V. Throughput

A. Econometric Throughput

Both ORA and SoCalGas used econometric models to forecast throughput for residential, commercial core (G-10), industrial core (G-20), and commercial/industrial noncore (G-30) customers. The table below sets forth the ORA and SoCalGas direct showing forecasts for each of these classes as well as the forecasts for the relatively new gas air conditioning and gas engine customer classes. At the request of the ALJ, ORA and SoCalGas also prepared forecasts based on the forecast period of 1999-2001. Those forecasts are also included in the table. The analysis which follows is generally based upon each party's initial showing, the ORA forecast for the years 2000-2002, and the SoCalGas forecast for 1999.

TABLE 1

ORA and SoCalGas Econometric Demand Forecasts

(MMdth)

Class	ORA	ORA	SoCalGas	SoCalGas
Years	2000-2002	1999-2001	1999	1999-2001
Residential	263.02	261.20	254.70	257.90
G-10	78.689	78.25	79.10	79.90
G-20	4.70	4.68	4.70	4.70
Gas Engine	1.60	1.60	1.60	1.60
Gas AC	0.12	0.12	0.12	0.12
Total Core	348.00	345.90	340.20	344.30
Comm/Ind G-30	147.00	146.40	146.90	141.50

ORA and the company disagree on the forecasts for residential demand, commercial core (G-10) demand, and commercial/industrial noncore (G-30) demand. For the residential class, ORA forecasts an average throughput over a three year BCAP period (2000-2002) of 263.02 MMdth while the company forecasts a 1999 demand of 254.70 MMdth, a difference of 3.3%. ORA forecasts G-10 demand of 78.689 MMdth while SoCalGas forecasts demand at 79.107 MMdth, a difference of less than 1%. For G-30 load, ORA forecasts a demand of 147.0 MMdth while SoCalGas forecasts a demand of 146.9 MMdth, again a difference of less than 1/2%. These differences are the result of differing assumptions regarding the length of the forecast period and the number of heating degree days. ORA relied on a three year forecast of demand for the period 2000-2002 while the company used a forecast of throughput for a single year, 1999.

The second major difference results from different estimates of heating degree days (HDD) which are a key factor in explaining historic gas demand, particularly residential demand. The majority of California's gas and electric utilities rely on historical averages to forecast heating and cooling degree days. ORA followed this practice and based its estimate on a 20-year average with a resulting estimate of 1358 heating degree days. SoCalGas proposes to move away from the traditional approach and bases its estimate on a trend analysis. The trend analysis produces an estimate of 1,222 heating degree days. SoCalGas' lower heating degree day estimate has the effect of increasing rates to all customers, raising residential rates by 4%, wholesale rates by 3%, and noncore rates by approximately 1%.

SoCalGas justifies the trend analysis on the ground that southern California has been experiencing a warming trend over the past 20 years which isn't captured through a 20-year average. ORA asserts that the 20-year average reasonably captures any warming trend since it results in an HDD estimate which is 10% lower than the estimate used in the Global Settlement. ORA believes that the SoCalGas model goes too far in producing an estimate that is almost 20% lower than the estimate used in the Global Settlement. SCGC has serious doubts about the accuracy of the linear trend analysis used by SoCalGas as it produces results significantly at variance with traditional methods and raise rates for all customers.

B. Non-Econometric Throughput

Eight categories of throughput on the SoCalGas system are forecast non-econometrically: (1) exchange contracts; (2) enhanced oil recovery (EOR); (3) wholesale (excluding SDG&E electric generation (EG)); (4) SoCalGas EG; (5) cogeneration; (6) SDG&E EG; (7) Distribuidora de Gas Natural de Mexicali (DGN); and (8) Rosarito. The following table compares the ORA and SoCalGas forecast for each of these categories for a BCAP period which extends through the year 2002. At the request of the ALJ, ORA and SoCalGas also prepared a forecast based on the years 1999-2001. These forecasts are included in the table.

TABLE 2

(MMdth)

	ORA	SoCalGas
CUSTOMER CLASS	2000-2002 1999-2001	1999 1999-2001
Exchange	9 9	9 9
EOR	49 49	49 49
Wholesale	94 90	94 95
EG	230 215	202 181
Cogeneration	85 85	84 77
SDG&E EG	48 44	44 37
DGN	5 5	3.6 4
Rosarito Adjustment	25 15	0
Total	545 512	485.6 452

As can be seen from the table, the major areas of dispute relate to the estimates for EG and cogeneration throughput on the SoCalGas system and EG throughput on the SDG&E system. While the table shows a significant difference between ORA and SoCalGas with respect to Rosarito throughput, this is the result of different ratemaking recommendations. SoCalGas proposes to exclude Rosarito throughput from the cost allocation process and instead recommends a revenue crediting mechanism. ORA, on the other hand, recommends including the throughput in the cost allocation process in order to develop a full cost of service rate.

Two factors account for the differences between ORA and SoCalGas. First, SoCalGas proposes to use a single year's forecast, 1999, for the entire BCAP period while ORA proposes using the average of a three year forecast for period 2000-2002. Second, ORA used a more recent forecast of electric demand. As an input assumption, ORA used the California Energy Commission's (CEC) Outlook 1998 forecast of electric demand for the years 2000-2002. ORA also used SoCalGas' forecast of gas prices for the same period even though recent experience indicates that the forecasts may be on the high side.

SoCalGas claims the ORA forecast is too high because it failed to take a number of relevant factors into account. Accounting for off-system generation, NOx emissions, and Qualifying Facility (QF) buyouts, SoCalGas provided a total EG forecast (SoCalGas and SDG&E) for the period 2000-2002 of 233 MMdth. This is lower than the ORA forecast of 324 MMdth, and is considerably lower than the company's own 1999 forecast for 300 MMcfd. SoCalGas reduces the throughput even further by assuming 13 MMdth is lost to bypass in the year 2000 and 21 MMdth is lost to bypass in 2001.

ORA contends that SoCalGas' forecasts are contrary to CEC forecasts which show EG gas demand increasing over the 2000-2002 period rather than decreasing. The CEC, in its 1998 Natural Gas Outlook, projects EG demand similar to SoCalGas for 1999. However, it forecasts EG demand growing by 13% on average for the period 2000-2002. A 13% increase over SoCalGas' 1999 forecast of 256 MMdth yields a forecast of 289 MMdth. This is consistent with the ORA forecast of 286 MMdth.

C. Revenue Risk

For the past five years, SoCalGas has been at risk for both noncore throughput variations and discounting. SoCalGas proposes to continue that practice over the upcoming BCAP period provided that its 1999 forecast of throughput is adopted. Given the forecast of EG throughput contained in its presentation, SoCalGas admits that its 1999 forecast is a "stretch" target which balances risk and reward. It takes the position that if a higher throughput is adopted it should be protected from the risk of a three year forecast through reinstitution of the 75/25 ratepayer/shareholder balancing account protection that existed prior to the Global Settlement.

ORA argues that the SoCalGas proposal is skewed in favor of shareholder rewards. SoCalGas' policy witness acknowledges that the 1999 forecast is designed to provide upward earnings potential. This is also evidenced by both the higher ORA forecast using either a 1999-2001 or 2000-2002 period and the CEC's estimate for EG gas demand for the 2000-2002 period.

D. Gas Price Forecast

In past BCAPs, the gas price forecast served an important role since it was used to set the core procurement rate. Core gas prices are now revised monthly for both SoCalGas and SDG&E to track market conditions. Because of this regulatory change, the gas price forecast is less significant. Its use now is as an input to the econometric and production cost models used to forecast core and noncore throughput. After reviewing the model's sensitivity to price changes, ORA relied on SoCalGas' gas price forecasts for the years 2000-2002.

E. Impact of the Joint Recommendation

The JR would resolve the throughput issue by adopting a higher level of throughput than that proposed by SoCalGas. It would also reinstitute 75/25 balancing account protection for noncore revenue. This is a reasonable compromise given the litigation positions of the parties. ORA and other parties take the position that SoCalGas' forecast is too low while SoCalGas takes the position that the ORA forecast is too high. The JR adopts a forecast considerably higher than the one contained in SoCalGas' rebuttal testimony. The 75/25 balancing account is reasonable since it will continue to place shareholders at some risk for discounting while protecting shareholders and ratepayers in the event the adopted forecast is significantly off the mark.