7. Forecast of Gas Sales

ORA forecast gas sales for 2004 through 2008 based on average weather conditions, expected customer growth, economic conditions and phase-in price levels. SCE forecast very small growth in the number of residential customers and no growth in the number of business customers. Given the limited amount of development occurring in Avalon, ORA concluded that SCE's forecasts of customer growth were reasonable. We adopt SCE's customer growth estimates.

The parties differed, however, on their forecasts of gas sales per customer. While both used similar statistical and mathematical models to forecast gas sales, SCE assumed that the rate increase it seeks will cause reduced gas usage. ORA's analyst presented a regression analysis of recorded gas usage that found no significant relationship between rates and usage for residential customers and only a minor effect for business customers. While the forecasts of both parties reflect a decline in sales for 2005, ORA forecasts a somewhat smaller decline. ORA's analysis was based on the methodology that SCE used for projecting gas sales in its 1986 gas case. We will adopt the ORA analysis of gas sales, in part because the rate increase we authorize today is smaller than that sought by SCE and presumably will not result in decreased usage to the extent forecast by the company.

7.1 Gas Sales Adjustment Mechanism

For the first time in its rebuttal testimony, Edison proposed that the discrepancy in forecast sales between ORA and Edison could be handled through the creation of a balancing account. Called the Gas Revenue Adjustment Mechanism (GRAM), the balancing account would be used to calculate the difference each month between authorized gas base rate revenues and actual gas base rate revenues. If the monthly measure showed an over-collection, the amount would be credited to the GRAM; if it showed an under-collection, the amount would be debited. Rates would be adjusted annually to reflect the GRAM balance, including interest.

Like ORA, we question whether the cost and administrative burden of monitoring and reviewing this balancing account is worth the small amounts of money that would be involved. The difference in SCE and ORA sales forecasts for 2005 is relatively small. The proposed balancing account was presented late in the game and has not been subjected to extensive analysis of costs and benefits. We decline to adopt a GRAM balancing account at this time.