5. Discussion

5.1. Ongoing Voltage Reduction Measures

We endorse PG&E's program to (a) operate distribution circuits at the lowest peak load voltage levels consistent with Rule 2, (b) modify voltage regulating devices to ensure that these devices maintain voltage within Rule 2, and (c) continue to pursue such opportunities during the normal course of substation related work. PG&E reports that the cost of this program is recovered within PG&E's current revenue requirement and rates. No party opposes PG&E's program. We find that PG&E's efforts are consistent with good utility practice, Commission expectations, and Commission orders.

No evidence shows any utility operating consistently or purposefully outside Rule 2 limits. Rather, the testimony shows that each utility strives to operate within the limits of Rule 2, and makes necessary corrections when brought to its attention. Nonetheless, we encourage each utility to continue all reasonable efforts to operate distribution systems at the lowest peak load voltage levels consistent with Commission orders and Rule 2. Further, each utility should explore all reasonable opportunities for additional voltage reduction within Rule 2 limits in the normal course of operations. This includes implementing all cost-effective measures to reduce voltage during normal substation related work, not only to reduce peak demand, but also to save energy. These efforts may be especially relevant for SCE and SDG&E recognizing, however, that the type of customer load served, circuit loading, or circuit configuration may limit those opportunities. While capacity reductions from this effort will be modest, making adjustments within Rule 2 ensures that all reasonable savings are achieved without jeopardizing service, safety and reliability.

5.2. Emergency Measures

We decline to adopt the specific proposals of either CEC or TURN to use manual or automatic voltage reduction capabilities during system emergencies. All other parties oppose these proposals, including PG&E, SCE, SDG&E, Farm Bureau, the Office of Ratepayer Advocates (ORA), the California Manufacturers and Technology Association (CMTA), plus the University of California and the California State University (UC/CSU).¹⁰ Opposing parties persuade us that the need for emergency voltage reduction has subsided, the potential benefits do not justify the risks, and that other reasonable alternatives are preferable.

5.2.1. Need Has Subsided

As PG&E points out, the Governor's request to study voltage reduction was initiated under far worse conditions than exist today. At that time, California was in the midst of a crisis, we were on the precipice of Summer 2001, and a significant number of rotating outages were predicted. Conditions are now materially different. For example, PG&E interconnected 10 generating plants totaling 1,301 MW during 2001. In addition, for the period of October 1, 2001 through December 31, 2001, PG&E has applications from 41 projects totaling 7,503 MW.

Demand has also declined. PG&E experienced a significant reduction in Summer 2001 peak load not only from relatively cool weather, but also from conservation and the economic slowdown. Some conservation is permanent (e.g., as a result of the replacement of some equipment with energy efficient appliances), and economic recovery will take time. PG&E does not expect loads in 2002 to exceed peak loads in 2000, even if Summer 2002 temperatures return to normal.

CEC cautions that California's energy supply situation will be in a precarious condition for some time to come if we do not make every effort to conserve and prepare. Any number of events, according to CEC, could again threaten blackouts, including loss of either a power plant or transmission line. No party, however, alleges that we remain in the dire situation we faced last July. Nonetheless, we consider emergency voltage reduction measures as further insurance against rotating outages.

5.2.2. Benefits Are Limited

The benefits of an emergency voltage reduction program are limited to the remote voltage reduction capability of SCE during its on-peak period. That is, absent remote adjustment capabilities, individual site visits to substations and feeders are required to adjust voltage by manually altering voltage regulator controls. The manual nature of this effort makes it extremely difficult to implement temporary changes only during limited hours of daily peak demand or emergencies. No evidence shows that this would be either cost-effective or practical. Only CEC recommends that rules be modified to permit each utility to utilize manual capabilities. We decline to adopt this recommendation.

Remarkably, the majority of rotating outages experienced by California in 2001 have been in periods other than summer on-peak. Nonetheless, voltage reduction may reduce peak demand. Benefits, however, are limited to no more than 160 MWs during SCE's summer on-peak.

5.2.3. Risks are Not Quantified but Potentially Great

While the potential benefits are somewhat predictable, SCE and others correctly point out that the risks are not. For example, it is very difficult to predict the actual effect of voltage reduction on an appliance, since utilization voltage (i.e., voltage at the appliance) is a function of service voltage (voltage provided by the utility at the customer meter), the internal wiring of a building, the load on individual circuits within a building, and the condition of all appliances on the internal circuits. What is known, however, is that the risk of equipment failure and system disruption increases as voltage decreases.

In particular, voltage reduction (from voltage otherwise planned during system peak) increases the level of current (over that originally forecast) for constant power loads, which in turn increases the risk of equipment overload and failure. Erratic operation of equipment, or equipment failure, can result in injury, property damage, and business disruption. Under existing conditions, PG&E states that approximately 80% of its system is already expected to operate at its current-carrying capacity during system peak. This makes the potential for problems from increased current a significant concern for the utility, as well as its customers.

Moreover, reduced voltage has the effect of increasing the exposure of sensitive customer equipment to voltage sags and nuisance-tripping. This can be particularly problematic for sensitive and expensive laboratory and hospital equipment, as well as tools and computers.¹¹ SDG&E reports increased customer complaints when service falls below 114 volts at the customer meter. In particular, SDG&E testifies that electronic devices (including computers, motor control equipment, and manufacturing process controllers) are sensitive to voltage, and may fail under substandard voltage conditions.

Existing low voltage situations may also be made worse. That is, utilities undertake reasonable efforts to maintain service voltage within Rule 2 limits, and no evidence is offered to the contrary. Distribution systems may not be perfect, however, and some customers may receive voltage below the Rule 2 level at some times. Low voltage problems are generally not identified or remedied until a customer reports a problem, since utilities have no independent means of detecting voltages at every customer meter. A low voltage condition for a particular customer may not be serious until voltage is reduced further by an emergency voltage reduction event. In this case, the emergency voltage reduction may result in erratic operation, equipment failure, injury, property damage, and business disruption.

Reduced voltage and increased current also limits the flexibility of system operators. As a result, their ability to perform outage restoration, and transfer load between feeders when outages occur, is compromised.

SCE states that it would initiate voltage reduction at distribution substations, and voltage reductions would affect entire service areas. SCE does not have the capability to remotely reduce voltage for only selected circuits or customers. (Exhibit 105, page 2.) As a result, voltage reduction would affect essential as well as non-essential customers. ¹² At least hypothetically, a voltage reduction plan might be devised that excludes essential customers. Such plan, however, may involve additional costs (e.g., to enable isolating voltage reduction to limited circuits or customers). Absent better and more complete information presented by a party advocating voltage reduction, we must carefully and cautiously weigh the benefits and risks.

Thus, as opposing parties state, emergency voltage reduction can cause injury, damage and disruption. This includes increasing the risk and incidence of injury, equipment damage, and equipment maintenance costs; shortening equipment life; causing equipment outages; disrupting or halting business operations; and reducing system operator flexibility. Voltage reduction also places essential customers at risk, thereby potentially jeopardizing public health, safety and welfare. UC/CSU states that no party knows whether, or the degree to which, one or more of these impacts will occur, and that the Commission should not proceed without more information about the effects. We agree.

5.2.3.1. Rule 2 and ANSI C84.1

We also consider the specific resulting minimum voltages, and the duration of voltage reduction events, in assessing whether or not to adopt the recommendations of CEC and TURN. Rule 2 of each utility, and the American National Standards Institute (ANSI) Standard C84.1¹³, are relevant in making that assessment.

Rule 2 addresses both service voltage (at the meter) and utilization voltage (at the appliance). In turn, Rule 2 requires that all customer-owned utilization equipment be designed and rated in accordance with voltages specified in ANSI C84.1, if the equipment is to give fully satisfactory performance. (Rule 2, Section C.2.a.)

The minimum and maximum ANSI C84.1 voltages, based on 120-volt systems, are specified in two ranges:

Line No	Range	Service Voltage	Utilization Voltage
1	A	114-126	110-125
2	B	110-127	106-127

Source: ANSI C84.1-1995 (R-2001),

Table 1 and Annex B

Minimum Range A service and utilization voltages are 114 and 110, respectively. Minimum Range B service and utilization voltages are 110 and 106, respectively.

Range A represents voltages for which electrical systems are designed and operated. Range A is the expected normal condition, and the "occurrence of service voltages outside of these limits [Range A] should be infrequent." (ANSI C84.1-1995 (R-2001), Section 2.4.1.)

Range B voltages "shall be limited in extent, frequency, and duration." (Id., Section 2.4.3.) When Range B occurs, "corrective measures shall be undertaken within a reasonable time to improve voltages to meet Range A requirements." (Id.)

CEC's proposal to reduce voltage to no less than 117 volts at the distribution substation means that most customers should experience service voltage (at the meter) of no less than 110 volts. The resulting utilization voltage for most customers would be no less than 106 volts, assuming a drop of 4 volts within the customer's wiring. (Id., Annex B, Note C; also Rule 2, Section 2.b). This is consistent with ANSI C84.1 Range B. Thus, it "makes sense to consider temporary operation in Range B when rotating blackouts are imminent."¹⁴

Nonetheless, there are important risk factors that must be weighed before adopting CEC's proposal. First, if emergency voltage reduction is authorized as recommended, the event might last as long as the entire SCE on-peak period, or up to 6 hours each summer weekday.¹⁵ Parties dispute the duration of "temporary," and whether or not ANSI C84.1 intends a Range B event to last up to 6 hours each summer weekday. ANSI C84.1 is not clear, but does state in part that operating outside Range A "should be infrequent," and Range B events "shall be limited in extent, frequency, and duration." We decline to find a program that authorizes Range B up to 504 hours per year¹⁶ to be consistent with events that are "temporary," "infrequent," and "limited in extent, frequency and duration." We agree with ANSI C84.1 Committee Member Conrad that temporary does not mean operating distribution systems in Range B through an entire season.¹⁷

Second, even if successful in maintaining utilization voltages no lower than 106 volts for limited periods, operation in Range B for extended times "nearly assures some customers will see voltages below Range B."¹⁸ Thus, voltages below Range B are nearly certain, with concurrent risk of injury, damage and disruption, if we authorize emergency voltage reduction, and it is used for long periods, including up to as often as allowed.

Third, there is an important difference between the National Electrical Code (NEC) and ANSI C.84.1 with regard to voltage drop. The NEC encourages building designers to limit voltage drop to 5% (6 volts on a 120 volt system) compared to 4 volts in ANSI C84.1. This creates a bias for problems because voltages in buildings are allowed to be 2 volts lower than permitted in ANSI C84.1. "This is one more argument for caution when lowering overall system voltages even below Range A."¹⁹

Fourth, there is no evidence that voltages on branches and ends of circuits that are properly balanced for Range A will necessarily be uniformly reduced during an emergency voltage reduction. Rather, some voltages may be below Range B during an emergency voltage reduction event, even when voltages started inside Range A, because systems are not designed for this level of operation, particularly over long periods. Thus, the emergency voltage reduction proposal places some customers at substantial risk of utilization voltage less than 106 volts. This includes not only non-essential customers, but also essential customers normally excluded from rotating outages.

Finally, the Chairman of the Accredited Standards Committee C84 states that service voltage of 114 volts is the minimum voltage to ensure acceptable performance of equipment.²⁰ That is, the Chairman of the responsible committee warns of unacceptable performance when service voltage is below 114 volts. Unacceptable performance creates risk of injury, damage and disruption.

5.2.3.2. Test Results

The CEC witness testified that tests performed under his supervision confirm that all appliances and commercial motors operate safely and efficiently at utilization voltages as low as 110 volts. (Exhibit 102, page 2; RT Vol. 7, page 423.) This is the low end of Range A, and does not itself support use of Range B during the recommended emergency voltage reduction program.

The CEC does not base its recommendation on testing of equipment below 110 volts at the appliance. (RT Vol. 7, page 427.) Rather, CEC's recommendation is based on ANSI C84.1 Range B, wherein utilization voltage may be reduced to 106 volts. The tests performed by the CEC's witness, however, included operating equipment at much lower voltages for the purpose of satisfying those concerned that voltage reduction will be harmful and damaging to equipment. (RT Vol. 7, page 427.) No damage to equipment was observed at voltages down to 100 volts during the tests, according to CEC's witness. (RT Vol. 7, page 424.)

These test results, however, have not been published, have not received formal peer review,²¹ and were not introduced as evidence. We are not persuaded that the test results are representative of all equipment that might be subject to emergency voltage reduction, or actual conditions faced by individual customers.

Test results were available to representatives of PG&E, SCE, and SDG&E, as well as one member of the ANSI Committee responsible for ANSI C84.1.²² Witnesses for PG&E, SCE and SDG&E all oppose CEC's recommendation, despite having access to the data. The representative of the ANSI Committee responsible for ANSI C84.1 made several cautionary statements, despite having access to the data.²³ Thus, this limited informal peer review does not demonstrate support for the conclusions reached by CEC.

The CEC witness also reported consulting with the manufacturer of all tested equipment to confirm that their equipment will operate as low as 105 volts. (RT Vol. 7, pages 424-5.) We are not convinced that this adequately supports the CEC's recommendation, since the actual range of utilization voltage may go below 105 volts during an emergency voltage reduction event.

The CEC witness concludes that all equipment will operate without damage or failure at 106 volts, or even somewhat lower. The only equipment harmed by voltage reduction, according to the CEC witness, will be (1) equipment that is improperly installed and without proper safeguards, or (2) equipment that was ready to fail anyway. (RT Vol. 7, page 446.)

We are not persuaded it is reasonable to define away the problem by concluding that only causes other than voltage reduction can be responsible for damage to equipment. Equipment damage and failure is accelerated with voltage reduction. Equipment installed improperly and without safeguards that does not fail under normal voltage conditions (Range A), but which fails as a result of emergency voltage reduction, will expose the customer to potential injury, property damage or business disruption. Equipment that was going to fail anyway, but whose demise is accelerated and results from an emergency voltage reduction, will expose the customer to potential injury, property damage or business disruption.

We agree with ORA that we do not live in a world wherein all equipment at all times meets current code (even if it met required code at the time it was installed), is properly installed, and has fully functional safeguards and safety devices. We decline to expose ratepayers to unknown but potentially large risks by concluding that all equipment is properly installed to meet current code requirements, and with proper safeguards or safety devices that are fully operational at all times. Moreover, even where a motor has an automatic sensor to shield the motor from harm during a low voltage event, the protection may be to disconnect the motor from the circuit, or otherwise terminate motor operation. The motor might be saved, but there may be large consequences (e.g., lost product, production time, lab results).

There may be many causes and reasons for injury, damage and disruption due to equipment failure other than from voltage reduction. We will not simply assume away any relationship to voltage reduction, however, given the clear, unambiguous and persuasive testimony from the majority of witnesses on this subject.

5.2.3.3. Stable Voltage

At least in theory, one approach to mitigating the risk of emergency voltage reduction is to require customers to protect themselves at their own expense. Customers may do this by installing their own special or auxiliary equipment (e.g., capacitors) to increase utilization voltage from 106 to 110 volts during an emergency voltage reduction event.²⁴ No party, however, offers any estimates of the scope or cost of such self-protection.

The relative reliability of California's electricity system has for decades allowed all, or nearly all, customers to rely on stable voltage within ANSI C.84.1 Range A. Expecting or requiring customer self-protection from an emergency voltage reduction program shifts an unknown amount of costs to an unknown number of customers. We decline to adopt this shift of burden absent information on the effect.

5.2.3.4. Liability

An essential element of CEC and TURN recommendations is waiver of utility liability for injury or damage resulting from emergency voltage reduction. Waiver of utility liability transfers the risk of voltage reduction from the utility to its customers. Specifically, it transfers resulting harm to the customers who experience injury, damage or disruption. We decline to adopt this approach for several reasons.

First, the shift of burden is inequitable. That is, affected customers are forced to bear an unreasonably disproportionate responsibility for system-wide load reductions. Second, the magnitude of the liability is unknown. We are not persuaded to adopt this recommendation without specific information about the resulting exposure. Third, if those customers are essential customers, it harms not only the customer, but also the broader public health, safety and welfare.

As an alternative to waiver of utility liability, Farm Bureau proposes that the Commission establish a fund or bond to cover losses. We have no estimate of the size of the fund or bond that would be necessary, and decline to adopt this approach without reasonable boundaries.

TURN asserts that a utility is generally held harmless from damages due to blackouts caused by events beyond its control, and, assuming the utility acts responsibly, should similarly not be liable for damages that result from a program consciously undertaken to avoid blackouts. According to TURN, this recommendation is premised on the assumption that there is a net societal benefit to "keeping the lights on." Other than opinion testimony either in favor of, or opposition to, voltage reduction, no evidence is presented that seeks to quantify the net benefit or tradeoff between voltage reduction and rotating outages. We are not persuaded that the net benefit of "keeping the lights on" by voltage reduction is greater than the net benefit of rotating outages (as discussed more in the section below).²⁵

TURN supports compensation to ratepayers for any damages they suffer by voltage reduction through use of Federal Energy Regulatory Commission (FERC) Account 925 (Injuries and Damages). In response, SCE points out that there may be an accounting mechanism to record the cost of damages, but the current rate structure and revenue requirement are not based on either implementation of a voltage reduction program (and the increased damages that might result), or the exposure having been shifted to ratepayers.

Simply recording costs in FERC Account 925 does not result in cost recovery. We are not inclined to adjust rates to permit this recovery, or permit recovery of an insurance premium for this coverage, given that we have no information on the amount of any adjustment. We similarly decline to authorize a memorandum account for future recovery without any information on the exposure.

Several parties point out that the concern regarding waiver of liability, and the limited information available, is evidence in-and-of itself that a voltage reduction program is not ready for implementation. We agree.

5.2.3.5. Risk of Voltage Reduction Compared
to Rotating Outage

While we decline to transfer the risk of voltage reduction from the utility to its customers, we recognize that absent an emergency voltage reduction program customers are exposed to an incremental increase in the potential of rotating outages. Rotating outages involve risks to customers, as well as the broader public health, safety and welfare. In balancing the risks of emergency voltage reduction compared to the risks of exposure to an incremental increase in rotating outages, we are persuaded by the majority of parties that rotating outages are preferable.

That is, not only the utilities, but also ORA, CMTA, Farm Bureau and UC/CSU argue against voltage reduction. As these parties contend, emergency voltage reduction involves unknown effects over potentially long durations without an existing or proposed warning system. Rotating outages, however, are announced before they occur, are of limited duration (e.g., 60 to 90 minutes), and customers have several options available to mitigate their exposure or damage from rotating outages.²⁶ Moreover, emergency voltage reduction, as proposed, exposes essential as well as non-essential customers to unacceptable risk, while, in contrast, essential customers are normally exempt from rotating outages. Every party in opposition to voltage reduction has weighed these alternatives, and elects rotating outages over voltage reduction. We think this is reasonable.

TURN argues that denial of an emergency voltage reduction program on the SCE system would be unfair to SCE ratepayers. TURN says such denial would require SCE ratepayers to suffer rotating outages even though these ratepayers have paid for a system that would allow voltage reduction, thereby permitting more customers to remain on line.

To the contrary, no evidence shows that SCE's capability to remotely change tap settings on transformer banks was implemented for the purpose of emergency voltage reduction. It is not unfair to SCE ratepayers that this capability not be used for an otherwise unintended purpose. Whether or not this capacity should be used for the additional purpose of emergency voltage reduction depends upon the benefits and costs. For all the reasons stated above, we find that potential benefits do not outweigh the costs, including risk.

TURN asserts that the concern about possible lost product, production time and other effects from voltage reduction must be weighed against the certain consequences from rotating outages. We agree, and conclude that the risk of injury, damage and disruption from voltage reduction exceeds those same risks from rotating outage.

5.2.4. Other Reasonable Alternatives Are Preferable

Investments in several alternatives promise more secure benefits with greater potential capacity savings and fewer risks. These include investments in new generation, energy efficient equipment, conservation programs, and expanded interruptible rate tariffs. CMTA asserts that additional participation in demand responsive programs and similar measures can provide far greater savings than an emergency voltage reduction plan. UC/CSU contend that energy efficiency and demand reduction programs allow customers to manage their loads in accordance with their understanding of usage and needs, and are vastly preferable to a top-down energy management approach that includes substantial unknowns. We agree.

TURN claims that demand reduction alternatives are more costly than voltage reduction. TURN states, for example, that interruptible programs cost $84/kW-year or more as an incentive for a few customers to interrupt load, while voltage reduction is much less costly. TURN says it understands why large customers want other ratepayers to spend ratepayer money on the large customers to interrupt load, but that this is not good public policy. Finally, TURN argues that demand reduction programs are highly uncertain, as illustrated by the performance of SCE's interruptible customers over the past 18 months.

To the contrary, the cost of voltage reduction may be low to ratepayers as a whole, but can be very high to those who suffer injury, damage or disruption. The cost may be very high to public health, safety and welfare if the affected customer is an essential customer. Moreover, we think the reliability of SCE's interruptible program is increased after the recent opportunity for those SCE customers who were not well suited for the program to opt-out. On balance, we think it is better to have demand reduction programs at known costs along with controlled interruptions as needed rather than the unknown costs, risks and effects of voltage reduction.

5.2.5. Decline Other Variations of Emergency Voltage
Reduction

TURN recommends that we order SCE to implement emergency voltage reduction not only during on-peak Stage 3, but also on-peak Stage 2. For all the reasons stated above, we decline to adopt CEC's recommendation to authorize an emergency voltage reduction program during power emergencies, when blackouts are imminent. Because we do not adopt the more limited CEC proposal, we similarly decline to adopt the more expansive TURN proposal.

CEC recommends that the Commission study opportunities for customers who own their own substations or distribution transformers to implement voltage reductions in response to system emergencies or energy conservation. No other party supports this recommendation. CEC has no estimate of costs or potential savings. We have no reason to believe the potential savings would justify costs incurred by both the Commission and customers, and the risks incurred by customers. We do not authorize utilities-entities we directly regulate-to implement emergency voltage reduction. We decline to study opportunities for customers-whom we do not directly regulate-to engage in the same, or similar, action, absent a compelling showing that an untapped opportunity reasonably exists.

Farm Bureau's alternate recommendation is to implement voltage reduction in 0.5% increments. No other party supports this recommendation. If benefits are linear, each 0.5% increment would mean up to a maximum of 32 MW in potential savings on SCE's system. Farm Bureau's recommendation, however, involves implementing voltage reduction until customers complain, and/or noticeable problems occur. Customer complaints and noticeable problems would most likely include complaints and problems based on injury, damage and disruption. We are not inclined to implement a program until injury, damage and disruption occur.

Further, Farm Bureau's recommendation is more like an experiment than an emergency voltage reduction plan. We are not persuaded that we should engage in such an experiment. Overall, we think the need has subsided, potential benefits do not justify the risks, and other reasonable alternatives are preferable, thereby making the Farm Bureau's alternate recommendation no more desirable than the core recommendation.

Finally, utilities and others recommend comprehensive and thorough laboratory and field testing, as well as formal input from ANSI, the Institute of Electrical and Electronic Engineers, manufacturers, consumer testing groups, and consumers before proceeding with a voltage reduction program. We do not think this is necessary. The need for emergency voltage reduction has subsided and the benefits are marginal. The cost and burden on parties to further assess the risks and benefits are unlikely to be necessary or reasonable, given the availability of other alternatives.

5.2.6. Conclusion

The current energy situation does not at this time justify adoption of one or more of the recommended emergency voltage reduction programs. Rather, numerous other programs now in place as a result of Executive Orders by Governor Davis, as well as initiatives by this Commission and utilities, have significantly moderated the need to authorize an emergency voltage reduction program. If authorized, such program would be effective only in SCE's service area beginning in Summer 2002. On balance, the potential benefits are modest, while the costs and risks, although not quantified, are potentially great. Other reasonable alternatives are available. The majority of parties oppose an emergency voltage reduction program. Both the Chairman of ASC C84 and a member of the ANSI C84.1 Committee caution against service voltages below 114 volts. As a result, we decline to adopt an emergency voltage reduction program. Nonetheless, should the need resurface, parties may propose specific emergency voltage reduction measures for further Commission consideration.

²⁴ For example, "where the operation of the applicant's equipment requires unusually stable voltage regulation or other stringent voltage control beyond that supplied by PG&E in the normal operation of its system, the applicant, at his own expense, is responsible for installing, owning, operating, and maintaining any special or auxiliary equipment on the load side of the service delivery point as deemed necessary by the applicant." (PG&E Rule 2.C.1.e.)