Occupany Sensor on VAV Box

2 posts / 0 new
Last post


Would anyone who has evaluated energy savings through the use of occupancy sensors controlling VAV boxes please share some experience, do's and don'ts, etc.
I believe the VAV boxes would have a minimum setback to 0% or 10% flow and possibly no reheat in the summertime. I have been trying to model this with little success. Is it possible to do using freeware or should I start constructing hourly reports. Thank you all for your assistance.


Kevin Kyte's picture
Joined: 2011-09-30
Reputation: 0

Hello Kevin,

I have not yet modeled this but here is what I am planning to do:

1. Determine the total floor area of zones with occupancy sensor
control. Let's call it "Area X"
2. Estimate the amount of time that any one zone would be unoccupied.
For purposes of this explanation, assume it is 15% (per the ASHRAE 90.1-2004
Table G3.2).
3. Multiply "Area X" by 15% to determine an equivalent floor area that
would be unoccupied continuously. Let's call this "Area Y"
4. Look at all of your zones used to develop "Area X" and determine the
relative proportion of areas that have north, east, south, west exposures
and are in the core of the building. Let's call the proportion of area in
the north exposure "N"
5. Multiply "Area Y" by each of these different proportions (e.g. "Area
Y" x "N") to determine how much unoccupied space needs to be represented.
Review your list of "Area X" zones to find the ones with approximately the
correct floor area corresponding to the requirements for the different
exposures - these will be normally unoccupied spaces. Depending on how the
zones were initially modeled, you may need to split some zones to make
specialized "unoccupied" zones (this could get pretty time intensive unless
you anticipate having to do this when the model geometry is assembled). The
purpose of going through the hassle of identifying zones in each of the
different exposures is to model the impact of occupancy sensor control on
the average space.
6. Within the normally unoccupied zones, set the minimum airflow in the
Airside Systems/Zone/Airflow tab to 0 cfm and change the occupancy, plug,
lighting, heating, and cooling schedules to use an unoccupied schedule (use
YEAR" schedules you have already made).

It seems like this approach could potentially work and would be fairly
accurate for a building where the aggregate unoccupied periods are
distributed randomly throughout the typical workday. If there are separate
mechanical systems, the procedure above would need to be replicated for each
system. This approach would probably work for a typical office but may not
work for an operation with mandatory lunch hours. In reality, the typical
zone would probably not float all the way back to its unoccupied setpoint
since the unoccupied periods during normal building hours will be fairly
brief (<1 hr?). Assuming that some zones are always at setback temperatures
does decrease conduction gain/losses, however, this component of space load
is generally small relative to solar, equipment, and lighting loads during
occupied periods in non-residential spaces. The most significant
uncertainty in this approach is the guesstimate of the fraction of time that
the typical space will be unoccupied.

If there are periods when many zones are likely to be unoccupied at the same
time, things can get a bit more complicated. Specialized schedules can be
made for occupancy, plug, lighting, heating and cooling, and you would need
to use the minimum airflow schedule option in Airside Systems/Zone/Airflow.

Is this what others have done? Does it seem reasonable?


Eric Studer's picture
Joined: 2011-09-30
Reputation: 200