Three Issues - Best Methodology, Date of Application, and Procedure for Implementing

Three major issues in this investigation require resolution by this Commission. First, what methodology for determining gas rates best conforms to the principle that core and non-core customers, each pay the transportation costs associated with providing gas to them? Second, what should be the effective date for this change in methodology? Third, how to resolve those issues, if any, that arise in implementing this change in methodology? We address each in turn.

Issue 1: What is the Appropriate Methodology for Calculating the Gas Procurement rates for SDG&E's Core and Non-Core Customers?

I.00-08-003 states that "It has always been the Commission's intent that the non-core procurement customers would be charged the cost of interstate pipeline capacity used by them while the core customers would pay for their interstate capacity." Both ORA and SDG&E agree that the cost of interstate pipeline capacity is now implicitly included in the price of gas that the utility buys at the California border. Under its current methodology for setting gas procurement costs, SDG&E averages the costs of border gas, which includes transport costs, and the costs of other gas purchases, which do not include transport costs, over the total forecasted core and non-core procurement volumes. This calculation yields a per-therm cost of gas, called the weighted average cost of gas (WACOG). Both the core and non-core customers currently pay this WACOG. Furthermore, the core customers pay reservation and transportation charges for interstate pipeline usage.

The procedure of putting all gas costs into a common pool spreads the transportation costs included in border gas over every therm of gas included in the WACOG. Thus, the simple averaging of total gas costs over total gas demand results in an inappropriate shifting of transportation costs from non-core customers to core customers. This outcome arises because gas purchases at the border, which include transportation costs, principally serve non-core gas customers, while other gas purchases, which do not include transportation costs, serve core customers (who separately pay for transport costs).

ORA, in its Comments of September 7, 2000, proposed two methods for correcting this faulty methodology that it describes as follows:

1. "Total WACOG - calculate procurement rates for both core and non-core customers based on the average cost of all gas supplies, including commodity, variable transportation, firm reservation costs."

2. "Border price - calculate non-core rates based on the cost of gas for purchases at the California border, which implicitly include interstate capacity costs. Core rates would be calculated using the existing methodology, with the exception that non-core border volumes would be excluded." (ORA, Comments, p.2.)

In addition, ORA, although offering both options, recommended that the Commission direct SDG&E to revise its methodology consistent with Option 2. Subsequently, in the ORA Opening Brief, filed on October 27, 2000, ORA again asserted its preference for Option 2, the Border Price Method. (ORA Opening Brief, p. 1.)

The SDG&E Opening Brief states, "After reviewing ORA's comments and conferring with ORA, SDG&E now recommends adoption of ORA Option 2 in this investigation." (SDG&E Opening Brief, p. 3.) SDG&E further states that this approach is consistent with prior Commission decisions since SDG&E holds long-term firm interstate pipeline capacity "only to meet core reliability needs." (SDG&E Opening Brief, p. 3.)

In addition, SDG&E included in its opening brief a detailed statement of Option 2. It says:

"ORA Option 2: Non-core customers are allocated the average of SDG&E's monthly gas procurement costs for purchases only made at the California border, which implicitly include interstate capacity costs. The noncore weighted average cost of gas (Noncore WACOG) is determined by dividing the total costs for all gas supply purchases at the California border by the total volume of those supply purchases."

"Core customers are allocated all remaining SDG&E monthly gas procurement costs not allocated to noncore customers. The core weighted average cost of gas (Core WACOG) is determined by dividing the total commodity, variable transportation, and reservation costs for all gas supply purchases from US basins, Canada, and at the California border, less the total noncore costs for all gas supply purchases at the California border, by the total core usage, adjusted for core storage withdrawals as applicable." (SDG&E Opening Brief, pp. 9-10.)

SDG&E characterizes this methodology as differing from the current methodology in that the "noncore is not assigned any of the commodity that flows utilizing these firm pipeline reservation charges - the non-core WACOG is based solely on border costs." (SDG&E Opening Brief, p. 10.)

Finally, in the LIF Reply Brief, LIF states that it, too, agrees "with ORA that Option 2 (or the Border Price Method) appears more favorable to residential CORE customers than the first option." (LIF Reply Brief, p. 1.) LIF states that its preference for Option 2 arises in part because it finds Option 2 "more favorable to residential CORE customers, yielding an additional $1.7 million to the unfair cross-subsidy versus $1.1 million under Option 1." (LIF Reply Brief, p. 1) Using information from a series of data requests, LIF identifies the impact of Option 1 and Option 2 on core residential, core commercial, and non-core customer classes. LIF presents its analysis is a series of tables showing the impact by month of Options 1 and Option 2. (LIF Reply Brief, pp.2-3.) Based on its analysis, LIF urges the Commission to adopt the Border Price Method.

Resolving Issue 1: Option 2, the Border Price Method, Best Meets Current Commission Policy

Option 2, the Border Price Method, as described by SDG&E, offers a methodology for determining the costs of procuring gas for the core and non-core customers that best meets the principle that core and non-core customer classes should each pay the transportation costs associated with providing their natural gas service. In particular, Option 2 assigns the firm pipeline reservation charges to the core customers, for whom SDG&E holds the reservations of pipeline capacity. Further, it avoids assigning to core customers transportation costs embedded in charges for "border gas" that these core customers did not consume. More specifically, it calculates two WACOG's. One WACOG is for the non-core customers, and is based solely on gas purchased at the California border. The other WACOG is for the core customers, and it contains all residual costs of purchasing gas, as well as the variable and reservation costs for the transport of gas, which are incurred to move gas bought in US and Canadian Basins to California.

As long as border purchases remain the sole source of gas procured for non-core customers, this methodology provides the right allocation of transportation costs to customer classes. Indeed, it conforms fully to the principle of assigning gas costs to the customer class for which the utility incurs

the costs. If, however, SDG&E begins to make purchases for non-core customers directly from gas basins, then SDG&E will need to modify this methodology to insure that commodity gas prices continue to reflect those costs associated with serving a particular customer class. Failure to do so may cause a discrepancy between prices and the cost of serving a class of customers.

Option 1, in contrast to Option 2's precise allocation of transportation costs, offers a rough approach to the allocation of transport costs. Under Option 1, all costs of gas, both procurement and transport, are dumped into one grand WACOG, identical for both the core and non-core customers. Under this approach, all customers pay transportation costs in proportion to the amount of gas that the customer class uses. This avoids the drawback of the current methodology, in which core customers pay both the transport costs for the gas that they use and a portion of the transport costs of gas bought for the non-core users. Option 1, however, treats all gas transport costs and gas purchases alike, ignoring the fact that SDG&E makes basin purchases and reserves transport capacity to serve core customers, but purchases border gas for its non-core customers. Thus, Option 1 fails to conform fully to the policy principle embedded in the prior Commission decisions, the principle of assigning gas costs to the customer for whom the utility incurs the costs.

In addition to consideration of the conformity of the pricing methodology with principles previously adopted by the Commission, it is also important to consider the impact of our decisions on rates. Concerning this issue, the calculations of LIF show us that both Options 1 and 2 lower the gas procurement costs and rates for core customers. (LIF Reply Brief, pp. 2-3.) In addition, LIF shows us that our choice of Option 2 leads to decreases in gas costs in monthly residential bills, and decreases greater than those generated by Option 1. These decreases, however, are low, ranging from a low of $.01 for March of 2000 to a high of $.40 in September of 2000.

An analysis of the stipulated facts filed by SDG&E and ORA on October 13, 2000 indicates that the use of Option 2 results in a decrease in core rates of $1.7 million when compared to the current methodology for the recorded period of February 2000 through October 2000. This reduction is $717,040 greater than that yielded by Option 1 for this recorded period.⁷ The choice of Option 2 will therefore likely result in lower gas bills of core customers, somewhat lower than those that Option 1 would produce. For core customers, under Option 2, gas prices would have dropped up to $.186 per Million British Thermal Units (MMBtu), or a little more than 3% in the peak month for which we have historic data. Had Option 2 been in effect from February 2000 through October 2000, core customers would have paid on average 1.1 % less for gas.

For non-core customers, Option 2 produces greater increases in gas prices than Option 1. Gas prices would rise up to $.636 per MMBtu, or a little over 11%, in the peak month during the period for which we have historic data. Had Option 2 been in effect from February 2000 through October 2000, non-core customers would have paid on average 5.0 % more for gas. Although this increase is substantial, the resulting rates remain reasonable.

In summary, we adopt Option 2 because it fully conforms to the long-standing policy of assigning costs to those customers causing the costs, while Option 1 offers only a rough conformity to this principle. Moreover, Option 2 results in modest reductions in the gas rates for core customers, and modest increases in the gas rates for non-core customers. In both cases, gas rates remain reasonable.

Issue 2: What Should Be the Effective Date for this Change in Methodology?

The SDG&E Opening Brief states that a change in methodology "should be prospective only, effective with the first monthly commodity price filing following an order to change methodology (SDG&E Opening Brief, p. 4)." SDG&E argues that any other implementation plan "carries the potential of disturbing the marketplace and the reasonable reliance of customers on current market prices in making their supply and consumption decisions." (SDG&E Response, Attachment, p. 4.)

SDG&E claims that the Order Instituting Investigation (OII) "acknowledges" that SDG&E's tariff filings were all in compliance with "the then-effective Commission decisions." It argues that if changes in methodology are adopted, they should be applied only prospectively. (SDG&E Opening Brief, pp. 4.)

ORA agrees with SDG&E that the cost reallocation "should be prospective, effective in the first monthly procurement advice letter filing after the Commission issues a decision in this proceeding." (ORA Opening Brief, p. 4.) ORA argues that a reallocation of costs would be "unfair to noncore procurement customers that relied upon a Commission approved tariff in making their procurement decisions. (ORA Opening Brief, p.4.)

ORA also states two other reasons why it supports a prospective application of this cost allocation methodology. First, ORA argues that higher prices may cause procurement customers to reduce consumption or switch vendors. Second, ORA states that there is no evidence that SDG&E "violated or failed to comply with any Commission decision or resolution." (ORA Opening Brief, pp. 4-5.)

In contrast, LIF supports the application of this revised methodology for setting gas procurement prices starting from the first protested advice letter. LIF notes that since the ORA protest of SDG&E's February 2000 tariff filing, all parties had "notice that SDG&E's gas procurement methodology was incorrect and providing a cross-subsidy to NONCORE customers." (LIF Reply Brief, p. 3.) LIF notes that since the issuance of the OII on August 3, 2000, "all NONCORE customers were put on notice of the flawed procurement methodology and should not be allowed to profit from this cross subsidy." (LIF Reply Brief, p. 3.) Further, LIF cites the size of the misallocation and the potential termination of the non-core procurement program as facts supporting a full crediting of core customers. (LIF Reply Brief, p.5.)

Resolving Issue 2: Make Border Price Method Effective as of February 2000

The arguments of SDG&E and ORA against the retroactive application of a revised methodology for pricing gas fail to persuade us. First, SDG&E's argument that adjustments to correct for the misallocation of costs will disrupt markets fails to consider that an adjustment need not disturb the prices previously paid by customers when making their consumption choices. In particular, balancing account mechanisms, used prospectively to adjust for historic over or under collections in utility accounts, provide rate stability at the time of a gas purchase. Therefore, this ratemaking treatment does not disturb prior customer consumption choices.

Second, SDG&E's argument that the tariffs were in compliance with the then-effective Commission decisions is not persuasive. SDG&E fails to cite, and our research cannot find, any Commission decision explicitly adopting a methodology that would require the averaging of all border gas prices, which include transport costs, with the basin-purchased gas prices. On the other hand, the Commission did adopt specific decisions, cited above, endorsing the principle that core and non-core customers should pay for the transport services that they use. Finally, SDG&E's claim that the OII acknowledges that its tariffs comply with Commission decisions is a misreading of I.00-08-003. We can find no statement in the OII supporting such a conclusion.

Similarly, ORA's arguments that a reallocation of costs would be unfair to non-core procurement customers fails to recognize the possibility of constructing a prospective cost allocation that does not disturb prior tariffs. Second, the arguments of ORA that a surcharge may cause procurement customers to reduce consumption or switch vendors carry little weight. If the pricing methodology had conformed to the principles embedded in Commission policy, non-core customers would have likely reduced consumption in response to higher gas costs, while core customers would have likely increased consumption in response to lower gas costs. Second, it is unclear that the avoidance of a switch in vendors, an apparent concern of ORA, serves a public purpose, and it is not a goal of law or regulatory policy. Indeed, Commission policy seeks to provide Californians access to gas reasonably priced, not to maintain SDG&E's market share.

LIF requests that the Commission apply the revised methodology starting in February 2000. (LIF Reply Brief, p. 3.) LIF correctly notes that our OII states that "The costs that may be reallocated include the costs covered by AL-1184G and subsequent advice letters as well as costs incurred from and after the date of today's decision." (OII, mimeo., p. 9.) Furthermore, the OII clearly ordered the creation of a tracking account starting with the date of the OII, and further stated in Ordering Paragraph 5 that "The costs recorded in the tracking account and all costs reflected in AL 1184-G and subsequent advice letters may be reallocated by the Commission between the core and the non-core to reflect the Commission's decision in this proceeding regarding the appropriate allocation of costs." (Id., p. 13.)

It is clear that the protests of the advice letters and the publication of the OII provided notice of the possibility that the Commission could reallocate costs from one customer class to another. Nothing could be more explicit (than Ordering Paragraphs 4 and 5 of the OII) in stating that the Commission's range of remedies would include the reallocation of costs dating from February of 2000. Therefore, we find no legal issues that would prevent a reallocation of gas transport costs dating to February 2000, when ORA filed its first protest.

Finally, a reallocation of gas transport costs constitutes the appropriate action to remedy the errors identified in this particular situation. No one disagrees with the proposition that the methodology used by SDG&E misallocated costs that led to an overbilling of core customers, starting as far back as 1986 and increasing in size as a result of a major change in the gas market in February 1998. (SDG&E Response, p. 1.) In any market, whether regulated or unregulated, customers expect the correction of billing errors once noticed and timely reported. ORA's protest of SDG&E's advice letters starting on

February 2000 served to notice and timely report the billing errors. Thus, an adjustment of costs starting from this date constitutes a reasonable action to protect consumers who expect that regulated prices reflect the costs of providing service. We therefore adopt this remedy, readjusting costs to February 2000, for the current situation.

Issue 3: How to Implement New Pricing Methodology for Allocating Transport Costs to Core and Non-Core Gas Users

SDG&E is a strong proponent of applying a new pricing methodology on a prospective basis. For this reason, it does not propose any strategy to implement the new pricing methodology with a starting date of February 2000. Nevertheless, SDG&E expresses opposition to a variety of methods for implementing a change in pricing methodology from February 2000. SDG&E expresses opposition to a "supplemental charge on the bill for each non-core procurement customer based on its historic consumption and to include a credit on the bill for each core customer based on the core customer's historic consumption." (SDG&E Opening Brief, p. 5.) SDG&E states that such an approach will be difficult to administer and expensive to implement and unfair to those who bought gas under the assumption that the gas prices fully reflect cost.

SDG&E also opposes the amortization of past under or over collections of costs in rates applicable to future purchases. SDG&E points out that non-core customers sign one-year contracts to purchase gas from SDG&E. SDG&E notes that those signing a contract would be committed to paying a surcharge over the entire remaining life of the contract. While SDG&E notes that it does not oppose a recalculation of prices based on a new methodology following adoption of this decision, SDG&E argues that both applying the new methodology and collecting past undercollections, even if amortized, will have a larger impact on customers, which it does not support.

SDG&E argues that a "retroactive reallocation over rates applicable to only future noncore procurement service" presents problems because under a Comprehensive Settlement now before the Commission, SDG&E would terminate its non-core procurement services and therefore sign up no new customers as of April 1, 2001. (SDG&E Opening Brief, p. 8.) SDG&E argues that this may provide only a truncated period for amortizing an undercollection.

ORA, despite its protest of all advice letters starting in February 2000, makes no proposal for implementing a change in methodology from this date.

LIF, in its Reply Brief, expresses support for "a supplemental charge that would be placed on each NONCORE procurement customer, while a credit would be allocated to CORE customer bills." (LIF Reply Brief, p. 4.) As an alternative, LIF suggests the creation of a "$1.7 million fund to assist CORE residential customers through bill payment plans, low-income ratepayer subsidies and other efforts designed to lessen the energy burden that SDG&E customers have experienced since summer 2000." (LIF Reply Brief, p. 4.)

In SDG&E's Comments on the draft decision, SDG&E renews its objections to ordering rebates that start with the first protested advice letter in February 2000. However, if the Commission should order rebates, SDG&E requests that it book the undercharges to non-core customers into the non-core PGA for subsequent recovery.

Resolving Issue 3: Implement Border Pricing Method With Next Gas Pricing Advice Letter; Rebate Overcollections and Surcredit Undercollections Over Next 12 Months

Since we will apply the revised methodology starting at February 2000 there are now two implementation issues before us for resolution. The first is when to order gas pricing to conform to the new pricing methodology. The second issue is how to amortize the over and undercollections that have accrued over the last year.

All parties agree that the new methodology should be implemented as soon as possible. Moreover, the sooner we implement the new methodology, the sooner gas prices will conform with our prior decisions. Thus, we will order the use of the new methodology starting with the first advice letter filing following the adoption of this decision.

Concerning the past over and undercollection, we estimate that we will need to shift approximately $1.9 million of transportation costs that accrued over the last year to the non-core customers and to rebate that same amount to core customers. We note that despite the early agreement of ORA and SDG&E on how to bring the pricing methodology into conformance with our prior decision, our OII ordered SDG&E to continue using the current flawed methodology until directed further. (OII, Ordering Paragraph 7, p. 13.) For this reason, the amount subject to reallocation has grown and continues to grow.

Concerning the core customers, we note that the overcollection accrued over the last 12 months. It is reasonable to amortize this overcollection via a surcredit in cents per MBTU to core customers for the next 12 months. We believe that this surcredit mechanism, supported by LIF, offers an economical way to return the overcollection to customers. Moreover, since a year's consumption of gas follows a typically set pattern, applying the surcredit for a year will approximately distribute these revenues to those who paid them.

Although SDG&E notes that as many as 40% of customers move in a year, this is not relevant. Moreover, there is no evidence on what percentage move outside SDG&E's service territory. Thus, a surcredit remains a reasonable way to adjust customer bills.

SDG&E objects to imposing a surcharge on non-core customers, arguing that applying both the new pricing methodology and recovering the past undercollection may impose a large cost on its customers. Although we are empathetic with the plight of these customers, we believe that it is appropriate to permit SDG&E to recover the costs that we order it to rebate to core customers. Thus, we find that it is neither unreasonable nor unfair to permit SDG&E to recover the misallocated transport costs from its non-core customers. SDG&E's proposal to book the undercollected transportation charges into the non-core PGA offers an appropriate mechanism for allowing it to recover its costs. Moreover, this pre-existing method for recovering under and over collections of gas costs offers a simple way to recover the undercollections. We will therefore permit SDG&E to recover this undercollection by booking this amount to the non-core PGA. Should SDG&E elect to forgo collection of these revenues for market reasons, it is free to do so, but such a decision precludes recovery at another time.

SDG&E also points out that a settlement now before the Commission for adoption may lead to a termination of its non-core gas business. (I.99-07-033.) At this point, it is unclear whether and when the Commission will adopt this settlement, and we do not believe that this potential action should dissuade us form permitting SDG&E to recover gas transport costs via a

surcharge. If, however, the Commission should adopt this settlement, SDG&E may propos via an advice letter an alternative method for recovering these gas transport costs.

Finally, we note that LIF endorses implementation of a change in the pricing methodology combined with an amortization of prior over-and-undercollections via a surcharge mechanism. We adopt this proposed method for rebating the overcollections, and we have nothing further to add to our previous endorsement of this plan as a reasonable strategy to implement changes in the pricing methodology. We note, however, that we permit recovery of undercollections via booking these costs to SDG&E's non-core PGA.

In summary, we order a change in the methodology for pricing gas effective with the next monthly advice letter. Second, we order SDG&E to file advice letters that will amortize the overcollection from core customers for the gas transportation costs that have accrued since February 2000 via a surcredit. Third, we permit SDG&E to file advice letters to recover from non-core customers the transport charges reimbursed to core customers via booking these changes to SDG&E's non-core PGA. The charges booked to the non-core PGA are limited to the amount needed to recover the transport costs rebated to core customers for the period commencing with February 2000.