Is it possible to create an accurate electricity budget for your building if you expect occupancy changes in the year ahead? Let’s take a look at one building in southern California to see how Gridium’s analytics shed light on how this building’s energy use changed once a new tenant moved in.
Snapmeter’s vs-expectation analytics, first bill period after new tenant.
In the bill period immediately following the move-in of the new tenant, Snapmeter’s vs-expectation analytics highlight and visualize the change in this building’s energy use. It looks like this building is now using about 15% more electricity with the new tenant, and of course the overnight use on Friday and the jump in day-time use on Saturday would warrant some further onsite investigation… perhaps that overnight use is related to the move in of the new tenant, too.
If we take a look at this building’s load curve during the following bill period, we see the impact of Snapmeter’s machine learning algorithms. The analytics will adjust to this building’s “new normal” of increased occupancy. The updated model no longer shows any statistically significant increase over expectation.
Vs-expectation analytics, second bill period after new tenant.
While the building has dropped excess use on Saturday and Sunday, the analytics now expect the building to top out during the work week at around 260 kW. This building’s operations team should also be congratulated for fixing the building’s Monday morning hard start.
All of this means building management could use changes in its load curve to inform next year’s electricity budget. Let’s imagine a new tenant is similar in floorspace and use-intensity–meaning no surprise Netflix servers or bitcoin mining supercomputers–then another uptick of 15% should be included in the accounting. Another option is to use the occupancy feature in Billcast to quantify a data-driven, weather-normalized link between occupancy changes and energy use.
Let us know of any questions.
