“Know Your Customers”

Why utility expectations of demand response from small customers are overblown

William B. Marcus UCEI Policy Conference December 10, 2007

JBS Energy, Inc.  Consulting firm serving consumers, environmentalists, government agencies, and renewable energy producers since 1984  Economic analysis of utility operations, plans, and rate design  Manufacture and sell Aquacalc (handheld computer for surface water measurement)  www.jbsenergy.com

Rate Design Can Be Controversial

Who Are Small Customers  Lower Use of Peak Power Relative to Average Energy Use  Lower Saturation of Air Conditioning  Smaller Dwelling Units

 Smaller Household Sizes  Lower Incomes

Data Sources  Aggregate Billing Data (Use by Tier, Month, and Climate Zone)  Disaggregate Billing Data (Borenstein study)  Load Research Studies (Hourly Data)

 Residential Appliance Saturation Survey (RASS)

Load Research  Data provides use by time period and peak loads for sample of customers  Stratify by size of customer  May be able to stratify by location (PG&E)  May contain other information (e.g., CARE, all-electric, etc.)

Load Research  JBS has analyzed load research data for all 3 California utilities in the past.  For all three utilities, smaller customers use less peak power than larger customers.  This fact is true after controlling for climate zones on the PG&E system (which had that information)  Voluntary TOU customers also use less peak power

average or peak hourly use (kW)

SDG&E Load Research Data 3.50 3.25 3.00 2.75 2.50 2.25 2.00

Average hourly use Average hourly on-peak use Average hourly off-peak use Average July-September peak load

1.75 1.50 1.25 1.00 0.75 0.50 0.25 1000 kWh

SDG&E Load Research Data 2 180.0% 160.0% 140.0% on peak of total kWh%

120.0% 100.0%

on-peak use as % of offpeak use

80.0%

peak load as % of onpeak use

60.0%

summer load factor 9average divided by peak)

40.0% 20.0% 0.0% 1000 kWh

RASS Data  Over 20,000 stratified sampled data points for 4 major electric utilities (3 IOUs + LADWP)  Type of Appliances  Appliance Usage  Income  Household Size  Housing Unit Type and Size  Link to Electric and Gas Billing Data

Why Use RASS Data?  RASS explains why load research shows that small customers don’t use much peak load.  RASS has income data from customer survey – don’t need census data inference.

What Can I Tell You  Not much that is current - research in progress  Data under confidentiality agreements from utilities that allow me only to use it in specific cases. When I present the work in a case, I can tell you more about it.  2002 study using 1994-95 RASS – “Economic and Demographic Factors Affecting California Residential Energy Use.” on JBS website.

Examples from 2002 Report on our website PG&E Residential Basic Use by Square Feet and Climate Zone 17,000 Zone P Zone R

15,000

Zone S

13,000

Zone T

kWh

Zone V

11,000

Zone W Zone X

9,000 7,000 5,000 3,000 < 750

7501000

10001250

12501500

15002000

20002500

25003000

> 3000

square feet Use rises as the dw elling unit size increases in all zones. Use in very large dw ellings is typically 250-350% of use in the smallest dw ellings

PG&E Residential Basic Use vs. Income by Climate Zone 14,000 Zone P

13,000

Zone R

kWh

12,000

Zone S

11,000

Zone T

10,000

Zone V Zone W

9,000

Zone X

8,000 7,000 6,000 5,000 4,000 3,000 0-10

10-15

15-20

20-30

30-40

40-50

50-75

75-100

> 100

Income ($'000, 1995) use at least doubles from the low est to the highest income in all zones. Use is more sensitive to income in Central Valley zones R, S, and W and less sensitive in Bay Area Zone T.

Air Conditioning Type by Income (SCE Zone 17 Basic) 100% 90%

% with type of cooling

80% 70% No AC

60%

Central AC

50%

Window AC

40%

Evap Cooler

30% 20% 10% 0%