4. PG&E's Existing System

In response to Appendix A of the ACR, PG&E provided testimony on system design, staffing, call center performance, deployment of resources, outage commissions, and reliability performance in general and during the December 2002 storms. PG&E's electric distribution system serves 4.7 million customers throughout Northern California, from Eureka in the north, to Bakersfield in the south, and extending from the Sierra Nevada Mountains in the east, to the Pacific Coast in the west. Approximately 50% of PG&E's customers reside in the San Francisco Bay Area. PG&E maintains 2.3 million poles, 93,000 miles of overhead circuits, 22,000 miles of underground circuits, 300,000 underground enclosures, and 2,500 distribution substation transformers. PG&E trims or removes approximately 1.8 million trees annually to ensure clearance between vegetation and overhead lines and maintain compliance with all rules and regulations.

Electricity consumed by PG&E's customers is delivered to PG&E's distribution system via high-voltage transmission lines. PG&E has almost 800 substations that connect PG&E's distribution system to the high-voltage transmission system. These substations are the central hubs of PG&E's distribution system. Each substation transforms the high-voltage electricity (e.g., 230 kilovolt (kV)) to a lower voltage electricity (e.g., 12 kV) for delivery to end-use customers. Primary and secondary distribution lines extend from each substation and deliver electricity overhead or underground. Local transformers that further lower the voltage level (e.g., 12 kV to 240 volts (V)) and switching equipment link the distribution lines in patterns that provide service drops to meters at customer premises.

PG&E's service territory is divided into 7 operating areas, 18 operating divisions, and 3,033 individual circuits. Service reliability and staffing varies by area and division due to circuit design, geography, customer density and other factors. Table 1, below, contains the 2002 Division and system values for the number of customers served, the number of circuits, the line miles of primary overhead and underground distribution circuitry, and the SAIDI and SAIFI performance indices.

TABLE 1

PG&E 2002 SERVICE RELIABILITY METRICS BY DIVISION

Division	Customers	Circuits	OH Miles	UG Miles	Total Miles	Excluding Major Events SAIDI SAIFI		Including Major Events SAIDI SAIFI
Central Coast	290,610	163	5,839	1,125	6,965	206.6	1.390	605.2	2.273
DeAnza	216,340	121	1,741	828	2,569	101.7	0.838	387.5	1.418
Diablo	292,653	128	2,239	2,093	4,333	124.1	1.372	203.4	1.729
East Bay	356,741	206	1,765	875	2,641	111.7	0.980	189.3	1.451
Fresno	380,985	285	10,867	2,064	12,932	161.0	1.323	218.2	1.518
Kern	216,083	228	7,035	1,162	8,197	151.3	1.201	184.9	1.335
Los Padres	187,360	76	4,578	876	5,454	121.0	1.177	162.4	1.490
Mission	361,329	156	1,823	2,326	4,150	65.5	0.823	99.0	1.084
North Bay	245,682	96	2,805	1,173	3,979	138.2	1.212	747.5	2.521
North Coast	362,934	186	10,455	1,587	12,043	223.1	1.178	1,098.1	2.330
North Valley	199,093	155	10,055	815	10,871	231.2	1.416	977.6	2.458
Peninsula	301,086	201	2,250	1,062	3,313	99.0	0.978	457.2	1.778
Sacramento	192,811	121	4,899	1,262	6,161	165.2	1.274	280.0	1.711
San Francisco	398,882	254	573	620	1,193	69.8	0.624	127.8	0.921
San Jose	380,913	155	2,360	2,156	4,516	108.4	0.931	193.6	1.219
Sierra	248,784	149	7,896	1,346	9,242	176.6	1.200	612.2	2.174
Stockton	262,745	191	6,215	1,427	7,642	181.2	1.314	326.3	1.879
Yosemite	225,733	162	10,958	866	11,824	137.5	1.245	215.1	1.497
System	5,120,764	3,033	94,361	23,671	118,033	139.2	1.112	381.9	1.670

As of December 31, 2002, PG&E employed a total of 3245 electric field personnel, a reduction of 14% from the 3699 field personnel PG&E employed as of 1998. All electric field personnel are available for emergency response. PG&E does not have separate staff dedicated to "reliability," "maintenance," or emergency response functions. Rather, individuals in these job categories perform work in a variety of functions and may be installing facilities to connect new customers one day and restoring power during an outage the next day.

As discussed above, the metrics the Commission considers when examining PG&E's electric reliability include SAIFI, SAIDI, CAIDI, MAIFI, the number of circuits that consistently perform poorly, and the MR&ROs per line mile. During emergencies and severe storms, PG&E's performance is also measured against a service restoration performance metric.

In addition to these metrics, PG&E internally evaluates its performance using metrics related to employee safety, customer service, and call center performance. MR&RO, defined as the total number of unplanned sustained outages that could be influenced by inspection and maintenance activities, was developed to measure the effectiveness of PG&E's inspection and maintenance practices. PG&E's average performance from 1999 through 2001 for the MR&RO and the number of circuits with greater than 12 outages per year improved relative to 1996 through 1998 performance for MR&RO or 1997-1998 performance for the number of circuits with greater than 12 outages per year.

Prior to 2002, PG&E tracked Occupational Safety and Health Administration (OSHA)-recordable events to measure employee safety performance. However, PG&E is re-evaluating whether the OSHA recordables metric remains appropriate since OSHA changed its reporting rules and definitions effective January 1, 2002. PG&E's OSHA recordables performance from 1999-2001 improved relative to PG&E's performance from 1996-1998. For 2002, PG&E used three measures of lost work days: (1) lost work days per case; (2) lost work days; and (3) lost work days per employee.

PG&E also tracks customer service quality through customer surveys and by measuring the effectiveness of outage management functions, such as providing information about an outage and providing accurate information on how long an outage will last.

PG&E has a network of three call centers that provide customer service through a combination of service representatives and Voice Response Unit (VRU) systems at all times. The VRU automated self-service functions include a general service/information menu for account balances, payment arrangements, and appliance service appointments. There is also a VRU menu for outages, which includes options to hear outage status messages, to report outages, and to request a callback. The call centers have over 900 inbound telephone lines for customers, with a capacity of handling over 500,000 calls per hour. In a normal week, the call center representatives and automated response systems handle about 260,000 calls.

PG&E uses a computerized OIS to identify, track, and communicate information for electric outages. PG&E uses its OIS in two key ways: (1) to assist in deploying resources to address outages, and (2) to provide outage information to customers via the call center. PG&E's OIS links field information (e.g., outage location, causes, response assignments, and estimates of restoration) to PG&E's Customer Information System (CIS), which is used in the call centers to communicate with customers. The OIS receives outage information from several sources, including customer calls, automatic system devices located on PG&E's facilities, field personnel, and Enhanced Outage Notification devices located in customer homes. The OIS addresses outages affecting multiple customers, such as during storm events. For single customer outages, trouble reports are managed through the Field Automation System.

When an outage report is entered into the OIS system, the system notifies the appropriate dispatcher. The dispatcher then sends a troubleman to the outage location and initiates an outage record in OIS, with an estimated time of arrival. Once this information is recorded in the OIS it becomes accessible to customers through the call center.

When an outage call is received, the call center systems check the OIS for outage record and status information. If there is no record of an outage, the customer is advised of this and can report the outage. If there is a record of an outage, the call centers provide the customer with the available outage status information including the estimated time of restoration (ETOR).

PG&E tracks the ETORs provided to the call centers for each of its outages so that it can compare the ETOR times in its OIS with the actual time the power was restored. PG&E does not record the ETOR information provided to customers when they call and therefore cannot compare the ETORs received by the customers to the actual time of restoration. PG&E states that the comparison of the first ETOR to actual restoration time provides a measure of how well PG&E's field and dispatch personnel estimate the time it takes to restore service, and comparing the last ETOR to actual restoration time measures how well the ETOR improves over time as better information becomes available.

PG&E's OIS database contains information concerning outages recorded since January 1, 2000. Comparable data from PG&E's prior outage database, CTAS, was not archived, and is not available. For the past three years, from 2000 through 2002, PG&E's ETOR was, on average, within two hours of the actual time of restoration. During non-major event conditions, including storms, PG&E either provided an ETOR, or restored service within four hours to 85 percent of all outages. In other words, customers who called PG&E to find out the ETOR of their service were provided an ETOR that, on average, was within four hours of the actual restoration time.

There were three major events in PG&E's service territory from 2000 through 2002. During these major events, PG&E either provided an ETOR, or restored service within four hours to an average of 73 percent of the outages. The difference between actual restoration time and the first ETOR during the December 2002 storms averaged 11.52 hours, while the differences between the actual restoration time and first ETOR during the November 2002 storms and the November 2001 storm were 6.25 hours and 5.14 hours, respectively.

During storm events, an emergency plan is used to supplement normal call center operations in responding to the storm. The emergency plans cover extending hours, adding service representatives and conducting daily conference calls during storms to review and assess operations performance. In addition, during storm events, one or more of PG&E's Operations Emergency Centers is activated to expedite the deployment of restoration resources and assist in outage communications.

As shown in Table 2 below, PG&E's SAIFI has been improving, from a high of 1.709 in 1996 to a low of 1.112 in 2002.⁹ PG&E's SAIFI from 1999-2002 was better than the average performance from 1996-1998.

TABLE 2

Year	SAIFI Major Events Excluded	SAIFI Major Events Included
1996	1.709	2.462
1997	1.639	1.700
1998	1.659	2.130
1999	1.477	1.481
2000	1.410	1.413
2001	1.439	1.559
2002	1.112	1.670
1996-1998 Average	1.67	2.10
1999-2002 Average	1.36	1.53

Similarly, Table 3 shows that while PG&E's SAIDI performance has fluctuated from year to year, PG&E's average SAIDI performance from 1999-2002 has improved slightly relative to PG&E's average SAIDI performance from 1996-1998.¹⁰

TABLE 3

Year	SAIDI Major Events Excluded	SAIDI Major Events Included
1996	178.1	347.0
1997	161.8	171.3
1998	180.0	317.0
1999	156.7	157.2
2000	167.9	168.4
2001	211.8	249.1
2002	139.2	381.9
1996-1998 Average	173.3	278.43
1999-2002 Average	168.9	239.03

⁹ Exhibit 13, Table 2-4, p.2-16 ¹⁰ Exhibit 13, Table 2-5, p.2-17.