Update on Step by Step Energy Methodology and Initial Econometric Results April 4, 2013 Presentation to ICAP WG

Transcription

1 Update on Step by Step Energy Methodology and Initial Econometric Results April 4, 2013 Presentation to ICAP WG

2 Step 1 Develop Econometric Forecasts of LBMP by Zone for , and Each year s forecast reflects forecast peak and energy Each year s forecast reflects the addition of AE2 and BEC measured with the econometric model Forecasts for each year were performed using historic gas prices for 2010, 2011 and 2012 which on average are similar to current forward gas prices 1

3 Electricity Price Is Largely Predictable Electricity price is determined by intersection of a supply curve and a demand curve The demand curve shifts hourly in reasonably well-understood ways, largely dependent on hours, weekdays and seasons The supply curve is largely fixed, but nonetheless shifts owing to: Random outages, planned outages, semi-planned outages Changes in fuel prices Unit additions and retirements Ambient conditions Transmission conditions Opportunities outside the NYISO for energy purchases and sales 2

4 The Data Observed Average Day-Ahead LBMP and Integrated Loads, Hourly, in eleven NYISO load zones from 11/1/ /31/2012, taken from NYISO website Daily Observed Gas Prices from the Transco Z6 (NYC and LI), Tetco M3 (Upstate), and Tennessee Zone 6 (HV) Interpolated Across Weekends and Holidays, Lagged Two Days Weather data from NOAA for Albany (HV), JFK (NYC and LI) and Syracuse (ROS) Starting dates for AE2 and BEC (7/1/11 and 6/1/12, respectively) 3

5 Changes to the Econometric Model Removal of Reserve Margin Variable Explicit Supply Shifters for AE2 and BEC More Detailed Gas Price Specification Lagged Two Days in the data to reflect known conditions Addition of Ln(Gasp) and (Ln(Gasp)) 2 terms in addition to the month and hour terms 4

6 The Model, Roughly AE2 Even without any explicit measures of outages or congestion, this model explains about 87 percent of the variance in market prices Econometric Model Output Published 5

7 Variables in The Econometric Model Dependent variable: llbmp log of LBMP Independent Variables Month x Zone and Hour x Month dummies Zonal Load (independent coefficients by zone) Zonal Load x Aggregate Load (independent d coefficients i by zone) Aggregate Load, Load 2 and Load 3 (Independent coefficients i by aggregates of load zones) AE2 and BEC dummies (independent coefficients by aggregates of load zones) Day of Week Dummies Max daily temperature and min daily temperature, by region 6

8 Results of Basic Econometric Forecasts LHV CCGT 1 Forecast Year Prediction Base Year Average LBMP Average Gas Cost Net Revenue/kW with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ Actual $5.39 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $98.17 Actual $5.64 $ with AE2 and BEC $3.80 $ with AE2 and BEC $3.80 $ with AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ Actual $3.80 $ Based on S&L Characteristics 7

9 Results of Basic Econometric Forecasts LHV LMS Forecast Year Prediction Base Year Average LBMP Average Gas Cost Net Revenue/kW with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $56.39 Actual $5.39 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $40.78 Actual $5.64 $ with AE2 and BEC $3.80 $ with AE2 and BEC $3.80 $ with ihae2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $62.90 Actual $3.80 $ Based on S&L Characteristics 8

10 Results of Basic Econometric Forecasts NYC CCGT 1 Forecast Year Prediction Base Year Average LBMP Average Gas Cost Net Revenue/kW with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ Actual $5.39 $ with AE2 and BEC $5.72 $ with AE2 and BEC $5.71 $ with AE2 and BEC $5.72 $ w/o AE2 and BEC $5.72 $ w/o AE2 and BEC $5.71 $ w/o AE2 and BEC $5.72 $ Actual $5.72 $ with AE2 and BEC $3.31 $ with AE2 and BEC $3.31 $ with AE2 and BEC $ $ w/o AE2 and BEC $3.31 $ w/o AE2 and BEC $3.31 $ w/o AE2 and BEC $3.31 $ Actual $ $ Based on S&L Characteristics 9

11 Results of Basic Econometric Forecasts NYC LMS Forecast Year Prediction Base Year Average LBMP Average Gas Cost Net Revenue/kW with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $75.46 Actual $5.39 $ with AE2 and BEC $5.72 $ with AE2 and BEC $5.71 $ with AE2 and BEC $5.72 $ w/o AE2 and BEC $5.72 $ w/o AE2 and BEC $5.71 $ w/o AE2 and BEC $5.72 $62.20 Actual $5.72 $ with AE2 and BEC $3.31 $ with AE2 and BEC $3.31 $ with AE2 and BEC $ $ w/o AE2 and BEC $3.31 $ w/o AE2 and BEC $3.31 $ w/o AE2 and BEC $3.31 $96.49 Actual $ $ Based on S&L Characteristics 10

12 Results of Basic Econometric Forecasts ROS CCGT 1 Forecast Year Prediction Base Year Average LBMP Average Gas Cost Net Revenue/kW with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $78.80 Actual $5.39 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $67.57 Actual $5.64 $ with AE2 and BEC $3.80 $ with AE2 and BEC $3.80 $ with AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $99.83 Actual $3.80 $ Based on S&L Characteristics 11

13 Results of Basic Econometric Forecasts ROS LMS Forecast Year Prediction Base Year Average LBMP Average Gas Cost Net Revenue/kW with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ with AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $ w/o AE2 and BEC $5.39 $31.70 Actual $5.64 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ with AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $ w/o AE2 and BEC $5.64 $31.45 Actual $3.80 $ with AE2 and BEC $3.80 $ with AE2 and BEC $3.80 $ with AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $ w/o AE2 and BEC $3.80 $18.51 Actual $5.39 $ Based on S&L Characteristics 12

14 What Resources Are Reflected in Historic Results and Econometric Forecasts Resource Status Historic Unadjusted Forecast AE2 In from July 11 In by virtue of modelling BEC In from June 12 In by virtue of modelling Astoria 2&4 In until April 12 Deemed in Far Rockaway 4 and Glenwood 4 & 5 In until July 12 Deemed in Dunkirk 3&4 In until September 12 Deemed in Danskammer In In HTP Out Out Nine Mile 2 Uprate In from July 12 Deemed out Athens SPS In In 13

15 What Resources Are in CARIS 2 and What is Objective for Forecast? Revenue CARIS2 Status Forecast Objective AE2 In In BEC In In Astoria 2 & 4 Out Out Far Rockaway 4 and Glenwood 4 & 5 Dunkirk 3&4 Out Out Danskammer In Out HTP In In Out Out Nine Mile 2 Uprate In In Athens SPS Out In 14

16 Rationale for Objectives New resources such as AE2, BEC, HTP and NMP2 Uprate will operate for forecast period (May 2014 to April 2017) and Athens SPS will continue to operate Danskammer will not operate in forecast period Astoria, Far Rockaway, Glenwood and Dunkirk 3 & 4 are mothballed and may or may not operate however, as raised at last meeting, adjusting to alternate reserve levels by increasing load and leaving in all older less efficient plants could push energy prices up unrealistically accepting CARIS2 status of the units is a reasonable way to avoid this effect 15

17 Step 2 Adjust the Econometric Results to the Objectives for the Resources CARIS 2 Objective Athens SPS HTP NMP2 IN Case (Case 1) BEC, AE2 Athens SPS, HTP, NMP2, CARIS 2 Matching History Case (Case 2) OUT IN OUT Danskammer, Dunkirk 3&4, Astoria 2&4, FR4, GL 4&5 Athens SPS, BEC, AE2, Danskammer, Astoria 3&4, FR4, GL 4&5 HTP Adjust Econometric LBMP Forecasts by Ratio of LBMP between Case 1 and Case 2 to Arrive at Econometric Forecasts Consistent with the Objectives Use Monthly/Hourly Adjustment Method 16

18 As a Result of Retirement the Forecast Excesses Are Relatively Modest Excess Capacity Percentages for the Objective Case NYCA 6.55% 5.81% 5.21% Zones G to J 4.74% 3.52% 2.60% Zone J 5.86% 4.49% 3.59% Zone K 7.37% 6.56% 6.32% Excess values are stated t relative to the LCR/IRM 17

19 Step 3 Produce MAPS LBMPs for Equal Reserve Case and Further Adjust Econometric Results % 2.60% 2.60% % 2.60% 2.60% % 2.60% 2.60% Peak and energy by locality will be increased to achieve lowest excess for any locality. The ratio of LBMP between the objective case and this case will be used to further adjust econometric results. 18

20 Step 4 Adjust LBMPs to Various Excess levels Adjust peak and energy levels l in MAPS to produce LBMPs at various reserve levels keeping level of excess uniform over localities Adjust econometric results to various reserve levels Adjust real-time LBMPs for scarcity pricing impact Compute net energy revenue at various reserve levels 19

21 Step 5 Net Energy Revenues Determine dispatch and net energy revenues for various technologies from LBMPs for various excess level scenarios using model that considers unit costs and start up costs Interpolate net energy e revenues e for excess levels not explicitly modeled using a linear interpolation between modeled scenarios 20

22 Recap Econometric LBMP results for , , and using forecast peak and energy and historic gas prices from 2010, 2011 and 2012 and reflecting operation of AE2 and BEC Adjust econometric LBMP results based on MAPS run with historic capacity to MAPS run with resource objectives Adjust econometric LBMP results based on MAPS run with resource objectives with loads adjusted to produce equal excess Adjust econometric LBMP results based on MAPS runs at various reserve levelsl Develop net energy revenue estimates at various reserve levels l and interpolate t net energy revenue if necessary 21