International Business Machine’s San Jose Cogeneration Project (IBM)

Docket No. 85-SPPE-2

Small Power Plant Exemption Granted March 5, 1986

Project Never Constructed

Staff Counsel: Gregg Wheatland

Hearing Officer: Garret Shean

Presiding Member: Commissioner Doug Noteware


Project Summary


SPPE Filing and Project Description

International Business Machine Corporation (IBM) filed its Application for a Small Power Plant Exemption (SPPE) on May 5, 1985. IBM wished to convert an existing standby gas turbine generator into a natural gas fired cogeneration plant that would provide both steam and 65 MW for use by its production facility in San Jose. The congregation plant’s electricity would replace power which IBM currently purchased from PG&E. Any excess electricity not used by the IBM plant would be sold to PG&E.

The concept of an industrial applicant building a powerplant to produce its own electricity became known as "self-generation." IBM was the first self-generation project required to pass a need test. (Years earlier, the Texaco Cogeneration Project, Docket No. 80-SPPE-1, had been "deemed needed" as a cogeneration plant without regard to self-generation.

Now this SPPE proceeding was required to determine how the demand conformance tests would be applied to IBM, since self-generation had not been contemplated in ER 5. A similar problem was being faced in the Chevron El Segundo case, another self-generator, Docket No. 85-SPPE-5.

The SPPE Process

SPPEs are governed by Public Resources Code section 25541. The CEC must make two separate findings in order to grant an SPPE - that "No substantial adverse impact on the environment or energy resources will result from the construction or operation of the proposed facility" and that "Generating capacity will not be added which is substantially in excess of the forecast of electrical energy demands adopted pursuant to subdivision (e) of section 25305)." After receiving an SPPE, projects must then obtain appropriate local permits.

CEC staff’s environmental analysis for an SPPE is presented in an Initial Study, which the CEC conducts as the CEQA lead agency for the project under Public Resources Code section 25519(c). Traditionally, a facility which qualifies for issuance of a negative declaration is considered to also satisfy the environmental finding for granting an SPPE.

The Initial Study and Environmental Issues

The cogeneration plant site was part of a 550 acre IBM complex in San Jose. This was an industrial area, with the nearest residence half a mile away. The project raised no environmental issues. CEC staff issued its Initial Study and proposed Negative Declaration for IBM on January 5, 1986, concluding that the powerplant would not have a significant effect on the environment. IBM never had to present any testimony of its own on the facility’s environmental effects.

Demand Conformance and Self-Generation

ER 5 (The 1985 California Electricity Report) provided a complicated set of need tests which facilities had to pass in order to achieve demand conformance. (See Gilroy Foods, Docket No. 84-AFC-4 and ARCO-Watson, Docket No. 85-AFC-1.) Commissioners were already in disagreement as to whether an SPPE under ER 5 had to meet the identical test administered to AFCs. (See Placerita, Docket No. 84-SPPE-1, and later, Champlin, Docket No. 85-SPPE-7.) At the same time, CEC staff was beginning to apply the need tests very strictly, seeking to avoid the approval of facilities that exceeded the forecast.

However comprehensive ER 5 attempted to be, it failed to anticipate the arrival of self-generation. All ER 5 tests were based upon need in a utility’s service area. But the utility’s needs were irrelevant to a self-generation plant such as IBM, proposed to replace utility power and service the applicant’s own industrial needs. Objectively, ER 5 had no test intended for a self-generation plant, SPPE or AFC. However, a need test was necessary, and the Commission had to come up with one that could be applied to IBM.

CEC staff took the preliminary position that self-generation did not matter. Staff proposed normal application of the ER 5 need tests, an approach which guaranteed that IBM would fail every test.

Meanwhile, IBM asserted that a self-generation plant was exempt from the demand conformance requirement as a matter of law. Applicant opposed being subjected to any need test, since they were all meaningless to a self-generator.

Given the gap between staff and applicant, the Committee had to design its own test. On November 13, 1985, the Commission approved a Committee recommendation that ER V’s Unspecified Reserved Need Test be assigned to IBM, utilizing the Committee’s own concept of "economic load following" instead of "physical load following" as specified in ER 5.

In practice this meant that the IBM need hearings would focus upon the economics of self-generation, with the Committee applying its own standard of "ratepayer indifference".

Applicant’s witness testified that, if the project were built, lost sales to IBM would cost PG&E $5.1 million annually, only an insignificant .0006 of PG&E’s current revenue requirements. Rates paid by average PG&E residential customers could increase 3 cents per month to make up for this loss. Meanwhile, IBM’s electricity sales to PG&E would be governed by the principles of a negotiated "real time pricing" agreement, under which IBM would be paid less than current avoided costs under Standard Offer No. 1. According to the applicant, "ratepayer indifference" was thus achieved. (Pages 10-11 of the CEC Decision.)

CEC staff witnesses applied the ER 5 Unspecified Reserved Need Test to IBM, concluding that IBM failed Condition 3 by operating as a baseload plant which did not match PG&E’s physical loads. As for the Committee’s proposed "ratepayer indifference standard" staff witnesses calculated PG&E’s annual revenue loss at $7.6 million, with ratepayers indifferent to IBM sales of surplus electricity to PG&E. (Pages 11-13 of the CEC Decision.) Staff took no position on the ultimate question of whether, as an SPPE, IBM substantially exceeded the ER 5 forecast. (Page 17 of the CEC Decision.)

PG&E testified that the IBM project by itself would not significantly and adversely affect PG&E and PG&E’s customers. (Page 16 of the CEC Decision.)

The Committee, Commissioner Noteware, Presiding, issued a Proposed Decision on February 24, 1985, that recommended granting the SPPE. This document was 49 pages long, with an exhaustive discussion on self-generation and demand conformance. The Final Decision was reduced to 20 pages, significantly modified and narrowed in response to staff comments and Commissioner concerns. The less said, the better.

As revised, the Committee and the Commission, exercising their discretion, ruled that, IBM exceeds the forecast and does not conform to Condition 3 of the Unspecified Reserved Need Test. However, under the concept of ratepayer indifference, these standards are not exceeded "substantially". Thus, given IBM’s unique circumstances, the self-generation plant would not add generating capacity "substantially in excess of the Commission’s latest adopted forecast." (Pages 14-17 of the CEC Decision.)

The Committee thus recommended granting the SPPE, declining to make any ruling upon IBM’s legal arguments, and urging the ER 6 proceeding to more fully address the issue of self-generation. (Pages 17-20 of the CEC Decision.)


At the March 5, 1986 Business Meeting, IBM’s SPPE was granted in a decision signed by three Commissioners that reflected the Committee approach. There were actually five votes in favor of the SPPE, but for very different reasons.

Commissioner Crowley filed a March 14, 1986 concurring opinion which argued that self-generation was inherently different from seller generation. She found IBM’s self-generation project beyond the scope of ER 5, and felt it should not have been evaluated under the policies of the Specified Reserved Need Test. Commissioner Crowley nevertheless supported the Committee’s efforts to design a fair and reasonable application of ER 5 for IBM, pending development of a new self-generation policy in ER 6.

Commissioner Gandara’s March 18, 1986 concurring and dissenting opinion criticized the Commission for pretending to apply ER 5 methodology to IBM, while jettisoning it in practice. He reaffirmed his Placerita dissent, calling for an uncomplicated SPPE test that merely looked at whether the project was substantially in excess of the forecast. Commissioner Gandara would have found that IBM met this basic statutory test without the unnecessary inclusion of concepts such as "ratepayer indifference."


The IBM Cogeneration Project was never constructed, apparently because PG&E offered IBM a bargain rate for its electricity, undercutting the project’s economics. Such special rate negotiations between a utility and a potential self-generator would become routine, as the utilities sought to hold on to their customers and discourage self-generation.

Ultimately, the Energy Commission’s self-generation need tests would even require these negotiations, so that utilities had every opportunity to compete. However, if self-generation remained in a company’s best economic interests, the CEC need tests would treat self-generation as special, allowing the construction of such plants without regard to a utility’s needs.

As a result of the many disputed SPPEs under ER V, (Placerita, IBM, El Segundo, and Champlin), subsequent Electricity Reports, starting with ER 6, have often included specific language directing that SPPEs be subjected to the identical need test required for AFCs. The Commission also developed a special self-generation need test. The first such test from the 1986 Electricity Report (ER 6) contained the following language:

"(1) With respect to that portion of the power sold to the utility, the Commission will apply the principles and conditions applicable to QFs.

(2) With respect to that portion of the power produced for internal use, self-generation projects will be presumed needed if the applicant demonstrates that it attempted to negotiate a rate with the utility it would otherwise purchase power from and cannot obtain a rate that in the applicant’s judgment is less than the value of the self-generation project (cost of power from the proposed project less intrinsic value to the self generator). This presumption may be rebutted by "clear and convincing" evidence."

The Richmond Cogeneration Facility, Docket No. 86-SPPE-1, would be the first project evaluated under the new self-generation need test.



In my opinion, IBM was a visionary project that, if constructed, would have established a major precedent for self-generation to serve the ever-growing electricity needs of large Silicon Valley companies. A host of IBM self-generation clones could have benefited the entire state during the energy crisis caused by de-regulation in 2000-2001. The debate over system necessity for the controversial Metcalf facility (Docket No. 99-AFC-3) may have been avoided if IBM and other similar projects had been built in the 1980s and 1990s exactly where they were needed most.

Instead, PG&E appears to have convinced IBM to remain its customer by negotiating special rates. IBM abandoned this project, and to my knowledge, self-generation never became a serious solution to Silicon Valley's enormous electricity demands. PG&E's short term view of IBM self-generation as destructive competition turned out to be a bad decision for everyone in the long term.