1. RTDF Overview of Data Analysis & Status of “Consesus Items”
John Dumas
Director Wholesale Market Operations
ERCOT
October 10, 2011

2. ERCOT Methodology To Calculate Change in PNM for Different Constant Floors
Consider only the NSPIN deployment intervals in which load exceeds the load at the time of the deployment
For each level of price floor, assume that floor price will set the new System Lambda during the deployment
Assume FIP price = $4/MMBtu. i.e. Peaker Operating Cost = 10*FIP = $40/MWh
Additional PNM is calculated for the SCED intervals where the difference between the new System Lambda and original System Lambda is positive.
Reduce PNM for SCED intervals where NSPIN not deployed and Original System Lambda >=$2999/MWHr. This is based on the assumption that online (non-QSGR and QSGR) NSPIN capacity always available to SCED.
October 10, 2011

3. ERCOT Methodology: Change in PNM for Different Constant Floors (12/1/2010 – 8/29/2011)
All NSPIN deployments for capacity considered.
Total NSPIN deployment hours = hours
Two scenarios considered
All NSPIN deployment for capacity hours where the load exceeds load at time of deployment = hours
All NSPIN deployment for capacity EXCLUDING February 2nd event and 1st week of August = 120 hours
Floor level $/MWh
Additional “PNM” when NSPIN deployed
Scenario 3 a) - ALL
Additional “PNM” when Non-Spin deployed
Scenario 3 b) – Exclude Feb 2nd & 1st week of August
Reduction in PNM when Non-Spin NOT deployed 72
$2,803
$2,129
$-8,656
180
$17,149
$12,480
$-8,337
250
$27,195
$19,513
$-8,130
650
$91,852
$62,310
$-6,948
1,000
$151,299
$101,645
$-5,913
2,000
$322,323
$214,511
$-2,958
October 10, 2011

4. ERCOT Methodology: Change in PNM for Different Constant Floors (12/1/2010 – 8/29/2011)
Theoretical additional PNM contribution calculated by taking the difference between the floor and the maximum of actual system lambda or $40. This assumes that the floor sets prices for the hours in which Non-Spin was deployed for capacity and load was high.
October 10, 2011

5. ERCOT’s Interpretation of Luminant’s Methodology To Calculate Change in PNM for Linearly Varying Floors
Consider only the deployment intervals in which load exceeds the load at the time of the deployment.
However, NSPIN capacity will not set the price for all these deployment intervals.
Consider the floor to be sloped for the NSPIN capacity portion of the Energy Offer Curve:
Online (non-QSGR & QSGR) : $250/MWhr to $1000/MWHr
Offline: $1000/MWHr to $3000/MWHr
NSPIN portion of the Energy offer curve will set price when the difference between the UPWARD room available to SCED discounting LSL injection from QSGR and offline Resources is LESS than NSPIN deployed.
Based on the MW amount of NSPIN required to meet load, calculate the New System Lambda using the sloped floor.
Additional PNM is calculated for the SCED intervals where the difference between the new System Lambda and original System Lambda is positive.
October 10, 2011

6. ERCOT’s Interpretation of Luminant’s Methodology To Calculate Change in PNM for Linearly Varying Floors
2 result presented based on different assumptions
SCENARIO 1: Snapshot of deployed MW at the time of deployment & LSL 30-min into the deployment  Keep constant throughout the deployment period
SCENARIO 2: Data posted on RTDF meeting page
BOTH uses HDL_BELOW_SWCAP rather than HDL
October 10, 2011

7. ERCOT’s Interpretation of Luminant’s Methodology To Calculate Change in PNM for Linearly Varying Floors
Clarification on Data posted to RDTF meeting page used in Scenario 2
Resource categorization by Offline, Online, and Quick Start type is done based on telemetry at the initial time of deployment.
Any resources not in the Online or Quick Start categories determined at the time of Non-Spin Deployment are considered as Offline resources.
Effects of telemetry error, shifting of NSRS responsibility to other resources or different Combined Cycle modes has not been taken into account.
October 10, 2011

8. ERCOT’s Interpretation of Luminant’s Methodology ASSUMPTION SET 1 results
All NSPIN deployments for capacity considered.
Total NSPIN deployment hours = hours
Two scenarios considered
All NSPIN deployment for capacity hours where the NSPIN capacity sets price = 84.5 hours
All NSPIN deployment for capacity EXCLUDING February 2nd event and 1st week of August where the NSPIN capacity sets price = 48.3 hours
Additional “PNM” when NSPIN deployed
Scenario 3 a) - ALL
Additional “PNM” when Non-Spin deployed
Scenario 3 b) – Exclude Feb 2nd & 1st week of August
$57,937
$37,521
October 10, 2011

9. ERCOT’s Interpretation of Luminant’s Methodology ASSUMPTION SET 1 results
84.5 hours
27.02 hours
16.25 hours
October 10, 2011

10. ERCOT’s Interpretation of Luminant’s Methodology ASSUMPTION SET 1 results
48.3 hours
9.17 hours
8.87 hours
October 10, 2011

11. ERCOT’s Interpretation of Luminant’s Methodology ASSUMPTION SET 2 results
All NSPIN deployments for capacity considered.
Total NSPIN deployment hours = hours
Two scenarios considered
All NSPIN deployment for capacity hours where the NSPIN capacity sets price = 67.9 hours
All NSPIN deployment for capacity EXCLUDING February 2nd event and 1st week of August where the NSPIN capacity sets price = 39.1 hours
Additional “PNM” when NSPIN deployed
Scenario 3 a) - ALL
Additional “PNM” when Non-Spin deployed
Scenario 3 b) – Exclude Feb 2nd & 1st week of August
$31,671
$23,104
October 10, 2011

12. ERCOT’s Interpretation of Luminant’s Methodology ASSUMPTION SET 2 results
67.9 hours
19.4 hours
5.5 hours
October 10, 2011

13. ERCOT’s Interpretation of Luminant’s Methodology ASSUMPTION SET 2 results
39.1 hours
7.3 hours
3.3 hours
October 10, 2011

14. Estimate Effect Of Different Constant Floors On System Lambda for Non-Spin Deployments (12/1/2010 – 8/29/2011)
Corresponding to the PNM additional revenue based on ERCOT methodology for the different floors, calculate the impact on System Lambda for 7 X 24 and 5 X 16 time blocks.
All NSPIN deployments for capacity considered – NO EXCLUSIONS.
Calculate the time-weighted ( by SCED interval duration ) average of system lambda from 12/1/2010 to 8/29/2011. This is the original price (system lambda) for the 7 X 24.
Calculate the change in system lambda for each SCED interval during non-spin deployment for each floor . Note : Similar to the PNM change we are only looking at SCED intervals where the GTBD >= Initial GTBD at time of Non-Spin deployment and applicable floor would have set price.
October 10, 2011

15. Non-Spin Deployments (12/1/2010 – 8/29/2011) Estimate Effect Of Different Constant Floors On System Lambda
Original System Lambda ($/MWHr) and corresponding increase in System Lambda for different floors
Increase to System Lambda for Various Floors
Original System Lambda
October 10, 2011

16. Non-Spin Deployments (12/1/2010 – 8/29/2011) Estimate Effect Of Different Constant Floors On Heat Rate (5X16)
Original Heat Rate (MMBTU/MWHr) and corresponding increase in Heat Rate for different floors (5X16). Assume FIP=4
Original Heat Rate
Increase to Heat Rate for Various Floors
October 10, 2011

17. RTDF Status on “Consensus Items”
Based on RTDF understanding of “consensus items”, developed draft NPRRs:
Offers should be at system wide offer cap (SWCAP) for Responsive Reserve & Regulation Up service portion of the Energy Offer Curve.
Draft NPRR: “EOC Requirements for Generation Resources Assigned Regulation Up and Responsive Reserve Services”.
No ERCOT system change
Implemented via QSE submission of Energy Offer Curve.
October 10, 2011

18. RTDF Status on “Consensus Items”
Deploy RUC units for capacity insufficiency at the offer cap; RUC for system capacity should be offered liked mothballed RMR.
Draft NPRR: “Requirement That Energy From Generation Resources Committed by the DRUC or HRUC Processes be Dispatched by SCED at the SWCAP”. 
Requires ERCOT system change.
This NPRR conflicts with NPRR 416 “Removal of the RUC Clawback Charge for Resources Other than RMR Units”
October 10, 2011

19. RTDF Status on “Consensus Items”
Online non-spin units and quick start units should always be offered in SCED and deployed by SCED.
Draft NPRR: “Standing Non-Spin AS Deployment in the Operating Hour for Generation Resources Providing Online Non-Spin AS in the Operating Hour”.
No ERCOT system change
Implemented via QSE telemetry.
ERCOT should deploy non-spin in increments smaller than a thousand MW.
No draft NPRR provided.
ERCOT already complies.  Plan to review current procedure to determine improvements (if any).
October 10, 2011

20. RTDF Status on “Consensus Items”
Prices should be at system wide offer cap when ERCOT deploys load resources or EILS – re-price Energy after Load deployments.
No draft NPRR provided.
ERCOT provided “Dynamic Price Adder” concept.
Requires ERCOT system change.
Some market participants have proposed including a “Make Whole” Payment provision to this concept.
Morgan Stanley has proposed extending the “Dynamic Price Adder” concept with “Make Whole” payment for ALL Reliability deployments.
October 10, 2011