OLA / SLA Update

1. OLA / SLA Update Betty Day Nodal Advisory Task Force

2. DAM Status • Current and Testing Target • Currently testing 5000 Energy-Only Bids/Offers & 5000 PTP Obligation Bids: • Received patch MMS5 P4 (delivered end of 2009) • Able to process 5K DAM Energy-Only Offers/Bids and 5K PTP Obligation Bids • Good Quality of Solution • Need better performance • Tuning configuration parameters • Identifying performance bottlenecks • Patch MMS5 P5 will be delivered mid-February 2010 • The MMS5 P5 to be the version that goes into the START of Market Trials Phase 4 • Expecting that with MMS5 P5 we will get further improvements especially in the Network Application processing of NOIE CRR Offers • Scoping the changes for MMS5 P6 – primarily in the area of more efficient transmission constraint management • Goal of 10,000 Energy-Only Bids/Offers & 10,000 PTP Obligation Bids Nodal Advisory Task Force

3. Bid Limitations in the CRR Market in ERCOT • In the ERCOT CRR Market there is a need to limit bids • In each CRR Auction, the total number of bids+previously awarded CRRs cannot go over 200,000 • Short term issue is dimension sizing in auction engine • Long term issue is serious doubts about length of time to solve extremely complex auction with large number of bids • Probable bid limits for first six monthly auctions • Number of Allocated PCRRs ~500 • Number of registered market participants currently ~80 • Number of bids per participant ~ 2400 • [(200,000-500)/80 = 2493] • Bid Limit is per month in Monthly Auction • Bid Limit is per year in Annual Auction Nodal Advisory Task Force

4. ERCOT Auction considerably more complex than other Nexant Markets • Multiperiod • Monthly solves over three models simultaneously • Annual solves over twelve models simultaneously • Larger number of contingencies • ~2000 for ERCOT • ~200 or less for CALISO and MISO • Both of these (protocol required) features have huge (~ 2 orders of magnitude) implications in the complexity of the auction • Protocol 7.5.5.3 (1) • “The auction must be a single-round, simultaneous auction for selling the CRRs….” • Protocol 7.5.1 (2) • “The CRR Network Model must, to the extent practicable, include the same topology, contingencies, and operating procedures as used in the Network Operations Model…” Nodal Advisory Task Force

5. Comparison of other ISOs Nodal Advisory Task Force

6. Text reason for Deviation from Defined High Sustained Limit/Low Sustained Limit John Adams Nodal Advisory Task Force

7. What issues are we trying to address? TPTF agreed to addressed the market concern of withholding capacity by allowing a QSE to telemeter its high capability, which ERCOT would use. TPTF addressed ERCOT’s concern about exaggerated capability by requiring that each adjustment of High Sustained limit be accompanied by a text reason explaining why the resource is capable of additional, or less capacity. It was the stated opinion of TPTF that such a text explanation; in combination with the enforcement programs of the TRE and Market Monitor would be sufficient to dis-incentivize exaggeration of capability; and any attendant reliability risk. Protocol requirement: 6.4.5 Resource Status (1) ERCOT shall use the telemetered Resource Status for all applications requiring status of Resources during the Operating Hour, including SCED, Load Frequency Control (LFC), and Network Security Analysis processes. QSEs shall provide ERCOT with accurate telemetry of the current capability of each Resource including the Resource Status, Ramp Rates, HSL, and LSL and a text reason for any Resource where a Ramp Rate is deviating from a standard Ramp Rate curve for the Resource, or the HSL is less than, or LSL is greater than, the normal high and low limits set in Section 3.7.1, Resource Parameter Criteria. Nodal Advisory Task Force

8. Developments • At the last NATF meeting, ERCOT agreed to work with Russell Lovelace of Shell to develop wording clarifying the requirement. • ERCOT worked with Russell to develop acceptable wording. • At the last BOD meeting; NPRR194 passed with wording which seems to make the issue moot. • Within the first 15 of each Season, each QSE shall provide ERCOT a seasonal HSL for any Generation Resource with a capacity greater than ten MW that will be operated during that Season. • The HSL for the designated Generation Resource shall be determined based on the Real-Time averaged MW telemetered by the Resource during the 30 minutes of constant output. • With this change; it seems unacceptable to ever telemeter an HSL higher than the seasonal HSL demonstrated by testing. Nodal Advisory Task Force

9. Model load frequency Discussion John Adams Nodal Advisory Task Force

10. What issues are we trying to address? ERCOT had considered supporting daily database loads in its Nodal design, and can support CIM case production at this periodicity. However, downstream acceptance of database loads at this interval is proving problematic. • Risks • Visits to CAISO and MISO have revealed that they believe ERCOT has underestimated the difficulties of supporting a Nodal Network model change because of its impact upon markets.Each of these markets load their market database once/quarter. • In the past, there was much less financial risk due to model errors, because prices were only influenced around 4 interfaces. W-N, N-S, N-H, S-H. • With Nodal; a model error anywhere in the system financially affects CRR holders, participants in the DAM, and participants in the Real time market. • Historically; ERCOT has encountered errors in network model loads, even after testing. Nodal Advisory Task Force

11. Developments & Next Steps • ERCOT • Has prepared a White Paper describing the issues and delivered to NATF and NDSWG. • Will develop a procedure to be used by TSP’s, QSE’s and ERCOT to coordinate NOMCR’s in supporting a non-daily database load process. • Will work with NDSWG to address TSPs concerns. • Will report to NATF, and if necessary other shareholder groups on status of this effort. Nodal Advisory Task Force