Andy Sukenik. Itron s Forecasting Brown Bag Seminar. September 14, 2010

Size: px

Start display at page:

Download "Andy Sukenik. Itron s Forecasting Brown Bag Seminar. September 14, 2010"

Lindsey Cameron
5 years ago
Views:

1 Technolog Choice Modeling And Sukenik Itron s Forecasting Brown Bag Seminar September 14, 2010

2 Please Remember In order to help this session run smoothl, our phones are muted. To make the presentation portion of the screen larger, press the expand button on the bottom toolbar. Hit the escape button to return to the regular screen. If ou need to give other feedback to the presenter during the meeting, such as, slow down or need to get the presenters attention for some other reason, use the feedback button on the upper right side of the toolbar and we will address it right awa. If ou have questions,,press the Q&A button on the toolbar and tpe our question in the popup box. We will tr to answer as man questions as we can. 2009, Itron Inc. 2

3 2010 Brown Bag Seminars Smart Grid: Implications for Forecasting and Analsis March 2, 2010 Sales Forecast Trend Review June 22, 2010 Technolog Choice Model September 14, 2010 Real Time Forecasting Issues and Challenges December 14, 2010 All at noon, Pacific Time All are recorded and available for review after the session. 2009, Itron Inc. 3

4 Long term Forecasting Challenges Energ Efficienc remains a hot topic in the media toda. Sources driving the Efficienc trends include: > Federal Efficienc Standards > Stimulus Program Funding (Energ Efficienc Tax Credits) > Future Carbon Emissions Polic > Utilit DSM Program Impacts > Naturall Occurring Efficienc Gains Forecasters face the common challenge of properl incorporating Efficienc trends into their models. This brown bag session presents an overview of modeling residential technolog efficienc options and examines a model for forecasting central air conditioning efficienc given alternative input assumptions. 2009, Itron Inc. 4

5 SAE Modeling Framework A modeling framework that captures both the impact of factors that drive short term energ use as well as longterm structural changes The framework accounts for: > Economic impacts households, h household h size, personal income, prices > Weather impacts > Structural changes house square footage and thermal shell integrit, > Equipment tstock changes appliance saturation ti and efficienc 2009, Itron Inc. 5

6 Statisticall Adjusted End Use (SAE) Model AC Saturation Central Room AC AC Efficienc Thermal Efficienc Home Size Income Household Size Price Heating Saturation Resistance Heat Pump Heating Efficienc Thermal Efficienc Home Size Income Household Size Price Saturation Levels Water Heat Appliances Lighting Densities Plug Loads Appliance Efficienc Income Household Size Price Cooling Degree Das Heating Degree Das Billing Das XCool XHeat XOther Sales = a + b XCool + b XHeat + b XOther + m c m h m o m e m 2009, Itron Inc. 6

7 XCool Variable,m,m CoolUse CoolIndex XCool = m,,m Tpe Sat = Tpe Tpe BaseYear Tpe Tpe Tpe Sat Eff Sat BaseUEC Index Structural CoolIndex Tpe BaseYear Eff = BaseYear m c BaseYear m b BaseYear m a BaseYear m m CDD CDD HHSize HHSize Income Income Price Price CoolUse,,,,, 2009, Itron Inc. 7

8 XCool Variable XCool,m = CoolIndex CoolUse,m CoolIndex Cooling End Uses -CAC -RAC -Heat Pump Cool = Structural Index BaseUPC Tpe Tpe Sat Tpe Sat Tpe BaseYear Eff Eff Tpe Tpe BaseYear CoolUse, m Price = Price, m BaseYear a Income Income, m BaseYear b HHSize HHSize, m BaseYear c CDD CDD, m BaseYear 2009, Itron Inc. 8

9 Residential SAE Input Data Available through EFG Residential SAE Standard Spreadsheets Residential Technolog level Spreadsheets 2009, Itron Inc. 9

10 Residential Standard Spreadsheets The SAE Residential Spreadsheets are updated when the EIA releases a forecast. The most recent of which is the 2010 Annual Energ Outlook k( (AEO) The standard spreadsheets contain the following data: > Demographic and Economic Trends > Electricit Prices > Building Shell Trends > End use Saturation Trends > End use Efficienc Trends 2009, Itron Inc. 10

11 Residential Space Cooling Saturation Trends Pacific Region Saturation Trends 2009, Itron Inc. 11

12 Residential Space Cooling Efficienc Trends Pacific Region Efficienc Trends 2009, Itron Inc. 12

13 Technolog Option Spreadsheets Displa the technolog level data that underlies the EIA s regional end use efficienc trends. Provide functionalit to adjust the end use model inputs to align with service territor specific: > Saturations > Efficiencies Facilitate the construction of alternative equipment stock efficienc scenarios. 2009, Itron Inc. 13

14 Technolog Level Detail Pacific CAC The Residential SAE Technolog Option Spreadsheets displa the technolog option marginal purchase share and efficienc assumptions that underlie the EIA Baseline marginal efficiencies. More Efficient More Efficient 2009, Itron Inc. 14

15 Tech Spreadsheets Efficienc Model Options Efficienc Model Options EIA. The EIA option uses the historical and forecasted Regional Efficienc data provided b the EIA. Frozen. The Frozen option holds the base ear efficienc level constant throughout the forecast period. Model Option parameters include: > Base Year Technical Potential. The Technical Potential option assumes all consumers from the ear 2006 forward purchase the most efficient technolog option available on the market. StockAccounting Accounting. The Stock Accounting Model option enables the implementation of customized marginal efficienc data and deca and replacement parameters. Model Option Parameters include: > Appliance Life (Average & Standard Deviation) > Deca & Replacement Rate 2009, Itron Inc. 15

16 Tech Spreadsheets Purchase Share Model Options Purchase Share Model Options EIA. The EIA option implements the technolog level purchase shares implicit in the EIA regional purchases data. Utilit Specific. The Utilit Specific option implements customized purchase shares from the base ear forward. Model Option parameters include: > Utilit Specific Purchase Share Model. The Model option uses a Logit function to compute the purchase shares for each technolog option as a function of Life Ccle Cost. Model Option parameters include: > Technolog Option DSM Rebate % > Real Electric Price > Discount Rate 2009, Itron Inc. 16

17 Marginal Efficienc Calculation 2010 Marginal Efficienc Calculation 33% * % * % * % * 23 = 14.3 SEER More Efficienti More Efficienti 2009, Itron Inc. 17

18 Marginal Efficienc Calculation 2010 Marginal Efficienc Calculation 33% * % * % * % * 23 = 14.3 SEER More Efficienti More Efficienti 2009, Itron Inc. 18

19 Marginal Efficienc Pacific CAC The Weighted Average of Technolog Options Efficiencies represents the Marginal Efficiencies inputs to the Stock Accounting Model. 2009, Itron Inc. 19

20 Stock Accounting Model In the Technolog Option Spreadsheets, to accuratel capture the overall impact of the Marginal Efficiencies, an abbreviated stock accounting model is used. A deca and replacement algorithm replaces older vintages of stock decaing with new vintages. Adjustable Parameters included in the Stock Accounting Model include: > Appliance Life > Deca Rate > Replacement Rate The Default values for these Parameters were reverse engineered to tie the SA model results in closel with EIA Sales/Customer estimates for each appliance. 2009, Itron Inc. 20

21 Average Efficienc Pacific CAC Central Air Conditioner Default Appliance Life = 18 Years 2009, Itron Inc. 21

22 Baseline Efficienc Scenario Assumptions Alternative efficiencies scenarios can be constructed to isolate the energ impacts associated with different levels of efficienc gain. Equipment Efficienc Scenarios include: 1. Frozen 2. Baseline 3. HighEfficienc PotentialCase 4. DSM Efficienc Incentive Programs 5. Carbon Legislation Price Scenarios 2009, Itron Inc. 22

23 Scenario 1: Frozen Efficienc Scenario Saturation Efficienc Frozen Efficienc Scenario -Hold 2005 Efficienc Levels constant throughout the forecast period. 2009, Itron Inc. 23

24 Scenario 2: Baseline Efficienc Scenario Standards Efficienc Scenario -Assume all consumers purchase the least efficienc available market option (13 SEER CAC) Baseline Efficienc Model Scenario Option Settings -The efficienc -Base path Year used = to 2005 drive the EIA s AEO 2010 Sales Forecast -Efficienc Model Option = Stock Accounting -Includes Efficienc -Purchase Gains Share associated Model Option with federal = Utilit Specific standards, government -CT_AIR#1 funding, (13 SEER) naturall Purchase Share = 100% occurring efficienc gains, and regional average DSM program activit. 2009, Itron Inc. 24

25 Scenario 3: High Efficienc Potential Scenario Standards Efficienc Scenario -Assume all consumers purchase the least efficienc available market option (13 SEER CAC) High Efficienc Potential Scenario -Assume all consumers purchase the most efficienc available market option (23 SEER CAC) Model Option Settings -Base Year = Efficienc Model Option = Stock Accounting -Purchase Share Model Option = Utilit Specific -CT_AIR#1 (13 SEER) Purchase Share = 100% 2009, Itron Inc. 25

26 Construct a DSM Program Scenario The goal is to use the Purchase Share Model to construct alternative efficienc paths associated with the following scenarios > DSM Program Rebate Incentive > Assume once the Federal Tax Credits expire, the utilit supplements the funding and maintains i the funding for the rebate program through h , Itron Inc. 26

Purchase Share Model Framework PurchaseSh are Purchase

Cost ] ) o 1 o ( c + b [ LifeCcle Cost ]) exp o 1 o Life

DiscountedOperatingCost o Capital Cost Raw Inputs -Raw

Discounted Operating Cost Raw Inputs -Base UEC -Real

27 Purchase Share Model Framework PurchaseSh are Purchase Share Logit Model o = O k = 1 exp o ( c + b [ LifeCcle Cost ] ) o 1 o ( c + b [ LifeCcle Cost ]) exp o 1 o Life Ccle Cost Calculation o LifeCcleC ost = CapitalCost + DiscountedOperatingCost o Capital Cost Raw Inputs -Raw Capital Cost b Technolog Option (EIA) -DSM Rebates Discounted Operating Cost Raw Inputs -Base UEC -Real Electric Price -Discount Rate (d) -Time Horizon for DOC Calculation (T) 2009, Itron Inc. 27

28 Implementing DSM Scenario Assumptions For this example, the DSM Scenario assumes the tax credit funding subsidizing the purchase of higher efficienc Central Air Conditioners is extended throughout the forecast period. The Purchase Share Option is set to Model The DSM % Rebate is extended through the balance of the forecast period. 2009, Itron Inc. 28

29 DSM Scenario LCC and Purchase Shares CAC Stimulus Scenario -Funding DSM Scenario -Funding , Itron Inc. 29

30 Scenario 4: DSM Program Efficienc Scenario DSM Program Efficienc Scenario -Assumes funding levels from the federal tax credit are extended through the balance of the forecast period. 2009, Itron Inc. 30

31 Scenario 5 Carbon Legislation Price Scenario Carbon Legislation in the form of a Carbon Tax or Cap and Trade polic will impact forward looking electric prices. There are both short and long term effects that t result from alternative prices forecasts. Short Term Impacts: > The Price component of the SAE End Use Usage equation captures the short term impacts driven b changes in utilization levels. Long Term Impacts > Theoreticall higher prices increase operating cost, providing incentive to purchase higher efficienc options. 2009, Itron Inc. 31

32 Incorporating Efficienc Scenarios into SAE Estimate the SAE model using our best estimate of actual historical efficienc to define XCool, XHeat and XOther coefficients. > The SAE model coefficients will adjust to tie the UPC indices into our Monthl Sales values. Lock the coefficients from the above model and run the alternative equipment efficiencies through the model Taking the difference in Model Results across scenarios can quantif the efficienc gains implicit in each scenario 2009, Itron Inc. 32

33 Summar SAE model structure provides the abilit to identif end use factors that drive energ use. Stock Accounting Models incorporate saturation and efficienc trends at the individual appliance level, supporting the forecast with a rich set of customized information. Direct adjustment of the UPC indices with respect to utilit programs is an alternative approach to making a post model dl adjustment for DSM. Incorporating AlternativeEquipment EfficiencScenarios into the SAE framework can isolate efficienc impacts from the model results and provide a measure of incremental energ savings associated with ihappliance Efficienc i DSM programs. 2009, Itron Inc. 33

34 Questions? Press *6 to ask a question or tpe in bottom, right corner. REMAINING 2010 WORKSHOP SCHEDULE AND REGISTRATION Fundamentals of Sales and Demand Forecasting Orlando, FL September Forecasting 101 San Diego, CA September Australian Energ Forecasting Workshop Melbourne, Australia October Fundamentals of Short Term and Hourl Forecasting San Diego, CA October OTHER FORECASTING MEETINGS 3 rd Annual Australian Forecasting User Group Meeting Melbourne, Australia October Itron Users Conference Orlando, FL October For more information and registration: Contact us at: , or forecasting@itron.com 2009, Itron Inc. 34

The Role of Modeling and Accounting in Estimation of Unbilled Revenue Presented by: J. Stuart McMenamin

Electric / Gas / Water The Role of Modeling and Accounting in Estimation of Unbilled Revenue Presented by: J. Stuart McMenamin Please Remember In order to help this session run smoothly, your phones are