Energy savings from photosensors_savings by liuqingyan



        Energy savings from photosensors and occupant sensors/wall

        switches on a college campus


                           Birt,         B.; Newsham, G.R.

                           September 2009

                           A version of this document is published in / Une version de ce document se trouve dans:
                           Lux Europa 2009, 11th European Lighting Conference, Istanbul, Turkey,
                           September 9-11, 2009, pp. 731-738

The material in this document is covered by the provisions of the Copyright Act, by Canadian laws, policies, regulations and international
agreements. Such provisions serve to identify the information source and, in specific instances, to prohibit reproduction of materials without
written permission. For more information visit

Les renseignements dans ce document sont protégés par la Loi sur le droit d'auteur, par les lois, les politiques et les règlements du Canada et
des accords internationaux. Ces dispositions permettent d'identifier la source de l'information et, dans certains cas, d'interdire la copie de
documents sans permission écrite. Pour obtenir de plus amples renseignements :
      Energy savings from photosensors and occupant sensors/wall switches on a
                                  college campus

                                      Benjamin J. Birt 1 , Guy R. Newsham1

We report the savings realized by the use of lighting controls across a college campus in southern Ontario, Canada.
The campus had an extensive lighting system of addressable dimming ballasts, controlled and monitored centrally.
Of the 2295 luminaires from which we recovered data, 87 were controlled by 62 photosensors; these were in rooms
that also featured a manual control system (wall switch/dimmer/preset scenes) and occupancy sensor. The site was
monitored from mid-April to mid-July, 2008. Substantial energy savings were realised by the control system, though
savings varied greatly by control type, room type, and by day. Manual use of wall switches and switching by
occupancy sensors contributed most to savings. Meaningful additional savings were contributed by dimming, either by
a photosensor or use of preset scenes. We provide a detailed breakdown of the savings by control and room type.
Note that in a large space occupant sensors/wall switches affected all luminaires, whereas photosensors affected only
those luminaires close to windows or skylights.

Keywords: Lighting Controls, Energy Efficiency, Daylighting, Occupancy

1. Introduction
Lighting is responsible for a large fraction of institutional building energy use. In 2006, lighting accounted
for 10% of the total energy end-use in the commercial/institutional sectors in Canada [1]. Therefore,
reducing the lighting load while still satisfying occupant needs will go a long way in reducing overall
energy use. There are several methods available to reduce the lighting load of a building. For example,
lights could be automatically switched off when a room is not in use, or dimmed when there is sufficient
natural light entering the room.

Lighting control systems can range from a very simple manual on/off switch that is operated by the
occupant, to a system of addressable ballasts, dimmers, photosensors and occupant sensors all
controlled by a central computer. Such systems are widely available, and are encouraged by energy
codes and green building rating systems [2], yet there are limited studies of the long-term performance of
such systems in the field. For example, Jennings et al. [3] found that private offices with occupancy
sensors saved 20-26% compared to the manual operation of a wall switch alone. In offices where
automatic daylight dimming controls were installed and properly commissioned additional savings of 21%
were obtained over a seven-month period. Galasiu et al. [4] measured a saving of up to 70% in an open
plan office with the combined use of an occupancy sensor, photosensor and a personal dimming control.
Had the various controls been installed individually they would have saved, on average: occupancy
sensors 35%, photosensors 20% and individual dimming controls 11%. Automatic and manual controls in
private offices in a study by Morrow et al. [5] saved 61% of the energy over a 10 hour schedule; 46% of
the savings were due to motion sensors and automatic dimming while the remaining 15% was due to
occupant switching and dimming.

2. Methodology
Data was collected continuously for 94 days from mid-April to mid-July, 2008 from seven buildings on a
college campus located in southern Ontario, Canada; up to 49 conventional working days were used in
the final analysis for individual ballasts. The buildings contained classrooms, both private and shared
offices, circulation areas and mixed-use areas. In total the campus had 2295 addressable luminaires with
dimmable ballasts that were monitored and controlled centrally. Luminaires were controlled locally by a
number of controls including occupancy sensors, photosensors, and manual wall switches (with dimmers).
A number of classrooms and offices also had the additional option of preset scene selection. The central
control software recorded every change of state of each dimmable ballast. The system had been in
operation for approximately one year before the commencement of the study. Ambient lighting was
provided by ceiling-recessed, lensed direct-indirect luminaires with two 3500K T5 1220 mm (4 ft.)
fluorescent lamps. In some spaces the lamps were 28 W, and in others they were 54 W HO (high output).
    National Research Council Canada – Institute for Research in Construction
The power draw of each ballast at any given time was determined from the recorded dimming values,
using dimmer level vs. power draw curves supplied by the lighting system manufacturer.

In this paper we focus on rooms containing ballasts controlled by a photosensor, there were 53 such
rooms with a total of 62 photosensor-controlled ballasts. The 53 individual rooms could be broken down
into 3 main categories: classrooms (36), corridors (9) and offices (8).

To calculate the savings due to the various controls, the operating schedule and baseline wattage for
each room were required to determine a baseline energy use. The operating schedule was determined
for each individual room by overlaying the daily power profile for all normal days in the measurement
period on the same graph and determining a start and end time that encompassed the room’s general
power profile (Figure 1). The start time was selected as the earliest time (to the nearest half hour) that the
lights were usually turned on in the morning (left hand side of Figure 1). Similarly, the end time was the
latest time that the lights were turned off towards the end of the day (right hand side of Figure 1).
Occasionally there were days when the lights were operated outside of these times; the usages were
contributed by either security walk-throughs, cleaning staff, or work being performed out of normal hours;
unusual active periods such as these were ignored in the analysis. The baseline maximum wattage,
WattsBaseline was the maximum wattage that the luminaire drew during periods with no daylight, which was
typically 80% of full output (as chosen by the host site). These values were used to calculate the
theoretical daily energy use over the course of the schedule time (equation 1).
          Energy Baseline = ( End time − Start time ) × Watts Baseline                                      (1)
This choice of baseline strikes a balance between the conservative option of treating each day for each
room separately, and the liberal option of developing a schedule for each room type that encompassed all
common start and end times for all rooms in that room type.

Figure 1. Determination of the start and end times of the schedule for each room. A suitable time was selected that
encompassed the first switch on times and the last switch off times.

Savings by each of the control systems was determined from a specially written program to analyse the
dataset of over 5 million records. Figure 2 shows an example of a ballast in a classroom, illustrating the
method used to determine the various savings. The black solid line is the actual power drawn by the
luminaire. The green dashed line represents the theoretical baseline operation of the luminaire if centrally
controlled on and off at the start and end of the schedule, and operating at the baseline wattage.

Due to the on/off nature of both wall switches and occupant sensors and the information we had from the
system, we were not able to unequivocally separate savings from these two sources. Savings due to the
occupancy sensor/wall switch are the blue (dark grey) shaded areas (determined using equation 2), and
were considered to accrue whenever the power value was equal to the minimum (1.8 W in this case)
during the baseline schedule. The occupant sensor/wall switch therefore saved the difference between
maximum and minimum power draw of the ballast. The daily saving was the integration over the schedule
                                                          Area Blue
        Occupant Sensor/Wall Switch Savings(%) =                        ×100                                 (2)

Figure 2. The power consumed by an example light fixture on a particular day. The green dashed line is the
maximum use based on the standard schedule for that room. The black line shows the actual use on that day. The
blue (dark grey) areas are the energy saved due to the occupant sensor or wall switch. The orange (medium grey)
area shows the additional savings due to using a preset scene. The yellow (light grey) shaded area highlights the
additional savings due to a photosensor.

Additional savings were made via the dimming of the luminaires. All rooms considered in this paper
dimmed the lights in response to available daylight entering the room via a photosensor. In classrooms
and some offices users also had the option of using preset lighting scenes to manually dim the lights
below their normal maximum. To determine only photosensor savings on each of the photosensor-
controlled ballasts, it was necessary to match each of the photosensor-controlled ballasts to other ballasts
in the same zone. Once matched, the additional daily scene savings from dimming the lights using the
preset scenes (equation 3) could be subtracted from the total dimming savings on the photosensor-
controlled ballast (equation 4).
         Scene Savings (%) =                  × 100                                                      (3)

                              AreaYellow + AreaOrange
        Total Savings (%) =                         × 100                                                 (4)
To determine either the daily total additional savings due to all forms of dimming, or the scene setting only
savings, the following criteria was followed. If the power draw on a per minute interval was greater than
the minimum (lamp on) and less than the maximum wattage then the savings will be due to some form of
dimming. Savings on a per minute interval was the difference between the maximum power draw and the
actual ballast draw. Integration of the savings over the scheduled period gave the daily savings.
Photosensor only savings for a zone with preset scenes being used was calculated by subtracting the
daily scene setting savings as calculated from other ballasts in the zone from the daily total savings as
calculated on the photosensor-controlled ballast. Preset scene usage in Figure 2 is characterised by the
static dimmed value and the savings are highlighted as the orange area. Dimming due to a photosensor
is characterised by a fluctuating curve, with yellow areas denote the savings.
3. Results
A brief description of the rooms from which data was obtained is in Table 1.

Table 1. Rooms and zones analyzed. Orientation is the direction that the window faces. ‘SKY’ - skylight installed, ‘-‘
- direction was unable to be determined, A and B - correspond to the same room that had two photosensors installed
(one of the photosensors was influenced by a skylight and the other by a window, differentiation between the two was
not possible). Asterisks indicate zones in which the photosensor-controlled ballast could not be matched with other
ballasts in the zone, and therefore a separation of scene savings and photosensor savings was not possible.
     Zone                                     Max Power       Ballasts per      Ballasts      Start     Finish
    Number     Room Type Orientation           Draw (W)          Room          Controlled     Time      Time
       1       Classroom      N                  44.6               9              2          7:00      19:00
       2       Classroom      N                  46.7              12              1          7:00      19:00
       3       Classroom      S                  42.5               5              1          7:00      19:00
       4       Classroom      W                  46.7               4              1          7:00      23:00
       5       Classroom     NW                  46.7               6              2          7:00      21:00
       6       Classroom      N                  46.7               9              2          7:00      21:00
       7       Classroom      N                  46.7               9              3          7:00      23:00
       8       Classroom      N                  46.7               9              1          7:00      23:00
       9       Classroom     SW                  47.8               6              1          7:00      23:00
      10*      Classroom    SKY                  47.8              25              4          7:00      23:00
      11*      Classroom    SKY                  47.8                                         7:00      23:00
      12       Classroom      N                  44.6              19                5        7:00      23:00
      13       Classroom      E                  44.6              12                1        7:00      23:00
      14       Classroom      E                  47.8              12                2        7:00      18:00
      15       Classroom     SW                  44.6               9                1        7:00      20:00
      16       Classroom     SW                  46.7               9                1        7:00      23:00
      17*      Classroom      B                  29.0              22                7        7:00      19:00
      18*      Classroom      A                  53.9                                7        7:00      19:00
      19       Classroom      E                  47.8              12                1        7:00      19:00
      20       Classroom      E                  38.7              11                1        7:00      19:00
      21       Classroom      E                  44.6              12                1        7:00      18:00
      22*      Classroom     SW                  47.8              18                1        7:00      18:00
      23*      Classroom     SW                  41.5                                1        6:30      19:00
      24*      Classroom     SW                  41.5                                1        6:30      19:00
      25*      Classroom     SW                  46.7                                1        6:30       0:00
      26*      Classroom     SW                  36.9                                1        6:30       0:00
      27       Classroom      W                  41.5              12                1        7:00      17:00
      28       Classroom      W                  46.7              12                1        7:00      19:00
      29       Classroom      W                  46.7              12                1        7:00      20:00
      30       Classroom      W                  46.7              7                 1        7:00      16:00
      31       Classroom      N                  32.8              12                1        7:00      21:00
      32*      Classroom     NE                  59.5              12                1        7:00      20:00
      33       Classroom      E                  40.6              12                1        7:00      23:00
      34       Classroom      E                  36.9              12                1        5:00      19:00
      35       Classroom      E                  40.6              12                1        7:00      19:00
      36       Classroom     SE                  47.8              12                1        7:00      20:00
      37*      Classroom      N                  42.5              12                2        7:00      19:00
      38*      Classroom      N                  36.9                                1        7:00      19:00
      39*      Classroom     NE                  43.5              12                1        7:00      19:00
      40*      Classroom     NE                  43.5                                1        7:00      19:00
      41       Classroom      E                  42.5              12                1        7:00      21:00
      42       Classroom      E                  42.5                                1        7:00      21:00
      43       Classroom      E                  46.7              12                1        7:00      22:00
      44       Classroom      E                  43.5              12                1        7:00      22:00
      45       Classroom     SE                  46.7              12                1        7:00      22:00
       46       Corridor         -           59.5             1              1        24 hr operation
       47       Corridor         -           41.5             2              1        24 hr operation
       48       Corridor         -           47.8             1              1        24 hr operation
       49       Corridor         -           47.8             2              1        24 hr operation
       50       Corridor         -           52.6             2              1        24 hr operation
       51       Corridor         -           47.8             2              1        24 hr operation
       52       Corridor         -           47.8             2              1        24 hr operation
       53       Corridor         -           36.9             2              1        24 hr operation
       54       Corridor         -           52.6             2              1        24 hr operation
       55*       Office          S           51.4             2              1        7:00    19:00
       56*       Office          N           46.7             1              1        7:00    23:00
       57*       Office          -           42.5             10             1        24 hr operation
       58*       Office          S           95.6             7              1        7:00    19:00
       59*       Office         SW           53.9              6             1        7:00    19:00
       60*       Office          E           47.8             6              1        7:00    20:00
       61*       Office          E           41.5             6              1        7:00    20:00
       62*       Office          S           94.1             4              1        7:00    20:00

3.1 Occupant sensor and wall switch savings
Average savings across all room types for the combination of occupant sensors and manual wall switches
was 52.9% (standard deviation 32.8%) over 49 days. Investigating the three room types separately:
classrooms (36 rooms, 45 separate ballasts) had the largest savings in energy, saving 60.3% (s.d.
30.3%). This is as expected with their potentially long daily schedule (evening classes were common) but
sporadic use during this schedule; average actual occupancy (period for which lighting was actually in
use) was 5.0 hr (s.d. 4.3 hr) over an average 13.5 hr (s.d. 2.1 hr) daily baseline schedule. As expected,
corridors (n = 9) exhibited the least savings, 23.5% (s.d. 25.0%), due to their frequent use. On average,
corridor lighting was on for up to 18.1 hr (s.d. 6.2 hr) over a 24 hr period. Offices (n = 8) on average
saved 44.2% (s.d. 33.2%). Office schedules and savings were more consistent, with average occupancy
of 7.6 hr (s.d 5.0 hr) per day in a possible 14.4 hr (s.d. 4.4 hr) schedule. The long schedule observed in
the office classification of rooms was likely due in part to luminaires in open offices or in general
circulation/use areas of the offices; for example, reception areas and printing rooms were required to be
on for longer periods due to the later finish time of some classes.

Figure 3 displays the total average daily energy savings due to an occupancy sensor and wall switch
combined for each control zone. The zone number corresponds to the number in column 1 of Table 1.
The error bars show the standard deviation using daily values from the use of occupancy sensors and
manual wall switches combined. To gain a better understanding of the source of the savings we
subdivided the total savings into three bins, as shown by the stacked bars in Figure 3: 1) savings between
the start time in the baseline schedule and the first time the lights were switched on (this includes cases
when the lights were off all day), 2) savings between the last time the lights were switched off and the end
time in the baseline schedule, and 3) any savings between these two times. The greatest savings were
due to waiting for the initial switch on for the day: average across all room types 28.3% (s.d. 34.1%). The
average savings due to the final switching of the lights off at the end of the day across all room types was
13.7% (s.d. 19.3%), with the average savings made in the middle of the period of the schedule was 10.8%
(s.d. 16.4%).

As stated earlier, with the information we have we cannot unequivocally attribute these savings to
occupancy sensors or manual use of the wall switches. Nevertheless, it is probably reasonable to
assume that the start of day savings are most likely attributed to the wall switches. There is a greater
likelihood that some of the savings during the middle of the day and at the end of the day were due to
occupancy sensors, though whether it was a majority or minority of the savings is unknown. Interestingly,
Pigg et al. [6] conducted a field study in private offices that had both wall switches and occupancy
sensors, and concluded that manual switch use was reduced when occupants knew that an automatic
system to switch off lights was present.
 Average Savings due to Occupancy
                                     100                                             Classrooms                        Corridors     Offices

      Sensor/Wall Switch (%)

                                                       1                11             21            31          41            51           61
                                                                                  Zone Number             Switch off   Other        Switch on
Figure 3. Average energy saving per ballast obtained using an occupant sensor and manual wall switch. The energy
savings are subdivided into three bins: 1) savings between the start time in the baseline schedule and the first time
the lights were switched on (“Switch on”); 2) savings between the last time the lights were switched off and the end
time in the baseline schedule (“Switch off”); and, 3) any savings between these two times (“Other”).

3.2 Additional dimming savings
Additional energy savings were also realised by dimming the luminaires either by daylight detection with a
photosensor or manual dimming using preset scenes. Classrooms were the predominant room type
where preset scene selection was available. Scene setting was also available in some offices but it was
not possible to separate scene savings from photosensor savings; ballasts in corridors were assumed to
not have preset scene functions available for use by the occupants. For classrooms, ballasts in each of
the zones that were not controlled by a photosensor were analysed to determine the additional savings
made from dimming the lights via scene setting. On average the use of preset scenes in classrooms
saved an additional 4.1% (s.d. 8.2%) of the daily baseline energy, Figure 4 shows these savings by zone.
A solid symbol is used for zones where the photosensor-controlled ballast could be uniquely matched to a
non-photosensor ballast in the same room. An open sensor is used when unique matching was not
possible, and a room average estimate is provided.

                                                                    Scene Savings - Matched Zone
                                Savings due to Scene
                                 Average Additional

                                                                    Scene Savings - Unmatched Zone
                                    Selection (%)




                                                         20     0  25   5     3010       35 15    40        45
                                                       Zone Number
Figure 4. The average additional savings made from the use of selecting different preset scenes in classrooms.

Figure 5, shows the additional savings due to photosensors, across all room types. For ballasts in
classrooms that were able to be matched to a particular zone the additional savings due to use of preset
scene selection has been subtracted from the total savings (solid symbol). In the remaining classrooms
the savings due to scene setting could not be uniquely determined, and therefore the savings shown are
for photosensors and scene setting combined (open symbols). Error bars show the standard deviation
across days. By room type, average additional savings were classrooms (photosensor only, n = 1523
analysis days) 9.8% (s.d. 11.2%); classrooms (photosensor + scene, n = 700 analysis days) 16.4% (s.d.
15.7%); corridors (n = 450 analysis days) 25.1% (s.d. 10.3%); and offices (n = 400 analysis days) 21.7%
(s.d. 17.6%). A small portion of the savings in offices may be attributed to the use of preset scene
selection. Note that these savings only accrue when the lights are on, and are therefore in addition to the
savings provided by occupancy sensors and wall switches.

                                               Classroom - Photosensor Only
                                               Classroom - Photosensor + Scene
  Average Additional Savings due to

                                      80       Office
          Photosensor (%)








                                           0        10            20             30   40   50           60
                                                      Zone Number
Figure 5. Average energy savings per ballast through the use of a photosensor. Error bars show standard deviation.

To extrapolate the savings per zone into savings per room, one must consider how many ballasts in the
room were controlled by each control type. Typically, all ballasts may be controlled by an occupancy
sensor/wall switch, but only those near windows may be controlled by a photosensor. At the study site, a
system reset had resulted in the loss of photosensor zoning, such that often only one ballast per room
was controlled by the photosensor. In rooms with multiple ballasts this meant that room-level savings due
to photosensors were drastically reduced from what they could have been. In this extreme case this
meant that photosensors only saved on average 1.7% (s.d. 1.6%) of the total lighting energy in
classrooms and 4.7% (s.d. 3.2%) in offices. This emphasizes the importance of matching controls to
space types, and on-going maintenance. This low total savings per room would have a dramatic impact
on the payback period of the use of such technologies. A more effective use of the output of the
photosensor in controlling several fixtures near the window would allow for more energy to be saved on a
per room basis.

Savings expressed as a percentage are highly dependent on the baseline chosen for comparison, which
was not straightforward in a campus setting. To aid in interpretation, Table 2 summarizes the average
energy savings from each of the control types in percentage terms, relative to the average baseline daily
schedule, for each room type. Table 2 also shows the average number of hours per day the lighting
system was actually in use (i.e. scheduled hours minus the savings due to the occupancy sensors and
wall switches). Savings due to an occupancy sensor/wall switch may only be referenced to the scheduled
hours, but the photosensor and scene savings may also be referenced to the actual hours of use. For
example, the average energy saving for a ballast connected to a photosensor in a classroom was 9.8%.
This might sound quite small, but it is relative to an average baseline schedule of 13.5 hr. This means
that the photosensor saved the equivalent of 1.3 hr (13.5 x 0.098) of full ballast output per day, or 26%
(1.3/5.0 x 100) of full output during the period the lighting was in use.
Table 2. The average energy savings from each of the control types in percentage terms, relative to the average
baseline daily schedule, for each room type. Also shown is the average number of hours per day the lighting system
was actually in use
Room Type         Occupancy sensor/wall switch      Dimming Savings, %       Schedule                       Hours
                         savings, % (s.d.)                  (s.d.)             hours,                       used,
                  Switch      Other        Switch    Photo         Scene      hr (s.d.)                    hr (s.d.)
                    on                       off    sensor         setting
Classroom          33.6        16.0         10.7      9.8*          4.0*         13.5                         5.0
                  (35.5)      (20.9)       (16.2)    (11.2)         (8.4)        (2.1)                       (4.3)
                                                     16.4**         4.3**
                                                     (15.7)         (6.5)
Corridor          5.9         4.3         13.3      25.1***           -          24.0                        18.1
                (10.6)       (7.6)       (16.0)      (10.3)                      (0.0)                       (6.2)
Office           24.0        11.9         8.7       21.7***           -          14.4                         7.7
                (33.1)      (15.7)       (17.4)      (17.6)                      (4.1)                       (5.0)
* Matched Zones, ** Unmatched zones, *** Photosensor and Scene setting combined.

4. Conclusion
The college campus had a wide range of room types, operational hours and uses. As a result, the energy
saved due to the control types present varied greatly depending upon room type and occupancy. Rooms
that were occupied sporadically during the day, such as classrooms, benefited the greatest from the
installation of manual wall switches and occupancy sensors. Individual lighting fixtures in rooms that were
on for large portions of the day (e.g. offices) had some of the largest potential to save energy via the use
of a photosensor.

Despite the detailed considerations described above, in simple terms this study demonstrates that
substantial energy savings were realized with multiple controls in a field setting over many days and room
types. Further, the host site was very pleased with the performance of the system and, using their own
cost-benefit analysis, viewed the lighting control system as a very good investment.

This research was sponsored by the National Research Council Canada (NRC), the Program on Energy
Research & Development, and Public Works & Government Services Canada (PWGSC), and by the
Lighting Research Office, Electric Power Research Institute (EPRI) with the support of the J.H. McClung
Lighting Research Foundation. Of the many generous people who facilitated our work, we are particularly
grateful to: Jeff Choma, Sandy Hancock, Rob Mongrain, Mat Lefler (Georgian College); Chris Rigatti,
Chris Ballard (Lutron Electronics Co. Inc.); Phil Santia (e-Lumen International Inc.); We also thank all of
the building occupants, who accepted our presence in their work/study places.

1. National Resources Canada. Commercial/Institutional Secondary Energy Use by Energy Source, End-Use and
    Activity       Type.             2009             [cited      6th       April   2009];     Available         from:
2. Canada Green Building Council, (2004) LEED green building rating system: reference package for new
    construction & major renovations: LEED Canada-NC version 1.0. Version 1.0 ed., Ottawa: Canada Green
    Building Council.
3. Jennings, J.D., et al. (2000) Comparison of control options in private offices in an advance lighting controls
    testbed. Journal of the Illuminating Engineering Society, 29(2), p. 39-60.
4. Galasiu, A.D., et al. (2007) Energy saving lighting control systems for open-plan offices: a field study. Leukos,
    4(1), p. 7-29.
5. Morrow, W., et al., High performance lighting controls in private offices: A field study of user behaviour and
    preference, in World Workplace. 1998: Chicago, Illinois.
6. Pigg, S., M. Eilers, and J. Reed. (1996) Behavioral aspects of lighting and occupancy sensors in private offices: a
    case study of a university building. ACEEE Summer Study on Energy Efficiency in Buildings, 8, p. 161-170.

To top