Auditing a university campus in Dallas served by Oncor distribution with a competitive retail provider. The campus has 15 buildings on a master meter. The retail provider bills demand on the campus non-coincident peak ? the sum of each buildings individual peak regardless of timing. The campus argues demand should be billed on coincident peak ? the total campus load at the single highest interval. The difference is substantial: non-coincident peak is 4,200 kW, coincident peak is 3,400 kW. Thats 800 kW in dispute.
Coincident vs non-coincident demand ? which does your utility bill?
Patricia ? this depends entirely on how the campus is metered. If theres one master meter at the point of delivery, that meter measures coincident peak by definition ? it sees the total campus load at each interval. If each building has its own meter and the retail provider sums the individual peaks, thats non-coincident billing. Which metering configuration does this campus have?
Patricia M ? theres one master meter at the main substation plus individual building submeters for internal allocation. The retail provider should be billing based on the master meter reading which is coincident by definition. But their bill shows demand that matches the sum of building submeters rather than the master meter. I think theyre using submeter data instead of master meter data for billing.
In San Antonio Ive seen retail electricity providers in the ERCOT market bill demand using interval data from the ERCOT settlement system rather than from the customers meter. ERCOT settlement data can differ from meter data due to loss factors and allocation methods. Check whether the retail provider is using meter data or settlement data. The source of the demand reading matters.
Michelle ? excellent point. I requested the raw interval data from both the master meter and the retail providers billing system. The master meter shows 3,400 kW coincident peak. The retail bill shows 4,200 kW. The provider confirmed they are using ERCOT settlement data with load profiling rather than actual master meter interval data. They claim its standard practice.
Patricia ? in ERCOT, if the customer has an interval data recorder (IDR) meter the retail provider MUST use actual meter data for billing. Load profiling is only used for non-IDR meters. If the master meter has IDR capability ? which any campus meter certainly does ? billing on load-profiled settlement data instead of actual IDR data is an error. The retail provider owes a correction.
Mitchell ? confirmed the master meter is IDR-equipped. The retail provider has been billing on load-profiled settlement data since the account opened 4 years ago. They should have been using actual IDR interval data. The demand overcharge is approximately 800 kW per month at $9.50 per kW ? thats $7,600 per month or $91,200 per year. Four years of overcharges potentially totaling $364,800.
Patricia ? $364,800 potential refund from a billing data source error. This is why auditing deregulated markets requires understanding the settlement process, not just tariff rates. The error isnt in the rate ? its in which data feeds into the billing calculation. Brilliant find.
Update: retail provider acknowledged the data source error and agreed to rebill using actual IDR interval data. Total refund for 48 months: $341,600. The difference from my $364,800 estimate was due to some months where the load profile data was actually close to actual meter data. Still the largest single finding in my auditing career.
This thread illustrates why demand charge auditing in deregulated markets requires additional expertise beyond tariff analysis. The error was not in the tariff application ? it was in the data source feeding the billing system. Key lesson: in deregulated markets, always verify that the billing demand reading matches the actual meter data. Mitchells point about IDR metering requirements in ERCOT is critical ? any account with IDR metering must be billed on actual interval data. Patricias $341,600 recovery is exceptional but the underlying error pattern is more common than most auditors realize.