Transformer Load Monitoring
Status Report and Benefits Summary
Ameren-UE – where they were
• Home grown application to predict transformer
loads based on interval data from data recorders
installed at 200 different locations based on
customer segment.
• TLM data from the predictive algorithms set
purchasing practices and construction guidelines.
• 2 years ago, Ameren-UE was considering
improving the predictions from their TLM
program due to suspicions that the system was
Ameren-UE – where they were
• Options to improve TLM output:
– Increase sample size – significant cost vs. unknown
– Use daily metering reading data – significant
development task vs. unknown benefit
• Ameren-UE expected benefit through lower
capital expenditures.
• CellNet suggested SLS concept after discussions
with distribution engineers and management.
Ameren-UE – what they achieved
• Prototype one district of 50,000 meters and manually
review data.
• Results of prototype:
– Identified 10+ overloaded transformers
– Identified transformer/meter mismatches in AM/FM
– Determined that TLM predictions were 70% too high
• Preliminary results were declared a success and
implementation plans for the full service were initiated
with a goal to complete prior to Summer, 1998
Ameren-UE – what they achieved
• Over 20 snapshots retrieved for both summer and winter peaks
• Immediate benefits in reduced capital budgets, improved customer
service and reduced outages:
– 20% savings in capital purchases of transformers ($12 million annual
budget reduced to below $10 million) due to purchase of smaller
– By no longer purchasing transformers with CSP option is saving $100 per
– Re-use of larger transformers to compensate for material shortages have
reduced customer down time
– Identification of 250+ overloaded transformers resulted in reduced
outages and lower labor costs
• Longer term benefits through change in purchasing standards and
“poor man’s SCADA”.
Ameren-UE – future plans
• Poor man’s SCADA
– Ameren-UE intends to push the SLS capability to the
rural area where substations are not automated.
– Snapshot information will first be used to analyze
circuit and transformer loads.
– More regular information (weekly?, daily) will be
obtained to monitor un-automated substations.
• Selective application of SLS
– Ameren-UE would like to begin using the load
information by substation, circuit, etc. to analyze their
system for phase imbalances, before and after
reconfiguration activities and line loss studies.
Other CellNet Users – Status and Plans
– KCPL has no transformer monitoring and uses simple sizing
guidelines. Overloaded transformers are addressed in response to
low voltage and outage conditions reported by the customers.
– KCPL has been trying to implement a distribution engineers
workstation for the last few years with no success.
– KCPL wants to implement SLS soon as they believe the benefits
are tremendous (“We don’t need to build a business case”).
– KCPL is limited by their AM/FM at this point as the transformer to
meter binding is only about 75% accurate.
– NSP has no plans for implementing SLS due to IT resources and
lack of focus in the DA area.
Other CellNet Users – Status and Plans
– PSE has a home-grown algorithm which analyzes the
monthly meter reads to identify overloaded
transformers. PSE uses simple sizing guidelines to
determine transformer sizes upon install.
– PSE is planning to implement SLS within a year as they
feel that the low failure rate of their transformers
indicates an overbuilt system.
– PSE also does not require a business case (“It’s a no
brainer”) to justify SLS.
General Business Case
• 20% reduction in annual capital budget allocated to
transformer purchases
• $100 per transformer reduction as CSP option is no longer
• 50% reduction in transformer inspections (assume 5% of
population is inspected at $50 per inspection)
• Reduced customer outage due to overloaded transformer
(one-time benefit based on 0.25% of transformers)
• Improved transformer/meter connectivity (one-time
System Load Snapshot
• The SLS application utilizes the TOU capability of
the CellNet system.
– Not available for any meter on a TOU billing rate.
(May not be available for any meter on a demand rate
as well, but this is a process issue as opposed to a
technical issue).
– Limited to 3 configurable intervals and two different
• Default configuration is 4:00-5:00PM, 5:00-6:00PM and 6:007:00PM during Summer and 5:00-6:00AM, 6:00-7:00AM and
7:00-8:00AM during Winter.
• Any three intervals which are a multiple of 5 minutes can be
System Load Snapshot
• Operation
– PPL identifies a peak day, or day of interest, for which
snapshot information is required.
– PPL notifies service provider by 8:00 AM on the day
following the day of interest.
– Service provider provides ASCII file two days after
notification with MeterID, TransformerID, kWh in each
of the three intervals, and a flag to indicate the
reliability of the data for each meter.
System Load Snapshot
• Limitations and Issues
– 20 to 30% of the data provided will have a flag indicating potential
missing or incomplete data.
– PPL must provide accurate transformer to meter binding
information to service provider.
– The service provider conducts the read for SLS data in the
afternoon after the regular billing meter read has completed.
Consequently, more than a few SLS requests in a month affects the
service provider’s ability to use the system for other purposes (both
maintenance and non-PPL revenue applications).
– The SLS application has the result of swamping the distribution
engineer with data. The first set of data is very important as it is
the first detailed look at the overall system loading. PPL needs to
ensure that adequate resources are available for the analysis of this
• SLS benefits are clear to all current CellNet
system users.
• Status of AM/FM systems and the accurate
binding of transformer to meter is the gating item
to implementation.
• 20% annual savings in capital budget with other
softer benefits are conservatively achievable.
• Other benefits beyond transformer load
monitoring are apparent, but not validated.
Transformer Monitoring – Misc.
• Most of the interviewees did not see value in metering at the
transformer beyond theft and outage management.
– Both PSE and Ameren-UE used 1% theft in their business cases and feel
that the preliminary analyses justify a higher theft number.
– Metering at the transformer would significantly improve their business
case for theft as the percentage identified would increase to almost 100%
and the percentage recoverable would increase as well.
– Reliable outage detection and restoration notification at the transformer
was seen as a huge benefit by all of the people interviewed. Each said that
they could build a viable business case on reliable data.
– Metering the transformer for volts, amps and power factor was not seen as
a large benefit and most viewed this as “too much information”.
• Larry Rushing – Manager for Distribution
– (314) 554-2412

Transformer Load Monitoring