IV. Analysis

We adopt a cost-based peaking tariff that adopts the basic components advanced by SoCalGas, but as amended to address concerns raised by ORA and others. We find that cost-based pricing mechanism proposed by SoCalGas, with the modifications suggested by ORA, is consistent with our policy of promoting economic bypass. ORA suggests basing the peaking rate using the non-coincident noncore peak day, rather than the coincident system peak-day. Partial bypass customers will pay tariffed rates for service at a rate comparable to current end-use customers. However, those customers will pay for the demand that they place on the system based on the highest monthly usage on a 12-month moving basis. This allows fair compensation to the utility toward system costs to the benefit of all ratepayers, but does not require that bypass customers reserve system capacity based on a peak-day requirement as SoCalGas suggests.

We are concerned that SoCalGas' market-based proposal produces perverse results, as we explain below. We are not convinced that either SoCalGas' market-based proposal or Watson's hybrid reflect market realities. There is no market, as such, for peaking rates. There is no other pipeline offering a peaking rate in competition with SoCalGas. SoCalGas' formula discourages large volume bypass, and Watson's proposal discounts very heavily for the large users and could impact captive customers unfairly.

Furthermore, SoCalGas's revenue cap produces certain perverse results. We test the SoCalGas revenue cap with a hypothetical example. We assume for the purpose of the hypothetical that the default tariff is 4 cents a therm and the customer's total annual volume is 100,000 therms or 100 Mtherms. Table 1 shows the results of the hypothetical under varying bypass volume, LRMC, and bypass rate. It is clear that SoCalGas' revenue cap formula produces perverse results because, as the utility' LRMC and bypass rate go down, the revenue cap stays the same, while the total cost for the customer goes up. SoCalGas' revenue cap has little to do with marginal cost or economic bypass and much to do with keeping the customer cost the same even though the competitive pipeline offers a lower rate and the utility marginal cost is lower. Furthermore, the revenue cap goes up as the customer's bypass volume goes down, hardly a result that SoCalGas would desire. It is evident therefore that SoCalGas' market based rate would provide customers with an incentive to bypass the SoCalGas system altogether and could, in certain situations, prove to be more punitive than the RLS tariff.

We, therefore, reject the SoCalGas proposal for a market-based peaking rate:

Table 1

SoCalGas Peaking Rate Revenue Cap

Total Vol. Therms	Bypass Vol. therms	Default Rate ¢/therm	LRMC ¢/therm	Bypass Rate ¢/ therm	Rev. Cap $	Rev. Cap ¢/therm	Total Cust Cost $
100,000	80,000	4	3	2	1600	8.00	3200
100,000	60,000	4	3	2	2200	5.50	3400
100,000	40,000	4	3	2	2800	4.67	3600
100,000	30,000	4	3	2	3100	4.43	3700
100,000	20,000	4	3	2	3400	4.25	3800

100,000	80,000	4	2	1	2400	12.00	3200
100,000	60,000	4	2	1	2800	7.00	3400
100,000	40,000	4	2	1	3200	5.33	3600
100,000	30,000	4	2	1	3400	4.86	3700
100,000	20,000	4	2	1	3600	4.50	3800

We also find a number of points in Watson's proposal unacceptable. To begin, Watson's arguments in support of more balancing options, are not reasonable. SoCalGas suggests balancing of nominations and burns to +/-5% on a daily basis and +/-1% on a monthly basis. Daily balancing requires the customers to manage their own gas supplies in a manner that does not adversely affect other customers on the system. All the interstate pipelines serving SoCalGas' market have daily balancing requirements, and some even have tighter provisions. ORA agrees with the daily balancing requirement advanced by SoCalGas.

On the other hand, Watson's proposal to allow for a full range of balancing choices would disproportionately benefit customers with erratic load profiles that do not align their deliveries with their consumption pattern over customers that make an effort to match their burns and deliveries. Monthly balancing would allow bypass customers to avoid imbalance penalties on the interstate pipeline and realize price arbitrage opportunities not available on the interstate pipeline. Under current natural gas market conditions, where the price of gas is very high, more relaxed balancing provisions might encourage the peaking customers to use SoCalGas' balancing as a price arbitrage tool which would impose additional burdens on captive customers.

Watson's proposed peaking rate tariff is also untenable because it advocates that local transmission charges would be collected through a quasi-demand charge that would only be paid in months that a customer takes peaking service. The monthly demand charge would be calculated by using the peaking customer's usage on the most recent coincident system peak day--which would advantage summer peaking customers such as electric generators. The end result is peaking customers' rate would approximate the tariff rate for full requirements customers.

A. Cost-Based Peaking Tariff Rate

Cost-based rates are grounded in good economic theory, rates established by the Commission are generally cost-based, ⁸ and there is sufficient precedent and history to support this rate mechanism. The cost-based rate we adopt for SoCalGas' peaking tariff will include the following components:

1. Customer Charge

The customer charge is designed to collect the total cost of the customer-related facilities through a monthly demand charge. SoCalGas uses the annualized cost of customer-related facilities adopted in the 1999 BCAP as a proxy for an assessment of the meter and associated facilities. SoCalGas then adjusts the revenue associated with the customer-related facilities by the Long Run Marginal Cost (LRMC) scaler to approximate the total cost of these facilities.

The monthly customer charge in SoCalGas' otherwise applicable tariff, GT-F for retail noncore customers, does not recover the full cost of the customer-related facilities. For GT-F customers, a portion of the customer-related costs is collected through the customer's volumetric transportation rate. For full-requirements customers, the utility has a reasonable expectation of recovering the customer-related costs. However, the utility cannot know the extent to which a bypass customer will take service from the utility. Therefore, SoCalGas believes, it is reasonable to permit SoCalGas to collect the full cost of the customer-related facilities as part of a monthly demand charge in the peaking rate tariff.

SoCalGas also points out that not all customers currently have a monthly customer charge as a part of their tariff. In particular, large electric generators and wholesale customers do not have a monthly customer charge, SoCalGas notes. It is therefore more equitable to adopt a consistent approach of using the annualized customer charge for all peaking service customers. SoCalGas proposes a monthly customer charge of $800 to $19,000 depending on the customer class. ORA agrees that SoCalGas should collect the full cost of customer-related facilities as part of a monthly demand charge. We will adopt the ORA proposal that the customer demand component of the rate should be equivalent to the currently authorized end-use customer rate for the specific customer class (e.g., electric generation, industrial, etc.) The customer demand component of the rate should be computed monthly based on the higher of either the current monthly usage or the highest monthly usage over the prior 12-month period.

We adopt ORA's customer charge proposal.

2. PPP Charge

SoCalGas proposes to collect the PPP charge from the partial bypass customer based on the customer's total natural gas consumption at the facility.

Since SoCalGas submitted its testimony, Assembly Bill (AB) 1002 became law. The CPUC implemented AB 1002 in December 2000. Under the provisions of AB 1002, customers of interstate pipelines are mandated to pay a volumetric public purpose program surcharge to the Board of Equalization. The customers of public utility companies are required to pay the surcharge as a separate line item on their bills effective July 1, 2001. Prior to July 1, 2001, the customers will continue to pay the costs of public purpose programs included in their volumetric transportation rates. We, therefore, find SoCalGas' proposal to charge its peaking rate customers a public purpose program surcharge based on their total usage including volumes delivered on interstate pipelines moot.

On the bills of its peaking tariff customers, SoCalGas shall show a separate line item public purpose program surcharge to be collected volumetrically in compliance with AB 1002. The surcharge will be based on the public purpose program surcharge rates adopted by the Commission in Resolution G-3303 and will only apply to the customer's volumes served by SoCalGas. The customer will pay the public purpose program on the volumes served by an interstate pipeline to the Board of Equalization.

3. Reservation Charge for Transportation

The coincident peak demand is the peak demand of the customer class at the time of the system peak. The noncoincident peak demand is the peak demand of the customer class during the year.⁹

The philosophy behind a peaking tariff is that it should collect the demand imposed by the customers at the time of the system peak. For reliability purposes, the gas utility designs its transmission and distribution system to meet the demands placed upon it on an abnormal peak day. In designing customer class rates, the utility's costs are allocated to various customer classes based upon marginal cost allocators based on coincident peak demand. A peaking rate should impose upon the customer who uses the service the extra burden for using the system for peak demand. Although the customer might use the peaking service to meet its own noncoincident demand, the utility is only concerned about whether it has to build additional capacity to meet the additional demands placed upon it in order to meet the needs of the customers served under the peaking tariff. To design a tariff based upon the customer class noncoincident demand makes little sense, since the customer class peak demand may not impose any additional demands on the system if it occurs at a time when the system demand is not at its peak.

Whether such a tariff might result in prohibitively expensive rates is a question that is difficult to answer at this time. Clearly, if the coincident peak were to fall in the future, such a result is possible. We are unable to compare and contrast the results of the ORA proposed methodology with those of SoCalGas because ORA presented no quantitative analysis of its proposal. The aim of the peaking tariff should be to assess a charge that results in a significant premium over the systemwide default in order to discourage noncore customers from exploiting the utility system at peak times while using interstate pipelines for base load requirements. If we were to divide the revenue requirement by the average annualized load of the noncore customer class, the premium would be zero. It is clear therefore that the denominator in the demand charge calculation should be a figure somewhat lower than the average demand of the noncore class. How much lower the figure should be really depends on the future demand profiles of the electric generators and other noncore customers. If we were to adopt the ORA proposal and use the noncoincident peak demand of the noncore class in the denominator, it is likely that we will arrive at a rate that is lower than the default rate based on the average demand. We believe that this is a highly probably scenario, considering that the electric generation demand patterns have changed in the last couple of years, resulting in a spike that is so high that if multiplied by 365 days, the figure could turn out to be higher than the area under the load curve.

We, therefore, adopt the SoCalGas method of using the noncore coincident peak demand for the calculation of the demand charge. Given the current volatile nature of the electric generation market, we will monitor carefully the effectiveness of the cost based peaking rate we adopt today. We take note of the fact that since the record was submitted in this proceeding, the electric generation market has undergone significant upheaval. If necessary, we will revise the charge in SoCalGas' next ratemaking proceeding.

4. ITCS Charge

SoCalGas proposes to collect all other non-fuel related costs in the monthly reservation charge. These other costs include transition cost accounts, such as the ITCS.

The ITCS is currently collected from all customers as a separate volumetric rate, to be applied to the actual, recorded, monthly throughput. We will continue this approach, as ORA recommends.

We agree with ORA. Currently, the ITCS is collected from all customers as a volumetric charge, and to make it a part of the demand charge would go against precedent.

5. Customer Class Versus Systemwide Peaking
Rate

ORA points out that SoCalGas proposes a systemwide peaking rate instead of a customer class specific peaking rate. In the absence of a record regarding the load profiles of various customer classes, we find the SoCalGas proposal adequate at this time.

6. Daily Balancing

SoCalGas proposes that peaking service customers would be expected to balance their nominations and burns to +/- 5% on a daily basis and +/- 1% on a monthly basis. The customer would also be expected to maintain uniform hourly deliveries and usage to the extent practical. If the customer anticipates significant variations in its deliveries or usage during the day, SoCalGas will attempt to accommodate the customer's expected load profile. In such cases, the customer and SoCalGas will establish a protocol that provides sufficient notification for the utility to meet the customer's load profile. As we previously discussed, we prefer daily balancing, because it requires customers to manage their own gas supplies in a manner that does not adversely impact other customers. Therefore, we will adopt the proposal put forth by SoCalGas.

7. Service Interruption Credit

SoCalGas currently offers its GT-F customers a Service Interruption Credit (SIC) as part of Rule 23. SoCalGas proposes to exempt partial bypass customers from the SIC provision. ORA agrees that partial bypass customers should not be afforded a SIC. We agree. It is self-evident that the utility would lose large amounts of money in the form of rewards for those customers who partially bypass its system.

8. General Tariff Provisions

SoCalGas uses an average system-wide rate for all noncore customers. SoCalGas proposes that the tariff be updated effective January 1 each year to reflect adjustments to PBR base margin and updates to the noncore balancing accounts.

Because we have made certain modifications to the SoCalGas proposal such as the inclusion of the ITCS as a separate volumetric charge, we will order SoCalGas to file an advice letter within 10 days of the issuance of this decision in compliance with the modified cost-based tariff we adopt today.

9. Applicability Provisions

Almost all parties except SoCalGas and TURN support the implementation of a peaking tariff on a facility-by-facility basis rather than imposing it on a customer basis, thereby subjecting total loads of generators with multiple facilities to the tariff. As ORA points out, there is no rationale for such an application. Under the SoCalGas proposal, if an existing generator develops a new power plant and decides to take service for that power plant from a competing interstate pipeline, then the generator will not be able to take peaking service for that plant from SoCalGas without subjecting all of the plants it owns to the peaking tariff. This provision is unreasonable and could prove to be so onerous that it might in fact promote bypass of SoCalGas' system by electric generators. Certainly, it may discourage development of new generation facilities.

The one argument that could be made in favor of a peaking tariff application to multiple generation facilities would be so the generator could shift demand from one facility to another to avoid peaking use all together. However, it is unlikely that such a shift could be made so that there would be no need for the generator to take peaking service. After all, generation in California peaks during the summer season. With the new electric industry structure, generators have an incentive to generate when wholesale electricity prices are high. Such behavior on the part of the generators has already shifted electric generation load profiles such that their shapes have more spikes than historical load profiles. We, therefore, conclude that it is not reasonable to allow a multiple generator peaking rate.

Watson and other parties suggest that new generation facilities should be exempt from a peaking tariff. We reject this proposal as discriminatory. There is no precedent for this Commission to set different rates for new customers who might move into a utility's territory even if they impose additional costs on the utility's system. Similarly, we cannot allow a customer to be exempt from a tariff just because the customer is new.