Research - Regional Technical Forum (RTF)

Distribution Efficiency – Quick Review
• Tier 1 study (19 circuits) completed in 2011
– Four circuits identified for pilot (Mill Creek, Clinton substations)
– Seven circuits had at least one unstudied adjacent circuit
Identified influential circuit parameters for CVR
Simplified Measurement & Verification Protocol for pilot
2012 savings target: 0 – 0.111 aMW (972 MWh) from pilot
Planned Tier 2 study, circuit prioritization
PacifiCorp’s 2012 Activities
• Engaged in industry research
– DSTAR ( Project P13-7 (still in progress)
• Distribution Systems Testing Application and Research
– NEETRAC ( Project 11-136 (still in progress)
• National Electric Energy Testing Research and Applications Center
• Engaged in Regional Technical Forum (RTF)
– Automated CVR Subcommittee & protocol development
• Review of applicable PacifiCorp standards
• Clinton and Mill Creek pilot projects
• Tier 2 study in Washington
DSTAR Findings
1. Demand :: Voltage relationship varies w/ feeder, time of
day, season, and over the long term
2. Estimating circuit CVR factors from lab values (bottom
up) cannot work (too little information)
3. Accurate long-term day on/ day off is extremely difficult
4. Increased frequency & severity of voltage sags
5. Software estimates may not suffice for M&V
NEETRAC Findings
1. Statistical analysis performed on pilot by Palmetto
Electric Cooperative (SE South Carolina)
Could not apply RTF’s Automated CVR Protocol #1 due to limited
13 interested utilities participating in project
2. Demand :: Voltage varies w/ temperature, humidity,
demand on previous day, and demand on same day of
previous week
3. Must be analyzed on circuit by circuit and day by day
RTF Progress
• Currently two custom CVR protocols exist
– Simplified VO M&V Protocol (Approved 5/2010)
– Automated CVR Protocol #1 (Approved 4/2004)
• Both are under review for compliance w/ new (6/1/2011)
RTF guidelines*, including:
– A maximum error introduced by sampling
– Sources stated for all assumptions
– Results are reliable and sufficient to meet the needs of planners and
“support regulatory processes related to the adoption and planning
of energy efficiency initiatives.”
*Guidelines are available at
Questioning Protocol Assumption #1
• Loads do not have to be distributed evenly?
Questioning Protocol Assumption #2
• End-of-line voltage varies linearly with load?
– Example from “Robustness of Simplified VO M&V Protocol
Measurement Period of 7 Days” by Utility Planning Solutions, PLLC
– If end of line voltage can be estimated by load, then several
calculations can be “simplified”
– The chart shown was derived from software, not actual reads
Questioning Protocol Assumption #2
End of Line C phase voltage
• Actual pilot measurements show low linear correlation
 15-minute
averages shown
 Large (~ 2v)
bandwidth for
any given load
 No model
predicts this
Substation MVA
Measurement & Verification Considerations
• Physics of individual appliance (lab test) well understood
• Physics of aggregate system not understood
– Variables exist on every level, and they change over time
• Substantial assumptions have to be made for M&V
– Pre- and post-reduction average voltage must be estimated for all
customer locations, all year long
– Energy response to voltage reduction must be estimated for all
customers, all year long (VO or CVR factor)
– To date, no way to determine accuracy of estimates over time
• Who can find a consistent, accurate, low cost method to
determine energy savings from voltage reduction?
Review of PacifiCorp Engineering Handbook
1C.2.1 Voltage Level and Range June 2004
– High level look at ANSI service & utilization voltages
1C.2.2 Steady State Voltage Level Nov. 1998
– Settings methodology (§ 4.1.3 and § A4) is CVR
1C.3.1 Voltage Balance Nov. 1996
– Cites 3% maximum voltage unbalance as target
– Accounts for economics, lowest total cost to customer and utility
1E.3.1 Distribution System Planning Study Guide Feb. 2000; minor updates Feb.
– Discusses current unbalance, voltage unbalance, LDC settings, and
capacitor location & settings in § 7.4, 7.8 and 7.11, and economics in
§ 8.3
Overall a good balance of economics and efficient system design
Simplified Protocol states that standards become “entrenched” after three
Clinton/ Mill Creek Pilot
• Using two 7-day test periods for M&V yields energy savings of
±0.54 aMW, before improvements were initiated.
• Example from 5Y610 (VO factor 0.446):
 Week of 6/25: average voltage = 122.8
 Week of 9/10: average voltage = 121.2
• How do we handle data issues?
VO energy change by
Simplified Protocol:
0.14 aMW
(Biennial target = 0.111 aMW)
– One missing value affects all calculations
– What if one of nine protocol thresholds is not met for a period of
time? Is the circuit non-compliant?
• At what point is the ‘cost-effective, reliable, feasible’
criterion of RCW 19.285 not met?
2012 Tier 2 Study
• Completed by SAIC, included 25 circuits
• Confirmed Tier 1 VO factor was too high (~30%)
• Reaffirmed Tier 1 finding that existing practices are very
• Company’s avoided cost updated to $86.74/MWh
• Nine additional viable circuits identified
• Other circuits failed to comply with ESUE* thresholds, or
had too low a benefit/cost ratio (BCR)
*ESUE = Energy Smart Utility Efficiency, a Bonneville Power Administration
program that includes system operational metrics for efficiency, is used in the
Simplified Protocol.
Anticipated Tier 2 Projects
• Six circuits from Orchard Substation (Yakima, WA)
Average cost $30k per circuit
22 phase balancing locations
One line regulator bank
Lower substation base voltages from 121 to 120 & 119
BCR = 1.15, saving estimated 0.10 aMW
• Three circuits from Sunnyside Substation (Sunnyside, WA)
Average cost $68k per circuit
4 phase balancing locations
One line regulator bank
Lower substation base voltage from 121 to 119
BCR = 1.48, saving estimated 0.08 aMW
• M&V alternatives are being discussed
Washington DE Plan
• 2013
– Stay engaged with the industry
• RTF hopes to have new protocol usable by 2013 Q3
– Continue biennial pilot with Orchard T-3797 and T-5035
• Consider M&V by software
– No further detailed circuit studies
• 2014
– Implement Sunnyside T-3570
Extra Slide

similar documents