University of Pennsylvania Carbon Footprint

Transcription

1 University of Pennsylvania Carbon Footprint October 5, 2007 T C Chan Center for Building Simulation & Energy Studies

2 University of Pennsylvania: Carbon Footprint 2 Table of Contents Table of Contents... 2 Team Members... 3 Acknowledgements... 4 Executive Summary... 5 I Introduction What is a carbon footprint? What is Carbon? Emission Factors Site and Source Emissions What is the Main Campus?...10 II. Carbon Footprint Energy consumption through steam, chilled water, electricity, and natural gas Transportation through University fleets of cars, vans, buses, and trucks Agriculture including fertilizer and agricultural waste Solid Waste disposal Refrigerant replacement Commuter Traffic by car, train, bus, bike, and walking Air Travel by faculty and staff Carbon Offsets Total Carbon Footprint...29 III. Conclusion An Institutional Action Plan...31 APPENDIX A Climate Neutrality Pledge APPENDIX B Campus Buildings APPENDIX C Carbon Calculations Clean Air-Cool Planet (CA-CP) Campus Carbon Calculator...38 Methods and Assumptions...39 Data Sources...41 Input and Output of Carbon Calculator...42

3 University of Pennsylvania: Carbon Footprint 3 Team Members University of Pennsylvania School of Design Faculty William Braham, Ph.D, FAIA Associate Professor of Architecture Ali Malkawi, Ph.D Professor of Architecture Muscoe Martin, AIA Adjunct Professor of Architecture Students Jaime Lee, MS 07 Sean Williams, MArch-MLarp 10 Emily Bernstein, MAch 09 Aroussiak Gabrielian, MAch-MLarp 10 Facilities and Real Estate Services Administration Anne Pappageorge, Vice President Facilities & Real Estate Services David Hollenberg, AIA, University Architect Dan Garofalo, Senior Facilities Planner Khaled Tarabieh, Project Manager

4 University of Pennsylvania: Carbon Footprint 4 Acknowledgements Facilities and Real Estate Services Chris Hanson, Data and Documentation Manager Kris Kealey, Urban Park Manager Carole Mercaldo, Accounting Manager Eric Swanson, Operations Engineer Karen DiMaria, Controller Maurice M. Sampson Taylor Berkowitz, Senior Planner Special Projects Bob Lundgren, Landscape Architect Anthony Sorrentino, Director of External Affairs Mike Coleman, Executive Director Operations Gerry McGillian, Director Technology Trades Business Services Division, Executive Vice President s Office Mary Armata Larry Bell Office of Risk Management and Insurance Ellen Solvibile, Claims Administrato Environmental Health and Radiation Safety James Crumley Kyle Rosatto Institutional Research and Analysis Christine Dougherty Tak Puang University Travel Office Susan Storb South Eastern Pennsylvania Transit Authority (SEPTA) Bharat Gohel

5 University of Pennsylvania: Carbon Footprint 5 Executive Summary This report presents the results of the first Greenhouse Gas Inventory, or Carbon Footprint, for the main campus of the University of Pennsylvania. It fulfills the initial terms of the Association for the Advancement of Sustainability in Higher Education (AASHE), climate neutrality pledge signed by President Amy Gutman in February of A Carbon Footprint is a measure of the impact of human activities on the environment in terms of the amount of green house gases produced, measured in equivalent units of carbon dioxide. As the evidence has Total Campus Emissions by Source grown showing the detrimental impact of green house gas emissions on climate change, so has the desire to reduce our individual and collective emissions. The purpose of this report is to analyze the sources of these emissions at the University. The total carbon footprint for the University of Pennsylvania, including projection to 2020, is shown in Fig. E.1 (see Fig for larger chart). The single largest source of greenhouse gas emissions is the purchased utility energies used for the environmental conditioning and electrical supply of campus buildings, which account for 90% of the carbon footprint. A key finding of this study is that from 1998 through last year (2006) the University has actually met and exceeded its Kyoto Protocol target, the emissions performance metric established at the UN meeting in This significant achievement is due to three different factors: 1. Improved efficiencies in the operations of TriGen, the University s steam supplier 2. New energy management techniques implemented by Facilities and Real Estate Services 3. Wind purchase offsets Trigen s 1997 steam production improvements are responsible for the initial decrease in Penn s carbon emissions. The effect of FRES s utility management techniques implemented in 2001 and the wind purchases begun in 2004, have further reduced the impact of campus growth and increasing energy usage. While the achievement of Kyoto targets is significant, it must be noted that the University s carbon footprint continues to grow annually and will exceed 1990 levels within 5 years, unless reductions are implemented. The University s carbon footprint can be reduced in three basic ways, 1. Efficiencies: reducing current and future fossil fuel energy consumption by buildings and systems. Conservation. changes in consumption habits and patterns 2. Renewables: switching to carbon-free and renewable sources of energy 3. Offsets: purchasing or producing carbon offsets either through TRECs or more direct projects. A balance of all three approaches will be required for Penn to achieve climate neutrality. Metric Tonnes eco2 500, , , , , , , , ,000 50, Year Projection Solid Waste Transportation On-campus Stationary Purchased Steam Purchased Electricity Wind Power Electricity Offset Kyoto Protocol Target Fig. E.1 Total Campus Greenhouse Gas Emissions, See Appendix A 2 The Protocol was not formally adopted in the United States, but had established a target of 7% below 1990 emission levels.

6 University of Pennsylvania: Carbon Footprint 6 I Introduction This report presents the first Greenhouse Gas Inventory, or Carbon Footprint, for the main campus of the University of Pennsylvania. The project grows out of the Sustainability Plan and Campus Audit that the TC Chan Center has been developing since 2005, and fulfills the terms of the climate neutrality pledge signed by President Amy Gutman in February of The pledge was organized by the Association for the Advancement of Sustainability in Higher Education (AASHE), and among its terms, it required a comprehensive inventory of all greenhouse gas emissions and an update of the inventory every other year thereafter, leading to an action plan for becoming climate neutral. This report concludes with a preliminary action plan. The report was commissioned by the office of Facilities and Real Estate Services (FRES) on behalf of the president, and the data was gathered with the cooperation of many sources across the University. 1.1 What is a carbon footprint? The preparation of Greenhouse Gas Inventories has become an increasingly formalized and recognized procedure for evaluating the impact of institutions and their operations on global warming. The purpose of this kind of inventory is to establish goals and identify strategies for the reduction of greenhouse gas emissions. The Kyoto Protocol established a level 7% below the emissions level of 1990 as an initial target for capping emissions, and many of our peer institutions have established reduction targets in relation to that standard, for example 10% or 30% below Kyoto. More recently such targets have been set as a first step to climate neutrality with initiatives such as the 2030 Challenge, and the President s pledge. 4 Climate Neutrality or Net Zero. The university will continue to need and use energy in a variety of forms, so the goal of climate neutrality is to achieve Net Zero climate impact. In simple terms that means radically reducing the consumption of fossil fuel based energies and/or switching to carbon neutral sources of energy, and producing or purchasing carbon offsets until fossil fuel sources can be completely eliminated. The concept and validity of offsets is discussed later in the report, but the trading of carbon credits can only be a tertiary strategy for achieving climate neutrality. This challenging questions raised by such an ambitious goal are how soon can net zero be accomplished, and at what cost? Scopes of Effect. Institutional sources of greenhouse gas emissions are conventionally divided in three different scopes. These distinctions identify operational boundaries for institutions to scope their sources of emissions and to provide accountability for prevention of double counting or conversely, double credits. There are three basic scopes, numbered in degrees of removal from institutional control. Scope 1 includes all direct sources of Greenhouse Gas (GHG) emissions from sources that are owned or controlled by an institution, including: production of electricity, heat, or steam; transportation of materials, products, waste, and community members; and emissions from unintentional leaks. Scope 2 includes indirect GHG emissions from imports of electricity, heat or steam generally those associated with the generation of imported sources of energy. The indirect nature of these emissions makes carbon accounting slightly more complex, though standardized procedures are rapidly being developed. 3 See Appendix A 4

7 University of Pennsylvania: Carbon Footprint 7 Scope 3 - includes all other indirect sources of GHG emissions that may result from the activities of the institution but occur from sources owned or controlled by another entity, such as: business travel, the commuting habits of community members, outsourced activities and contracts, and emissions from waste generated by the institution when the GHG emissions occur at a facility controlled by another company, e.g. methane emissions from landfilled waste. Credits included in Scope 3 also include carbon offsets purchased from other institutions or companies, such as wind or green electricity credits. To assess Penn s carbon footprint, data gathering focused on the following eight categories, going back in each case to Where data was not immediately or completely available, conservative estimating procedures were used (and are documented) to complete the inventory. Scope 1 and 2 1. Energy consumption through the use of steam, chilled water, electricity, and natural gas. 2. Transportation through University fleets of cars, vans, buses, and trucks. 3. Agriculture, including fertilizer and agricultural waste 4. Solid Waste disposal 5. Refrigerant replacement Scope 3 6. Commuter Travel by car, train, bus, bike, and walking. 7. Air Travel by faculty and staff 8. Offsets, such as wind or green electricity credits While the current report has gathered data for all three scopes, but scopes 1 and 2 represent the largest sources of emissions and the ones most directly affected by University policy or action. Scope 3 emissions are significantly more difficult to assess precisely or to certify for any kind of reduction. Also, one institution s scope 3 emissions are another organization s scope 1 or 2 emissions, leading to double counting. The WRI/WBCSD GHG Protocol considers the Scope 3 emissions to be optional when preparing an overall GHG inventory, as do similar protocols such as the U.S. Environmental Protection Agency's Climate Leaders program. A number of our peer institutions have already decided to only target scope 1 & 2 emissions. Nevertheless it remains important to track all sources, keeping in mind that scope 3 emissions like commuting and air travel require a different kind of action planning. 1.2 What is Carbon? Since the Kyoto Protocol of 1990, greenhouse gas emission efforts have focused on the reduction of six atmospheric gases with a recognized greenhouse effect on the global climate: CO 2, CH 4, N 2 O, HFC and PFC, and SF 6. While each of these gases has a different Global Warming Potential (GWP), they are commonly indexed to an equivalent amount of Carbon Dioxide, called eco 2, ecarbon, or simply carbon, so that simple comparisons and evaluations can be made. In common usage and in this report carbon is referred to as the emission to be reduced, though the reduction applies to the whole range of gases. The carbon equivalencies of the different greenhouse gases are shown in the table below, which also compares the 1996 and 2001 values. As the science has developed, the equivalencies have become more precise, but these are estimates used for the purposes of standardizing the accounting of effects.

8 University of Pennsylvania: Carbon Footprint 8 Carbon Equivalents Gas 2001 IPCC GWP 5 Carbon Dioxide 1 Methane, CH 4 23 Nitrous Oxide, N 2 O 296 HFC-23 12,000 HFC-125 3,400 HFC-134a 1,300 HFC-143a 4,300 HFC-152a 120 HFC-227ea 3,500 HFC-236fa 9,400 Perfluoromethane (CF4) 5,700 Perfluoroethane (C2F6) 11,900 Sulfur Hexafluoride (SF6) 22, Emission Factors An emission factor is a normalized measure of the amount of carbon that can be attributed to the consumption of a single unit of energy in a particular process. This can actually vary quite a bit for different fuels and processes, reflecting both the efficiencies of conversion and transmission, and the inherent dirtiness of different fuels. Because of the international nature of the climate agreements, carbon is typically reported in metric tons (MT) of carbon equivalent, which is 1,000 kg or 2,205 lbs. For calculations and analysis, emission factors are normalized to the appropriate units of energy, kwh or mwh for electricity, and MBtu for thermal sources, for example in tons eco2/mbtu or MT eco2/kwh. For comparisons and summaries, this report will convert emission factors to MT eco2/mbtu to make immediate evaluations possible. In some cases, the amount of eco2 per MBTU is sufficiently small, that it will be reported in kilograms, kg eco2/mbtu, which is simply 1/1000 of a ton. The basic emissions factors used in this report are normalized to MBTU and listed in the following table for comparison purposes. 5 Intergovernmental Panel on Climate Change, Climate Change 2001: The Scientific Basis (Cambridge, UK: Cambridge University Press, 2001.

9 University of Pennsylvania: Carbon Footprint 9 Energy Source MT eco2/mbtu site energy kg eco2/mbtu site energy Utilities Electricity PECO Electric Steam (heat) Tri-Gen Gray's Ferry Natural Gas Site Boiler Diesel (Distillate) Oil Generator Transportation Gasoline Car and light Truck Diesel Truck and Bus Jet Fuel Air Travel As mentioned previously, the ecarbon emission factors for direct scope 1 sources are relatively precise, and largely derive from the physics of combustion of different fuels. Indirect, scope 2 emission factors, however involve estimates of the mix of fuels or processes involved in the energy imported through centralized utilities. This is especially complex with purchased electricity that draws from a regional grid that includes multiple power plants, each with unique emissions patterns, and is itself interconnected with other regions. Similar questions occur even with the steam that the University purchases from a single provider, who produces steam in a multi-step process with standby equipment that can substituted or added as needed. This report has used the Clean Air / Cool Planet Campus Carbon Calculator to organize and calculate the emission factors. 6 This tool has been used at many campuses across the country, and largely automates the carbon accounting standards jointly established by the World Business Council for Sustainable Development and the World Resource Institute (WBCSD/WRI). Nevertheless, choices have to be made in the identification of local fuel mixes or efficiencies, and for each emission calculation, the sources of information and assumptions have been noted. 1.4 Site and Source Emissions An important distinction in the tracking of energy usage and emissions is the difference between the energy consumed on-site, delivered to building or end-use itself, and that consumed at the source. The utility reports and charges for energy delivered on-site, but emissions are a product of the fuels burned at the plant, or source, to provide the delivered energy. Although greenhouse gas emissions occur at the source, energy use is reported and understood in terms of site or delivered energy, so emissions factors convert site energy to source emissions. The difference between site and source energy can be considerable and is caused by both the inefficiencies of conveying power, whether it is through wires or pipes or conveyed in vehicles, and the inherent inefficiencies of combustion and conversion. Electricity, in particular, involves an inefficient conversion process, loosing 60-75% of the initial fuel value to waste heat. With respect to greenhouse gas emissions, this means that for every unit of electricity used, roughly three units of emissions are produced. There are similar conversion and transmission inefficiencies for each of the centrally distributed fuels. 6 A non-profit action group located in New Hampshire and that largely helps organize schools and institutions in the Northeast.

10 University of Pennsylvania: Carbon Footprint What is the Main Campus? The University of Pennsylvania and its various institutional entities own or control a great deal of real estate. By one count, the University owns, manages, or leases over 900 buildings around the world. For this first carbon footprint, it was decided to focus on the main campus in West Philadelphia, specifically the educational complex of buildings operated by Facilities and Real Estate Services. That represents both the heart of the University and the largest concentration of its buildings approximately 250 though in principle the carbon footprint should ultimately be extended to include all the real estate controlled by the University. Fig Campus Map showing buildings included in the carbon footprint Even with that tight focus, it is something of a challenge to identify the precise limits of the main campus. Our operational definition includes all buildings and facilities located in West Philadelphia that serve direct academic and residential functions for the faculty, staff, and student body, and that are owned and operated by the University of Pennsylvania, a final total of 141 buildings. The largest exclusions resulting from that definition are the Hospital of the University of Pennsylvania (HUP) and various remote holdings. HUP is a large network of hospitals and buildings that has its own facilities staff, and a carbon footprint of this operation will have to be prepared independently of this report. However, the portions used for teaching and education in the School of Medicine are included. Among the

11 University of Pennsylvania: Carbon Footprint 11 discontinuous holdings that were excluded are the New Bolton Center of the Veterinary School, the Morris Arboretum, the Boat House, and the Penn Club in Manhattan. According to our operational definition (and common understanding) the main campus does not include commercial or franchised retail ventures, even if the buildings themselves are owned and operated by the university. The footprint does include housing and activity centers for students, but not the student housing owned by other entities, such as fraternities or independent operators. It also does not include Penn owned buildings operated by outside entities such as the Wistar Institute or the Steinberg Dietrich Conference Center. A complete list of buildings included from the carbon footprint is in Appendix B, while the map illustrates the basic extent of the main campus Campus Overview 7 During the school year, there were 19,492 students and 6,525 faculty at the university. The University is the largest private employer in Philadelphia with over 14,000 staff employees, and the second largest in Pennsylvania. Full-time equivalent campus population has increased fairly steadily during the past decade at an annual rate of about 2%. The University of Pennsylvania s campus consists of 269 acres in West Philadelphia, with about 150 buildings housing the 12 schools of the University as well as a variety of residential halls, libraries, offices, performance centers, athletic facilities, and retail spaces. Penn s buildings total nearly 12 million gross square feet, 25% of which is office space, 20% residence, 17% labs, and the remaining 38% divided between instructional and study spaces, athletics, assembly, food services, healthcare and support. Campus buildings range in size from 875 SF in area to 384,000 SF. The smallest buildings, those with an area less than 10,000 SF, make up 11% of Penn s campus. Those with areas from 50,000 SF to 100,000 SF comprise the largest percentage of total buildings, constituting one quarter of the campus. Only 5% of campus buildings are over 250,000 SF in area. The largest buildings include four residential high-rises, two laboratory and research facilities, and the recently constructed Huntsman Hall. 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 Food Services 105, % Assembly 268, % Athletics 323, % Study 466, % Office 1,607, % Students Faculty Staff Figure Full-time equivalent campus population Re side nce 1,332, % Healthcare 86, % Lab 1,121, % Support 358, % Figure 1.5.3: Campus usage distribution by square footage, 2005 Other 511, % Instruction 412, % 7 With some modest updates, the campus overview was reproduced from the Sustainability Plan, Phase I Report, 2005.

12 University of Pennsylvania: Carbon Footprint 12 The age of campus buildings range from the newly constructed to those with significant historical status, including several buildings over 150 years old. 19% of the campus was constructed prior to the 20 th century, many of which have been included on the National Register of Historic Places or designated as historic by the Philadelphia Historical Commission. Of the remaining buildings, 55% were completed after the end of World War II, including a large amount of construction that occurred during the 1960 s and 1970 s. Recently constructed buildings include Huntsman Hall (2000), the Pottruck Center (2001), Levine Hall (2003), the McNeil Center (2005), Skirkanich Hall (2006), and the School of Veterinary Medicine Training and Research Building (2006). Overall, the campus growth has averaged 1.1% per year over the last 15 years, and will be entering another period expansion with the development of the postal lands along the Schuylkill, east of the main campus.

13 University of Pennsylvania: Carbon Footprint 13 II. Carbon Footprint 2.1. Energy consumption through steam, chilled water, electricity, and natural gas In the scope 1 and scope 2 categories of emissions the University of Pennsylvania consumes energy for three purposes, heating, cooling, and direct electrical usage. In meeting these demands the University maintains three central utility systems, an electrical grid to supply power, a steam system for heat, and a chilled water loop for cooling that is powered by purchased electricity. In addition, it purchases small amounts of natural gas to heat many of the smaller buildings and small amounts of diesel fuel to run emergency and peak-shaving generators. As the Figure indicates, the University uses similar amounts of electricity and steam in energy terms, and the total amount of usage has increased nearly steadily since 1990 with a notable leveling trend after Total Utility Energy Use 8,000,000 7,000,000 6,000,000 5,000,000 MMBTU 4,000,000 3,000,000 2,000,000 1,000, Electricty Steam Fig Total Utility Energy Consumption by Utility Type, MBTU

14 University of Pennsylvania: Carbon Footprint Sources and eco2 Emissions Steam. The centralized steam heating system at Penn connects to about 70% of the buildings and provides approximately 90% of the total heat for the campus. The steam is purchased from Tri-Gen s Gray s Ferry combined heat and power (CHP) plant across the Schuykill River in South Philadelphia. The annual energy imported as steam used for heating typically represents about 46% of the total imported utility energy. In 1997 the Gray s Ferry plant was upgraded with more efficient equipment and its fuel switched from distillate oil to natural gas, which is the cleanest of the fossil fuels in carbon emissions. Equally importantly, it became a combined heat and power facility at that time, which is inherently cleaner since it produces both electricity and steam. 8 The normalized Carbon emissions for steam production are 43 kg eco2/mbtu, or.043 MT eco2/mbtu. 2,000, Energy Consumption (MBtu) 1,500, ,000, , Degree Days Fiscal Year 0 Steam Cooling Electricity Heating Degree Days Cooling Degree Days Fig Total campus energy consumption by utility source, What is notable about steam consumption is the degree to which it is climate driven, rising in cold winters and dropping in warm ones. Figure shows the University s total energy consumption of steam, electricity for chillers, and direct electrical power plotted with annual heating and cooling degree days, which are a common measure of seasonal temperatures. When the winter is warm, steam consumption drops visibly, and vice versa. Cooling also shows some temperature effects, but it somewhat less climate driven. What is notable is the steady increase in direct electric power consumption. 8 It burns natural gas to generate electricity at an efficiency of about 35% and then uses the waste heat from that process to provide steam for distribution for heating purposes, for a total plant efficiency of about 70%. See Appendix C for efficiency emission assumptions.

15 University of Pennsylvania: Carbon Footprint 15 Natural Gas. For those buildings not connected to the central steam system, mostly smaller buildings acquired incrementally over the years, heating is provided by individual natural gas heating systems. The annual energy imported as natural gas for heating typically represents about 0.1% of the total imported utility energy. The normalized Carbon emissions were 27 kg eco2/mbtu, or.027 MT eco2/mbtu. Electricity for cooling and direct power usage is purchased from PECO, a subsidiary of Exelon Corp. This electricity is delivered to the campus in bulk supply through 6 Penn owned and operated substations spread throughout the campus. The internal electrical grid is operated and maintained by the University. In addition, several buildings are connected independently to PECO s grid and billed for that usage directly. The annual energy imported as electricity is about 54% of the total imported utility energy. Of that amount, the chiller plants typically represents about 12%, while direct electrical power usage represents about 42%. 9 Electricity comes from power plants within the Mid-Atlantic Area Council (MAAC) region of the country, which includes most of Pennsylvania, New Jersey, Delaware and Maryland. Power plants in this region use multiple fuel sources including nuclear, coal, oil, gas, hydroelectric, wind, biomass, and waste material Total eco2 Emissions by Utility 500, , , ,000 Tons eco2 300, , , , ,000 50, Year Purchased Electricity Purchased Steam On-Campus Generation Fig Total eco2 emissions by utility source, It is not yet possible to identify the total energy consumed for air conditioning on the campus, which in addition to the electricity required to run the chillers includes the considerable power required to operate the air handlers and pumps within each building. Those distinctions will be part of the Phase III campus Audit results.

16 University of Pennsylvania: Carbon Footprint 16 generation. The precise mix of sources changes from year to year, depending on many factors and introducing even more variability into the emission estimates. In 2006 the regional mixture of sources was 45.1% coal, 40% nuclear, 8.8% natural gas, 2.8% distillate oil, 1.4% biomass, 1.2% hydroelectric, 0.7% other sources. Of these sources biomass, hydro, and nuclear power are carbon free generation sources, while natural gas is the cleanest burning fuel and coal is the dirtiest. With the large nuclear mix in the MAAC, PECO s electricity is a bit lower in carbon emissions compared to other regions. The normalized Carbon emissions for electricity from this region were 160 kg eco2/mbtu, or.160 MT eco2/mbtu. Figure illustrates both the difference in emissions between steam and electric usage and the steady and projected growth in electric power consumption. In 1990, the two sources produced equal amounts of emissions, but after the 1997 improvements at the Gray s Ferry plant and the steady increases in electric consumption have made it the dominant source. Wind Power. In 2004 the University decided to begin purchasing 40,000 Mwh of dedicated wind power from Community Wind, about 10% of total electric purchases at the time, which allowed the company to expand its Pennsylvania facility. In 2006, PECO brought new wind capacity online and the University increased its wind purchases to 112,000 mwh. This comprised about 26% of Penn s electrical consumption in FY2006. Technically the wind purchase is a scope 3 carbon offset, achieved remotely and indirectly, and is discussed and accounted for in the offset section of the report. Diesel Generators. In addition to the energy purchased from PECO, Penn maintains the capacity to generate electricity for peak energy and emergencies with diesel (distillate oil) generators. The peak shave generator is scheduled to operate 182 hours a year. In 2006 it burned 37,620 gallons of diesel fuel, generating 445,539 kwh electricity. The annual energy imported as diesel fuel for generators typically represents about 0.1% of the total imported utility energy. The normalized Carbon emissions were 73 kg eco2/mbtu, or.073 MT eco2/mbtu Projections and Growth Rates The projections visible in the years on charts beyond 2006 are based on some very simple assumptions, and are only intended to describe the future trajectory of emissions if nothing is changed. They are based on steady state rates of growth extrapolated from the 10 to 16 years of data that was gathered. None of the expected variability in populations, budgets, markets, weather, and so on, have been included. Nor have any of the more precise plans for growth in the campus or its systems been incorporated in these projections, though such refinements would be part of any more detailed action planning. The specific growth rates used in these projections are as follows: operating budget, 6.03 %/yr; full-time students 0.50 %/yr; faculty, 0.22 %/yr; staff, 2.8 %/yr; building space, 1.11 %/yr; building energy consumption, 1.11 %/yr; direct electrical usage, 1.87 %/yr. What the projections do tell us is that there are two related kinds of growth in the consumption of utility energies occurring on the campus: the growth in heating, cooling, and lighting associated with construction of new buildings, and the growth in direct electrical consumption within those buildings. The first is largely a function of the building design and operation, while the second is a function of technology, along with the everyday practices, purchases, and choices of the faculty, staff, and students that occupy them.

17 University of Pennsylvania: Carbon Footprint Observations and Strategies Immediately apparent in the charts of energy consumption and utility emissions are a number of effects: first the variability of emissions due to annual differences in weather, then the drop in steam emissions beginning in 1997, followed by the dip in electric emissions after 2001 as a result of new management strategies. The effect of the wind purchases is illustrated in the offsets section, but equally visible is the effect of unchecked growth. At present, Penn consumes roughly equivalents amounts of energy for steam and electrical power every year, however electricity accounts for over 68% of total emissions, with steam only contributing about 19%. These two utilities are quite different in their patterns of usage, their tendency for growth, and their amenability to management. The steam required for heating is very much a function of yearly weather and the efficiencies of individual building skins and systems, which can be significantly improved over time. Electricity used for chiller plants also responds to the yearly weather, though much less dramatically and is somewhat more connected to internal power consumption and the operation of building systems than the efficiency of the building skin. However as Figure illustrates direct electrical consumption for HVAC systems, lighting and plug-loads has increased steadily in recent years, rising nearly 2% a year above the increases in building area. Current projections even show electricity consumption exceeding heating between 2010 and There are also some quite significant differences in strategy suggested by these observations. Heating, cooling, and lighting can all be reduced by a concerted program of improvement in the efficiencies of building skins and HVAC systems. This can be done from the top and from the bottom, as a part of the regular cycle of maintenance and renovation across the campus, and as part of a distributed, projectspecific program of improvements. Harvard s sustainability fund has shown just how much innovation can be unleashed with the bottom-up method, but the regular program of deferred maintenance should also target energy efficiencies and develop accounting techniques to repay them. Conversely, the ever increasing plug loads directly reflect the growth in research and educational technology, equipment purchases and usage habits of all the many individuals across the campus. It does not respond much to management directives and will require new strategies and real changes in behavior and practice. Laboratories. Perhaps the largest single source of direct power consumption is the HVAC and plug-load of research labs, though this too is distributed widely across schools, programs, and individuals. Although laboratories constitute about 17% of campus facilities, they consume nearly 55% of the annual electric purchase. This situation is not unique to Penn, and is a well recognized aspect of contemporary research. The electrified tools of modern science, from fume hoods to refrigerators, are integral to research, so simple appeals for conservation do not address the complex nature of their usage.

18 University of Pennsylvania: Carbon Footprint 18 Unlike offices or residences there is little comparative national data available for laboratories. The EPA and DOE have an EnergyStar type program called Labs21, which has developed a modest database of performance data that shows overall energy intensities ranging from 200 to 400 kbtu/sf and above. 10 These average intensities further indicate the necessity of understanding laboratory energy-use patterns. This data also suggests that Penn s labs are likely within normal operating parameters for contemporary labs. So while the lab buildings rank high in energy use per square foot, they are representative of the special challenges of this whole class Building Electric Usage by Space Type, kwh kwh, kwh, Special kwh, kwh, Recreation Residential Assembly kwh, Off/Class The Federal LABS21 program has studied the Fig Electricity Usage by Space Type, kwh, best practices in this area, and the lab facility managers should be organized to pursue efficiencies across the spectrum. kwh, Lab kwh, Library 10

19 University of Pennsylvania: Carbon Footprint Transportation through University fleets of cars, vans, buses, and trucks. Accounting for the miles driven by University owned vehicles is a difficult task. Given the decentralized nature of the University there is no common motor-pool, gasoline station, or vehicle department. Even the Transit Operations department does not have its own fuel facilities or maintenance crew. All fuel is purchased from local gas stations individually by the various schools and departments. All vehicles are also purchased independently by school or department. There is no centralized accounting of maintenance or fuel costs. Additionally, most vehicles on campus drive so few miles in a fiscal year that the budgets for fuel and maintenance often come from petty cash, discretionary funds, or are rolled into other budget items. While there is a highway tax rebate available to the school, due to its status as a University, for many departments this savings is too small to be worth applying for. Several vehicles on campus burn little more than one or two tanks of gas in a year. In order to target the total fuel consumption for the campus we obtained data from the Office of Risk Management which tracks car registrations for insurance purposes. This provided the best estimate for the total number of vehicles and a yearly mileage estimate for each vehicle. Unfortunately the registrations did not specify the model of the car or whether it was powered by a diesel or gasoline engine. The University of Pennsylvania currently has 151 vehicles registered with the state. These include 15 buses, 50 trucks, 66 SUV s, 15 sedans, 2 motorhomes, 1 motorcycle, and 2 unlabeled vehicles. In the past 10 years a total of 365 different vehicles have been owned and operated by the University. These include 22 buses, 125 trucks, 147 SUV s, 47 sedans, 4 motorhomes, 4 motorcycles, and 16 unlisted vehicles. University Owned Fleet by Vehicle type, 2006 CO2 Emmisions by Vehicle type, 2006 Bus Unknown Sedan Unknown Sedan Winnebago Motorcycle Bus Winnebago SUV Truck SUV Motorcycle Truck Fig University fleet by Vehicle Type. Fig eco2 Emissions by Vehicle Type Sources and Emissions Of these vehicles it has been assumed that the buses are all diesel powered. According to a 2005 survey 54% of the trucks owned by the facilities division of Penn were diesel powered. Facilities has owned nearly half of the trucks on campus (54 of 125) thus this percentage of diesel powered vehicles is applied to the remaining trucks on campus. It is assumed that all other vehicles, sedans, suv s, motorcycles, and motorhomes are gasoline powered. Applying these assumptions to the campus fleet we come up with 19%

20 University of Pennsylvania: Carbon Footprint 20 Diesel powered and 81% gasoline, with only the truck category being mixed. Since the Gasoline powered vehicles represent a cross section of makes and models, but are not specifically known, the general CAFÉ standards for fuel economy are applied to the total miles driven to obtain an estimate of gallons of gasoline consumed. Diesel vehicles are a little more difficult to estimate as they are not as strictly regulated by the government. A generally agreed upon average mileage for buses is 6-8 mpg. Diesel powered trucks get slightly higher mileage between mpg depending upon the size of the truck. Factoring in the miles driven and the type of vehicle we come up with an average mileage of 12 mpg for the diesel fleet. The mileage data obtained through this strategy demonstrates a fairly stable driving environment on the campus. Apart from the first 2 years of record availability there has been no increase in the amount of fuel consumption from University Fleet vehicles. Miles Driven by Fuel Type 2006 CO2 Emissions by Fuel Type 2006 Diesel Gas Diesel Gas Fig Miles driven by Fuel Type, Fig eco2 Emissions by Fuel Type, The University has said that it is not in the transportation business and as such runs only what it needs to provide a safe, and functional campus community. The contributions to the overall carbon footprint from the fleet are minimal, constituting only 0.2% of the total footprint Observations and Strategies Penn benefits from being a dense urban campus, allowing significantly lower emissions from fleet vehicles than other institutions. Despite the relatively small portion of emissions from this source there is still room for improvement. The possibility of centralizing the motorpool facilities creates enough volume to open up the opportunity for alternative fuels that would be cleaner and more eco-friendly for the environment. Though a relatively small contributor to the overall carbon footprint, this is a high profile one which could both save the University money and generate publicity.

21 University of Pennsylvania: Carbon Footprint Agriculture including fertilizer and agricultural waste Agricultural emissions can come from a variety of sources, but as an urban campus, the only emission in this category are the fertilizer s applied to the grounds. After the application of any nitrogen-containing fertilizer, some percentage is released as nitrous oxide. No historical information was found concerning fertilizer use, but it was determined that in 2006, 600 pounds of 20% nitrogen fertilizer was applied by landscaping sub-contractors. These numbers were simply extended to other years. In 2006, fertilizer constituted % of the total emissions. The more common source of agriculatural emissions is livestock, which will be an item for documentation when the carbon footprint is extended to the New Bolton Center. eco2 Emissions from Fertilizer Kg eco Year Fig eco2 Emissions from Nitrogen Fertilizer.

22 University of Pennsylvania: Carbon Footprint Solid Waste disposal Institutions have several methods for managing solid waste. The two most common are incineration and landfilling. Waste that is incinerated releases greenhouse gases when combusted and waste sent to landfills releases methane as it decomposes. And within those two processes, there can be varying degrees of greenhouse gas impacts: (1) a mass burn incinerator, (2) a refuse-derived fuel incinerator, (3) a landfill with no methane collection, (4) a landfill that collects methane emissions for flaring, (5) a landfill that collects methane emissions for electric generation, or (6) the waste can be recycled. As reported in the Phase I Sustainability Report, rates of recycling have fallen in recent years, though efforts are being made to enhance them. At Penn, non-recycled waste disposal is handled by Browning-Ferris Industries (BFI)32 and Waste Management, Inc. BFI manages trash dumpsters, and Waste Management disposes trash packed in the 20 compactors around campus. Recycled waste is also handled by BFI, with a new program being added by Precision Fig Solid Waste Disposal, Hydraulic. Overall, Waste Management handles slightly more waste than BFI Emissions The solid waste at Penn is sent to different kinds of landfills, with varying degrees of methane recovery, but it has not been possible to identify the precise distribution. Estimates were based on the landfill numbers on their respective websites. BFI gave no landfill descriptions, so basic landfill is assumed. Waste Management uses 283 landfills, of which 10 were Biogenerative, 95 were Methane collecting, and 17 were Power generating Waste Incinerators. The trash is then divided proportionally according to those types and the corresponding emissions were determined. As Figures & 3 indicate, the methane regulated landfills have very little greenhouse gas effect. The total solid waste emissions in 2006 totaled 5,836 metric tons eco2, or about 1.8% of the total emissions. Though the total contribution is small, a simple strategy for reducing this source of emissions is to have the University s solid waste directed to regulated landfills.

23 University of Pennsylvania: Carbon Footprint 23 Solid Waste Disposal by Landfill Type by Year 7,000 6,000 5,000 Short Tons 4,000 3,000 2,000 1, Refuse Derived Fuel (RDF) Incinerator Landfilled Waste with CH4 Recovery and Flaring Year Landfilled Waste with no CH4 Recovery Landfilled Waste with CH4 Recovery and Electric Generation Fig Solid Waste Disposal by Landfill Type. Metric Tonnes eco2 Emissions by Landfill Type 7,000 6,000 5,000 Metric Tonnes eco2 4,000 3,000 2,000 1,000 - (1,000) Year Refuse Derived Fuel (RDF) Incinerator Landfilled Waste with CH4 Recovery and Flaring Landfilled Waste with no CH4 Recovery Landfilled Waste with CH4 Recovery and Electric Generation Fig Emissions by Landfill Type.

24 University of Pennsylvania: Carbon Footprint Refrigerant replacement The purchase and disposal of HCFCs and PFC refrigerants are tracked and reported by the University. However, it was not possible to determine how much refrigerant is actually released at Penn. Accidental release of amounts below 50 kg do not need to be reported, and so are not tracked centrally. Amounts in this range do not represent a significant source of greenhouse gas emissions.

25 University of Pennsylvania: Carbon Footprint Commuter Traffic by car, train, bus, bike, and walking. Without an explicit survey of commuting habits, indirect forms of evidence were gathered to provide estimates of the numbers of commuters pursuing different modes of travel to and from the University. The most concrete information available are lists of parking passes going back to 2001 that identified a home zip code for each pass. These were used as a basic measure of the number of automobile commuters, and distances were calculated based on the zip codes. While there are other parking opportunities around the campus, and the number of regular car drivers likely exceeds this number, this estimate provides a simple measure of the automobile commuting directly enabled by the University. The number of passes represent about 18% of the faculty and staff, and about 10% of students. The only other piece of evidence are results reported in a 1995 student project that cited two ridership surveys, one from 1989 and the other from 1995, both of which suggested automobile faculty and staff commuting rates of 40% or higher. Given the changed nature of the campus, of West Philadelphia, and of traffic in the area, it seems possible that auto commuting has decreased, but a much more thorough study would be required to confirm that result. The numbers of commuters using public transportation is even more difficult to assess, and really highlights the challenge of precisely documenting these kinds of scope 3 emissions. The University does provide discounted rail and transit passes, and tracks both the total numbers of transit passes and the numbers of rail passes by travel zone. In addition, partial ridership data was obtained from Septa for the main rail stop and some bus stops, and this was used to gauge the adequacy of the University pass numbers. In both cases, the actual ridership appears to exceed the pass numbers, and anecdotally it is likely the ridership on Commuting Modes Number of Commuters Public Transit Drivers Walk, Bike, or Other Fig Commuting Modes

26 University of Pennsylvania: Carbon Footprint 26 other bus lines, the trolley, and the subway surface lines greatly exceed these modest estimates. However, in the absence of a proper survey of more concrete information, these modest numbers were used. In the period for which data was available a steady patter of decline in the rates of automobile commuting was evident and so that basic pattern was extended back to 1990 and forward to The decline of student commuting emissions and increase in those of the faculty staff which are visible in Fig are a result of the increasing staff population. eco2 Commuting 6,000 5,000 4,000 MT eco2 3,000 2,000 1, Student Commuting Faculty/Staff Commuting Fig eco2 Emissions from Commuting Commuting Emissions The total emissions from the automobile, bus, and train commuting captured by the parking and rail passes constitutes just less than 2% of the total campus emissions. Of that automobile emissions constitute that overwhelming majority, so the effect of further increasing the use of public transportation or the carbon-free modes like walking and biking can be substantial. Despite the uncertainties, commuting is a highly visible aspect of life at the campus and the University if fortunate to be connected to an extensive travel network.