Waveform Data to

Power System & Apparatus Monitoring

Presented by
Wilsun Xu, P.Eng, F.IEEE

University of Alberta
Edmonton, Alberta, Canada
May 2022

Based on a paper published in IEEE Transactions on Power Delivery, vol. 37, no. 1, pp. 3-17, Feb. 2022, “Synchronized
* Waveforms – A Frontier of Data-Based Power System and Apparatus Monitoring, Protection, and Control”,
doi: 10.1109/TPWRD.2021.3072889. 1
Outline:

1. Status and characteristics of synchronized waveform

(sync-wave) data

2. Three platforms of sync-wave applications

3. Strategies to develop sync-wave based applications

4. Synchrophasor versus sync-wave data - a brief comment

5. Conclusions & main takeaways

2
1. Status and characteristics of sync-wave data
Defining synchronized waveform (sync-wave) data – three characteristics:
• Voltage or current waveform data (sampled at least 64 samples/cycle, or 3.8kHz),
• With (explicit or implicit) precision time information for the data samples,
• The information is sufficient to align waveforms recorded at multiple locations to an
acceptable accuracy (to be established by a standard).

Example of sync-wave data

o 256 samples/cycle, i.e.

15.9kHz sampling rate

o 1mSecond GPS timestamp

accuracy

You can get this and other data from the

PES PQ Data Analytics WG website:
https://grouper.ieee.org/groups/td/pq/data/
3
1. Status and characteristics of sync-wave data
Devices with sync-wave measurement capabilities (SMU) are already available

Portable PQ monitor Stationary PQ monitor Gapless SMU Relay-based SMU Merging Unit

Three industry trends driving the need for waveform data:

• Increased adoption of power electronic (PE) apparatuses in
power systems
• More complex system dynamics (e.g. inverter-caused SSR)
• The move to online apparatus condition monitoring

SMU – sync-wave measurement unit (a generic name to facilitate description here) 4

1. Status and characteristics of sync-wave data
• Types of data
• Forms of data
(for eventual synchronized analysis)
o Raw waveform data
o Derived data (i.e. indices)

• Scheme of data collection and

transfer Differentiate
o On-demand such as download three concepts about the data:
o Event driven
o Real-time streaming o data with precision time information,
o synchronized recording of data,
• Central location for synchronized o synchronous transfer of (real-time) data.

analysis
o It does not mean control center only
o It can be a substation or even an engineering office
5
1. Status and characteristics of sync-wave data
Classification of applications:

1) Offline analysis
3 4
2) Online monitoring
(no automatic action) 2 1
1
3) Real-time P&C 2 3 4
(protection & control) 3 1,4
2
2
3 4
How sync-wave data is used is highly 1
dependent on the type of applications. 1,4
Real-time streaming of the data to 2 3
control center is only one of the 3 4
possible approaches
1
2
6
2. Three platforms of sync-wave applications

Platform No.1: Special purpose sync-wave platforms

• For (real-time) protection & control applications

• Extremely high reliability requirement due to automatic control actions
• Customized, dedicated SMU network is the most acceptable approach
• Consistent with current industry practice

Platform No.2: Multi-Use Sync-wave Platforms

• For online monitoring and offline analysis applications
• Real-time streaming of data is NOT necessary
• Thus a lot more options are available to construct such a network

7
2. Three platforms of sync-wave applications
Platform No.3: Mobile Sync-wave Platforms Using Portable SMUs
• For offline analysis, e.g. troubleshooting, model validation, forensic analysis etc.
• Can be deployed at almost any locations with little infrastructure support
• A very important tool to support university research including emulating PMUs

At substation
Installation of two SMUs
(Portable PQ monitors)
of example 2
I&V
probes
SMU
At pad-mounted
V & I probes connect to
SMU (not shown)
service transformer

8
2. Three platforms of sync-wave applications
Need to research and develop general-purpose data analytics algorithms
• Most useful sync-waves are those that Continuous stream of data arrive at SMU
contain changes or disturbances (called
abnormal waveforms here) Abnormal waveform detection
• Focusing on abnormal data reduce
capacity requirements on infrastructures Abnormal waveform extraction

• Need to develop general-purpose Pattern recognition

abnormality detection & pattern
recognition algorithms
Indices calculation (app specific)
• It is also useful to research application
specific data analytics algorithms (such Send to central location for
as extracting SSR indices) synchronized analysis & decision making

SMU seems to be the best location to perform the above analysis

9
3. Strategies to develop sync-wave based applications
Unique strengths of sync-wave data
• Waveform data is not new to power system engineers
• What is new is that waveform data from multiple locations can now be analyzed together
due to their being able to time-aligned properly

Values of multi-location data:

• Help to solve location related problems, e.g. which
inverter triggers instability?
• Support multi-port network/component characterization:
e.g. inertia of a regional power system
• Enhance information using multiple data: e.g.
differential protection and fault location
An example multiport problem

10
3. Strategies to develop sync-wave based applications

Application Application
How to Utilize Sync-wave Data
domains:

System Oriented Apparatus Oriented

Applications Applications

`
Strategies:

1 2 3 4 z 5
Participants Multiport Information Utilize disturbances Utilize disturbances
identification system content outside the from the
& ranking characterization enhancement apparatus apparatus
Application
examples:

Detecting Estimating Multi-sensor Disturbance- Incipient

destabilizing inertia fault based online fault
generators for an area location apparatus testing detection

See the paper for details 11

4. A brief comment - Sync-wave versus Synchrophasor
Color to B&W
• Synchrophasors are calculated from Information (spectral) leakage
waveforms, i.e. a derived form of Using DFT on
sync-wave Using DFT synchrophasor
to get to get oscillating Results
• Information is lost when Synchrophasor components

transforming waveform data into a Raw Precision time

single index waveform
data
Precision time

• Anomaly in a waveform cannot be Using DFT to get

spectra of oscillating Results
captured by phasors components
This is the future
• Since many applications don’t
require real-time streaming of
waveform data, the main advantage Why tie up our hands with a processed data?
of phasor – less demand on
Why limit our imagination to one complex number?
communication – does not really
exist We deserve more!

12
• 6. Conclusions & Main takeaways
• Waveforms are the most authentic and granular data revealing power system
behaviors. They provide much more information than the phasor data
• The main strength of sync-wave is to enable integrated analysis of multi-location
data, thus sync-wave is especially useful for solving problems involving:
o Interactions of multiple components (e.g. ranking, contributor identification)
o Multiport systems or subsystems (e.g. characterizing an area instead of a component)
o Cross-referenced information extraction (e.g. differential analysis)

• Real-time streaming of sync-wave data is not necessary for many applications. It

is needed mainly for a dedicated platform serving a specific control function
• Two other platforms, multi-use (on-demand access) platform and mobile platform
are likely to be more useful, at least at the early stage of sync-wave adoption
• Sync-wave data can support both system and apparatus oriented applications
13
I welcome any questions and comments

A more detailed presentation including recording can be found from

https://www.naspi.org/node/931

NASPI: North American SynchroPhasor Initiative

14