Integrating Solar and Net Zero Homes. EFG May 2016

Transcription

1 Integrating Solar and Net Zero Homes EFG May 2016

2 Roseville Electric Utility Municipal Utility located east of Sacramento CA. 57,000 customer (50,000 residential customers) 340 MW annual system peak High concentration of rooftop solar, 6% of single family homes have rooftop solar Adding approximately 700 residential customers per year due to greenfield development of new homes

3 California Policy Impacting Forecasts Zero net energy homes Rooftop solar adoption SB350 - Doubling of Energy Efficiency savings in existing commercial buildings by 2030 Electric vehicle adoption

4 Zero Net Energy Homes

5 Zero Net Energy Homes All new home construction after 2020 will be net zero energy All new commercial after 2030 will be net zero energy Roseville is adding between 600 and 1000 new homes per year. An increasing number will be net zero 100% will be net zero starting in 2020.

6

7

8 Zero Net Energy Homes-impacts Homes will use no net energy over the course of a year. Size of rooftop solar will increase on new construction homes. Distributed solar generation will vary depending on the season. Need to forecast which months will be negative and which will be positive. Solar generation is low during system peak peak load may continue to grow while system energy does not.

9

10 Residential Forecast

11 Residential Heating Variable

12 Residential Cooling Variable

13 Residential XOther Variable Xother includes the economic variables HHSize, HHIncome, and Electricity Price

14 Building the Solar Variable Currently Roseville has approximately 2,360 residential solar customers; with a total installed capacity of 8.3 MW (the current average size is 2.7 kw). Assumes that 50 new residential solar systems will be added each month, representing new construction. Additionally, 10 retrofits of existing housing stock will occur each month. Average solar system size of retrofits will be 3.5 kw. Average size of new residential solar will slowly increase from 2 kw to net zero (5.5 kw) for new homes by Given the assumptions regarding additional customers and average system size a monthly installed capacity forecast (MW) is generated. Solar is generation is then used as a variable in the Residential Use Per Customer regression model.

15 Residential Xsolar Variable

16 Residential Average Use Model Model Statistics Iterations 1 Adjusted Observations 94 Deg. of Freedom for Error 89 R-Squared Adjusted R-Squared AIC 6.47 BIC Log-Likelihood Model Sum of Squares 3,027, Sum of Squared Errors 54, Mean Squared Error Std. Error of Regression Mean Abs. Dev. (MAD) Mean Abs. % Err. (MAPE) 2.32% Durbin-Watson Statistic Ljung-Box Statistic Prob (Ljung-Box) Skewness Kurtosis 3.44 Jarque-Bera Prob (Jarque-Bera) Variable Coefficient StdErr T-Stat P-Value mstructrescyc.xheat % mstructrescyc.xcooljul_sep % mstructrescyc.xcool_shoulder % mstructrescyc.xother % mstructrescyc.xsolar %

17 Adjusted Residential Sales Forecast Year ResMWh Growth Res MWh Res MWh Minus PV Growth Res MWh Minus PV Res MWh Plus EV Growth Res MWh plus EV , , , , % 441, % 441, % , % 421, % 421, % , % 431, % 431, % , % 444, % 444, % , % 436, % 436, % , % 440, % 440, % , % 431, % 431, % , % 434, % 434, % , % 430, % 430, % , % 425, % 426, % , % 421, % 422, % , % 418, % 418, % , % 416, % 416, % , % 413, % 414, % , % 411, % 412, % , % 409, % 410, % , % 407, % 408, % , % 403, % 404, % , % 400, % 401, % , % 397, % 399, % , % 395, % 397, % , % 392, % 394, % Residential equation: ResMWh=(Forecast average use * forecast customers) + EV

18 Residential Year Res MWh Growth ResMWh Res AvgUse (kwh) Growth Res Avg Use ,332 9, , % 9, % , % 9, % , % 9, % , % 9, % , % 8, % , % 8, % , % 8, % , % 8, % , % 8, % , % 8, % , % 7, % , % 7, % , % 7, % , % 7, % , % 7, % , % 7, % , % 7, % , % 7, % , % 6, % , % 6, % , % 6, % , % 6, % Decrease is driven by solar adoption and new zero energy homes.

19 Commercial Forecasts

20 SB 350 Assumptions SB350 - Doubling of Energy Efficiency savings in existing buildings by 2030 Increasing energy efficiency in the small and medium commercial classes. The class as a whole will reduce load by 2.5% per year starting in Additional data on the implementation of SB350 will likely change the way this law is forecasted. Power resources wanted the law included in the forecast using the California Energy Policy Report energy efficiency number (doubled). Forecast is a worst case scenario on SB 350 implementation (18% reduction in energy use by 2030).

21 Small Commercial Variables

22 Small Commercial Average Use Model Model Statistics Iterations 1 Adjusted Observations 91 Deg. of Freedom for Error 79 R-Squared 0.96 Adjusted R-Squared 0.96 AIC 6.67 BIC 7.00 Log-Likelihood Model Sum of Squares 1,352,591 Sum of Squared Errors 54,947 Mean Squared Error Std. Error of Regression Mean Abs. Dev. (MAD) Mean Abs. % Err. (MAPE) 1.67% Durbin-Watson Statistic 1.99 Ljung-Box Statistic Prob (Ljung-Box) 0.25 Skewness Kurtosis 2.83 Jarque-Bera 1.91 Prob (Jarque-Bera) Variable Coefficient StdErr T-Stat P-Value GS1XHeat % GS1XCool % GS1XOther % *Additional monthly binary variables included

23 Small Commercial Forecast Year GS1MWh Growth GS1MWh GS1 AvgUse (kwh) Growth GS1 AvgUse ,613 14, , % 14, % , % 13, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 14, % , % 13, % , % 13, % , % 13, % , % 12, % , % 12, % , % 12, % , % 11, % Decrease is driven by SB350 assumption Average use and class totals are decreasing If solar becomes more common in the GS1 class this forecast will need to be updated

24 Medium Commercial Total Sales Model *Additional monthly binary variables included

25 Medium Commercial Forecast Year GS2MWh Growth GS2MWh GS2 AvgUse (kwh) Growth GS2 AvgUse , , , % 289, % , % 280, % , % 278, % , % 294, % , % 295, % , % 299, % , % 303, % , % 300, % , % 299, % , % 299, % , % 299, % , % 299, % , % 298, % , % 298, % , % 299, % , % 292, % , % 285, % , % 278, % , % 272, % , % 265, % , % 259, % , % 252, % Customer forecast is flat Sales move based on SAE trends and SB350 Assumptions

26 System Forecast Largest impact is SB350 and net zero energy homes drop forecast in Residential, GS1, and GS2 Year System MWh Growth SysMWh ,303, ,275, % ,225, % ,223, % ,243, % ,228, % ,241, % ,231, % ,245, % ,241, % ,240, % ,235, % ,231, % ,227, % ,225, % ,223, % ,210, % ,198, % ,183, % ,170, % ,157, % ,145, % ,130, %

27 Conclusions The SAE model accounts for the energy efficiency and solar growth in the residential, small commercial, and medium commercial classes. State laws are driving changes to the long term forecast. The unknowns in the long term forecast continue to grow.

28 Questions?