Ben Rickelman & Katie Krause

Abstract:

In this research project we attempted to learn how effective the sensory lighting system in Mazzucheli is. The reason we chose to research this is because there was some doubt as to whether or not the lighting system installed in Mazzucheli is energy efficient, as it is supposed to be. We noticed that the lights tended to turn themselves on and off pretty frequently, which did not seem to be energy efficient, and a lot of the times the lights were on were not appropriate. For instance, it had been observed that the lights were often on in the middle of the night or very early in the morning. To research this, we collected data using a light sensor to determine when the lights in sensor controlled rooms are on. Then we analyzed the data to determine the cost effectiveness of these detection controlled lights to determine their practicality. We found out through data collection that for a large amount of time the lights in the Mazzucheli building are on during off-hours due to the sensory system. This energy equated to over 84 dollars wasted in one month due to this inefficiency. Clearly the energy saving measures implemented in the Mazzucheli building need revision because they are creating more waste instead of cutting down on it.

Introduction:

National Wildlife journal produced an interesting article a few years ago shedding light on the issue of light efficiency. Mark Van Patten wrote the article, and in it he states: “To glimpse the future, there’s no better vantage point than in the nation’s colleges and universities. The values instilled there set the course for succeeding generations…The National Wildlife Federation’s Campus Ecology® program has peered into the future and we like what we see…America’s colleges and universities are already well embarked on an alternate path, one emphasizing conservation and efficiency…Eighty percent of the schools surveyed have already introduced lighting efficiency upgrades to save energy; 20 percent have plans to do so…A deeper trend is also at work…environmental concerns are now integral to the culture and values on America’s campuses. That’s another way of saying that an environmental ethic has taken root at the institutions where tomorrow’s leaders are being trained.”
In another article, “Shedding Light on Efficiency,” Ken Lovorn states: “LEED, ASHRAE 90.1 and soaring energy bills mean designers need to take lighting more seriously in the overall energy savings plan. Utilizing controls schemes is the way to go…to get into real reduced energy usage levels, more sophisticated lighting techniques must be employed. One method is the use of controls.” One of the lighting controls he mentioned is daylighting sensors, which is what Mazzuchelli has installed. This option reduces the level of artificial lighting based on the amount of available day-lighting. It utilizes a lighting level sensor to measure the amount of lighting in an area.
The purpose of our watershed project is to determine whether the lighting in Mazzucheli is effective. There are some obvious defects in the system that we have already observed. The lights in Mazzucheli turn on when the sun is not very bright. If the sun is bright the lights are not supposed to turn on. The problem with this is that the lights are often on when there are no classes in the classrooms. Another problem is that when teachers and students leave the classrooms they do not bother to turn the lights off. In some cases it may be because they think the lights will turn off automatically.
Another problem is that the switches could be confusing to some people because they are not typical light switches and there are as many as eight buttons in some of the classrooms that need to be pushed in order to turn off the lights. Some people just forget to push them all so they will stay on all day and night. This is definitely a waste of energy. Already we are beginning to believe that different lighting systems in the rooms would have been a better choice.
To determine if the automatic lighting system was efficient we set up an experiment using a hobo light sensor. These helped us to determine when the lights were on or off that were controlled by the day light sensor system. Analyzing this data helped us determine if these sensors were the best choice for Edgewood. Looking at the data provided by the light sensor will enable us to determine if our hypothesis that the Mazzucheli lighting system is inefficient is correct.

Methods:

We monitored a light in the plant room that seems to turn on by itself during the night. The plant room is a small green house like room that faces Lake Wingra. The room receives a lot of sun light and supposedly should be an ideal place for the lighting system. To record this apparent waste of energy we taped a Hobo P-light monitor to the side of the light to collect the data of what time it was on and off during the next few days. The monitor was shielded by a plastic cover and aluminum foil to block out sunlight that would contain our data. We then plugged the Hobo P-light collector into the computer and extracted the data with the use of the Boxcar program. This allowed us to analyze the patterns currently present with the lights in the plant room.
We determined how much power and thus, money is wasted by the lights not being used effectively in the Mazzucheli building by finding out the energy consumption of the building and how much it costs to run currently. We then repeated the light monitoring process with the Hobo in another room in the Mazzucheli building. The room we used was a typical classroom facing Lake Wingra on the second floor.
After we collected this data we analyzed it to determine how much the light sensors are causing the lights to stay on for unnecessary amounts of time. This was done by looking at graphs provided by the Hobo that illustrated when the lights in the rooms were on or off. A time period was then established when it would be highly unlikely that the lights in the room would be on for any practical purpose. There were several distinct patterns noticed such as that the lights stayed on for long hours during the nighttime routinely. This could not be attributed to human forgetfulness due to the fact that the lights would turn off for short periods of time during the night but then turn on again. These time periods of the light being detected on during off hours by the Hobo were used to determine when energy was being wasted for a set amount of time in the plant room.
During the 30 day trial we set a 12 hour time period from 6pm to 6am as a time when the light should not be on. By looking at the printout of the Hobo light collector for the plant room it was determined that of the 360 hours of the 12 hour period from 6 pm to 6am the lights were on 234 of the possible 360 hours or 65% of the time. There are 512 watts burned per hour in the plant room so this means there was 119,808 watts burned during this 234 hour period. This equates too approximately 120 kwh or over the 30 day time period 4 kwh/day wasted in the plant room.
This same process was repeated in a more abstract form concerning the 4 other classrooms in the Mazzucheli building. It was also noticed that these classrooms lights were routinely on all during the night too. We didn’t actually have time to collect the data from those classrooms using the Hobo. So we used the 65% marker found during the 6 pm to 6 am window as a rule to get an estimate of how many the lights were on in the rest of the classrooms. Though this data is not concrete like the plant room it gives you a much larger picture of the problem when the entire building is looked at.

Results:

Data from the Hobo light sensor that determines if the lights are on or off in the plant room of the Mazzucheli building. If the line is at the value of 1 it is represented by being at the top of the graph when the lights in the room are on. If the value is at 0 the line is represented by being at the bottom of the graph when the lights in the plant room are off. The data using the Hobo sensor collected from December 06 to the 22nd shows several distinct patterns of the light sensor system in Mazzucheli.

1. The lights in the plant room are actually on more at night than during the day in the plant room
2. Some nights the plant room lights turn on at midnight and stay on till around 8 am
3. There are also multiple examples of how the lights turn on for various intervals in late afternoon and kick back on around midnight.

Looking at the measurements provided by the Hobo light sensor we were able to see on a graph when the lights in the plant room were on. We set a time period of 6 pm to 6 am of when it would be unnecessary for the lights to be on because no one would be using the room during this time. There is a small chance that there is someone using the plant room after 6 pm which presents a small margin of error, but there needs to be one continuous time period we use for the entire project so the data is consistent. Each light in the room controlled by the light sensor burns 32 watts. There are 16 lights in the room so this means the light sensor controls 512 watts in the plant room. We counted up the hours shown on the Hobo lights in plant room graphs to determine the amount of watts burned unnecessarily due to the light sensors. During this 30 day trial measuring a window of 12 hours each day, it was determined that the lights in the plant room were on 234 hours of a possible 360. This means the lights in the plant room were on 65% of the time during the time span of 6 pm till 6am. This means there was a total of 119,808 watts per hour burned unnecessarily during this period (234 hours * 512 watts per hour). This means that approximately 120 kwh was consumed in the plant room during this 30 day trial due to a faulty lighting system. This in turn equates to 4 kwh/day being burned unnecessarily in the plant room.

A bill of Mazzucheli’s energy consumption helps illustrate the amount of money saved when the light sensor problem was fixed during winter break. Looking at the energy bill for the period of 12/15/2004 to 1/17/2005 helps illustrate the amount of energy used in the Mazzucheli building. This also helps demonstrate how much money could be saved with an effective system.

 Mazzucheli MG&E Bill

As noted before the plant room burned 4 kw/h per day. In the billing period of 1/17/2005 (Table 1) this was approximately 1% of the kWh/day in the Mazzucheli building. This was arrived at by dividing 4 kWh/day by the amount for the entire building which is 425.5. It should be taken into account that the plant room is a smaller room of the Mazzucheli building and that it uses two thirds of the amount of lights in a typical classroom controlled by sensory system. There are 24, 32 watt lights in the 4 other Mazzucheli classrooms that are controlled by the lighting system. This means there would be 32 watts times 24 lights per 4 classrooms to equate out to 3072 watts per hour for the rest of the sensory light system.

Though we did not record the amount these lights were malfunctioning within a certain time window it would still be helpful to use an approximate number to figure out a cost. Using the 234/360 hour mark found in the plant room as in inefficiency marker as an indicator we could get a general number on the amount of electricity wasted in the rest of the classrooms due to the sensory system. 3072 watts * 234 hours equals approximately 719 kwh or 24 kwh/per day for the rest of classrooms using sensory light inefficiently. This 28 kwh/per day wasted together combined with the plant room means each month Mazzucheli wastes that much energy due to the lighting system per day.

Unnecessary consumption of energy due to lighting system

Discussion:

In analyzing the data we have collected in this project we concluded that Delighting sensors the Midwest Sustainable Collaborative company installed were in fact not only inept in being energy conservation devices, but in fact exacerbated the situation of wasteful use of lighting in the recently renovated Mazzucheli building. The sensors installed in the building claimed to be able cut energy use, for lights by 70%, saving 14,000 kWh and $900 each year. Instead they burned more energy due to their tendency to stay on during the night and turn on and off at random points during times no one was in the building, which illustrates the inefficiency with these supposedly long term cost saving devices. The 28 KWh/day (Table 2) used unnecessarily discounts Midwest Sustainable Collaborative claim that this is an energy efficient device.
There were several noticeable trends that make the use of these Day lighting devices troubling. Lights in the Mazzucheli building would turn on around 10-12pm at night when generally no one is using or cleaning the building. They would then proceed to stay on for a few to several hours during the night. The supposed cost saving effectiveness of these devices which are claimed by their installer Midwest Sustainable Collaborative to pay themselves back in seven years is seriously questioned. As shown in Graph 1 a sizable percentage of the energy used in the building was in fact squandered by the lighting system. In fact it would seem normal lights with no sensors would be more effective if applied appropriately, and would be more cost efficient in the long run since they would not turn themselves on during late night hours.
In fact illustrated by Table 2 and Graph 2 the monetary impact of this lighting system builds up very rapidly. Though it might not seem like such a waste when looked at for just a day over time this wasteful use of the lights adds up. This proves our hypothesis that we predicted that the lighting system was inefficient and ineffective is correct. Though we had no idea at what degree or cost the system was inefficient. We found examining the data present in Table 2 conclude that in fact the system is hindrance to energy conservation instead of enabling it.
In light of the ineffectiveness of the system there should be a revamping of the lighting system in Mazzucheli with different conservation methods implemented. Depending on human ability to remember to turn off the lights instead of an automated system seems the best course of action. If there were signs reminding people to conserve and please turn lights off there would be much less energy wasted then the current system. Though this system is not perfect it has to be more effective than large amounts of lights turning on and off throughout the night which is happening with the lighting system.
There are several problems with our study that need to be addressed. First we did not get a large enough sample of more rooms in Mazzucheli being observed with the Hobo light sensor. More data from different areas would have helped strengthen our conclusion and analysis. This happened due to lack of time and trying to figure out exactly how to approach our hypothesis with little to no knowledge in the field of energy conservation. Another problem that presented us is that we have only one month’s worth of data of Mazzucheli’s energy consumption after the Day lighting problem was fixed during winter break. It would be more reliable and concrete if we could compare several months of before and after the fix to illustrate the contrast in energy consumption.
Energy conservation is a goal we should strive for with everything we do at Edgewood. Besides the monetary gain of using energy more wisely there is the ideal of maintaining a sustainable environment. We consume a lot of energy each day and our natural environment is struggling to cope with the demand we put on it in both the short and long term. A sensible approach to conservation should be taken that makes a positive impact instead of a frivolous and ineffective system.

Works Cited

1) Shedding Light on Efficiency. Lovorn, Ken. Consulting-Specifying
Engineer. Jul 2003, Vol. 34 Issue 1, p38

2) The Future Is Now. Van Patten, Mark. National Wildlife. Aug/Sep 2001, Vol. 39
Issue 5, p9

top