previous versions at the Internet Archive. Yahoo! is not affiliated with the authors of this page or responsible for its content.
Microsoft PowerPoint - MarkInglis07202004 ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Leveraging Load Research and Customer
Consumption Data to Improve Circuit
Utilization
AEIC
Annual Load Research Conference
St Louis, Missouri
Mark Inglis
Nexus Energy Software
July 20, 2004 ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Presentation Agenda
1. Why Is Distribution Utilization Important?
2. Overview of Methodology, Required Data & Systems
3. Case Study: Idaho Power Company
4. Case Study: East Coast Utility ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Why Is Distribution Utilization
Important?
Priority 1:
Increase
Reliability
i.e. Increase
Distribution
Capacity
Priority 2:
Reduce Capital
and O&M Costs
i.e. Increase
Distribution
Utilization The Irreconcilable Conflict
Or is it?
Perhaps both are possible, with a more precise and
granular understanding of system utilization ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
What are we Discussing Today?
Using load research data, combined with other
information around the utility, to estimate
hourly load patterns for distribution assets.
The goal is to improve distribution asset utilization
This is an important, often overlooked, use of
the valuable load research data.
We are presenting our approaches and findings
at utilities around the U.S. and would be happy
to provide a complementary presentation at
your utility ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Circuit Devices
Customer Transformer: a device (usually single phase) that steps the voltage
down from the circuit to a level appropriate for delivery into a customers premises.
There are anywhere from 1 to 10 customers attached per transformer. Some
transformers connect to two phases, while others will connect to three.
Primary Transformer: a transformer (usually three phase) that is used to step the
voltage of a feeder down to provide power for a downstream section of a feeder. This
can occur when large portions of a feeder require lower voltage levels.
Capacitor: an electrical device used to store electrical energy, and to release it back
into the power system when required. Capacitor perform a stabilizing function in an
electric circuit.
Fuse: a device that will heat up, melt and electrically open the circuit after a period
of prolonged abnormal current flow.
Recloser: an automatic, high-voltage electric switch that will automatically return to
normal after disturbance has passed.
Regulator: a device used to maintain constant voltage levels across a distribution
network
Switch: a device that alters the direction of current flow on a circuit
Conductor: the wire that conduct the current down a circuit. Conductors are usually
treated in sections and are rated to carry at specific level of current. For Wire Vision,
it remains to be determined whether it will be sufficient to sectionalize conductors
only when the conduct size changes. ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Hierarchy
AA
AB
AC
AD
AE
AF
AG
AA
AM
AQ
AI
AK
AL
AO
AN
AT
AV
AW
AGB
AJ
AR
AH
AP
AGA
AS
AU
AGA-SW
AGB-SW
AW-SW
AB-LT
AC-LT
AC-LT
AF-L
T
AF-L
T
AF-L
T
AI-LT
AJ-LT
AM-LT
AM-LT
AM-LT
AQ-L
T
AR-LT
AR-LT
AS-LT
AS-LT
Node AA
Node AGA
Node AGB
Node AGH
Node AM
Node AN
Node AS
Node AT
Node AU
Node AV
Node AW
Node AP
Node AO
AG-L
T
AU-LT ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
What Do Our Business Users Do?
Distribution System Operators:
Monitor conditions on the circuits;
Perform load switching to ensure reliability;
Manage planned outages and conduct planned switching;
Respond to faults, failures and maintenance requests.
Distribution System Planners:
Run short-term annual load flow analysis;
Identify short-term activities needed to maintain reliability;
Run long-term forecasts and load flow analysis;
Model new system additions needed to maintain reliability or
serve new load;
Monitor progress of planned activities. ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
How is this Estimation Currently
Done?
Wide variety of approaches, but basic approach:
Meter peak demand upstream for a substation or
feeder
Allocate this peak to the devices below the meter
using nameplate capacity
Do load flow analysis on the results ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Everyone talks about asset management, so
what is it in simple terms?
Asset Utilization
Asset Condition
Operating Expenses
Capital Costs
Performance /
Reliability
Asset
Management
System
Inputs
Measures
Results ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
So gaining a better understanding system utilization
requires more hourly or sub-hourly metering, right?
Can you tell where your customers are on your
distribution network?
Is billing meter data from your customers accessible?
Do you have at least some existing interval metered
customers or SCADA metered points?
Do you believe that statistics and sampling are real?
Do you believe that metering is not as easy, affordable,
or reliable as it may appear?
Not Necessarily
The answer to the metering question is not all or none,
but somewhere in between
(although much closer to none than you may think) ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Overview of Methodology
Segment
customer
billing
data by
customer
type,
region,
and weather
zone
Develop
weather-
sensitive
load
profiles for
each
segment
using
load research
samples
Develop
individual
hourly load
forecasts
for every
customer
using
weather
variables
Aggregate
load results
from GIS
hierarchy
and
compare
and
calibrate to
SCADA data.
Compare
hourly load
forecast at
any feeder
point and
phase with
GIS
conductor &
device ratings
Segment
Profile
Forecast
Aggregate
Analyze ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Generic Wire Vision Data Flow
Diagram
CIS DATA FILES
Customer ID
Meter ID
Monthly Loads
Zip Codes
Billing Cycles
NETWORK CONNECTIVITY DB
Device ID
Section ID
Device Parameters
Section ID
FromNodeID
ToNodeID
Connectivity Defns
Device Definitions
Node Definitions
Node ID
GIS location
Node IDs
Section IDs
TX CONNECTIVITY Files
Meter ID
Transformer Device ID
Device IDs
Meter ID
s
HISTORICAL WEATHER
Scenario ID
Weather Station ID
Hourly weather values
Zip codes covered
SCADA OUTPUTS
Circuit ID
Section ID
Hourly loads
SHORT-T. WEATHER
Scenario ID
Weather Station ID
48 hour weather values
Zip codes covered
MEDIUM-T. WEATHER
Scenario ID (up to 10)
Weather Station ID
3 month weather values
Zip codes covered
Scenario ID
Weather values
Zip codes covered
LOAD RESEARCH
Meter ID
Hourly loads
Zip codes
Meter IDs
Hourly loads
Zip codes
OUTPUT by VMETER
Meter ID
Scenario ID
Profiled Hourly Loads
LOAD VISION
ENGINE
WIRE VISION
EXTREME WEATHER
Scenario ID
Weather Station ID
1-year weather values
Zip codes covered
Circuit ID
Section ID
Hourly Loads
EXISTING CUSTOMER
GROWTH FORECASTS
Existing Meter ID
Load Growth %
NEW LOAD GROWTH
FORECASTS
New Load Base
Load Growth %
Section ID or Node ID
Meter IDs
Monthly loads
Zip codes
Meter IDs
TX ID
Device ID
Section ID
Node ID
Meter ID
New Load
Section ID
New Load
Wire Vision Outputs
Circuit navigation
Section & device loads
Device loading statistics ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Using Load Research Data:
Segmentation
The finer the aggregation, the more important
is correct segmentation
Transformer vs. System
Segmentation must be practical
Best measures:
Rate Class
Geography
Seasonal consumption patterns
Load Factor
Others? ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
Using Load Research Data:
Diversity
The finer the aggregation, the more important
is correct diversity:
Using a sample of 200 customers to create a profile for
a small transformer will bias results towards too flat
Using a sample of 20 customers to create a profile for
a large circuit will bias results towards too peaky
Our approach is to employ monte-carlo
simulation using load research data and profile
segmentation to derive diversity factor at time
of peak. ©2004 Nexus Energy Software . All rights reserved. Confidential Not for Distribution
What does this actually look like in
practice?