MIT owns 140 buildings, which fall into the three categories of academic, residential, and service.1 To determine how much energy was wasted after business hours six academic buildings were surveyed. Each of the surveyed buildings was reviewed twice; once on a weeknight and another during the weekend. Rooms, including service closets, were checked for visible signs of power usage. Specifically, computer monitors which were left on, full overhead lighting, and partial overhead lighting. A room with full overhead lighting on, and not obviously occupied was counted as on. A room with partial lighting or no lighting and a monitor which was on were counted as 15% on. All told, some 900 plus rooms were visited.
| Building | Date | Day | On | Occ. | Off | Part. | Total | % On |
|---|---|---|---|---|---|---|---|---|
| 3 | 2002-11-23 | S | 33 | 9 | 113 | 3 | 158 | 21% |
| 3 | 2002-12-10 23:30 | T | 38 | 10 | 98 | 14 | 160 | 25% |
| 14N | 2002-11-23 20:00 | S | 3 | 1 | 103 | 2 | 109 | 3% |
| 14N | 2002-12-10 23:10 | T | 3 | 2 | 92 | 2 | 99 | 3% |
| 16 | 2002-11-23 19:00 | S | 19 | 3 | 82 | 12 | 116 | 18% |
| 16 | 2002-12-10 23:00 | T | 26 | 1 | 99 | 16 | 142 | 20% |
| 34 | 2002-11-23 | S | 5 | 0 | 8 | 3 | 16 | 34% |
| 34 | 2002-12-10 23:20 | T | 8 | 2 | 7 | 1 | 18 | 45% |
| 35 | 2002-12-01 21:00 | S | 19 | 4 | 83 | 7 | 113 | 18% |
| 35 | 2002-12-10 23:45 | T | 17 | 6 | 75 | 6 | 104 | 17% |
| 54 | 2002-11-26 21:00 | T | 14 | 19 | 181 | 2 | 216 | 7% |
| 54 | 2002-12-06 20:00 | F | 26 | 7 | 214 | 8 | 257 | 11% |
| E51 | 2002-11-26 22:00 | T | 26 | 8 | 60 | 6 | 100 | 27% |
| E51 | 2002-12-01 22:00 | S | 25 | 15 | 66 | 12 | 118 | 23% |
| Total | 131 | 44 | 641 | 47 | 910 | 15% | ||
Clearly, pacing up and down halls is not the most accurate means of determining energy usage. There are many sources of error, several of them attributable to the surveyor. Bathrooms were not counted, and some doors may have gone unnoticed. In addition, it is difficult to determine if a room with an opaque door is in use; particularly ones sealed with weather stripping such as those of laser labs. Other errors are inherent in the assumptions made by simply noting light eminating from rooms. Many appliances, such as photocopiers, do not normally generate noticable light and are unaccounted for. Rooms vary in size and function and therefore have different numbers and types of light fixtures and appliances. Finally, a room may even harbor others within.
Despite the numerous sources of error, the survey would seem to be a good measure of public awareness of energy conservation. Specifically, due to the exclusion of bathrooms, photocopiers, etc. the survey measures personal awareness and responsibility. Individuals appear to feel a lack of responsibility for institutional energy use such as bathroom lighting, lighting in Athena clusters, and photocopiers.
There were too few buildings sampled to indicate trends across the campus, although one may speculate about a few. While the limited number of rooms in building 34 is a suspect sample it is probable that other EECS building surrounding building 34 will have similar energy consumption patterns. Buildings 36, 37, and 38 would most likely fall in the orange, or 30-40% on, range. Likewise there are several buildings off of the Infinite Corridor with a similar blend of class rooms, offices, and labs as building 3. Buildings 4, 5, and 6 may also fall in the yellow, or 20-30% on, range.
It is interesting to note that while the percentage of rooms with lights on was slightly lower over the weekend, there was no significant difference between weekdays, weekends, or holidays. The average campus rate of 15% is not perfect, but seemingly respectable. It is also significantly lower than the expected 25 to 40%. However this rate does not include the majority of appliances and the actual rate is probably significantly higher. There is clearly room for improvement, with some regions of campus in more need than others.
Energy conservation solutions fall into two categories, social and technical. The purpose of social solutions are to convince people to think about energy differently, and to be concientious in its use. Technical solutions may either serve as a stop gap, to remove the need for people to think about energy, or to use energy more efficiently when it is used; by a concientious consumer or otherwise.
Innovative approaches are required for social solutions to be effective. Every child in america grows up with parents that hound them to turn lights off when not in use. There was even a broad movement to conserve during the energy crises of the late 20th century; one may still come across stickers from the era on light switches. And yet, 15% of the lights on campus are left on, as well as untold quantities of office and laboratory equipment.
There are four classes of technical solutions to conserve energy with lighting. The first is to remove the switch and have the lighting be controlled automatically, either by a timer or sensors. A typical motion sensor uses 484 kJ of electricty per week.2 If this motion sensor were to turn off an 85W flourescent, the same amount of light as four 100-watt incadescent lamps, for 93 hours a week, during the weekend and 9 hours each night, it would save an astonishing 28.4 MJ of electricity. A second solution is to switch to an alternative energy source such as wind. If the energy used is clean it should be immaterial if a lamp is left on when not in use. This is essentially true except that lamps have a finite life span; leaving a lamp on would still reduce its life. One might also simply rely upon the sun for lighting in locations where it is not needed after hours, also effectively removing the switch. In many locations, including MIT, there is insufficient direct sunlight much of the year. Direct sunlight could be supplemented by light collected in concentrators and delivered by fiber optics. Finally, one may use more efficient lighting. While energy is still used, the quantity is diminished.
Below is an independently compiled graph of some common lamp technologies. The graph shows the lifetime and lighting efficiency of the lamps. 3
Below is table of most available lamp technolgies compiled from a variety of sources. 4-13 It lists the best values for each attribute of a lamp, although there may be no single lamp possessing all of these values.
| Lamp type | LM/W | $/LM | Min. LM | Lifetime (khour) | Color | Drawbacks |
|---|---|---|---|---|---|---|
| Fluorescent, T8 | 95 | 0.007 | 10 | white | Hg, UV, Lead, asphalt, PCBs(pre-1979) | |
| Incadescent | 20 | 0.0007 | 300 | 1.5 | yellow-white | |
| LED, White | 25 | 0.295 | n/a | 100 | white | Gallium, Silver |
| Mercury | 50 | 0.001 | 5000 | 24 | Blue-white | Hg, UV |
| Metal Halide | 90 | 0.0007 | 3200 | 20 | White | Hg, UV, Halogens |
| Quartz Halogen | 20 | 0.0003 | 200 | 3 | White | UV, Halogens |
| Sodium, HPS | 125 | 0.0004 | 4400 | 24 | Hg, UV | |
| Sodium, LPS | 180 | 4500 | Yellow-gold | |||
| Solar | n/a | n/a | n/a | white | UV | |
| Xenon | 400(15)? |
For comparison laser efficiency appears to be approximately 700 lumens/watt.