|
local
transformers, and power cables, the company found it could
use the AMR information to predict equipment failures that
could cause power outages. The ability to model loads on its
distribution network is expected to help the company reduce
outages and gain lower operational costs.
Situation
Exelon Corporation is one of the largest electric utilities
in the United States with 2006 revenue of U.S.$15.6 billion,
income of $3.5 billion, and one of the industry's largest
portfolios of energy generation capacity. The company, which
serves more than 5 million customers, was created in 2000
by the merger of PECO Energy Company of Philadelphia and Unicom
of Chicago (owner of Commonwealth Edison).
In 2004,
Exelon's PECO Energy Company completed deployment of a fixed
network automated meter reading (AMR) system from vendor Cellnet
to provide daily automated meter reading services for some
1.7 million electric and 500,000 gas meters in Philadelphia.
The AMR system is based on a radio-frequency fixed network.
Each meter is equipped with an AMR module that every 5 minutes
transmits data to local control units which relay the data
to the control center.
Rather
than sending human meter readers to homes and businesses,
the Cellnet AMR system automatically transmits meter readings
to the utility. The system also delivers data that can flag
meter tampering, and is designed to receive and process special
meter information packets, such as "last-gasp" outage
messages when power is lost, and "power-up" restoration
messages.
The AMR
data is so rich in information, that Exelon sought additional
ways to pull value from the system by using data from the
electric meters to reduce outages, as part of its "smart
grid" practices, an industry term referring to the automated
gathering of distribution data that can be used to enhance
service and conserve energy. To accomplish this it needed
to create a database to help analyze information that might
help it predict and prevent points of failure in the delivery
system.
Solution
Working with Microsoft® Utility Partner Itron, Exelon
launched a pilot project to determine whether AMR data could
be used to reduce the occurrence of and severity of power
outages. Itron created its Distribution Asset Analysis (DAA)
meter data analysis program, using Microsoft SQL Server®
2005 database software as the relational data store, running
on the Windows Server® 2003 Enterprise Edition operating
system. The solution was deployed on a single Intel-based
computer.
"We
have been pleased with the performance and ease of use of
SQL Server 2005," says Eric Miller, Vice President of
Software at Itron. "The native capabilities of SQL Server,
as well as its integration into the full Microsoft reporting
and workflow tools make it a great choice for high-scale business
intelligence applications like DAA."
An enterprise-grade
database is required because the AMR information transmitted
every 15 minutes from 1.7 million electric customers generates
more than 7 terabytes of data per year. To support richer
modeling of the distribution network to support smart grid
initiatives, the full DAA deployment will work with three
years of data, generating database loads of some 21 terabytes
and 750 million rows.
The application
integrates with SQL Server 2005 to create complex analytical
models that represent the entire power distribution network,
with special emphasis on modeling the loads on four distribution
elements:
Unit
substations. Large stand-alone transformers serving between
1,000 and 2,000 customers.
Step-down
transformers. Step-down transformers are large transformers
typically found on utility poles or in underground power vaults
that reduce primary voltage from a typical 13,200 volts to
2,400 volts.
Local
transformers. Found on utility poles or in underground
locations, local transformers provide neighborhood distribution
for small collections of users.
Power
cables. The modeling included the cables used throughout
the distribution system, from unit substations to business
and residential delivery.
For the
pilot project, a subset of one year's worth of AMR readings
that already had been collected by the AMR system, was analyzed.
Predicted failures from the analysis were then compared against
what actually happened.
"By
creating a circuit model of the full distribution network,
we're hoping to gain better insight into how the distribution
system is operating," says Glen Pritchard, Consulting
Engineer, Meter Reading Technologies, at Exelon. "Basically
we've designed a family tree showing power distribution. The
circuit starts at the substations and then it splits many
times until it gets to the customer. If a transformer has
five customers on it, the model sums up the load for each
of the customers on that transformer. We can easily extend
the model to include fuses, and look at the loads that five
transformers behind a fuse are creating."
Benefits
The pilot project demonstrated to Exelon that analyzing AMR
data using Itron's DAA meter data analysis application supported
with SQL Server 2005 could provide the utility with a number
of smart grid benefits, including the ability to predict and
prevent power failures, avoid collateral damage, reduce operational
costs, and help model customer needs.
Predict
and Prevent Power Failures
The pilot project confirmed that new value could be drawn
from the same AMR data already collected for billing, and
that AMR data can bring value to an organization's outage
management system (OMS). Exelon found that integration of
AMR data into OMS business processes should help the company
to:
- Improve
planning and engineering
- Increase
crew productivity through predictive maintenance
- Reduce
outages
- Enhance
customer satisfaction
"Utilities
normally don't associate AMR information with reducing operational
outages," says Pritchard. "But our pilot project
demonstrated that we could predict overloads developing within
the distribution system, and such information can be used
proactively to intervene to prevent distribution failures."
Examining
the information collected on the SQL Server 2005 database,
Exelon identified five transformers that appeared to be overloaded.
When the predictions were tested against real-life incidents,
two of the transformers had already failed.
"Had
we been working in real-time mode, as we hope to be in the
future, the DAA meter data analysis running on the SQL Server
database would have flagged these overloaded transformers
for attention," Pritchard said. "This would have
given us the warning we need to act proactively, avoiding
the expense of doing maintenance in emergency mode, and relieving
our customers from the disruption of a power failure. We're
also modeling cable sizes and loads to predict and prevent
line failures."
The modeling
information will enable Exelon to improve planning and engineering,
and to increase crew and dispatch productivity by reducing
the disruptions that require crews to work in emergency mode.
The smart
grid solution is flexible enough to support a spectrum of
modeling options.
"We
can use the DAA meter data analysis solution hosted on SQL
Server to create rich scenarios, such as: What happens if
we have five consecutive days of 95-degree temperatures?"
says Pritchard. "We can use historic data and our modeling
to create a wide variety of scenarios to gauge the sensitivity
of different elements of the distribution system. We're finding
more and more innovative ways to draw value from the AMR information.
All of this helps reduce the chance or duration of an outage,
which greatly enhances customer satisfaction."
Avoid
Collateral Damage
Adding to the problems of a transformer, cable, or other element
of the power distribution delivery system failing is the fact
that such occurrences can sometimes cause collateral damage
to unrelated utilities, such as telephone and cable infrastructure
especially when the equipment is co-located in duct banks
and other underground utility spaces.
"When a transformer or cable fails, you can end up with
manhole fires that can cause considerable damage to unrelated
cables or circuits," says Pritchard. "Even if the
neighboring cables aren't damaged by flames, there can be
collateral damage. Using DAA and our SQL Server database to
predict and prevent equipment failures helps eliminate the
collateral damage that can occur."
Using
predictive modeling of the power distribution system to prevent
local power failures can also help prevent triggering wider
outages that can occur when local failures trigger cascading
events that can result in widespread blackouts.
Reduce
Operational Costs
The DAA meter data analysis should help reduce operational
costs by providing a clearer view of what is happening with
the power distribution network. "Using SQL Server to
analyze our AMR data is giving us the information we need
to work smarter, with a much more precise model of what is
happening with our power distribution," says Pritchard.
"We anticipate this will lead to lower operational costs
as we are able to more precisely direct our maintenance efforts
to avoid power failures rather than reacting to failures after
the fact."
Prior
to storing the AMR information on a SQL Server database, it
was a time consuming job to sit at a keyboard assembling information
for specific trouble spots. "Without SQL Server we would
have to gather data from the AMR system meter by meter, and
then analyze what we found to gain a single result in time,"
says Pritchard. "SQL Server and our DAA application automate
the process and provide data on a system wide basis to provide
a rich model of what is happening. This information helps
us work smarter. And working smarter reduces operational costs."
Model
Customer Needs
The same data modeling that helps predict and prevent power
failures can help Exelon plan for future growth. "Collecting
our AMR data on SQL Server for analysis enables us to gain
a degree of system status that we've never had before,"
says Pritchard. "This gives us information that will
help us monitor growth in customers and energy demands so
that we can on a very granular basis, ensure that our infrastructure
isn't being overloaded. We have gained analytic information
on customer usage patterns to create modeling assumptions
that show how much growth can be accommodated, and at what
point different parts of the distribution system need to be
updated."
Applying
analytics to AMR data provides a spectrum of opportunities
for utilities.
"Increasingly,
our utility customer will be creating and implementing many
new BI applications that leverage the wealth of data provided
by advanced metering systems," says Miller. "The
Microsoft platform makes sense both for software providers,
and for utility IT personnel developing specialized applications
for their business users."
In summary,
Exelon has gained visibility into its operations by importing
AMR data into its DAA meter analysis application supported
by a SQL Server 2005 database, and creating distribution system
models that help it to predict and prevent outages.Microsoft
Server Product Portfolio
For more
information about the Microsoft server product portfolio,
go to:
www.microsoft.com/servers/default.mspx
Microsoft
SQL Server 2005
Microsoft SQL Server 2005 is comprehensive, integrated data
management and analysis software that enables organizations
to reliably manage mission-critical information and confidently
run today's increasingly complex business applications. By
providing high availability, security enhancements, and embedded
reporting and data analysis tools, SQL Server 2005 helps companies
gain greater insight from their business information and achieve
faster results for a competitive advantage. And, because it's
part of Windows Server System, SQL Server 2005 is designed
to integrate seamlessly with your other server infrastructure
investments.
For more
information about SQL Server 2005, go to: www.microsoft.com/sqlserver
For more
information about how Itron DAA software can help you optimize
distribution system efficiency and reliability, e-mail us
at knowledge@itron.com
Itron
wishes to thank our strategic partner, Microsoft, for enabling
us to publish this case study document in Itron News.
|
