So, a property management firm manages a 140,000 square foot office building that has a main electrical meter apart from the utility meter. FacilityConneX sucks up the data from the main meter and runs analytics. “So what”, you ask, “what could be interesting about the data from the main meter? Why don’t I just get interval data from the utility and analyze it in a spreadsheet?”
Well, more often than you would expect, it’s VERY interesting and this story is about one of those times.
One thing is generally true – without probable cause, no one has time to ask their utility account-representative for interval data, wait for it a few weeks, send a reminder note when he forgets, then plug it all into a spreadsheet to see if there is any treasure to be found (yes some of you can access utility data online but it’s probably still a time-consuming task to process it).
That’s why this relatively simple but time-consuming task never gets done – unless it’s automated. And that’s what we do – our analytics work 24/7. They never get tired or bored and they’re sure to find the treasure.
What did our analytics find? See if you can spot the gold before I spill the beans. Here’s a load profile for a summer month. The building is unoccupied 10pm to 5am and weekends. It’s all electric – mechanical cooling and resistance heating. The load drops to 220 kW at night and rises to 550 kW during the day – a very repeatable occupied/unoccupied profile. See the gold yet? Here’s another clue – the lighting and HVAC is shut off (theoretically) during unoccupied hours.
***spoiler alert*** If the heating and cooling and air handlers and lighting is all off, what the heck is the 220 kW load that never ever goes away?! Forty percent of the normal daytime load never shuts off. Could the plug load be that much? We don’t think so, but even if it was, plug loads should be shut off at night too!
Sorry to say, this will be a 2-part blog. Our client hasn’t solved the mystery yet but we have made the following suggestions:
- Suck it up and roll out of bed one night, bring an electrician, pull the covers off electrical panels and measure circuit amps until you know where all that mysterious power is going. I’ve been shocked to arrive at some of my facilities late at night only to find a lot of lighting and equipment on that was supposed to be off – yes, your Building Automation System can and does lie to you – loads that are scheduled to be off may be on and you’ll never know unless you go see for yourself…
- …Unless there is enough sub-metering to reveal in more detail where that persistent night-load is going. That’s our second suggestion, add sub-metering to large air handlers, resistance heat loads, and lighting panels.
- And run analytics that continuously check the “status” of equipment against the power that it draws. The status could be lying. The meter will tell the truth.
Conservatively speaking, if half of the 220 kW persistent load can be shut off nights and weekends (4316 hours/year), that’s a savings of 475,000 kWh and $66,000 annually ($0.14/kWh).
And were not done yet. There’s more treasure to be found! When you re-sort the interval load data from largest to smallest you produce a “Load Duration” curve that shows how many hours the load was greater than any specific load. For example, the front of this curve is very steep – it shows only 8 hours at a load above 600 kW. If the load had been managed (limited to 600 kW) during those 8 hours, the billing demand would have been reduced by 100 kW at a value of nearly $1,000 (Demand charges are about $10/kW). Do that every month and you save $12,000 annually.
Luckily, most modern building automation systems have load management routines included in their basic packages so implementation of load management is easy and inexpensive.
Finally, when is the billing demand established? This utility is a bit of a rogue in that it allows the billing demand to be set anytime of the day – not just during peak hours (the way every other utility does it). Either way, it’s nice to know when you are setting your billing demand and we provide a chart that does just that:
It’s quite evident that, during the heating season, billing demand is being set during the early morning hours when the building is transitioning from unoccupied setbacks to normal occupied temperature settings. Surprisingly, this is happening even though the controls contractor made a deliberate effort to soften the transition by spreading it out over a 4-hour period. A demand management routine could augment this gradual transition and prevent high loads from spiking the billing demand.
One meter, a handful of analytics, $78,000 in potential annual savings – that’s some serious treasure.
…Part II of this blog when we identify the mysterious persistent load and turn it off!