I'm working on a complex multi-tenant office building in Phoenix where APS provides service under Schedule E-32 with time-of-use rates. Building has 12 tenants with different operating hours and the landlord is struggling with how to allocate TOU charges fairly. Current method just divides everything by square footage, but the law firm on the 8th floor runs 24/7 while the insurance office on 3rd floor is 9-to-5 only. The on-peak charges are $0.12847/kWh vs off-peak at $0.06234/kWh - huge difference. Anyone developed methodologies for TOU allocation based on actual usage patterns rather than just square footage?
APS Schedule E-32 time-of-use allocation methods
Carla, we've dealt with similar TOU allocation issues with Seattle City Light. The most defensible method is to install submeters on each tenant space and track actual usage during peak vs off-peak hours. Barring that, you can use occupancy schedules and lighting/HVAC load profiles to estimate each tenant's contribution to peak hour usage. The square footage method is lazy and unfair when rate differentials are that large.
David makes a good point about submetering being ideal, but that's expensive for existing buildings. I've used a hybrid approach where you allocate base building loads (elevators, common area HVAC, emergency lighting) by square footage, but allocate tenant-specific peak usage based on documented operating hours and connected loads. For that law firm running 24/7, they should bear much more of the on-peak charges.
Norm, that hybrid approach sounds promising. The challenge is getting accurate connected load data from each tenant. The law firm has a server room that I estimate draws about 15 kW continuously, plus their after-hours HVAC override adds another 25 kW during peak summer hours. Meanwhile the insurance office probably adds zero load during APS peak hours (3-8 PM). Current allocation method has insurance office paying $890/month in peak charges - seems grossly unfair.
Carla, I've used engineering estimates for similar situations with Xcel Energy TOU rates here in Colorado. You can calculate theoretical load profiles based on tenant type, square footage, and operating schedules. Office space typically runs 2-3 watts per square foot base load, then add HVAC loads based on hours of operation. It's not perfect but much more equitable than straight square footage allocation.
This is a great discussion. I had a similar case with Indianapolis Power & Light where a 24-hour call center was getting subsidized by daytime-only tenants under a square footage allocation. We developed load profiles based on actual spot measurements and tenant-reported operating schedules. Call center's utility allocation went up 40% but it was much more accurate and defensible.
Greg, that's exactly the kind of reallocation I'm expecting here. The law firm is definitely being subsidized by the 9-to-5 tenants. I'm going to propose load profiles based on connected equipment surveys plus documented operating hours. Even if it's not perfect, it has to be more fair than the current method.
Carla, make sure whatever methodology you propose is clearly documented and reproducible. I've seen landlords get sued over utility allocation changes, even when the new method is more accurate. Having engineering backup and clear documentation is crucial for defending your approach.
Susan's absolutely right about documentation. Down here in Mississippi we had a case where the new allocation method was challenged in court. Having detailed engineering calculations and clear methodology documentation saved the day. Tenant ended up paying attorney fees because the new method was demonstrably more accurate.
Thanks everyone for the input. I'm developing a detailed engineering study with load profiles, operating schedules, and allocation methodology. The goal is to make it bulletproof from both technical and legal perspectives. Will update this thread with results once I present it to the landlord.
Carla, I'd be very interested to hear how this turns out. We're seeing more TOU rates being implemented across the country and this allocation issue is going to become increasingly common. Your methodology could be a template for similar situations.