Upload
randolph-montgomery
View
212
Download
0
Embed Size (px)
Citation preview
1
UFE Workshop
Sponsored by
COPS
October 19, 2004
2
AgendaAntitrust Admonition
Morning Session - Ed Echols1. Review of meeting and minutes from September2. Discuss the Protocol timelines regarding UFE3. Update on ERCOT staff and PWG work on protocol revision to calculate
Distribution Losses based on actual load for the operating day4. Determine if there is general consensus on a problem statement5. Group development of solutions list to resolve the stated problem (Dependent on
consensus around problem statement – no solutions are needed if a problem is not definable) Prioritize the list Build out high level requirements needed to accomplish the solution Afternoon Session - Carl Raish
6. Develop a market participant consensus on a how to measure and make UFE monitoring available to the market List items desired for analysis
What data is needed to perform that monitoring/analysisWhat data is available
Is it important to update previous UFE reports from 20027. Schedule next meeting
3
Interval DataMetered Accounts
Profiled Energy UsageNon-Interval Data
Non-Metered Accounts
Distribution Line Losses
ERCOT Wide Transmission Line
Losses
UFE +/-
Net Load (Generation) for
Settlement Interval
(Includes Actual Losses in
the UFE Zone)
GAP - - - - - - >
CALCULATION OF UFE
Net Generation
compared to
Retail LoadBuild-up
4
DISTRIBUTION UTILITY• Inaccuracy of method used to calculate distribution losses.
• Unrecorded services.
METERING AGENT • Incorrect meter data.
• Inaccuracy in calculation of un-metered service consumption.
• Meter reading errors (gen. or end use meters)
• Errors in estimation of meter readings.
ERCOT SYSTEMS• Inaccuracy of load profiles on a settlement interval basis.
• Incorrect aggregation of retail load or zonal generation.
• Inaccuracy in method used to calculate transmission losses.
• Incorrect assignment of customer to profile type.
•Incorrect assignment of customers to UFE zone.
•Theft
Contributors to UFE
5
FactorNumber Description
ResponseAverage Min Max
Associated with Methodology: 1 Inaccuracy of load profiles on a settlement interval basis 54.78 27.0 732 Inaccuracy of method used to model transmission losses 6.67 1.0 273 Inaccuracy of method used to model distribution losses 8.78 2.0 27
Incorrect aggregation of metered load data :4 Customers assigned to wrong UFE area 2.35 0.0 105 Double counting of meters 1.26 0.0 56 Bad load meter data (e.g. multipliers, C.T.s, etc) 4.67 0.1 107 Theft 3.07 0.1 7
Incorrect aggregation of generation metered data:8 Incorrect netting of generation in a zone 1.00 0.0 29 Double counting of meters 0.33 0.0 110 Estimated meter readings 2.22 0.0 9
11Bad generation meter data (e.g. multipliers, C.T.s, etc) 1.56 0.0 3
Inaccuracy in estimating loads:12 Inaccurate estimation of meter readings 6.33 2.0 1013 Inaccurate calculation of un-metered service consumption. 4.39 2.0 7
14Inaccurate or missing estimate of unrecorded services (services that are being provided before meter installation) 2.44 0.0 5
Total Percent: 99.9
February 08, 2000 Survey 2 Results
Factors that Contribute Toward UFE(Percent of Cost Contribution per Settlement Period)
Utility Survey of UFE Contributing Factors
6
• Discussion began in Settlement Technical group in October 1999 and continued into March 2000 with a “UFE sub-team” formed in late December 1999.
• February 3rd and 10th, 2000 Settlement Technical Meetings focused on narrowing zone options and refining allocation data. Also started analysis of old control zone metering points to determine option feasibility for market open.
• February 24, 2000 Settlement Technical Meeting - Group developed alternative hybrid recommendation for individuals to present to their respective companies for future decision making.
• March 03, 2000 - Settlement Technical group detailed two options to be forwarded to the Retail Users Group for voting.
History of UFE Decisions
7
What work has already been done that improved settlement accuracy?
Modifications have been made to aggregation algorithms to improve load estimations
ADU vs Usage Factor Extended look-back period for IDR estimation PRRs to extend NIDR data available for estimation to 12
months (currently 6 months) Adjusted BUSIDRRQ profile Removed gap validation for usage data loading
Analysis of 2002 UFE presented to BOD & available on website
8
Is there a consensus problem statement the market wishes to address in the UFE Task Force
9
What is needed to assist the market in Monitoring Measuring and Analyzing UFE
10
Some ERCOT Suggestions
Update ERCOT report currently on the website to include results for calendar year 2003 … determine where we are now with changes already implemented
Consider improvements for ESIID-to-substation assignment Potential impact to current settlements if incorrect ERCOT test settlements at substation level have
shown problems … mismatch of load with substation telemetry
ERCOT validations based on zip code continue to show many suspect assignments
Suggest geo-coding service address into latitude and longitude and perform validations on location/distance measures
11
Some ERCOT Suggestions (continued) Evaluate loss calculations and profiling error
Compare settlement version of annual kWh to substation telemetry at TDSP level … removes profiling error and substation assignment error from comparison
Substations are metered (~50%), observable (~40%) and unobservable (~10%)
Drill down are there systematic differences by time-of-use? By voltage level?
Are substations primarily serving business load or residential load more accurate (profile related issues)?
Are substations in some weather zones more accurate than other zones (also profile related issues)?
12
Some ERCOT Suggestions (continued) Compare UFE at the substation level
If previous comparisons are good enough, compare UFE across substations.
If UFE is uniform across all substations, we probably do not need UFE zones.
If UFE is not uniform, can we identify zones where it is uniform?
Is there enough benefit in terms of settlement accuracy to justify the metering investment to implement UFE zones?
Could we consider continuing with current method of UFE calculation allocate to substations by their telemetered UFE share?
13
Some ERCOT Suggestions (continued)
Improved settlement estimates of load at the substation level would be beneficial to ERCOT operationsCould produce better estimates at
unobservable substationsImproved transmission planningImproved congestion managementTNT / State Estimator more accurate
14
MCPE * UFE
2003 2004 (J an 1 - J ul 14)
Net 170,888,230$ 30,863,873$ Positive Intervals Only 236,198,494$ 69,401,081$ Negative intervals Only (65,310,264)$ (38,537,209)$ Absolute Value 301,508,758$ 107,938,290$
NEGATIVE UFE does not result in Money Changing Hands… it is the means to balance what Resources brought to the market with who should pay for that service.
Negative UFE is an accounting treatment to balance what actually was delivered by reducing what Aggregation processes over calculate – ie Negative energy was never there in the first place and cannot be charged to any entity
Absolute value analysis of UFE dollars will be overstated if Negative UFE is not excluded.