Regional Technical Forum Automated Conservation Voltage

advertisement
Regional Technical Forum
Automated Conservation Voltage Reduction Protocol #1
Meeting Agenda and Notes:
December 12, 2012, 2:04 pm – 4:20
Attendees:
Bob Fletcher, Erin Hope-BPA, Steve Brooks-BPA, Tom Wilson-PCS UtiliData, Don
Jones Jr., Pacific Power, Wyatt Pierce-Pacific Power, Ross Taylor- Avista Senior
Distribution Engineer, Mark Kendall – RTF, Adam Hadley-RTF, Takala, Gus , Gowan,
Kevin, Ray Hisayasu, - PSE Transmission, Tom Leinhard-Avista,, Eric Brateng-PSE,
Charlie Grist – RTF, Mike Baker- SBW, Gary Grayson,- Idaho Power, Chris Milan –
BPA, Marc Peterson – Idaho Power,
Agenda:
1) Subcommittee introductions and experience with CVR - Kendall
2) Subcommittee purpose, (e.g. custom protocol for Automated CVR), - Kendall
3) Summary of the RTF Guidelines for Custom Protocols - Kendall, Baker
4) Some background on the RTF's Automated CVR Protocol #1, and Voltage
Optimization Protocol - Fletcher and other subcommittee members
5) PCS UtiliData's proposed update to CVR #1 - Wilson
6) Define data and protocol descriptions we need to complete a custom protocol Fletcher and Subcommittee
7) Next steps, homework, schedule next meeting - Fletcher and Subcommittee
Notes:
Introductions:
The meeting was convened by Kendall at 2:04 and introductions were conducted. A
number of regional distribution engineering experts were in attendance. Kendall
introduced Fletcher as the RTF contract technical specialist.
Overview and Purpose:
The overall purpose of the subcommittee is to develop a Standard Protocol for
Automated Conservation Voltage Controls that conform to the RTF’s Operational
Guidelines for development and Maintenance of RTF Savings Estimation Methods.
The meeting today is to familiarize the group with the current CVR Protocol #1 and
the Simplified Voltage Optimization Protocols of the RTF and describe how those
will need to be updated to meet new Guidelines.
Guidelines Introduction:
Mike Baker was introduced as the RTF’s representative regarding conformance with
the guidelines. Baker provided an overview of Standard Protocol Measure
requirements. Mike defined a standard protocol, defined why the CVR technology
did not fit the RTF custom protocol guidelines and proceeded to identify what is
needed for the RTF to approve a standard protocol. Baker identified the data
standards, the pre and post measurement required to provisionally prove the
protocol is accurate on a sample of projects, the measure and procedure
specification standard, the energy estimation methods (statistical, engineering….),
the baseline definition and defining it, the efficient case definition and verification
requirements, and all the documentation required of the practitioner and for the
protocol itself. See the Guidelines for Development and Maintenance of RTF
Savings Estimation Methods at the RTF web site for detail. A very high level
summary was presented in a PowerPoint presentation.
CVR Protocol #1 and Optimized Voltage Control History and Characteristics:
Fletcher prepared and presented two power point slide shows (posted on the
Subcommittee Web Page):
 Simplified VO Protocol 2010 Discussion (slides 1-15)
 CVR No.1 Protocol 2004 Discussion (slides 1-6)
Other documents were provided to the group with the meeting announcement for
their review and reference:
 Voltage Optimization Protocol 2010
 Automated CVR Protocol 2004
 Automated CVR Performance Estimation Methods 2004
 Guidelines for RTF Savings Estimation Methods
 Guidelines for MV Federal Projects
 IPMVP Guidelines 2002
Discussion included many technical aspects of dealing with various elements of
distribution feeders. For example, the protocol is lacking instruction on how to
select and document the site for measurement of line voltage (substation, end of
line, average…) when there are line drop compensators, power factor correction or
other switching on distribution lines being treated with CVR. The protocol needs to
address those issues.
The following questions were raised during the discussion that address what is
needed for the protocol to be well understood, savings estimation reliable and
transparent and methodology clear enough to be assured of replicability. The
automated CVR protocol questions, and issues that need to be addressed to meet the
guidelines criteria include:
1. Does the CVR Protocol No 1 meet minimum RTF guidelines for Custom SiteSpecific savings estimate and cost-effectiveness calculation methodology?
2. Is the CVR Protocol No 1 applicable to a wide variety of distribution systems
configurations and operation practices (residential feeders, commercial,
industrial, mixed feeders, inside industrial facilities)?
3. Does the CVR Protocol No 1 adequately ensure comprehensive distribution
system cost-effective efficiency? This was raised as not being the point of the
protocol by some others.
4. Does the CVR Protocol No 1 adequately ensure that distribution system
safety and facility standards are meet? The protocol must operate within the
safety requirements and constraints of the utility.
5. What changes are necessary to simplify the automated CVR protocol
requirements and calculation methods?
6. Verification of statistical methods to adequately evaluate savings
7. Simplified statistical measure savings calculation methods clearly apply to
the field data collected
8. Distribution load flow analysis requirements
9. Facility data requirements of lines, equipment, and loads
10. Characteristic load models (ZIP) requirements
11. Historical load data and temperature data requirements
12. Distribution system efficiency impacts
13. Location and number of voltage EOL measurements need to be specifically
defined.
14. Location and number of temperature sensors
15. Voltage measurement locations (e.g., Veol, Vsub, Vreg, Vcap) defined in the
protocol
16. System average voltage or end-of-line average voltage measurement
locations or averaging method from data collected
17. System performance requirements (i.e., Vmax, Vmin PFmin, VDsec-max,
Vsource-max, VDpri-max, Control Bandwidth)
18. Does the standard protocol apply to wide variety of automated CVR volt/var
systems and vendors equipment?
Proposed Protocol Reviewed:
Tom Wilson of PCS UtiliData reviewed the draft protocol text itself, by category. He
discussed the sections that addressed: the purpose of the protocol (scope), sunset
criteria for the protocol, definition of key terms, eligible project definitions, required
practitioner knowledge, data collection requirements, post period re-verification
triggers, baseline performance, method description, the model and savings
estimation method, the control groups required, relationship to other RTF or other
protocols, and the recommended methods and/or tools for conducting the
verification and calculation. He provided a draft protocol in the current RTF format
which is posted at the RTF subcommittee web page.
Discussion raised the following questions regarding the proposed protocol
language.
 The best practice for the protocol is not clearly described and it appears that
the protocol itself may be the best practice,
 The data collection description does not specifically describe how one
chooses the "other" months for the data collection period,



The calculator algorythms are not specifically described and the calculator
transparency is not described,
The skills required of the practitioner do not appear to be "widely
represented" in the region as indicated for a Standard Protocol,
Use and incorporation of time series data in the algorythms and calculator
are not clearly enough described that we can assure that practitioners will
yield the same result subsequent times or across practitioners.
Written comments from Subcommittee (PacifiCorp):
Section 1, Purpose, should include a brief scope definition that differentiates the
applicability of the protocol to utilities that (a) already have automated CVR systems
in place and seek more accurate M&V, (b) utilities that are evaluating the potential
cost effectiveness of an automated CVR system but have not yet implemented one.
This section could point the reader to appropriate resources for estimating the
savings that would cost justify the implementation of an automated CVR system (I.e.
those in category (b), who cannot use this protocol to justify a future installation of
an automated CVR system).
Section 3, Definition of Terms, provides a somewhat misleading definition of CVR.
The most general purpose of CVR is to use less energy in aggregate, recognizing that
some utilization devices (LCD and plasma televisions, for instance) use more energy
at a lower voltage. A more accurate description might be:
Conservation Voltage Regulation/Reduction (CVR): CVR is the operating
practice of controlling distribution feeder voltage with the goal of using less
total electrical energy and reducing electrical demand. This is accomplished
by providing a utilization voltage near the low end of the acceptable range
and/or rated voltage, where many utilization devices operate at their peak
efficiency.
Section 3, Definition of Terms, further propagates existing confusion over the
difference between CVR and VO factors. The RTF has previously made the
distinction (see May 4, 2010 RTF Meeting) that VO factors apply to end-use
utilization, while CVR factors include both the end-use and distribution system
savings. This section explicitly references the utilization device’s efficiency curve,
which should apply to a VO factor, not a CVR factor. If the protocol is to use the term
“CVR factor,” then it should explain why “VO factor” is not the more appropriate and
consistent term.
Section 3, Definition of Terms, does not define the terms “heating regime hours” and
“cooling regime hours,” which are used in Section 7.
Section 4, Eligible Projects, appears to describe a fully implemented pilot project. If
the protocol is not intended for pilot projects, this would be a good location to
mention refinements in calculations that improve upon the M&V methods and
assumptions used prior to 2011. That is to say, a utility with a system that fits this
definition must have been using some other method to calculate savings until now,
and that method is evidently not as accurate as the method described in the
protocol.
Section 5, Required Knowledge, describes an extremely specialized individual. In
fact, given the ongoing research into the physics of CVR by institutions such as EPRI,
NEETRAC, PNNL, and others, probably no person could legitimately claim to have
“full understanding” of the first bullet. The remainder of the bullets appear to lean
heavily on the characteristics of the submitter, and would not describe a party who
is reading the protocol for guidance. A more reasonable description of attributes
might be:
Knowledge and Skills of Practitioner
The practitioner who has lead responsibility for applying this protocol to an
automated CVR system should have appropriate knowledge of the following:





The application of CVR to distribution systems and the underlying
physics of relationships between operating voltage levels and energy
consumption
The use of engineering time series analysis
The use of Robust Statistical procedures used to analyze non-Gaussian
data
Distribution feeder and substation operations
This protocol, and specifically its requirements and procedures
The practitioner must also be able to successfully:


Operate an appropriate time series load flow application, such as
Gridlab-D, OpenDSS, or MatLab
Inspect and interpret raw feeder energy and voltage data
Section 6.b, Metering and Data, should include a requirement to record the number
of operations on the regulating device(s), for the purpose of identifying any increase
in equipment maintenance costs or reduction in equipment life.
Section 9.b, Program Savings, is missing a line break in the third bullet, between
GWh and CVRf.
Section 9.c, CVRf Estimation, appears to describe a method which yields a value with
kWh/volt units, which does not match the Definition of Terms in Section 3. The 24hour sum of the difference between the integrated demand profiles should be
divided by an energy value (such as the total inactive temperature-corrected
energy) to get a unitless “energy” value. The mean difference in end of circuit
voltages should be divided by a voltage value (such as the mean inactive end of
circuit voltage) to produce a unitless “voltage” value. Then the two unitless values
could be divided to produce the CVR factor that is described in the Definition of
Terms.
Section 10, Model, states that energy usage has been compensated by temperature.
The surrounding discussion (Sections 4 and 6) suggests the compensation accounts
for hourly temperature values. It should be observed that the circuit load on a cold
day can be substantially less than the circuit load on the third or fourth cold day in a
row, even if the temperature is the same each day. If this is accounted for in the
model, it should be stated here. If it is not accounted for in the model, it should be
listed as one of the additional factors that will improve accuracy.
Next Steps and Schedule:
A couple members of the group want to bring Kevin Schneider of PNNL into the
discussion. He is knowledgable, has applies CVR and conducted evaluations and
protocol analysis. One person brought up that other manufacturers such as GE,
Siemens, Beckwith Electric, ABB or others should be involved. The next steps
include:
1. Develop four separate protocols for subcommittee review:
a. Update Automated CVR Protocol #1 for feeders with existing conditions
b. Automated CVR for feeders that are optimized to IEEE standards
c. Optimized feeders (reconductoring, capacitance, voltage drop
compensation, phase balancing…)
d. Update the Voltage Optimization Protocol
2. RTF staff posting the presentation(s) and discussion. On a Subcommittee web
page.
3. UtiliData and Fletcher identify some responses to the questions raised and
relate them specifically to the pertinent sections of the draft protocol.
4. Have protocol language editing Subcommittee meeting for each protocol.
5. Develop calculator specifications and calculator for testing on the 31 NEEA
feeder data. (RTF to collect from RW Beck).
Download