Evaluating progress on the University of Pittsburgh’s climate action plan through robust GHG inventory data on building emissions

The global target to limit climate change below the 1.5°C threshold is a 45% reduction in anthropogenic CO₂ emissions below 2010 levels by 2030, and reaching carbon neutrality around 2050 (IPCC, 2018). Universities have a unique position to lead in climate action implementation. Their research capacity and role in education for sustainable development make them apt to lead the decarbonization transition. Universities additionally impact local development, as their campuses are often interconnected with local municipalities and communities (Goddard and Vallance, 2013). Universities recognize this responsibility, with many making efforts to track and limit their carbon footprints (CFs). As 2030 and similar carbon reduction targets quickly approach, universities are now at a pivotal moment where they must transition from planning for deep decarbonization to enacting it.

Greenhouse gas (GHG) inventories are one of the most widely used tools for GHG accounting and management (US EPA, 2015). They quantify emissions; identify hotspots, major GHG sources, and reduction opportunities; and track progress. GHG inventories are also a requirement for signatories of the Second Nature Climate Leadership Network, a public commitment to advance campus climate action with over 800 signatories of total US higher education institutions (Barron et al., 2021). GHG inventories are used to establish a university’s CF and facilitate strategic planning to reduce the university’s GHG emissions, which often takes the form of a Climate Action Plan (CAP). Climate action planning at the campus scale is a roadmap to a university’s climate neutrality goals, typically including identifying and quantifying GHG emissions goals, establishing actionable priorities, and setting reduction milestones and target dates (Spirovski et al., 2012). Both GHG inventories and CAPs create opportunities for student engagement and hands-on learning about sustainability, and some CAPs include sustainable courses and degree programs to further education for sustainable development (Button, 2009).

Some universities have detailed the GHG inventory process and its findings, including Louisiana State University (Moerschbaecher and Day, 2010) and the Norwegian University of Technology and Science (Larsen et al., 2013). Multiple studies compare GHG inventories between universities. One study analyzed factors that affect GHG emissions by looking at data from 135 universities, finding that GHG emissions are most influenced by the institution’s size, amount of laboratory space, whether there is a medical school, and commuting habits (Klein-Banai and Theis, 2013). Robinson et al. (2018) analyzed six GHG methodology guidelines and found double-counting and resource cost to be common challenges. Despite most universities following the GHG Protocol, GHG inventories are often customized to a university’s size, resource availability, and goals. Highlighting this variability, De Montfort University focused its GHG inventory on consumption-based emissions and included procurement, which other universities often omit (Ozawa-Meida et al., 2013).

The current literature on CAPs at universities is mostly qualitative and focuses on the early stages of development. The University of New Hampshire detailed the process of establishing a CAP and recommended student involvement and cultivating partnerships that align with a university’s educational mission and identity (Cleaves et al., 2009). Mazhar et al. (2019) identified inconsistent boundaries in measuring and reporting, hyperfocus on technical aspects of carbon management, and the static nature of CAP documents as potential areas of improvement in the carbon management planning process at universities in the UK. In addition, some studies combine elements of GHG Inventories and CAPs. Spirovski et al. (2012) investigated how South East European University’s five GHG inventories informed the development of their CAP. At Ernst-Moritz-Arndt-Universität Greifswald, Germany, a pilot project explored potential opportunities and challenges in implementing sustainable actions to achieve carbon neutrality (Udas et al., 2018). Multiple other studies look at universities’ GHG emissions and propose reduction measures (Mustafa et al., 2022; Thomas, 2005; Vásquez et al., 2015).

Although studies have separately explored GHG inventories and CAPs at universities (full list provided in supplementary material), to the authors’ knowledge, no papers compare long-term GHG inventories and their interaction, trends and results before and after implementing a CAP. This case study builds on the existing body of literature by analyzing one CAP priority area (buildings) and their GHG emissions over 15 years of GHG inventory data before and after the implementation of the CAP. The University of Pittsburgh (Pitt) has a unique opportunity to leverage over a decade of GHG inventories by comparing them with Pitt’s CAP to evaluate data trends and improvements and ultimately determine if the university is on track to meet its carbon neutrality goals. This study aims to demonstrate how long-term GHG inventories can both inform an initial CAP and track progress over time.

As a Second Nature Climate Leadership Commitment signatory, Pitt completed nine GHG inventories over 15 years (Cicco et al., 2024). A collaboration of Pitt faculty, staff and graduate students complete the GHG inventories using guidance set by the Greenhouse Gas Protocol through the Sustainability Indicator Management and Analysis Platform (SIMAP) software (University of New Hampshire Sustainability Institute, 2017). In 2020, Pitt announced its goal to be carbon neutral by 2037, Pitt’s 250th anniversary (Monahan, 2022). Informed by the GHG inventories, Pitt published its first Climate Action Plan (PittCAP) in 2022 (University of Pittsburgh Office of Sustainability, 2022).

The largest contributor to Pitt’s GHG inventories’ emissions has consistently been building energy use, specifically electricity and heating. In FY 2022, purchased electricity accounted for 37% of Pitt’s CF; steam was the second-largest contributor at 28% (Geremicca et al., 2023). Relatedly, the PittCAP expects 9.4% of Pitt’s GHG emissions reductions needed to reach carbon neutrality to come from increases in energy efficiency of the existing building stock on Pitt’s campus (Figure 1). The large impact on Pitt’s CF and the urgency required to improve energy efficiency by the 2037 goal have made buildings a focus of the University’s sustainability efforts.

The PittCAP denotes the Office of Facilities Management (FM) as the lead collaborator in achieving PittCAP’s existing building efficiency reduction goal (EBERG). FM plays a crucial role in campus sustainability by overseeing the operation of Pitt’s building stock. However, aligning clear, measurable sustainability performance targets proved challenging because PittCAP’s goals were broader in scope and used different metrics than those guiding FM’s operations. Because Pitt FM operates at the individual building level, establishing building-specific GHG reduction goals provides a more actionable unit of analysis. Therefore, this research presents a method to set building-specific targets, based on the broader PittCAP goal, to effectively communicate sustainability targets in a more tangible and attainable way between GHG inventory teams and FM. These targets provide context for the CF of each building in comparison to the PittCAP goal and serve as indicators for prioritizing buildings for energy efficiency interventions.

By exploring the PittCAP’s goals for building-related emissions and analyzing major trends and influences in Pitt’s nine GHG inventories, this work demonstrates how GHG inventories can inform a CAP and evaluate its progress. In addition, this research disaggregates the CAP’s energy efficiency target to the building level to recommend project prioritization and validate these recommendations with Pitt FM. The following strategy was developed to evaluate trends in building GHG emissions through Pitt’s GHG inventories and progress toward the PittCAP carbon neutrality goals:

• Investigate the historical data on steam and electricity documented in the GHG inventories and summarize all trends and major influences that resulted in fluctuations throughout the years.

• Calculate CFs at the individual building level starting in FY 2019, using building energy use data paired with emissions factors from SIMAP.

• Set a building-specific goal for each existing building on Pitt’s campus based on the EBERG defined in the PittCAP of 22,200 MT CO₂e annually by 2037 (Figure 1).

• Create yearly emission reduction targets based on each building’s 2037 goal to indicate whether each building is on track to meet the PittCAP 2037 carbon neutrality goal.

This study was conducted in two steps to achieve the goal of investigating how long-term GHG inventories inform a CAP and can subsequently be used to create tangible next steps and track progress on a CAP goal. First, historical GHG inventory data on building energy use and GHG emissions were summarized (Section 2.1). After developing a strong understanding of historical trends, this data was investigated quantitatively through the lens of the PittCAP. Individual buildings’ CFs were calculated to set goals for each existing building on Pitt’s campus based on the PittCAP EBERG. This process and methodology are further explained in Section 2.2.

Steam and electricity data were summarized to identify key drivers of building energy use and associated GHG emissions on Pitt’s campus. These trends were developed over all GHG inventories to show fluctuations in energy use and to explore the causations as explained by the inventories. Building natural gas was excluded from this summary, despite its inclusion in Section 2.2, due to its consistently smaller contribution to overall emissions, 4.9% of total emissions in FY 2021 compared to steam (27.6%) and electricity (51.8%), and its limited focus in GHG emission reduction strategies and documentation in the GHG inventories (Geremicca et al., 2022). It also closely follows broader building energy use emissions trends.

After analyzing historical building energy data to investigate trends in energy use and GHG emissions before and after the implementation of the CAP, the GHG inventories were additionally analyzed to compare Pitt’s buildings’ emissions specifically to the PittCAP EBERG. The PittCAP EBERG is an annual reduction of 22,200 MT CO₂e from the FY 2019 GHG inventory baseline by FY 2037 (University of Pittsburgh Office of Sustainability, 2022). No explicit emissions reduction goals for steam, natural gas or electricity are stated, but they are expected to decrease to meet the overall building emissions goal.

Electricity, steam, and natural gas consumption data that were used in the GHG inventories to calculate building emissions were disaggregated for 87 buildings for FYs 2019, 2020, 2021 and 2022. FY 2019 was the selected benchmark for this analysis because it marks the beginning of the PittCAP. FY 2023 was not included in this analysis due to a lack of building-level data. Pitt FM recorded the data in kilowatt-hours (kWh), pounds (lbs) and cubic feet (CF), respectively. Although this energy consumption data can indicate campus energy efficiency, energy use alone does not describe the full climate impact of each building.

CFs for each building were evaluated by multiplying the consumption of each energy source by an emission factor. These emission factors were obtained from SIMAP, a web-based tool designed for universities to calculate their GHG emissions. SIMAP is widely used by universities for GHG inventorying, standardization, emission factors, and GHG protocol alignment (Second Nature, 2025). The GHG emissions from electricity, steam, and natural gas in MT CO₂e were then summed to determine the total CF of an individual building for each FY. After analyzing all buildings, the total emissions of all 87 buildings for each year were summed.

Once CFs were determined for each building from FY 2019 through FY 2022, building-level emissions targets were created to quantify progress toward the PittCAP EBERG. To determine the emissions targets, it was assumed that each building would contribute to the PittCAP EBERG (22,200 MT CO₂e) proportionally to its contribution to the total building-related emissions in the GHG inventories. Díaz et al. (2013) similarly calculated emission reduction targets for a building from country or sector-wide emissions objectives. However, in their methodology, building GHG emissions targets were based on floor space and building use type to estimate a building’s fractional emissions contribution to overall building sector emissions. In this study, because the PittCAP governs a smaller number of buildings and data is available for the GHG emissions of each building, we could calculate each building’s contribution to campus emissions directly and categorization based on floor space and use type was not necessary. Díaz et al.’s (2013) equation (9) was used to calculate individual building emissions targets, $Gi*t,$ from the PittCAP EBERG of 22,200 MT CO₂e annual reduction from FY 2019 baseline by 2037:

The definition of Díaz et al.’s variable i was modified from representing a building type to representing a specific building on Pitt’s campus. The variables are defined in this study as follows:

$Gi*t$ = individual building emissions targets for building i in year t;

$βit$ = allocation factor of the emissions objective to the specific building i in year t;

$∝t0,t$ = $GTot*tGTott0=$ fractional reduction factor between baseline year t₀ and year t calculated from the target value for total emissions in year t,$GTot*t$⁠, divided by the baseline value for total emissions in the baseline year t₀, $GTott0$⁠; and

$Gt0=∑i = 1IGi(t0)$ = the sum of all building emissions in the baseline year t₀.

Although Díaz et al. states that different values of $βit$ can be selected for any political or social considerations, the authors chose to follow the same allocation method, where emissions are allocated in proportion to a historic emission baseline, as used in equation (8) (Díaz et al., 2013):

This equation was used to calculate all building emissions targets for meeting the 2037 PittCAP EBERG. Next, linear interpolation was used to set annual emissions targets from FY 2019 to FY 2037. This process was repeated for the 87 buildings, which include diverse building typologies: 22 college/university (25%), 20 laboratory (23%), 18 residence hall/dormitory (21%), 10 parking (11%) and 17 other (20%).

Section 3.1 summarizes major influences and resulting trends on steam and electricity use and emissions documented in the GHG inventories. The usage data is also normalized against the following explanatory variables to isolate building energy efficiency: heating degree days (HDD), cooling degree days (CDD), and campus building area (CBA). Section 3.2 shows the CFs for existing buildings starting in FY 2019 and their progress on building-level goals based on the PittCAP EBERG. Yearly emissions targets indicate whether each building is on track to meet its FY 2037 goal, and results for FY 2022 are summarized, showing which buildings have the largest CF and their contributing progress toward the PittCAP goal.

In FY 2008, Pitt’s first GHG inventory revealed that purchased electricity was the largest contributor to overall GHG emissions, accounting for 138,700 MT CO₂e and 51% of total emissions. Electricity emissions and usage have decreased roughly 48% and 2% from FY 2008 to FY 2023, respectively. Steam was the second-largest category in FY 2008 at 55,100 MT CO₂e. While steam emissions have decreased 10% in FY 2023 from FY 2008, demand is 40% higher. From 2008 to 2023, the major influences on building energy use and emissions trends are (1) transitions to cleaner fuel mixes, (2) building energy efficiency, (3) renewable energy credit (REC) procurement, (4) the global COVID-19 pandemic and (5) changes in SIMAP emissions calculation methods:

• Transition to cleaner fuel mixes: In FY 2008, Pitt conducted its first GHG inventory to benchmark steam use and emissions supplied by the Bellefield Boiler Plant (BBP) before shifting some steam production to the more efficient Carrillo Street Steam Plant (CSSP), which became operational in FY 2011. The BBP used a fuel mix of 50% natural gas and 50% coal for steam production, which produced 55,100 MT CO₂e GHG emissions in FY 2008 (Bilec and Aktas, 2010). The CSSP was built to reduce GHG emissions by switching from coal to 100% natural gas and to improve air quality with ultra-low NO_x technology. In FY 2011, the CSSP provided nearly half of the University’s steam demand, with additional production transferred to the CSSP in future years [Figure 2(a)]. Also in FY 2011, the BBP was retrofitted to operate on 100% natural gas, eliminating coal from the fuel mix and increasing the BBP’s efficiency. As illustrated in Figure 2(a), although steam consumption increased 31% in FY 2011 from FY 2008, the switch to the CSSP as Pitt’s main steam source, increased plant efficiency, and elimination of coal from the BBP’s fuel mix resulted in a 6% decrease in steam emissions in FY 2011 (Bilec and Ketchman, 2013). In addition, Pitt has been able to deliver more steam to buildings for fewer GHG emissions long-term, keeping total steam GHG emissions below the FY 2008 baseline for nearly every inventory, despite a steady demand increase [Figure 2(a)].

In FY 2008, the regional electricity fuel mix primarily consisted of coal (72.8%), nuclear (22.3%), natural gas (2.7%), and small contributions from renewables, oil and other sources (Bilec and Aktas, 2010). Since FY 2008, the region’s primary electricity-generating fuel source has shifted from coal to nuclear and natural gas. In FY 2014, GHG emissions were reduced by 15% from the previous FY due to this shift [Figure 2(b)] (Bilec and Hasik, 2016). In FY 2023, the fuel mix was natural gas (43.0%), nuclear (33.3%), coal (16.6%), renewables (6.1%), and small contributions from oil and other sources (Cicco et al., 2024). This continued shift to an electricity fuel mix with lower associated GHG emissions has been one of the main drivers in reducing GHG emissions since FY 2008:

• Building energy efficiency: As shown in Figure 2(a), steam emissions trends mirror consumption from FY 2014 onwards. This consistency reflects minimal changes in efficiency, fuel sources and each plant’s share of campus steam supply. Consequently, GHG emission variations are primarily driven by changes in demand. FY 2019 is an outlier in this trend, as steam emissions decreased despite demand increasing by over 13% (Bilec et al., 2020). Electricity emissions have not typically mirrored demand [Figure 2(b)] due to annual changes in the fuel mix and REC procurement. The main influences on energy demand highlighted in the GHG inventories are HDD, CDD, and CBA.

To better understand the significance of building energy efficiency on energy use trends, Figure 3 normalizes steam and electricity usage against these explanatory variables (left y-axis), thereby isolating energy efficiency. Notably, in FY 2017, steam demand decreased by 34% (Bilec and Gardner, 2018). This significant decrease in demand is attributed to a reduction in HDD, with normalized steam usage staying relatively constant in Figure 3. However, the increase in steam demand in FY 2019 in Figure 2(a) contrasts with a 25% decrease in normalized steam usage in Figure 3, partially due to the addition of housing facilities that do not use steam, but also pointing to increased building energy efficiency. Normalized electricity usage increased from 0.0132 kWh/ft²/CDD in FY 2008 to a record high of 0.0204 kWh/ft²/CDD in FY 2023. This increase in FY 2023 is due to the addition of energy-intensive laboratories to the building stock (Cicco et al., 2024). Although current electricity and steam emissions are below the FY 2008 baseline, the recent increase in normalized electricity usage in Figure 3 indicates a need for improvement in the University’s building energy efficiency efforts as outlined in the PittCAP:

• REC procurement: Pitt recorded unbundled RECs for the first time in the FY 2019 GHG inventory, reflecting a purchase of 36,230 MWh of RECs, equivalent to 16.8% of that year’s electricity usage (Bilec et al., 2020). Along with changes in the electricity grid fuel mix previously noted, RECs contributed to a 47% reduction in electricity emissions compared to 2008, despite a 9% increase in electricity use. The significant effect of RECs on GHG emissions was also demonstrated in FY 2021, when GHG emissions increased despite electricity demand decreasing [Figure 2(b)], partially due to a large reduction in REC procurement.

• COVID-19 pandemic: The global COVID-19 pandemic caused a campus shutdown for 3.5 months of FY 2020, leading to a 4% decrease in steam demand as many buildings were closed. Pitt’s switch to virtual instruction and partial shutdowns due to the pandemic explains the continued decrease in normalized steam demand in FY 2021 (Figure 3). Interestingly, Figure 3 indicates a large increase in normalized electricity demand in FY 2021 that contradicts the decrease seen in normalized steam use. This discrepancy is mostly due to operational changes to buildings during the pandemic that lingered into FY 2022. These findings are consistent with other studies that found an increase in energy demand of buildings due to updated operational guidelines for HVAC systems during the pandemic (Zheng et al., 2021).

• Changes in SIMAP emissions calculation method: Throughout the 15 years Pitt has completed GHG inventories, SIMAP and GHG Protocol best practices have changed multiple times. When best practice recommendations change, Pitt updates its methodology accordingly, which can also cause changes in emissions results. In FY 2011, electricity emissions fell 2% despite a 7% rise in use, largely due to a shift to a supplier-specific fuel mix for emissions calculations (Bilec and Ketchman, 2013). Emissions calculation methods changed again in FY 2020 (Bilec et al., 2021), causing a similar discrepancy.

Despite both steam and electricity emissions being below their FY 2008 baselines, the steady increase in steam demand in recent inventories and electricity use approaching baseline levels in FY 2023 is inconsistent with the priority of reduced energy consumption and emissions in the PittCAP. It highlights the need for additional measures to meet the University’s carbon neutrality goals. Normalized steam demand fluctuated across all GHG inventories but ended only slightly below FY 2008 levels in FY 2023 (Figure 3), while normalized electricity use is at a record high in the most recent inventory. This demonstrates that the success Pitt has had in reducing building-related emissions is mainly through switching to cleaner-burning fuel sources and increasing REC procurements. Pitt will likely be unable to meet emissions targets without reducing energy demand, which will become even more challenging with expected campus growth. To continue progress toward carbon neutrality, additional measures to increase efficiency must be taken.

This analysis compares Pitt’s buildings’ GHG emissions to the PittCAP EBERG at a building level. The first buildings analyzed using the method described in Section 2.2 were Allen Hall and Lothrop Hall. Allen Hall is primarily used for education and features classrooms, student study spaces, and offices; Lothrop Hall is a residence hall. Between 2019 and 2020, sustainability initiatives in Allen Hall included efficiency efforts such as changing older lightbulbs to LED bulbs, but the building has not undergone major renovations. In FY 2019, Allen Hall contributed 0.40% of Pitt’s total building GHG emissions with 593 MT CO₂e. Lothrop Hall’s contribution to FY 2019 total building GHG emissions was 1.4%. Using this paper’s equivalent distribution method, the FY 2037 goals for both Allen Hall and Lothrop Hall require a 15% reduction from each building’s baseline, or 90 MT CO₂e and 312 MT CO₂e, respectively.

The “business as usual” lines in Figure 4 (Yellow Line) show Allen and Lothrop Halls’ expected annual CFs without implementing PittCAP reduction strategies. The “CAP goal” lines in Figure 4 (Green Line) illustrate incremental annual building emission reductions until 2037. The FY 2037 reduction goal only requires a 0.8% reduction annually in emissions from the FY 2019 baseline in both buildings.

Across Allen and Lothrop Hall, electricity emissions were lower in FY 2022 than in FY 2019, likely due to a 46% decrease in CDD. However, increased steam emissions in Allen Hall outweighed the decreased electricity consumption, leaving Allen Hall above its FY 2022 emissions target – and not on track to meet its FY 2037 reduction goal [Figure 4(a)]. In contrast, Lothrop Hall saw a decrease in both steam and electricity consumption, leading to a 19% reduction in GHG emissions that surpassed its FY 2037 reduction target [Figure 4(b)]. However, Lothrop Hall’s FY 2021 steam use decrease is likely due to decreased residence hall occupancy during the COVID-19 pandemic, as mentioned in Section 3.1. As a result, Lothrop Hall’s FY 2021 and FY 2022 reductions are likely to be unmaintainable in future years. Academic buildings like Allen Hall were less affected due to similar heating and cooling schedules and increased ventilation requirements.

This analysis was extended to the 87 buildings in the University’s FY 2019 GHG inventory. The 2037 GHG emissions reduction goals and contributions to the EBERG for all studied buildings are provided in the supplementary material. To meet the PittCAP EBERG, each building’s necessary decrease in GHG emissions is approximately 15% by FY 2037, which translates to an annual reduction of 0.8%. Figure 5 illustrates the goal line and the total emissions for all buildings in FY 2019 through FY 2022; it also projects GHG emissions reductions to FY 2037. In all studied years, total building emissions achieved the reductions required to surpass yearly targets, remaining collectively on track to meet the FY 2037 PittCAP EBERG.

As presented in Figure 5, FY 2020–FY 2022 total building GHG emissions surpassed yearly reduction goals with emissions decrease from baseline of 5.0% in FY 2020 (7,287 MT CO₂e), 4.1% in FY 2021 (5,986 MT CO₂e) and 4.0% in FY 2022 (5,820 MT CO₂e). As shown in Figure 6, of the 87 buildings analyzed, 21 buildings already reached their 2037 emissions goal (24%), 28 are on track or ahead of schedule (32%) and 38 are not on track to meet their FY 2037 GHG emissions reduction goal (44%).

The 10 largest GHG emissions contributors to total building emissions are indicated by yellow markers in Figure 6. Of these 10 buildings, five are laboratories, one is a residence hall, one is a university building used for education and office space, and 2 are considered “other” (a gym and an events center). One building has reached its 2037 goal, four are on track, and five are not on track (Table 1). In order of emissions contribution, the five that are not on track to contribute their equivalent share of GHG emissions reduction are Biomedical Science Tower 3, Benedum Hall, Chevron Science Center, Wesley W. Posvar Hall and Langley Hall. This information can help the University target which buildings are underperforming and inform decisions about where to prioritize energy efficiency upgrades. The energy efficiency strategies outlined in the PittCAP include upgrading lighting, heating, cooling, controls and plug load systems to increase building energy efficiency. Some of these upgrades have already begun, reflected in some buildings’ energy use reductions from FY 2019 through FY 2022.

In conjunction with other considerations (including energy use intensity, building square footage and estimated reduction potential), the results of this analysis have already been used at Pitt to inform building energy efficiency prioritization, which is overseen by Pitt’s FM office. This CF analysis helped FM bridge the gap between energy use and GHG emissions, directly connecting their work with the PittCAP. The combination of GHG inventory trend data, individual building trends and which buildings have the highest CF and contribute most to the PittCAP reduction goal are all important additions to university decision-making processes to select and prioritize upcoming energy efficiency projects.

One limitation of this analysis is the lack of access to building-specific data prior to FY 2019 and in FY 2023. Analyzing trends over only four years is difficult, especially with two of those four years occurring during the COVID-19 pandemic. The COVID-19 pandemic affected building occupancy and ventilation rates, which are linked to energy consumption and related GHG emissions. Data from FY 2020 and FY 2021 were directly affected by the pandemic, and FY 2022 had lingering effects, which could contribute to reductions that cannot be sustained with university buildings at full occupancy. As more building-specific data becomes available in the coming years, the authors hope to frame a more complete picture of energy usage, GHG emissions, and relative trends.

Another limitation lies in the assumption that steam, electricity and natural gas emissions decrease proportionately in the projections. Future work for this study could include scenarios where steam and electricity decrease or increase in different ratios. Specifically, climate change scenarios could be used to predict HDD and CDD in future years; building energy use could be projected the same way. Other scenarios could also be evaluated for buildings, including varying reduction goals based on use type, past renovation information, and/or building scale of GHG emissions (i.e. focusing only on energy-intensive buildings).

This study offers a replicable model for universities that have established their CF and climate goals and now need actionable targets and tracking tools. While many institutions set high-level goals and conduct GHG inventories, this work demonstrates how to translate those inventories into practical, measurable metrics that track forward progress and can forecast future scenarios. This is essential for universities like Pitt that are at a pivotal moment from planning and goal setting to measuring and evaluating progress.

A transferable lesson is the importance of strong collaboration between FM and CAP policymakers. Establishing a mutual interpretation of the CAP goals at the building level allowed more productive discussions on building decarbonization strategies that incorporated FM’s insights on energy efficiency retrofit potential. Another lesson learned is the need to separate goals and tracking for existing and new buildings. Their reduction strategies differ, and focusing on existing buildings from a set baseline year clarified current energy use emissions without conflating it with changes in CBA. Looking specifically at existing buildings in this study allowed the authors to consider existing building stock and meeting the demands of future growth as separate problems with separate solutions. In addition, long-term, consistent tracking of GHG emissions is critical. Robust documentation in over a decade of detailed reports enabled the authors to critically analyze trends and causations.

Finally, this case study provides insight for other universities by sharing both successes and setbacks in Pitt’s decades-long sustainability journey in tracking and reducing emissions. Showing how certain choices have impacted emissions, like prioritizing switching to cleaner fuel sources before building energy use reductions, can help other HEIs decide how to prioritize their own decarbonization strategies.

Universities present a distinctive opportunity to teach the importance of sustainability through practice. In this case study, over a decade of GHG building emissions data at the University of Pittsburgh was analyzed and summarized, including building-specific goals. Yearly GHG emissions reduction targets for each existing building on Pitt’s campus were set based on emissions reduction requirements defined in the PittCAP. This study fills the knowledge gap of evaluating progress on a CAP through a robust set of quantitative and qualitative GHG inventory data on building emissions. Furthermore, CFs for 87 buildings were calculated over four years to help university decision-makers understand the emissions implications of their energy efficiency work.

One of the main findings of this study is that although building-related emissions (from steam and electricity) have decreased since Pitt’s first GHG inventory in FY 2008, the largest contributor to this decrease was not energy use reductions, but from switching steam and electricity generation sources away from coal and to cleaner-burning sources including natural gas and nuclear.

FY 2023 steam demand is above the FY 2008 baseline year, and electricity demand is only slightly less. The largest contributing factors to energy use fluctuations identified in the GHG inventories are HDD/CDD trends, CBA, and occupancy changes due to the COVID-19 pandemic. When normalized over HDD/CDD and CBA, electricity use was higher in FY 2023 than in FY 2008 and steam use was only slightly less. This shows that Pitt has a great opportunity to further GHG emissions reductions through future energy efficiency upgrades to buildings.

In addition, the 87 buildings were evaluated on their progress toward their individual 2037 PittCAP goal (15% total reduction per building). Twenty-one buildings have already reached their 2037 emissions goal, 28 are on track or ahead of schedule and 38 are not on track to meet their emissions goal in FY 2037. In addition, the 10 largest building contributors to university steam and electricity consumption were identified. Of those 10, one has reached its 2037 goal, four are on track and five are not on track. The five high-impact buildings that have not shown adequate emissions reductions between FY 2019 and FY 2022 would be potentially good candidates for energy efficiency upgrades.

The results of this study will aid in energy efficiency project prioritization and communication of PittCAP emissions goals across the university. This study aimed to advance the fields of GHG inventories and climate action planning by showing how long-term GHG inventories can help inform an initial CAP and then be used to track progress over time – demonstrating within one cascade how researchers can help guide a university in making these decisions. Ultimately, this work contributes to the field of sustainability in higher education by showing how universities can move from goal setting to implementation, and by providing lessons learned that will be increasingly relevant as institutions confront the challenge of achieving their decarbonization commitments.

The authors would like to thank Can Atkas, Vaclav Hasik, Kevin Ketchman and Hailey Gardner for their contributions to the University of Pittsburgh’s greenhouse gas inventories. This work was supported by the University of Pittsburgh’s Mascaro Center for Sustainable Innovation and Office of Sustainability.

During the preparation of this work, the author(s) used ChatGPT to improve the readability of certain sections. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the publication’s content.

The supplementary material for this article can be found online.