Marginal Cost, Revenue Allocation, and Residential Rate Design for San Diego Gas & Electric Company ERRATA Prepared testimony of William B. Marcus JBS Energy, Inc. 311 D Street West Sacramento California, USA 95605 916.372.0534 on behalf of San Diego Consumers Action Network California Public Utilities Commission Application 11-10-002 August 15, 2012
55
Embed
Marginal Cost, Revenue Allocation, and Residential Rate Design … · 2019-11-20 · Marginal Cost, Revenue Allocation, and Residential Rate Design for San Diego Gas & Electric Company
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Marginal Cost, Revenue Allocation, and Residential Rate Design for San Diego Gas & Electric Company ERRATA
Prepared testimony of
William B. Marcus
JBS Energy, Inc. 311 D Street
West Sacramento
California, USA 95605
916.372.0534
on behalf of
San Diego Consumers Action Network
California Public Utilities Commission
Application 11-10-002
August 15, 2012
i
Table of Contents
I. INTRODUCTION AND SUMMARY .................................................................... 1
II. MARGINAL CUSTOMER COSTS ........................................................................ 2
A. INTRODUCTION ...................................................................................................... 2 B. THEORETICAL CONSIDERATIONS: ONE-TIME HOOKUP COST METHOD VS.
RENTAL METHOD ............................................................................................................ 4 1. Interchangeability of Use across Customers .................................................... 5 2. Discussion of Marginal Cost Theory as it Relates to Customer-Hookup
Equipment ................................................................................................................... 6 3. Unreasonable Barriers to Entry ....................................................................... 7
4. Achieving Commission Goals Regarding Cost Causation, Timing, and
5. OTHC: An Approximation .............................................................................. 10 6. Discussion of the Over- vs. Under-Charge Issue ........................................... 10
7. Conclusion ...................................................................................................... 11 C. CUSTOMER COUNTS ............................................................................................ 11
D. MARGINAL CUSTOMER-RELATED DISTRIBUTION O&M COSTS .......................... 12 E. CUSTOMER ACCOUNTING AND CUSTOMER SERVICE O&M ................................. 13
1. Meter Reading ................................................................................................ 14
2. Revenue Offsets from Tariffed Service Charges ............................................. 15 3. Accounts 907-910 (Customer Service and Information) ................................ 15
F. CUSTOMER COST RESULTS ..................................................................................... 21
III. MARGINAL DISTRIBUTION DEMAND COSTS ......................................... 23
A. ADDITIONAL VEGETATION MANAGEMENT FUNCTIONALIZED AS DISTRIBUTION
IV. MARGINAL GENERATION CAPACITY COST AND MARGINAL
ENERGY COST.............................................................................................................. 23
V. DISTRIBUTION REVENUE ALLOCATION .................................................... 24
I. RESIDENTIAL CUSTOMER CHARACTERIZATION ................................... 26
A. REASON FOR ERRATA .......................................................................................... 29 B. RELATIONSHIP OF USAGE TO INCOME, SIZE AND TYPE OF DWELLING, AND
1. Income ............................................................................................................ 31 2. Single vs. Multi-Family................................................................................... 33 3. Square Footage............................................................................................... 36 4. Air Conditioning ............................................................................................. 40
Table 13: Un-Weighted number of RASS Surveys by Climate Zone .............................. 52 Table 14: SDG&E Summer Tier Groups used in CEC RASS Analysis .......................... 53
List of Figures
Figure 1: Average Income by Summer Tier Group and Climate Group ......................... 31 Figure 3: Percent of Single-Family and Multi-Family Households within Tier Groups and
Climate Zones SDGE ........................................................................................................ 33 Figure 5: Percent of Single-Family and Multi-Family Households within Income Groups
Figure 6: Percent of Single-Family and Multi-Family Households across Income Groups
SDGE ................................................................................................................................ 35 Figure 7: Percent of Households by Square Footage ........................................................ 36
Figure 9: Percentage in Tiers 2-5 (Average Summer Monthly Use) by Square Footage of
Dwelling SDGE ................................................................................................................ 38 Figure 11: Average Income by Climate Group and Square Footage SDGE .................... 39
Figure 13: Average Summer Monthly Usage by Air Conditioner Type and Climate Zone
Group SDGE ..................................................................................................................... 41
Figure 15: Single-Family Pool Groups by Summer Tier Groups and Climate Group
SDGE ................................................................................................................................ 43 Figure 17: Average Income by Swimming Pool Group and Climate Group SDGE ........ 44
Figure 18: CEC Climate Zone Map .................................................................................. 53
Figure 19: SDG&E Electric Baseline Zone Map .............................................................. 54
Prepared Testimony of William B. Marcus on behalf of SDCAN 1 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
I. Introduction and Summary 1
This testimony is presented on behalf of San Diego Consumers’ Action Network 2
(SDCAN) by William B. Marcus, Principal Economist of JBS Energy, Inc. Mr. Marcus 3
has 34 years of energy experience and has appeared before this Commission on many 4
occasions, and has filed testimony or formal comments before about 40 federal, state, 5
provincial, and local courts and regulatory bodies in the U.S. and Canada. His 6
qualifications are attached. 7
SDCAN proposes major changes to SDG&E’s proposed marginal customer costs. We 8
use the One-Time Hookup Cost (OTHC) or New Customer Only (NCO) method, to 9
has been previously adopted by the Commission in a number of recent cases. 11
However the largest change to customer costs involves customer service O&M costs, 12
where SDCAN specifically removes non-distribution energy efficiency costs and non-13
marginal costs of demand response, the California Solar Initiative, and the Self-14
Generation Incentive Program. SDCAN also reduces meter reading costs to reflect AMI 15
deployment (consistent with SDG&E’s inclusion of AMI meters in customer-related 16
capital costs), and functionalizes only 2% of vegetation management costs as customer 17
costs instead of SDG&E’s 12-13%. 18
As a result of the changes to marginal customer costs, SDCAN proposes a reduction to 19
residential distribution charges of almost 18% relative to current rates. Essentially, 20
SDG&E dramatically “overshot” in the last case, and the changes recommended here – 21
particularly to remove costs in Accounts 907-910 that were improperly included in 22
marginal costs – have a significant effect. 23
Finally this testimony provides evidence regarding the load and usage patterns of 24
residential customers in support of MRW’s testimony on residential rate design on behalf 25
of SDCAN. Load research data indicate that apartment dwellers and small users have 26
less peaked load patterns than single-family residents. A review of data from the 27
Prepared Testimony of William B. Marcus on behalf of SDCAN 2 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Residential Appliance Saturation Survey (RASS) indicates that small users have lower 1
incomes, live in smaller units and have fewer peak-oriented appliances (central air 2
conditioning and swimming pools) than larger users. 3
II. Marginal Customer Costs 4
A. Introduction 5
Table 1 presents a comparison of SDG&E’s, DRA’s, and SDCAN’s class average 6
Marginal Customer Cost results. We present a more-detailed comparison (by schedule), 7
below. 8
Table 1: Comparison of Marginal Customer Cost Results 9
SDG&E
(RECC1)
DRA (OTHC
or NCO2)
UCAN
(OTHC or
NCO)
SDG&E >
UCAN
Residential Class 259.05 78.22 89.85 169.21
Small Commercial Class 600.15 292.99 261.11 339.04
Commercial & Industrial Class 2,187.08 1,286.57 1,059.72 1,127.37
Agricultural Class 728.19 438.30 309.42 418.77
Lighting Class (Cost Per Lamp) 19.64 13.12 3.91 15.73
1 RECC is the Real Economic Carrying Charge, which is discussed below.2 OTHC is the One-time Hookup Cost (also called New Customer, Only or NCO), which
is discussed below. 10
SDCAN has identified four issues regarding SDG&E’s calculation of marginal customer 11
costs. 12
First, SDG&E continues the practice of submitting the rental method, which is discussed 13
below, as its preferred method of calculating marginal customer costs. As DRA points 14
out, this is contrary for long-held Commission policy and should not be adopted. In its 15
place, the Commission should adopt the one-time hookup cost method (OTHC), also 16
called “new customer only”, or NCO. DRA also supports the use of the OTHC method. 17
Second, the OTHC calculation relies on forecasts of customer counts. Although it 18
continues to prefer the rental method, SDG&E did provide workpapers with OTHC 19
calculations in its November 2011 filing. SDG&E’s calculation used unreasonably-20
optimistic estimates of customer counts, which should be adjusted downward. DRA is 21
Prepared Testimony of William B. Marcus on behalf of SDCAN 3 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
making this recommendation, but, whereas DRA uses a lower estimate of customer 1
growth to make its recommendation, SDCAN uses estimates of construction units within 2
SDG&E’s service territory calculated from SDG&E’s July, 2011 DRI forecast (and 3
recommended by SDCAN in Phase 1 of the Sempra GRC) and arrives at a slightly 4
different result than DRA. 5
Third, SDG&E has included all of Account 908 costs in its calculation of customer-6
related O&M, which highly inappropriate. SDCAN removes about 90% of Account 908 7
from consideration of marginal cost to account for the following: (1.) costs of energy 8
efficiency programs are not even distribution-related and (2.) other activities that we 9
remove (e.g., demand response, Self Generation Incentive Program, California Solar 10
Initiative) are not marginal costs even if included in Account 908. A portion of the 11
Account 908 costs (for the Commercial, Industrial and Government program) are also 12
assigned to non-residential customers. 13
Fourth, SDG&E functionalizes too much of the distribution O&M costs related to 14
vegetation management as customer-related (and too little as demand-related). SDCAN 15
recommends that the Commission increase the amount of tree trimming costs that are 16
functionalized to primary and secondary lines (distribution demand) from approximately 17
88% to 98% and reduce the functionalization to service drops (customer costs) from 12% 18
to 2%... 19
DRA also makes recommendations regarding (1.) Customer Service Costs; (2.) 20
adjustments to SDG&E’s Transformer, Service and Meter (TSM) costs to reflect 21
applicant contributions; (3.) adjustments for residential infill; and (4.) the definition for 22
Small Commercial Class. As noted above, SDCAN provides an alternative 23
recommendation on customer accounting and customer service costs to DRA’s, with 24
additional reasoning beyond that contained in DRA’s case. We have no position in this 25
testimony on the DRA’s second and third recommendations. As for the fourth point, 26
while we believe DRA’s position is well-reasoned, we did not account for it in our 27
marginal cost calculations, since such a definition change for the Small Commercial 28
Class would have no effect on revenue allocation at least at the level between residential 29
and non-residential customers. 30
Prepared Testimony of William B. Marcus on behalf of SDCAN 4 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
SDCAN discusses in turn each of its issues and recommendations, as follows, and then 1
shows a comparison of results for each rate schedule, below. 2
B. Theoretical Considerations: One-Time Hookup Cost Method vs. Rental 3
Method 4
There are two alternate methods for treatment of customer-related investments in 5
calculating marginal customer costs. The first method is the one-time hookup cost 6
(OTHC) method, also called “new customer only” or NCO. This method has been 7
adopted by the California PUC in a number of recent cases identified below. The OTHC 8
method multiplies the total cost of investment in a new hookup by the number of new and 9
replacement customers added to the system. 10
The alternative method multiplies the cost of a new investment by the real economic 11
carrying charge rate (RECC) and then multiplies this cost by the total number of 12
customers in each class. This approach is equivalent to developing a rental charge for the 13
customer access equipment and is referred to as the “rental method”. The rental method, 14
developed in the 1970s and 1980s, is based on an environment where a competitive rental 15
market for customer access equipment exists but where purchase or up-front payment for 16
that equipment is prohibited. Instead of simulating a competitive market,1 it prohibits 17
purchasing equipment, or paying for it up front in hookup charges, and, thus, simulates a 18
market with extreme barriers to entry by relevant participants in that market. The 19
Commission has not adopted the rental method in recent years. 20
While SDG&E continues to attempt to convince the Commission of the theoretical 21
superiority of the rental method in determining customer costs in this case, SDCAN 22
recommends that the Commission continue its practice of adopting the OTHC method for 23
calculating marginal customer costs in this proceeding, which aligns with DRA’s 24
position. The Commission has adopted the OTHC method in prior litigated cases for all 25
of the major electric and gas utilities in the state with the exception of SDG&E’s electric 26
department, where settlements have simply averaged revenue allocation numbers from 27
four different cost studies including both Rental and OTHC. The OTHC method has 28
been adopted in three PG&E BCAPs and two litigated PG&E electric cases, the 1996 rate 29
1 Marginal cost pricing is theoretically designed to simulate the operation of a competitive market.
Prepared Testimony of William B. Marcus on behalf of SDCAN 5 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
design case for SDG&E, and the 1996 SDG&E gas BCAP, and the 1999 consolidated 1
SoCal and SDG&E BCAP.2 2
1. Interchangeability of Use across Customers 3
SDG&E offers the following in support for its preference for the rental method: 4
SDG&E feels that the “rental” method sends the most accurate price signal to all 5
customers with similar hookups, not just new customers (under NCO).” In the 6
practical application of customer electricity rates, all customers pay a “rental” 7
for the distribution demand-related equipment and other services necessary to 8
maintain an account. The rental method follows the same process by applying the 9
annualized investment cost and ongoing costs required to maintain the account of 10
all customers. 11
In other words, SDG&E contends that because the cost of non-customer-related electric 12
distribution equipment is calculated for purposes of revenue allocation on the basis of the 13
rental method—through the RECC—it is proper to apply the rental method to customer-14
related equipment. 15
The conceptual lumping of customer-related equipment with other, non-customer related 16
equipment when making marginal cost calculations using the RECC is incorrect, 17
however. This is because the use of customer equipment, unlike that of equipment above 18
the line transformer, is not interchangeable across customers. 19
When one customer reduces his or her load there is more capacity on equipment that is 20
above the line transformer available for other customers to use as they increase their load 21
without the utility having to increase capacity. This is the reason that it is correct to give 22
customers ongoing price signals for their use of above-the-line-transformer distribution 23
equipment through the implementation of a rental charge (i.e., the application of the 24
RECC). 25
On the other hand, once customer-hookup equipment is installed it is uniquely and 26
indivisibly used by the customer at that location; other customers would not gain access 27
11 System 100.0% $11,136 $1,025,312 $1,036,448 $1,036,448 0.0% 11
12 12
13 Distribution Revenue Requirement ($000): $1,036,448 13
14 14
15 Non Marginal Revenue Requirement Components ($000): 15
16 Lighting Facilities Charges: $4,600 16
17 Standby Revenue: $4,183 17
18 Distance Adjustment Fees: $2,353 18
Note:
(1) Updated Allocation of Total Distribution Revenue: allocation of the current distribution revenue requirement based on the marginal Distribution Allocation Factors presented in this Application.
(2) Current Total Distribution Revenue Allocation: allocation of current distribution revenue requirement based on the current class distribution allocation percentages reflected in current rates; rates .
effective January 1, 2012, pursuant to SDG&E Advice Letter 2323-E.
(3) Distribution Revenue Requirement: the $1,036,448,000 Distribution Revenue Requirement reflects the current distribution revenues being collected in rates effective January 1, 2012,
excluding revenues that have separate allocation treatment such as Self Generation Incentive Program (SGIP) and Demand Response costs.
(4) Lighting Updated Total Distribution Revenue Allocation: as stated in footnote 3 of the testimony of William G. Saxe (Chapter 3), circuit and substation load data is not available for the
lighting class. For this reason, the Updated Total Distribution Revenue Allocation for lighting is set equal to its Current Distribution Revenue Allocation, using the Goal Seek Factor in Cell O26.
Updated Distribution Revenue Allocation
SAN DIEGO GAS & ELECTRIC COMPANY - ELECTRIC DEPARTMENT
2012 GRC PHASE 2 (A.11-10-002)
UCAN's DISTRIBUTION REVENUE ALLOCATION
Uncapped Distribution Revenue Allocation by Customer Class
Prepared Testimony of William B. Marcus on behalf of UCAN 25 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Current rates and SDG&E’s and SDCAN’s uncapped distribution allocation results are
compared in Table 10.
Table 10: Comparison of SDG&E and SDCAN Distribution Revenue Allocation
Current Rate SDG&E UCAN SDG&E > UCAN
Residential $573,261 $534,119 $471,369 $62,750
Small Commercial $119,152 $131,836 $126,129 $5,708
Medium/Large C & I $330,455 $357,687 $426,319 -$68,633
Agricultural $5,189 $4,414 $4,240 $174
Lighting $8,391 $8,391 $8,391 $0
SDCAN’s allocation to residential customers is $63 million lower; the small commercial
allocation is nearly $6 million less, with larger customers receiving increases. It is also
likely that a lower allocation to residential customers will mechanically cause the CARE
discount to be reduced, thus reducing costs to non-residential classes.
It is likely that caps will be required with this type of allocation change. However, when
considering caps, it is important to look at the total rate changes (given that commodity
and CTC allocations are increasing for residential and decreasing for large customers,
and to look at changes both in percentage terms and in cents per kWh.
Prepared Testimony of William B. Marcus on behalf of UCAN 26 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
I. Residential Customer Characterization
To provide support to the work by SDCAN witness Laura Norin of MRW and
Associates, we are providing information on differences in load pattern by size of
customer (from SDG&E’s residential load research sample) and on economic and
demographic factors that affect customer usage in the SDG&E service territory (from the
Residential Appliance Saturation Survey or RASS data base). The work done here is
similar to work that JBS Energy has done for all of the California utilities on several
occasions, as well as for utilities in Nevada.
Our findings from SDG&E’s load research data are that smaller customers have better
load patterns than larger ones. This finding is consistent with SDCAN’s finding in
previous cases dating back to 2000. The RASS analysis shows that usage, while not in
lockstep with income, has a significant association with income; in particular that the
richest customers on average use more energy. This association arises in part because of
strong correlations between income and the square footage and type of dwelling and the
presence of energy-consuming equipment such as central air conditioning and swimming
pools.
Prepared Testimony of William B. Marcus on behalf of SDCAN 29 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
A. Reason for Errata
These errata correct the analysis of the California Energy Commission (CEC) Residential
Appliance Saturation Survey (RASS) for the SDG&E service territory by using SDG&E
Baseline zones instead of CEC Title 24 climate zones to climatically group customers.
This change affects none of our conclusions, and affects the quantitative results by only a
few percentage points in most cases.
In general, because the mid climate zone using baseline quantities was larger and
included portions of cooler CEC climate zones, both the cool zone and mid zone had
slightly less energy use per customer because the customers used more than average for
the cool zone and less than average for the mid zone. We stand by the general
conclusions presented in testimony but wish to accept SDG&E’s help in assuring that this
analysis is correct.
During rebuttal to Mr. Marcus’ testimony, SDG&E pointed out that the RASS portion of
our analysis used California Energy Commission (C EC) Title 24 climate zones to group
customers instead of SDG&E baseline zones, even though SDG&E provided SDGE
baseline zones for each customer.
We appreciate SDG&E telling us that a variable for the baseline zones was assigned to
each customer, a field that we overlooked in the nearly 800 data fields contained in the
dataset, as it was given the name UTILSDGE. The Title 24 Climate zones were identified
three times in both sets of consumption data (gas and electric) and additionally in the
RASS data using fieldnames such as “T24CZ”, and correspond closely to the baseline
zones, so the effect on the results is minimal. The late delivery of the dataset also hurried
our initial review.
The following updates the original testimony section titled “Relationship of Usage to
Income, Size and Type of Dwelling, and Appliances” beginning on page 29 and the
associated “Attachment E: Methodology for Analysis of Residential Appliance Saturation
Survey”.
Prepared Testimony of William B. Marcus on behalf of SDCAN 30 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
B. Relationship of Usage to Income, Size and Type of Dwelling, and
Appliances
We next examine the reasons why small customers use less energy and have better load
patterns than larger customers. We also examine relationships of consumption, among
single-family and multi-family customers by income.
At a high level, consumption is not in lockstep with income. However, there are
relatively strong correlations between consumption, size of dwelling, whether the
dwelling is single and multi-family, saturation of energy consuming appliances such as
central air conditioners and swimming pools, and income. As a result, the proposals by
SDG&E will give disproportionate rate breaks to large customers who are more likely to
have central air conditioners and swimming pools that contribute to peak loads and who
tend – on average - to be more affluent, while raising rates to CARE customers and many
other smaller customers who own less peak-heavy equipment.
We divided the SDG&E system into three climate zones groups – Cool, Mid, and Hot,
based on the SDG&E baseline zones and associated weather stations that each customer
was assigned to. The cool zone was SDG&E zone 1: the coastal baseline zone The Mid
climate group was the SDG&E inland (SDG&E zone 2) and mountain (SDGE zone 4)
baseline zones which had similar baseline quantities. The Hot Zone Group was SDG&E
baseline zone 3: low desert). We have not reported results for SDG&E’s hot zone, due to
a statistically insignificant number of RASS survey responses (only 20 respondents).
We broke the customers in each climate zone into groupings based on the average use of
the four inner summer months (June-September 2008). Each grouping was roughly
based on the average monthly summer quantities in the Cool and Mid zones (less than
130% of average basic baseline, 130-200%, 200-300%, and over 300%) rounded to the
nearest 10 kWh per month.
Our definition of which tier group a customer falls into is based on a monthly average of
the four peak summer months. In our analysis, a customer is in a Summer Tier Group if
the monthly average of the four summer months’ consumption falls within the Summer
Tier Group range. These groups roughly correspond to usage in each tier (though there
Prepared Testimony of William B. Marcus on behalf of SDCAN 31 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
may be some small amounts of spillover into the higher tier in the warmest summer
months).
We cross-tabulated and analyzed income by tier grouping, and by whether customers
were single-family and multi-family in each of the climate zones. We also analyzed the
saturation of central air conditioning and swimming pools by income and by tier
grouping and analyzed the relationship of the square footage of dwellings to tier grouping
and income.
More methodological information is contained in Attachment E.
1. Income
In the SDG&E zones, usage (measured by Summer Tier Group) increases with income in
the cool and mid climate zones.
Figure 1: Average Income by Summer Tier Group and Climate Group
The percentage of customers with income under $30,000 who had Tier 4 or 5 usage
(average monthly use above 200% of baseline in those four summer months) was 8% in
the cool zone and 6% in the mid zone. By comparison the percentage of customers over
$100,000 with Tier 4 use was 41% in the cool zone and 48% in the mid zone.
Prepared Testimony of William B. Marcus on behalf of SDCAN 32 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 2: Income Percentages by Summer Tier Group and Climate Group SDGE
Prepared Testimony of William B. Marcus on behalf of SDCAN 33 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
The reason is clear. Higher incomes are associated with larger dwellings, more saturation
of central air conditioning, and more swimming pools, as shown below. We start with an
examination of usage, income, and type of dwelling as related to square footage.
2. Single vs. Multi-Family
Multifamily customers use considerably less than single-family customers as shown in
the two figures below. Over 70% of multi-family customers use less than 130% of
baseline on average while very few use more than 200% of baseline.
Figure 3: Percent of Single-Family and Multi-Family Households within Tier Groups and Climate Zones SDGE
Prepared Testimony of William B. Marcus on behalf of SDCAN 34 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 4: Summer Average Monthly Kwh by Single-Family and Multi-Family Households
Multifamily customers use about 45% to 48% less than single-family customers in both
of the major climate zones. This phenomenon can be expected because of the smaller
size of the dwellings and common walls that reduce heat gain and loss, as well as income
differences that may affect usage.
There also are large differences in income between single-family and multi-family
dwellers. While a majority of households in all income groups live in single-family
dwellings in SDG&E’s service area as a whole, the proportion rises from 32% to 87% as
income rises.
Prepared Testimony of William B. Marcus on behalf of SDCAN 35 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 5: Percent of Single-Family and Multi-Family Households within Income Groups SDGE
Figure 6: Percent of Single-Family and Multi-Family Households across Income Groups SDGE
On the SDG&E system as a whole, 55% of single-family dwellers earned more than
$75,000, compared to 20% of multi-family households. Both climate zones showed a
disproportionate percentage of households under $30,000 in multifamily units as
expected.
Prepared Testimony of William B. Marcus on behalf of SDCAN 36 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
3. Square Footage
Figure 7 shows the percentage of dwellings by square footage. The more urbanized cool
area has more dwellings under 1500 square feet than the suburban inland area.
Figure 7: Percent of Households by Square Footage
Average usage generally increases with square footage. (Figure 8).
Prepared Testimony of William B. Marcus on behalf of SDCAN 37 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 8: Average Summer Monthly KWh Usage by Climate Group and Square Footage SDGE
Figure 9 computes the percentage of customers with usage in each tier with dwellings of
a given size. For those in dwellings less than 1000 square feet, 77% in cool zones and
95% in mid zones were at or below Tier 2 levels. Only 6.1% of those in cool zone
dwellings over 2500 square feet and 34% in mid zones were in the Tier 2 range. In these
large dwellings, 53% in the mid zone and 65% in the cool zone had average summer
usage that fell into Tier 4 or Tier 5.
Prepared Testimony of William B. Marcus on behalf of SDCAN 38 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 9: Percentage in Tiers 2-5 (Average Summer Monthly Use) by Square Footage of Dwelling SDGE
Prepared Testimony of William B. Marcus on behalf of SDCAN 39 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
There is a strong correlation between square footage of dwellings and income. Of those
in dwellings over 2500 square feet, 47 to 75% (depending on climate zone) earned more
than $100,000. Very few people earning over $100,000 lived in dwellings under 1,000
square feet – 13% in the more urbanized cool zone, and 9% in the mid zone (Figure 10).
Figure 10: Square Footage within Income Groups by Climate Zone SDGE E
Figure 11: Average Income by Climate Group and Square Footage SDGE
Prepared Testimony of William B. Marcus on behalf of SDCAN 40 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
4. Air Conditioning
Appliance such as air conditioners and swimming pools also affect summer peak usage
and saturation of these appliances is correlated with income.
The average income of a central air conditioning user is higher in all climate zones. See
Figure 12.
Figure 12: Average Income by Air Conditioner Type and Climate Group SDGE
Relative to having no air conditioner, a central air conditioner increases average monthly
summer usage by 74% in the cool zone (an increase of 284 kWh per month) and about
51% in the mid zone (an increase of 248 kWh per month).
Prepared Testimony of William B. Marcus on behalf of SDCAN 41 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 13: Average Summer Monthly Usage by Air Conditioner Type and Climate Zone Group SDGE
5. Swimming Pools
Swimming pools also are correlated with energy use and income. Customers must have
and pay for the energy it uses before they are counted as having a pool. Pools in common
areas are grouped with those without a pool. It should be noted that virtually no one in a
multifamily dwelling has a pool. Thirteen percent of households have pools. They use
more energy and have higher incomes than other households. Pool users tend to fall into
higher tier groups, and their usage is higher.
Prepared Testimony of William B. Marcus on behalf of SDCAN 42 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 14: Pool Ownership across Income Groups SDGE
As expected, there are very few swimming pool owners at the low end of income; it rises
to 17-19% for incomes over $75,000.
Prepared Testimony of William B. Marcus on behalf of SDCAN 43 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 15: Single-Family Pool Groups by Summer Tier Groups and Climate Group SDGE
In the cool to mid climate zones, a pool owner has usage that is 86-103% higher than a
household without a pool, an increase of 376 kWh per summer month in the cool zone
and 551 kWh per month in the mid zone. (Figure 16) The increase in usage with a
swimming pool appears larger than with Edison and may be correlated with other factors.
Prepared Testimony of William B. Marcus on behalf of SDCAN 44 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Figure 16: Average Summer Monthly Kwh Usage by Pool Group and Climate Group SDGE
As shown in Figure 17, average incomes of pool owners are 26-33% higher than of those
without swimming pools.
Figure 17: Average Income by Swimming Pool Group and Climate Group SDGE
Prepared Testimony of William B. Marcus on behalf of SDCAN 45 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
6. Conclusion
The RASS data provided by SDG&E provides support for the contentions that lower
users who will be charged more by a customer charge are of lower income, are more
likely to live in apartments and smaller dwellings in general, and do not have as much
peak-oriented energy consuming equipment (central air conditioners and swimming
pools). This information will be used by the MRW witnesses in support of their
residential rate design recommendations.
Attachments
Prepared Testimony of William B. Marcus on behalf of SDCAN 46 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Attachment A: Qualifications of William B. Marcus
Principal Economist, JBS Energy, Inc.
William B. Marcus has 28 years of experience in analysis of electric and gas utilities.
Mr. Marcus graduated from Harvard College with an A.B. magna cum laude in
economics in 1974 and was elected to Phi Beta Kappa. In 1975, he received an M.A. in
economics from the University of Toronto.
In July, 1984, Mr. Marcus became Principal Economist for JBS Energy, Inc. In this
position, he is the company’s lead economist for utility issues.
Mr. Marcus is the co-author of a book on electric restructuring prepared for the
National Association of Regulatory Utility Commissioners. He wrote a major report on Performance Based Ratemaking for the Energy Foundation. He analyzed restructuring
and stranded costs in eight states and provinces for environmental, consumer, and
independent power clients.
Mr. Marcus has prepared testimony and formal comments submitted to the Federal Energy Regulatory Commission, the National Energy Board of Canada, the Bonneville
Power Administration, the U.S. Bureau of Indian Affairs, U.S. District Court in San
Diego, Nevada County Municipal Court, committees of the Nevada, Ontario and California legislatures and the Los Angeles City Council, the California Energy
Commission (CEC), the Sacramento Municipal Utility District (SMUD), the
Transmission Agency of Northern California, the State of Nevada’s Colorado River Commission, two arbitration cases, environmental boards in Ontario, Manitoba, and
Nova Scotia; and regulatory commissions in Alberta, Arizona, Arkansas, British
Columbia, California, Colorado, Connecticut, District of Columbia, Hawaii, Manitoba, Maryland, Massachusetts, Nevada, New Jersey, New Mexico, North Carolina,
Northwest Territories, Nova Scotia, Ohio, Oklahoma, Ontario, Oregon, South Carolina,
Texas, Utah, Vermont, Virginia, Washington, Wisconsin, and Yukon. He testified on issues including utility restructuring, stranded costs, Performance-Based Ratemaking,
resource planning, load forecasts, need for powerplants and transmission lines,
environmental effects of electricity production, evaluation of conservation potential and programs, utility affiliate transactions, mergers, other revenue requirement issues,
avoided cost, utility business risks, regulated retail return margins, and electric and gas
cost of service and rate design.
From April, 1982, through June, 1984, Mr. Marcus was principal economist at
California Hydro Systems, Inc., an alternative energy consulting and development
company. He prepared financial analyses of projects, negotiated utility contracts, and
provided consulting services on utility economics and resources.
From July, 1978 through April, 1982, Mr. Marcus was an economist at the CEC, first in the energy development division and later as a senior economist in the CEC’s
Executive Office. He prepared testimony and economic studies on purchased power
pricing, transmission projects, renewable resources and conservation programs, and
managed interventions in utility rate cases.
Prepared Testimony of William B. Marcus on behalf of SDCAN 47 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
From 1975 to 1978, Mr. Marcus was a research analyst at the Kennedy School of
Government, Harvard University.
Mr. Marcus served on the 1991-92 SMUD Rate Advisory Committee, which made cost allocation and rate design recommendations to the SMUD Board. He has served on
several advisory committees for local governments in California.
Prepared Testimony of William B. Marcus on behalf of SDCAN 48 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Attachment B: Responses to SDCAN Data Requests 8-6 to 8-11
Prepared Testimony of William B. Marcus on behalf of SDCAN 49 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Attachment C: Excerpts from Prepared Testimony of Kathleen Cordova in Sempra GRC, Phase 1 (Exhibit SDG&E-15)
Prepared Testimony of William B. Marcus on behalf of SDCAN 50 SDG&E 2012 Test Year General Rate Case Phase II (CPUC App. A. 11-10-002)
Attachment D: Supporting Tables for Calculation of Distribution Revenue Allocation
Attachment E: Methodology for Analysis of Residential Appliance
Saturation Survey
San Diego Gas and Electric (SCE) provided the dataset used in this analysis. The
California Energy Commission (CEC) produced the dataset used in this analysis from data collected in the 2009 Residential Appliance Saturation Survey or RASS and
provided the dataset to SDG&E. SDG&E also provided gas and electric consumption
customer keyed to the RASS responses. SDG&E provided JBS with a comma separated value file with the orinal CEC RASS survey responses converted to string values. These
string values had to be converted back to the original CEC RASS ordinal values for
proper analysis. JBS combined the RASS survey with the billing consumption data. Each sample data point is weighted by the CEC to compensate for the sampling methodology;
we used the same RASS weighting factors in all of our analysis.
The statistical package SPSS (now owned by IBM) was used for the analysis to create a
combination of weighted crosstab and summary tables. Additional SDG&E related variables were created from the existing RASS variables and added to CEC RASS Survey
and Billing dataset. The RASS data was filtered to exclude incomplete electric
consumption data for the year June of 2008 through July of 2009, average monthly consumption during the summer months of June to September under 50 Kwh, bill days
for June 2008 under 25 days, household sizes equal to zero, seasonally occupied
residences, net metered homes, and all-electric homes.
All Electric homes include those with no gas consumption identified in the consumption
variable “yr1useg”that provides one years’ therm usage.
The single-family and multi-family group was created (SFMF variable) with two groups.
Mobile homes were grouped with Multi-Family households.
Air Conditioning (AC) Groups (ACGroup variable) were created from the RASS
variables for identifying Central AC, Room AC and Central Evaporative Coolers: No
AC, Room AC or a Central Evaporative Cooler, and Central AC.
The swimming pool groups (PoolPay variable) were created from the RASS “pltype” for
swimming pools to identify locations where the customer has and pays for the pool
energy (not a common area pool).
Square Footage and Income Groups were created by re-categorizing the more numerous
CEC groups into a smaller number of categories. Average incomes were assigned the
average of the income range (IncAvg variable) with the exception of the $150,000 and
more category, which was given the same average value of $175,000 as the CEC used. When used in this analysis, this value may understate some of income averages using
the highest income category values.
SDG&E Baseline Zones 1-4, were combined into three Climate Zone Groups: Cool, Mid,
and Hot. The Cool zone represents the immediate cool coastal area baseline zone, the Mid zone combines coastal influenced warmer coastal areas zones and mountain
baseline zones, and the Hot Zone which includes the El Centro dessert baseline climate
zone only. Hot Zone results were not reported in this analysis as the small number of surveyed customers might distort the results. The following table presents the number
of surveyed customers by climate zone. Note that due to the data validation and filtering
discussed above, the number of surveyed customers used in this analysis would likely
be lower than those reported below.
Table 11: Un-Weighted number of RASS Surveys by Climate Zone
SDGE Baseline Climate Zone Group Case Summaries
SDGE Baseline Zones N % of Total N
1 - Coastal 2313 56.70%
2 - Inland 60 1.50%
3 - Desert 20 0.50%
4 - Mountain 1688 41.40%
Total 4081 100.00%
The KWh consumption of the four summer months of June through September 2008 was
added into the SumKWh69 variable (data in the first monthly consumption variable in
the dataset not from June in 2008 was excluded). The average of the four months was placed in the SumKwhAvg variable. This was used to establish the upper tier level of
consumption in each Climate Group for the four summer months using SDG&E’s tier
criteria. Table 12 shows the derivation of Climate Zone Groups and Summer Tier Groups by Climate Zone Group and Figure 18 is a Map of CEC climate Zones. While not
sharing the exact geospatial boundaries, the CEC Title 24 climate zones correspond well
to the SDG&E Baseline zones.
Table 12: SDG&E Summer Tier Groups used in CEC RASS Analysis
SDGE Schedule DR
Summer: June to September Tier Groups Tier 2 Tier 3 Tier 4 Tier 5bCEC Title 24
Zone Baseline Zone Tier Group
Baseline
Allowance
Monthly
Baseline
Summer
Baseline Total 130% 200% 300% 500%
5 Santa Maria 5 0 0 - - - - -
6 Los Angeles 6 Cool 9.6 293 1,171 381 586 878 1,464
7 San Diego 7 Cool 9.6 293 1,171 381 586 878 1,464