NORTHWEST ELEMENTARY SCHOOL, PASCO COUNTY, FLORIDA
Northwest Elementary School, located on the west coast of Florida,
was the location of the first school site that was evaluated.
A 58,000 ft2 main building, comprised of classroom pods, administrative
spaces, a media center, and a cafeteria, became the subject of
the study. The work was sponsored by the Florida Department of
Education and is more fully described in a source report (Floyd
et al. 1995).
Primary lighting for the school was 2 x 4 luminaires with T10
lamps/magnetic ballast or T8 lamps/electronic ballast. The connected
facility lighting load is approximately 87 kW. The test building
was unusual in that it already contained an efficient lighting
system. Pasco County also has one of the most aggressive energy
management programs of any district school board in the state.
Even before installation of the occupancy senors, lighting was
effectively controlled by facility staff so as to prevent waste.
Given these factors, it was expected that the evaluation in Northwest
Elementary would provide insight into the minimum savings that
could be expected from the technology if properly applied in a
Technicians audited the school on December 21, 1994 and subsequently
drew up a plan for instrumentation to monitor its energy use.
The facility was instrumented on February 25, 1995, also using
a similar before and after monitoring protocol. Fifteen minute
electrical demand data were taken for six months prior to the
lighting controls being modified to accommodate occupancy sensors.
Data in the baseline period revealed that lighting made up approximately
24% of total electrical energy use at the school (70 kBtu/ft2).
A total of 46 passive infrared (PIR) occupancy sensors were installed
and carefully adjusted in terms of location, time delay, and sensing
sensitivity from August 7, 1995 to August 15, 1995. The installation
was performed by a team of two electricians, a Research Engineer
and the Energy Coordinator from Pasco County. Approximately 33
classrooms, seven offices, and a cafeteria were equipped with
occupancy sensors. In several offices wall sensors were used.
The remainder of the spaces (classrooms, cafeteria, and larger
offices) received ceiling mounted sensors. The broad coverage
of the ceiling mounted PIR sensors minimized the need for multiple
occupancy sensors in all but five areas. Dual technology sensors
were considered, but not utilized due to their higher cost.
Classroom occupancy sensors were located in a corner near the
teachers desk to minimize false "offs" when only the
teacher was in the classroom. All occupancy sensors were set to
a 10 minute time delay, which has worked well in most situations.
Shorter time delays may improve savings, however, false "offs"
may also increase. Past installation experience has shown that
unless the occupancy sensors are properly located, aimed, and
tested by experienced personnel, poor savings and occupant dissatisfaction
The analysis of the comparative pre- and post-retrofit periods
as shown in Figure 3, indicated an average savings of 10.8% (96
kWh) on school days of the pre-retrofit lighting energy with greater
reductions to total energy due to reduced load on the air conditioning
system. Most of the savings occurred during the evening hours
so that monthly peak electrical demand was unaffected. There are
some 200 school days per year, not including holidays, weekends
and summer recess. The school day extends from 7:00 AM to 3:45
PM, although office and janitorial activities often extend beyond
the formal school day schedule when much of the savings were found
Based on the monitoring, an annual direct lighting energy savings
of 26,620 kWh was estimated. To this was added an estimated additional
7,260 kWh in reduced HVAC costs (Rundquist et al. 1993). At the
facility's electricity rate ($0.05/kWh) annual monetary savings
is estimated at $1,694. The data did not evidence any reduction
in peak electrical demand from the retrofit, so no credit was
taken for this portion of monthly energy costs.
The cost for the sensors, wiring and relay packs for the project
was $4,067 or about $88 per control. Installation labor was valued
at $2,000 (125 man-hours). Including costs of installation and
set-up, the payback of the occupancy sensor retrofit was approximately
3.6 years with a 28% simple rate of return from the investment.
This performance is considered excellent given that the building
already had an efficient lighting system which was responsibly
controlled prior to the occupancy sensor installation. The project
results indicate that with proper installation and adjustment
(which was found to be critically important to user acceptance
and performance) occupancy sensor technology can provide economically