Final Report: Social Feasibility of Energy-Efficiency Retrofits and Educational Campaigns for Sustainable Energy Use in Pre-existing College Residence Halls

EPA Grant Number: SU833515
Title: Social Feasibility of Energy-Efficiency Retrofits and Educational Campaigns for Sustainable Energy Use in Pre-existing College Residence Halls
Investigators: Neff, Rob , McConnell, Virginia D.
Institution: University of Maryland - Baltimore
EPA Project Officer: Nolt-Helms, Cynthia
Phase: I
Project Period: June 1, 2008 through May 30, 2008
Project Amount: $10,000
RFA: P3 Awards: A National Student Design Competition for Sustainability Focusing on People, Prosperity and the Planet (2007) RFA Text |  Recipients Lists
Research Category: Pollution Prevention/Sustainable Development , P3 Challenge Area - Materials & Chemicals , P3 Awards , Sustainability


On college campuses, there are actions individuals can take to reduce energy use, and there are technology changes that the institutions can make that may or may not be beyond the control of individuals. We examine both types of changes, and attempt to measure the effects on energy use that result from each using an experimental design framework. A group of apartments on the UMBC campus offer a unique opportunity for assessing different policies for reducing energy use because students pay for all energy use in each apartment. Actual energy use can be recorded over time for each apartment and students have the incentive to reduce energy use because they have responsibility for the energy bill. The two policies that have been undertaken this year in the apartment buildings are the installation of programmable thermostats and an education program for students about how they can reduce their energy use.

There are several innovative aspects of this study which fit with the goals of the P3 program of moving toward sustainable development. The first is that the study will be able to quantitatively assess the effects of the policies that are undertaken. Too often, the effects of actions to reduce energy use are based on theoretical or suggested outcomes, and the actual in-use performance is not tested with data from the field. This project will compare the suggested effectiveness of several different types of policies, a technology change and an education campaign, with the actual reductions in energy use before and after the policies are implemented. It is vital that the actual effectiveness of measures to reduce energy use are known. Because resources are limited, the most effective and cost-effective measures need to be undertaken first in order to move toward communities with lower overall energy consumption.

The second innovative aspect of the study is that it uses an experimental design to isolate the effects to the two policies on energy use in the UMBC apartments. We have several apartment buildings that have no changes, several that get programmable thermostats, and several others that get both programmable thermostats and an education campaign that informs students about how to reduce their energy consumption. Because of this design, and the fact that we can read the energy use in apartments over time, we can isolate the effectiveness of each of these policies in terms of their effects on energy used. This will allow us to determine those policies that are most effective at reducing energy, and the results will help inform policies on campus and in the broader community.

The education campaign has two important features. It will support the effective use of the thermostats, because it will emphasize that the thermostats are set at levels to minimize energy use, throughout the day. Students will be educated about why trying to override the thermostats is likely to increase their energy bill and also increase energy use on campus and contribute to greenhouse has emissions. The education component will also describe other actions students can take to adopt a more energy conscientious lifestyle. This educational campaign allows students to learn about energy efficiency, how they may adopt certain methods into their lifestyles to decrease the amount of energy used. We will be able to observe how much additional difference the education campaign makes in reducing energy usage because there will be two groups equivalent in other ways, but one is getting the education component and the other is not. The different educational tools will reveal which are more successful in teaching citizens the importance of energy efficiency.

Throughout the project, the team has worked closely with Residential Life and Facilities Management on campus.

The purpose of the study is to identify the actual effectiveness of different policies to reduce energy use. The study sought to determine what changes residents of the apartment buildings on the UMBC campus could make to reduce the energy use in those apartments. One technology change that had been suggested was to put in programmable thermostats. Programmable thermostats are argued by the Energy Star program of EPA to reduce energy use by automatically reducing temperatures at times when people are away from home and late at night when people are sleeping. In a dormitory setting, they can be programmed to maintain temperature ranges at optimal levels for comfort and for energy efficiency. It has been found that it is very expensive to have large swings in temperatures – turning manual thermostats way down at night, for example, may actually increase energy use because of the large amount of energy required to heat up the rooms to comfortable temperatures in the daytime.

The other policy we examine is an education program for the students in the apartments. Students will not only learn about how the programmable thermostats may reduce their energy bills, but they will also learn about other ways that might change their behavior to reduce energy use. The education component involves distributing a brochure to the students, and there will also be a pizza party in each apartment that gets the education program at which students from our research team will present information about how to reduce energy use, and why it is important. They will also answer questions and take comments about the thermostats. The brochure has been developed and will be distributed in early April. It will be interesting to see if this program has any effect on energy use in the apartments. There is mixed evidence from the literature in economics and in environmental studies about whether energy information campaigns are successful.

The location of the study is a group of 6 apartment buildings behind library on the UMBC campus. At the beginning of the study period, all of the apartments had simple manual thermostats. Electricity use in each of the apartments is monitored separately and our team is able to read the energy use each week from the units. Residents of each apartment are billed separately for energy use.

Summary/Accomplishments (Outputs/Outcomes):

There is a control group, and two test groups.

Control group apartments - 2 apartment buildings, 10 apartments each, 40 student residents in total – no treatment.

Test group apartments 1 – 2 apartment buildings, 10 apartments each, 40 student residents in total. Programmable thermostats installed.

Test group apartments 2 – 2 apartment buildings, 10 apartments each, 40 student residents in total. Programmable thermostats installed, plus separate education program on use of the thermostats and on how students can influence their energy use and GHG emissions. Also on how energy use contributes to greenhouse gas emissions, global warming and sustainability.

Programmable thermostats have been installed in the two test groups described above. The thermostats used were made by Totaline, and they are the P286-1200 Multistage Programmable Thermostats. These thermostats are designed to reduce energy use.

The managers at Residential Life in consultation with the research team and energy experts at Facilities Management set the thermostats the following way to optimize energy savings, but still provide a level of comfort we hope will be acceptable to students. However, determining how students respond is a critical part of this project.

Settings: Monday, Tuesday, Wednesday, Thursday, Friday
6:00 am 68 degrees fan auto
9:00 am 63 degrees fan auto
5:00 pm 68 degrees fan auto
11:00 pm 63 degrees fan auto

Saturday, Sunday
7:00 am 68 degrees fan auto
11:00 pm 63 degrees fan auto

Installation of the thermostats was delayed from the original target date. The original time of the installation of the thermostats was to be over the January break. But EPA was delayed in distributing the money to campus until late in December (December 21, 2008) and the campus did actually make money available to the project managers until January 25, 2008.

Education campaign.
Education component will be administered toTest Group 2 by the students on the research team. This has also been delayed due to delay in EPA funding and also to delays on campus in having money available and thermostats installed. A flyer has now been developed that will be distributed to each student in the apartments in early April. In addition, there will be a pizza party around that time at which students from the research team will present information about how to reduce energy use, and why it is important. They will also answer questions and take comments about the thermostats.

Generally, we asked students about their own level of energy efficiency and if they felt that UMBC should be doing more to decrease the amount of greenhouse gases released into the atmosphere. Specifically, we inquired how much control the students felt they had over the old model of thermostats that were replaced with the new programmables.

Surveys were given to students in the dorms before the thermostats were put in. The survey is designed to assess the students’ use of energy, cooperation in the apartment on energy use, and attitudes about sustainable energy policies.

The distribution of the initial survey has also been delayed for the reasons given above. They have been distributed and collected in March. They were given to students by the R.A.s in the apartments, collected by R.A.s. The P3 team members worked with Residential Life on the UMBC campus to install thermostats, and to make the links with the R.A.s.

First survey: Determine whether all three groups are similar, in terms of

  • Inclination to change thermostat, temp settings, etc.
  • Appliances used in apartments
  • Environmental awareness, concern
  • Who pays the bills, student or third party, such as parent

Post treatment surveys: observe differences in behavior, especially with regard to use of thermostat, comfort in room, and effects of education campaign on Treatment group 2.

Energy Use.
Energy readings will be made at regular intervals in the fall leading up to the installation. Students on the research project team are reading the energy use monitors weekly for each apartment in the study, starting in late February. Energy readings each week will continue through the spring semester. A key element of the project will be to compare energy use across the three groups.

Overview and Time line






Control group

Treatment group 1

Treatment group 2


No changes (manual thermostats)

Programmable thermostats installed

Programmable thermostats installed and education provided

Identification of groups

2 apartment buildings, 20 apartments each

2 apartment building, 20 apartments

2 apartment building, 20 apartments

Timing of treatments

No changes

Thermostats initially projected to be installed in January 2008 Delayed. Actual installation, March 2008.

Thermostats initially projected to be installed in January 2008 Delayed. Actual installation, March 2008. Education in early April 2008 continuing through to final survey.

Energy monitoring

Monthly through the fall of 2007 and January and February 2008. Weekly from late February through April.

Monthly through the fall of 2007 and January and February 2008. Weekly from late February through April.

Monthly through the fall of 2007 and January and February 2008. Weekly from late February through April.

Surveys – before study is started (same for all groups)

Preliminary survey – administered in February 2008

Preliminary survey – administered in February 2008 before any treatment

Preliminary survey – administered in February 2008 before any treatment

Post – after treatments (different for all three groups)

Post survey - control– late April 2008.

Post survey - Treatment 1 late April 2008

Post-survey – Treatment 2 late April 2008

Surveys were distributed to 240 student residents in UMBC’s Terrace Apartments. Of those surveys, 115 were returned to the researchers on this project. The surveys provided information regarding how much time students spend awake in their apartments on any given day, how often the temperature was changed in the apartment over a 24 hour period during the fall and winter months, and the approximate temperature range in the apartment. Students were also questioned about the level of agreement of the temperature in the apartment with their suitemates, the control they felt they had over the heating and cooling of the apartment, and how much of the total apartment energy bill they paid (if at all). By asking the students about the level of difficulty students had paying their energy bills, it was thought that they might be more receptive to adopting energy efficient techniques into their lifestyles in order to reduce the amount paid per month for the energy bill. Other questions requested information regarding the students’ level of energy efficiency as compared to their suitemates, and whether energy efficiency ranked high or low on the apartment occupants’ priorities. Students also answered questions about whether they thought UMBC should be doing more to reduce energy consumption and greenhouse gas emissions, and how often they recycled to determine how environmentally aware they may be.

Many of these questions were asked using a Likert scale of 1 to 5 (lowest to highest) to determine the respondents’ level of agreement to the questions and rank the students accordingly. These answers allowed the researchers to assess the percentage of the student residents that agreed or disagreed to our statements.

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

52% of the residents indicated that they knew what their portion of the energy bill for their apartment was each month. 17% of the residents responded that they did not pay for their energy bill, while 33% said that they paid only a portion, and 50% stated that they paid all of their energy bill each month. 17% of the students felt that they were less energy efficient than their flatmates, while 33% of them stated they were equal in energy efficiency, but the other 50% responded that they were more energy efficient.

Figure 6.

17% of residents responded that they think that UMBC should be doing more to reduce energy consumption and greenhouse gas emissions. 33% felt that UMBC should not be doing more, and 50% were unsure about this.

The measurement energy use to date in the six apartment complexes (60 apartments total) resulted in the graph below. There was a wide range of energy usage in kilowatts. For example, two apartments of 4 residents each used a total 250 kilowatts a week, while one apartment with the same number of residents used over 900 kilowatts. The average amount of energy used was 450 kilowatts, while the median amount was 448 and the mode was 478. This close correlation indicates a normally distributed data set. So while there were extremes like the one listed above, the data was demonstrable of a bell curve.

Figure 7.


The project has already yielded some interesting results. There is a good deal of variation in energy use across the apartments, and students have different perspectives on the importance of energy use, and the way they pay for energy use. The results of the full study where the research team will be able to examine whether the programmable thermostats make a difference in energy use, the response of students to the programmables and the effectiveness of the education campaign will provide important insights about energy reduction policies. Both the approach used here and the insights gained will be useful to broad groups of citizens. We intend to write a journal article on our results, and to present them on campus to a variety of different groups. See the Phase II proposal for more on this.

P3 Phase II Project Description

  • Challenge Definition and Relationship to Phase I
  • Phase II of the project will be an extension of the study of energy use begun in Phase I. We will examine the effects of different policies to reduce energy use including a policy that uses an energy STAR device, programmable thermostats, and an education program that raises student awareness about the amount of energy they use and the effect on greenhouse gas emissions and sustainability. The team of research is made up of students and faculty from the Department of Geography and Environmental Systems and from the Economics Department at UMBC.

    We will use an experimental design as developed in Phase I of the project. The number of apartments will be increased, and an additional way of making student more aware of their own energy use will be included. This specific addition to the project is that one group of students who receive both the thermostats and the education campaign will also get an electricity monitor that can be used at an outlet, and will inform students about the actual energy use. This provides an additional way for students to become aware of their energy use. We will be able to determine if this component of the awareness campaign has an effect on energy use compared to the general education program.

    The analysis in Phase II will address the following issues:

    1. Does the installation of programmable thermostats have any effect on actual energy use in apartments on the UMBC campus? What is the magnitude of the effects across apartments?
    2. Do students accept the temperature levels established by the programmable thermostats? What aspects do they object to, if any?
    3. What is the cost-effectiveness of the programmable thermostats? How many dollars were spent per kwh saved?
    4. Does an education campaign along with the thermostats provide any additional savings in energy use? Do students respond better to lower temperatures in their apartments with the education campaign?
    5. Does the ability by the students to use an individual electric monitor to measure actual electricity use at one of their outlets add to their awareness of the energy use? Does it result in less over all energy use? If so, by how much?
    6. How can these results be generalized to college campuses and to the larger public? Are programmables a cost-effective way to reduce emissions?

    Because the implementation of Phase I was delayed due to funding lags, the effects of the different policies on energy use could not been fully evaluated in Phase I. The basic evaluation method from Phase I will be extended using the same experimental design. The number of apartments will be increased. There are enough apartment buildings on campus to extend the project slightly over Phase I. Each apartment building has 10 apartments, and each apartment has 4 students. Each apartment is metered separately, and students are billed for their energy use in the apartment.

    Numbers of Apartments and Students in Each Assessment Group

    Numbers of Apartments and Students in Each Assessment Group


     Number of apartment buildings

    Number of apartments

    Number of students

    Control group – no treatment




    Treatment group 1 – programmable thermostats

    3 (2 already have thermostats)

    30 (20 already have)


    Treatment group 2 – programmable thermostats and education prog




    Treatment group 3 – programmable thermostats, education program and electricity monitor for each student




    Surveys will be administered to all students, first in the beginning of the academic year, in September of 2008. Then follow-up surveys will be given to the students in March. The surveys will provide information about students energy use, and attitudes and awareness about the impacts of their behavior on the environment. The results of observations on changes in energy use of the different groups and the responses of the surveys will provide information about which policies are most effective, and how policies can be made more effective.

  • Innovation and Technical Merit
  • This study employs an experimental design framework to look at actual energy use in college apartments. The experimental design allows us to compare similar groups, controlling for a single change in one policy that is likely to affect energy use. This approach allows us to accurately measure the energy use resulting from a particular policy. This approach has not often been used in evaluating energy reduction policies.

    Also, this policy measures actual energy use in the apartments for the programmable thermostats. There are engineering estimates of the effectiveness of these devices, but there is no data on their actual effectiveness in use. There is a statement from Energy Star documents that programmable thermostats will save money, by one estimate an average of about $150 year in most homes Exit . How much they actually save in different settings may vary and this actual performance is important to measure. This will be an important aspect for the evaluation of energy reduction policies in the future, for changes to emissions inventories for greenhouse gas reduction policies, and for identifying the most cost-effective controls going forward.

  • Relationship of Challenge to Sustainability
  • Policies to reduce energy use, especially energy use derived from fossil fuels, are critical to reducing greenhouse gases. Reducing greenhouse gases is essential for mitigating the threat of climate change. Climate change is the most important threat to sustainability for the planet today. The UMBC campus, by signing on to the Presidents Climate Commitment Initiative and the state of Maryland, through recent legislation to reduce greenhouse gas emissions are committed to reducing fossil fuel energy use. This project will help identify ways to do reduce those emissions. Technical or more passive fixes such as the thermostats will be integrated with policies that will make those controls more effective by understanding the behavioral responses.

  • Measurable Results (Outputs/Outcomes), Evaluation Method, and Demonstration/Implementation Strategy
  • The results of the study will be measured by direct measurements of energy use in the dorms, and by student surveys. The evaluation approach is to use an experimental design to assess the changes due to a single policy, as described above. There will be results about the use of energy across the different groups of apartments. The changes in energy use will be linked to pre-existing attitudes and preferences on the part of students in the apartments. We will be able to see if students who pay their own bills, or who claim to be environmentally aware, are more likely to accept the programmable thermostats, or respond more to the education campaign. We will also be able to assess whether the electricity monitoring devices given to students in the one education group have the effect of raising environmental awareness, particularly for those who were not previously aware of their energy use and its implications or cost.

  • Integration of P3 Concepts as an Educational Tool

Supplemental Keywords:

RFA, Air, climate change, Air Pollution Effects, Atmosphere