IV. Hearing

The factual account given above was confirmed at the public hearing in this case.

Other significant matters developed on the record are discussed here.

1. MRCH prepared an exhibit attempting to show that actual meter readings were correct based upon connected load data. Cross examination established that complainant's calculations were unreliable.

PG&E presented no connected load calculations in support of its billing. It has consistently relied solely upon MRCH's billing history to support its underbilling estimate.

2. Complainant testified that its review of the sometimes erratic meter readings for the four years from 1992 to 1996 persuades it that the suspect wire to the adapter must have been intermittently making a connection so that the meter would sometimes register usage from the loose wire.

The PG&E account representative who prepared the bill testified that the representative made no adjustment for intermittent meter registration, based on the opinion of a company distribution planning engineer that such registration would not be possible.

A company supervisor of meter personnel testified, however, that a meter could intermittently register in the given circumstances so that bills varying from true usage might occur.

3. Complainant detailed its unsuccessful efforts to have PG&E produce the adapter or wires claimed to have caused any underbillings and lend or sell the underregistering meter to claimant so that the parts and meter could be independently examined.

4. Complainant presented an expert witness with many years of service with PG&E who testified that, when he was employed by the utility, it was company practice to promptly meet with the customer when a like situation occurred, to produce the faulty parts and materials, to test the meter in the presence of the customer, and to allow the customer to respond.

While PG&E's records state that MRCH's meter was tested upon installation in 1990, the test result documents were unavailable. A more recent test was done of the meter and capability of the system, but no evidence was introduced to prove that the adapter and its connecting wires were ever inspected after being installed and reportedly tested.

An adapter is a device having three cylinders where the three wires to the system are inserted. Once inserted, the wires are held in place by three screws, permitting current to pass through to the meter measuring customer usage.

As inspection results were unavailable, complainant's expert suggested that the adapter and its connecting wires be removed under the joint scrutiny of PG&E and himself, and the parts sent to a laboratory for analysis. It was reasoned that such a test would show when the disconnection of the loose wire from the adapter occurred. Each time that a wire is inserted into the adapter, the affixing screw makes a crimp in that wire. Thus, if there are two crimps in the wires, it can be assumed that one crimp was made upon installation in 1990 and the second was made when the registration problem was discovered in 1996. If only one crimp was found in a wire, it can be assumed that the wire was not properly installed in 1990, and the one crimp was made when the PG&E employee rewired the adapter in 1996.

It was complainant's view that the offending wire was never properly installed but nevertheless randomly connected itself without benefit of the affixing screw so that meter reads could not be taken as accurate at any time.

The laboratory report showed two crimps in two of the wires and one crimp in the wire claimed by PG&E to be loose.

PG&E responds that the test means nothing because its meter technician's recollection is that he changed the disconnected wire in 1996, replacing it with a new wire. Thus, PG&E reasons that the fact that only one crimp was found in the suspect wire confirms the testimony of PG&E's meter technician that he replaced that wire in 1996.