Frequently the most immediate financial return we can get from gathering data from things (and making decisions on that data) is from existing installed assets, rather than new build projects. The problem is that those assets often have not been designed or installed with monitoring in mind. This talk examines such a project in the Electricity Distribution business, and the critical challenges imposed by dealing with the practical problems of what is already installed, rather than assuming a clean sheet of paper.

1. The Impact of the Internet of Things on our Lives.
The Challenge of Legacy Assets
Richard Lucas, BSc(Eng) MBA
MD of ASH Wireless
ASH Wireless is an electronics design consultancy
Specialising in wireless and sensors

2. 2
An Ofgem funded Network
Innovation Competition project

3. 3
The problem
Assets have
nominal thermal
rating
Ratings = OC
Ratings K amps
Diverse range of
environments
Small changes in
environmental
factors can result
in very different
actual ratings
HV network
LV network
Assumed thermal
ratings can lead to
capacity being
under-utilised
or
unnecessary risk
Objective is to
maximise power
through
transformer

4. 4
Celsius as part of the smart future
Celsius
monitoring
Smart meter
data
Thermal Ratings Tool
Reinforce
Retrofit
cooling
Extra capacity
Lower bills for customers

5. 5
Step 1: Fit thermal monitoring
Internal
temperature
Asset
Environmental
factors
External
temperature
Thermal
coefficient
Learning
Thermal
Ratings
Tool
Deliverable
More
capacity
Benefit

6. 6
Step 2: Retrofit cooling
Retrofit
cooling
Internal
temperature
Asset
Environmental
factors
External
temperature
Full
capacity
Benefit
Enhanced
Thermal
Ratings
Tool
Deliverable
Retrofit cooling
specifications,
installation
methodologies
and buy order
Learning

7. 7
Partners and roles on project
Facilitate customer
focus groups
Develop customer
communication
materials
Lead the customer
survey engagement
Analyse trial data
Develop
methodologies to
understand
relationship between
asset temperature,
load characteristics
and surrounding
environment
Determine impact of
cooling technologies
Develop tool and
spec for low cost
temperature sensor
Recommendations
for BAU rollout
Supply complete
retrofit monitoring
solution
Provide ongoing
support throughout
installation,
commissioning and
operation of the
retrofit thermal
monitoring
workstream
Peer review of the
analysis
methodology of the
retrofit temperature
sensor part of the
project
An investigative
study on the impact
of Celsius on the
lifetime health of
network assets
Work with ASH,
Ricardo-AEA and
Electricity North
West to develop
retrofit thermal
monitoring solution
Participate in
evaluation and
selection of retrofit
cooling techniques

8. 8
Monitoring site selection and timescales
To enable trials to
take place during
all seasons and to
trial all cooling
techniques
Four year project
Subset of 520
substations –
enough sites to
adequately trial all
techniques
100 cooling
technique sites
Enough
substations to
represent 80% of
GB substation
population
520 substations

9. 9
Celsius
£5.5
million
Investment
Up to £583m
across GB by
2050 Financial
benefits
Awarded: 9th December 2015
Go live
Monitoring
installation
Mar 2017
Monitoring
trial
Mar 2018
Thermal
ratings
tool stage
1
Oct 2018
Retrofit
cooling
installation
Jun 2018
Cooling
trial
Jun 2019
Thermal
ratings
tool stage
2
Jan 2020
Closedown
Mar 2020

10. So the answer’s simple
1. Gather data on the power throughput &
characteristics, asset temperature,
2. Work out the relationships, and how you can change
these with retrofitted cooling technology
3. Decide how long you can postpone reinforcing or
replacing assets,
4. Save lots of money
But there are some practical problems…………..

11. 11
Legacy Assets
 A significant percentage of
electrical substations are
more than half a century old
 Most are not installed with
measurement equipment for
voltage, current, power,
power quality, temperature
of key assets
 Critical infrastructure – any
work must not intrude on
continued operation.
 Access to the sites is
restricted for safety reasons

12. 12
?How to fit monitoring sensors when…
 No internet connection
available
 No mains power available
 Can’t install wiring, mount
equipment, etc without
intruding on critical
operations
 Challenging RF environment
for radio connections,,
multipath, multiple
monitoring points needed

13. KeLVN
- Design Requirements
 Monitoring system can be installed
non-invasively;
 All equipment magnetic or cable-tie
mounting;
 All equipment battery powered for
duration of data gathering project
(3.5 years).
 All equipment wireless, only leads are
those required to actually take
measurements.
 Daily reports to back end. 30 minute
measurements of V, I, P, Q, THD,
Temperature

14. Backhaul Choices
• Clusters of sensors over a
small area
• Spread over a large part of
the country
• Mobile network is clear winner
• LPWAN
• Possible, but patchy network
support
• 2.5G GPRS chosen
• Coverage, component cost

15. Local Wireless Sensor Network
• Requirement for large substations (30m x 15m), sensors
may be in cabinets
• Path loss at 2.4GHz expected less than 90dB
• Avoid stationary nulls by using antenna diversity
• Energy-efficient 802.15.4 transceivers used, no front-
end-modules
• Operation needs to allow for more than one hub
• Large coverage
• Check installation from outside the substation

16. Local Wireless Sensor Network
• Simple, non-paired network access, minimised energy
• Sensors transmit the last 6 hours of measurements
(1 packet) every 30s
• Time-jittered to avoid multiple collisions from time-
aligned sensors
• A hub listens for 60s for reports every 6 hours
• Hub transmits a daily report to a back-end database
• Hub operates for 3.5 years on a Lithium D-cell
• Other sensors use AA or D cells for >3.5 year life

17. TT T
T
Voltage Sensor
Temperature Sensor
GPRS
Backhaul
Mobile
data link
Hub 1
Hub 2
(optional)
Hex 1
PMU
Hex 2
Ind.
V,I,T
Low Power
Radio
Network
Single
Temperature
Sensors
Back
End
KeLVN
- Substation Monitoring

18. Battery Life
• The Hub is an interesting challenge:
• Lithium D-cell, 10AH
• Over 3.5 years life
• Distribution of where the capacity is used:
KeLVN Hub battery usage
Daily GPRS data transfer
Sensor receiver window
Quiescent: ‘off’

19. Key learning points
• The installation and environment defines
• Equipment design (e.g. magnetic mounting)
• Protocol (e.g. multiple hubs allowed)
• installation procedure (fast, non-invasive)
• GPRS modem auto-connect modes are not reliable,
process needs detailed design
• Alarms
• Not required in Celsius
• Low latency for alarms is managed with a hub
software extension, and external power to the hub

20. Example of how legacy assets can
be made to yield valuable data
Retrofitting monitoring equipment to legacy assets:
• Ease of installation is primary consideration
• Use of a local wireless sensor network eases installation
• Optimise air interface to manage trade off between
latency, data rate, battery life

21. Additional material

22. 22
Site selection map

23. 23
Celsius technology
Hub Wireless sensor

24. 24
Celsius technology – trial fit
LV board with three sensors

25. 25
Celsius technology – trial fit
Transformer

26. 26
Trial site data
-5
0
5
10
15
20
25
30
35
-20
0
20
40
60
80
100
120
140
30/09/2016 00:00 30/09/2016 12:00 01/10/2016 00:00 01/10/2016 12:00 02/10/2016 00:00 02/10/2016 12:00
TemperatureDegreesC
RealPowerkWperPhase
Magda Rd 750kVA Transformer Load and Temperatures
P1 kW P2 kW P3 kW Top Oil oC Inlet oC Outlet oC Bottom Oil oC