Retired from TVA last year and now auditing accounts in the Tennessee Valley. TVA demand billing is genuinely different from every other utility Ive encountered. TVA uses a three-part demand structure: 1) Maximum demand ? highest 30-minute peak in the month 2) On-peak demand ? highest 30-minute peak during on-peak hours 3) Excess demand ? the difference between maximum and on-peak if maximum exceeds on-peak. Each has a different rate per kW. Most auditors from outside TVA territory apply single-demand analysis and miss the three-part structure entirely.
TVA demand billing ? unique structure most auditors dont understand
Vernon ? I audit TVA accounts through MLGW in Memphis and the three-part structure confused me for months. The excess demand charge is particularly tricky because it penalizes accounts whose peak occurs off-peak. A factory that runs a night shift and peaks at 2am pays both on-peak demand AND excess demand because the overall peak exceeds the on-peak peak. Its a penalty for off-peak peaking which seems counterintuitive.
James ? its counterintuitive but the logic is that TVA must maintain infrastructure for the maximum demand regardless of when it occurs. The excess demand charge pays for that capacity. But heres where the audit opportunity exists: if a client can shift their off-peak peak to occur during on-peak hours, the excess demand charge disappears because maximum demand equals on-peak demand. No excess.
Vernon ? shifting peak INTO on-peak hours to avoid excess demand charges? That goes against everything I learned about demand management. Usually we try to shift load OUT of on-peak hours. In TVA territory you might actually want to peak during on-peak?
Edward ? it depends on the math. The on-peak demand rate is higher per kW than the excess demand rate. But if you can eliminate the excess demand charge entirely by equalizing on-peak and off-peak peaks, the total demand bill might be lower even though on-peak demand increases slightly. You have to model both scenarios with the actual three rates. There is no universal rule ? its account-specific.
Vernon ? this is eye-opening. Ive been auditing MLGW accounts in Memphis for years and never properly modeled the three-part demand interaction. Ive been focused on reducing on-peak demand exclusively. If some of my clients have significant off-peak excess demand, I may have been missing the bigger savings opportunity on the excess demand side.
Frank ? exactly the realization I hope this thread produces. Pull the demand detail from MLGW bills and look at three numbers: maximum demand, on-peak demand, and the difference between them. If the difference is large, excess demand charges are likely a significant portion of the bill. Addressing that gap ? either by reducing off-peak peaks or by slightly increasing on-peak demand to equalize ? can produce counterintuitive but genuine savings.
Vernon brings invaluable utility-side perspective to TVA demand billing. The three-part demand structure is unique to TVA and its local power company distributors. Key lesson for all auditors: never assume demand billing works the same way across utilities. TVA accounts require a fundamentally different analytical approach than accounts served by investor-owned utilities. Vernons point about modeling both scenarios rather than applying universal rules is the correct methodology for any non-standard tariff structure.