Transcript Advice from DOE-EPRI Framework Experience

Early Lessons Learned
from DOE-EPRI
Framework Experience
Melissa Chan
MA DPU Grid Modernization Working
Group
December 17, 2012
D I S P U T E S & I N V E S T I G AT I O N S • E C O N O M I C S • F I N A N C I A L A D V I S O RY • M A N A G E M E N T C O N S U LT I N G
Overview of the DOE Smart Grid Investment Grant Program
99
$3.4 B
2015
recipients
in federal assistance
($7.9 billion investment)
program ends
Our analysis involves linking installed assets to monetized benefits.
Assets
Assets are divided into four categories:
• Advanced Metering Infrastructure
(AMI)
• Customer Systems
• Distribution
• Transmission
Benefits
Benefits are divided into four categories:
• Economic Benefits
• Reliability Benefits
• Environmental Benefits
• Energy Security Benefits
Page 2
Four Primary Analysis Focus Areas
I will discuss our experience collecting utility data on cost and performance to analyze
impacts in these areas for the DOE Smart Grid Investment Grant Program.
Customer Response to
Time-based Rates,
Information Feedback and
Customer Systems
• Advanced Metering
Infrastructure
• Pricing Programs and
Customer Devices (smart
thermostats and in-home
displays)
• Direct Load Control
Operations and Maintenance
Savings from Advanced
Metering
Distribution System
Reliability
• Feeder switching
• Monitoring and health sensors
• Meter Reading
• Service changes
• Outage management
Operations and Maintenance
Savings from Distribution
Automation
•
•
Automated and remote operations
Operational Efficiency
Energy Efficiency in
Distribution Systems
• Voltage optimization
• Conservation voltage
reduction
• Line losses
3
Application of the EPRI-DOE framework
»
»
The EPRI-DOE framework is not a predictive model
› Develop an experimental design for smart grid asset deployment
› Establish a good performance baseline for analysis
› Set up methods to collect quality data for analysis
Data collected by the U.S. DOE Smart Grid Investment Grant program
may be used to inform future decisions
› Expected timeframe for data reports is at the program end, 2015
› Preliminary reports that summarize some utility data will be available on an
annual basis
Page 4
Smart Grid Impact on Customer Response
» Customer response to time-based rates, information feedback, and customer
systems (smart thermostats and in-home displays)
›
›
›
Collect hourly electricity usage data aggregated from the utility’s meter data management
system (MDMS)
Example analyses
‒ Hour-by-hour customer demand changes due to time-based rate
‒ Monthly conservation achieved due to information feedback
Example challenges
‒ Defining appropriate treatment and control groups and “tagging” the specific
customers in the (MDMS)
‒ Tying impact to benefit, such as understanding how change in peak customer demand
for one utility affects peak generation plant dispatch
» Lessons learned
›
Experimental design is needed to understand outcomes of customer pilot programs
Page 5
Smart Grid Impact on Utility Operations and Maintenance
» Utility changes in operations and maintenance costs due to automated meter reading,
enhanced meter functionality, and distribution automation
›
›
›
Compare service costs and activities with and without technology
Example analyses
‒ Savings due to automated feeder switching
‒ Truck rolls avoided due to changes in meter operations
‒ Savings due automated meter reading
Challenges
‒ Collecting granular cost data if the utility has not segregated costs (e.g., all equipment
failures are tracked but not categorized by type)
‒ Collecting granular activity data if the utility does not track activity by type (e.g., feeder
switching is not tracked separately from other distribution operations costs)
» Lessons learned
›
›
Create cost and activity tracking systems to meet desired data collection needs
Baseline can be difficult to estimate if bookkeeping is not categorized according to the
analysis need
Page 6
Smart Grid Impact on System Reliability
» Improvements in system reliability due to smart grid assets
›
›
›
Understand how meter outage management systems, automated sectionalizing reclosers
or other feeder automating technologies
Example analyses
‒ Changes in reliability indices (SAIFI, SAIDI, CAIDI)
‒ Changes in restoration time and number of customers affected by a major event
Challenges
‒ Collecting feeder specific reliability indices in the case that distribution automation
technology is not installed system wide
‒ Understanding historical variance in reliability indices
‒ Estimating status quo restoration for major events
» Lessons learned
›
›
›
Utilities may need to set up data management to collect reliability data for specific feeders
Major event response is easier explained qualitatively than quantitatively
Baseline can be difficult to estimate given considerable variation in reliability indices, which
can be due to location of outages or frequency of tree trimming
Page 7
Smart Grid Impact on Distribution Efficiency
» Changes in distribution efficiency – losses, power factor, and peak distribution load
due to voltage and VAR control
›
›
›
Compare service costs with and without technology (automated capacitor banks, voltage
regulators, load tap changers)
Example analyses
‒ Analyze hourly real and reactive loads for feeder groups that receive new equipment
‒ Estimate changes in losses before and after new equipment is installed
Challenges
‒ Defining appropriate treatment and control feeder groups or designing a statistical
analysis of treatment feeder group hourly load data
‒ Normalizing loss or power factor data for comparison with post-test loss or power
factor data
» Lessons learned
›
›
Average power factor and average losses are not useful metrics, power factor and losses
during the distribution peak are most useful
Analysis of treatment feeder group hourly load data can lend insight into capacitor
operations and optimization
Page 8
Some early lessons from the Smart Grid Investment
Grant program
» Experimental design is critical, particularly for customer facing programs
» Understanding operations and maintenance cost savings may be difficult to
»
»
»
understand because a utility may need to change its activity and cost tracking to
estimate changes due to technology
Treatment and control group data collection can be difficult because a utility may
need to change its data management in order to specifically collect data for
customers receiving customer facing programs
Isolating the affect of automation equipment on specific substations or circuits can be
challenging because a utility may need change how it tracks equipment failures or
maintenance, restoration, and operations activities
Not all impacts can be quantified, some may have to be discussed qualitatively
Page 9
Contact information
» Melissa Chan
» Navigant Consulting
» 781-270-8386
» [email protected]
Page 10