Performance contracts between utilities and energy services companies (ESCos) are typically structured so that payments to the contractors are based on measured energy savings. The verification plans written into such contracts (and sometimes outlined in the utility's Request for Proposals [RFP]) often specify how the savings will be measured (see "Performance Contracting: Advice to Utilities," HE Nov/Dec '92, p. 13, and "Performance Contracting: An ESCo Perspective," HE Nov/Dec '92, p.23). Yet the details of savings measurements can be contentious, and utilities are increasingly adding tests to the verification plans to ensure quality control. Two contracts recently signed by ESCos--one with a utility and the other with a federal power marketing agency--illustrate a trend towards stricter verification plans.

A year ago, Portland General Electric (PGE), of Portland, Oregon, signed a pay-for-performance contract with SESCO, Inc., of Lake Forest, New Jersey. The verification plan specifies the method of savings evaluation: calendar-year pre- and post-weatherization utility bill data of treated homes will be compared to a control group with the same biases as the treatment group. The ESCo is allowed some latitude in choosing houses, but if there is more than a 10% difference in the energy consumption of the control and treated groups before weatherization, the extremes (outliers) will be removed from the samples. The contract not only spells out which weather normalization software (PRISM) evaluators will use, but even mandates the data screens and software defaults, according to PGE's senior product development analyst Rick Weijo.

To assert quality control beyond measured savings, the utility will also inspect one of the treated buildings, and a rigorous checklist of specifications must be met before SESCO receives payment. In addition, PGE will conduct surveys to determine customer satisfaction bonuses for SESCO of up to $20 per home.

The utility wrote most of these verification measures into the performance contract at the RFP stage, in order to ensure accuracy and quality of work, and that customer satisfaction reflects well on the utility, according to Weijo. "Specificity in an RFP keeps quality control issues clear cut," he said, and "there's less to negotiate." He warned that RFPs without explicit verification methodology can open the door to lengthy and contentious negotiations. "It's also a nightmare if you end up with different verification procedures for each contract."

Outlining verification procedures in the RFP does not always guarantee smooth negotiations, however. In January 1991, Bonneville Power Administration issued an RFP for a pilot energy conservation program to test the viability of using performance contracting. The RFP stated that for residential contracts, pre- and post-weatherization billing records for participants and a control group would be evaluated, weather normalization would be applied, and persistence of savings would be measured.

When a SESCO proposal for the contract was rejected by Bonneville, SESCO intervened in the federal power marketing agency's rate case in an effort to force further negotiations. In the hearings that followed, SESCO argued that Bonneville's proposed savings verification methods were "unnecessary restrictions and requirements." The company also objected that "Bonneville sought instead [of performance contracting] to require SESCO to adhere to a rigid schedule of installing certain measures, whether or not such measures are cost-effective," according to the Bonneville Administrator's Record of Decision.

SESCO claimed that specifying its intended audit procedures or describing how the company would choose which buildings to treat would force the ESCo to release proprietary information. But Bonneville found it unacceptable to set up a performance payment scheme based on an evaluation without this information, testified Pam Brandis, lead evaluator of Bonneville's demand-side bidding programs. She said that evaluators would need to know how houses were selected for weatherization in order to select houses for a comparable control group.

"Unless Bonneville is provided with the threshold criteria SESCO will use to determine ECM [energy conservation measure] installations, savings due solely to comprehensive treatment are not guaranteed," Brandis testified. "Coupled with a flawed program, a flawed measurement process would require Bonneville to make substantial payments for fictitious goods."

Bonneville peppered its testimony with descriptions of problematic SESCO programs elsewhere. It revealed the experience of Western Massachusetts Electric, a division of Northeast Utilities, where regulators deemed a SESCO program not cost-effective. A 1988 evaluation in Massachusetts Department of Public Utilities document 88-250 revealed sloppy installations, such as undersized pipe wrap and water heater blankets, and gross overestimates of savings by the contractor, among other problems.

Bonneville also related the experience of Hennepin County, Minnesota, where a shared savings program was implemented from 1984 to 1986 by SESCO vice-president Richard Esteves, who was head of Sentinel Energy of Minnesota, Incorporated. The program left homeowners (who invested in it) irate and a financing company bankrupt when savings failed to materialize (see "Shared Savings Miss the Mark," HE Nov/Dec '88, p.6).

Ultimately, Bonneville's administrator determined that SESCO's complaints were outside the scope of the rate case and no decisions were made regarding SESCO.

In the course of the rate case hearings, Bonneville defended its rejection of a bid from another ESCo, Free Lighting Corporation, for the same RFP. (Although the two organizations are independent of each other, Edward Esteves, brother of Richard Esteves, is president of Free Lighting.)

Bonneville also cited the experience of Orange and Rockland Utilities, Inc. (O&R), a northern New Jersey and southern New York utility. A 1991 evaluation by the State of New York Department of Public Service, the regulatory commission for O&R's New York division, found a severe lack of quality control in the LightSavers Program implemented by Free Lighting. "[F]ield evidence...indicates that [compact fluorescent] bulbs are being installed where it stretches credibility to suggest the bulbs will be used the minimum [3] hours per day, wrote Carol D. Taylor, energy efficiency chief at the Department of Public Service. The evaluation found 60 bulbs installed per home in some cases. Crews were not equipped with ladders or other tools "necessary to provide the service," and some had even asked residents to borrow theirs, the evaluators reported. The department was most concerned with the questionable validity of the savings estimates and the lack of utility oversight that the evaluation revealed.

SESCO and Free Lighting Corp. have now filed suit for unspecified charges against Bonneville in the Ninth Circuit Court of Appeals. That case is still pending. SESCO's argument is that Bonneville forsook cost-effective conservation opportunities when it rejected the SESCO bid, favoring BPA's own programs, some of which have been shown not to be cost-effective. "They sent us a letter with three ultimatums on it and we sent a letter specifically accepting each of them and they terminated negotiations because they said we hadn't unequivocally accepted them," said Esteves. (Bonneville argued in the rate case that by not signing its letter on the signature line, SESCO hadn't fully accepted the ultimatums.)

Eventually Bonneville signed an energy services contract with Lakeland Utility Conservation Incorporated (LUCI) of Lake Hopatcong, New Jersey, another independent corporation that has a close working relationship to SESCO. Although not based on measured savings, that contract also contains a strict verification plan. Designed to eradicate the possibility of fraud or poor quality installations, the plan involves an evaluation that includes inspections with video and still cameras and interviews with customers.

An independent verification contractor will visit a "statistically significant random sample" of the homes LUCI retrofits. Independent inspectors will use a long checklist of quality standards to ascertain, for example, if installers used the right kind of pipewrap and if it fit properly, or if they replaced every showerhead rated over 2.5 gpm with a 2.0 gpm or less model. The evaluation sample will be extrapolated to calculate the payment for each group of houses treated. If only 60% of the houses in the sample are retrofitted according to the explicit procedure prescribed in the contract, then 60% of the treated units will be paid for by the agency. Correction can be done group by group, followed by a second evaluation to measure the improvement, which would earn increased payment. The plan includes subsequent persistence checks by a verification contractor to ensure that materials and installations remain intact before any latter payments are granted.

So what do the ESCos think of the trends in performance contracting? Esteves is not willing to say that verification is getting tighter across the board, but said, "I think that there have been significant improvements in the measurement process and procedures that are available...A number of the contracts that I have seen are improved." Esteves continued, "On the other hand there are a lot of contracts that go by the name of performance conservation that are still just counting the number of ...widgets and that are based on engineering estimates of savings."

Not all ESCos mind the trend towards stricter verification procedures and some even welcome it. "We definitely feel that there needs to be more accountability and more incentive for performance contracting," said Steve Morgan, executive director of Citizens Conservation Corporation, an ESCo in Boston, Massachusetts. "Our position is that we are ready to accept any measurement plan that a utility imposes on itself," said Esteves. "We have not found any utility that's willing to accept us on that basis."

-- Karina Lutz