Impact of the COVID-19 pandemic on the carbon footprint of a Philippine university

The Philippines entered its most prolonged lockdown in 2020 when the coronavirus disease (COVID-19) became a pandemic. Additionally, there has been a shift from physical to online classes at all education levels. Against this backdrop, the restrictions imposed on the education sector could have environmental impacts, including on the carbon emission structure. Here, we compare the carbon footprint before and during the pandemic, determine how the pandemic changed the activities that directly affected carbon emissions, and present reduction methods to minimise emissions in the new normal. We calculated emissions before and during the pandemic to achieve these goals, using the data obtained from University of the Philippines (UP) Cebu. The total CO2 emissions of UP Cebu in 2019 were estimated to be 1420.7 tCO2e, which did not significantly differ from the 2018 emissions. In 2020, the total CO2 emissions were estimated to be 555.8 tCO2e, equivalent to a 60.9% decrease from the 2019 emissions. The per capita emissions in UP Cebu for 2019 and 2020 were estimated to be 0.9 tCO2e and 0.3 tCO2e, respectively—both below the national average. The pandemic caused a significant decrease in emissions per activity, except for fuel-related emissions which increased by 305.8%. In the post-COVID-19 world, especially when in-person classes return, UP Cebu must consider concrete strategies to curb its emissions. Specific decarbonisation methods for each activity were simulated and discussed. The results and reduction strategies presented are relevant to UP Cebu and other higher education institutions in the Philippines and Asia with the same characteristics.


Introduction
The Philippines had its first confirmed coronavirus disease  case in January 2020 [1], prior to the World Health Organization declaration of COVID-19 outbreak as a pandemic. To prevent the spread of the disease, governments have implemented lockdowns that affect social mobility, industrial activities, and economic development. The imposed public restrictions and altered economic activities have decreased global greenhouse gas (GHG) emissions [2,3], mainly CO 2 which declined by 14%-17% [4,5]. Despite the significant fallout in emissions, pandemic-related impacts on GHG concentrations are relatively small on a global scale [6].
The education sector has been highly affected by the COVID-19 pandemic and many have been forced to shift from face-to-face learning to online classes. The shift to online-based learning is regarded as one of the most sustainable approaches for reducing GHG emissions in the education sector [7], although the consequential increase in residential footprint is not accounted here. This is exceptionally true for universities with a high mobility and international students, where reductions in carbon emissions are significant owing to reduced travel [8]. An online education system saves 90% more energy and emits 80% less energy than the traditional on-campus setting, owing to reduced movement and lower energy consumption in household and academic operations [9]. For example, an Indonesian university observed up to 55 times lower carbon emissions while a Japanese university observed up to 84 times lower, both of which are attributed to the lack of daily commute [10]. Moreover, the transition from office jobs to work-from-home arrangements also contributed to the expected 7% decrease in GHG emissions [11] as private vehicle use and public transport were reduced [12]. In the Philippines, tagged as having the most prolonged lockdown globally, the shift to online learning began in the first quarter of 2020. For example, the University of the Philippines (UP) Cebu suspended classes from 13 March 2020 then slowly transitioned from in-person to online classes. When the lockdown was lifted in July 2020, some staff returned to the office but the students continued to study online until the end of the year.
With the potential for emission reduction on campuses due to COVID-19-related restrictions, it is pertinent to clarify how these factors affect the carbon footprint of universities to inform ongoing efforts and policies further, particularly considering a post-COVID-19 world [13][14][15]. This study sought to understand how university carbon footprints in the Philippines compares before and during the COVID-19 pandemic. Furthermore, how did the pandemic change activities that directly affect carbon emissions? Based on these findings, what practices can universities do to reduce carbon emissions post-COVID-19?

UP Cebu
UP Cebu is a public university and a constituent university of the UP System. It has two campuses on Cebu island, the Lahug campus and the South Road Properties (SRP) campus. The Lahug campus serves as the main campus for classes, offices, and other related activities, while the SRP campus serves as a campus extension. We chose to investigate UP Cebu because there had been a comprehensive carbon footprint inventory for the 2018 base year before the COVID-19 pandemic [16]. In this study, the inventory was continued for 2019 and 2020. The term 'pre-pandemic' indicates the period from January 2018 (start of base year inventory) until March 2020, while 'during pandemic' indicates the period from April to December 2020.

Inventory boundaries and calculations
Based on the 2018 emission inventory [16], the following activities in UP Cebu produced greenhouse gases: fuel consumption (scope 1), electricity consumption (scope 2), official travel (scope 3), student and staff mobility (scope 3), solid waste disposal to landfills (scope 3), and untreated wastewater disposal (scope 3). Information on the university population, days present on campus, etc are listed in table 1. Carbon emission was calculated using the basic equation, E = A × EF × GWP where E is the ton carbon emission equivalent (tCO 2 e), A is the activity output, EF is the emission factor, and GWP is the global warming potential. The intensities of each activity are indicated in table 2. Table 6 of the supplementary material lists the emission factor values. A GWP of 28 and 265 were used to convert CH 4 and N 2 O into CO 2 equivalents, respectively [17].

Fuel
The university has four vehicles which were either fuelled by gasoline or diesel. The said vehicles were used to move students and staff between the two campuses from January 2019 until March 2020. When the lockdown was implemented and the movement of public transportation in Cebu was minimised, the university decided to use the vehicles to transport staff from campus to a designated pick-up/drop-off area, and vice versa. UP Cebu kept a record of the fuel consumption per vehicle annually. From this, the carbon footprint was estimated by multiplying the fuel consumption and emission factor for the specific fuel type used. Although the Philippine government mandated a minimum of 2% biodiesel blend, EF for such a fuel blend was not specified. In this study, we used EF for diesel as specified by US EPA (table 6, supplementary material).

Electricity
Emission from electricity was calculated using the monthly electricity consumption data obtained from the monthly electricity bill. Total consumption per campus was determined by adding the electricity consumption measured per building. The EF used in the emission calculation (table 6, supplementary material) was obtained from the Philippines' Department of Environment and Natural Resources and is specific to the Luzon-Visayas grid where the electricity provider of UP Cebu belongs to. The Luzon-Visayas grid is powered by a mix of 49% coal, 25% natural gas, 13% geothermal energy, and the remaining 13% from hydropower, diesel/oil, wind, and solar power [18].

Official travels
Although UP Cebu did not directly calculate the distance travelled (D) per official business trip, it kept a record of the staff 's travel origin and destination. To calculate for the activity, the approach of Santos et al [19] was applied where D = D 0 × 2. Here, D 0 is the distance from origin to destination (km) as determined in Google Maps, and the ×2 accounted for the return trip. In cases where Google Maps provided several search results, the longest route was chosen. The approach assumed that only one fuel type was used. The calculated distances per trip were summed to determine the total distance travelled annually. Different EFs were applied depending on the type of vehicle used. As indicated in table 6 (supplementary material), the EF for air travel is further categorized into short, medium, and long-haul flights.

Student and staff mobility
The distance travelled by students and staff for their daily commute was calculated similarly to section 2.2.3. An online survey was organized to obtain data on the origin, modes, and commute frequency. The mode of transportation identified by the respondents was used to differentiate the emission factor used in the emission calculation. In the case of students, their movement from other provinces or islands to Cebu during the start of the semester was also accounted for. The respondents were also asked in the survey to indicate their movement during the suspension of classes and lockdown due to COVID-19 so as to determine pandemic-related emission. The survey method, including the type of questions asked and assumptions, was adopted from the protocol of Varón-Hoyos et al [20]. The survey questions are enumerated in the supplementary material. The survey was disseminated through emails, social media postings, heads of offices, and organizations. The questionnaire was made available from September to October 2021. For students, 9.0% and 8.5% response rates were obtained regarding commutes in 2019 and 2020, respectively. For staff, 30.4% and 33.2% response rates were obtained for commutes in 2019 and 2020, respectively. To calculate the total distance travelled per year per person, the distance travelled per day was multiplied by the total number of days spent on campus (table 1). The student and staff population data were used to calculate the average distance travelled per year. The average distance was then used to calculate the total distance travelled by the entire university population. This study assumed that the survey results represent the entire population.

Waste
UP Cebu did not have data on the quantity of solid waste and wastewater produced during 2019 and 2020. Regarding solid waste sent to landfill, waste volume was estimated by multiplying population and waste produced per person which is assumed to be 0.05 kg d −1 [21]. Emission was then calculated following the Electricity consumption Low-based on measured electricity consumption by the electricity provider Scope 3 Official travels Medium-based on collected data but assumed a similar route for the return trip Scope 3 Student and employee mobility High-based on data projected from the per capita value determined in the survey with 8.5%-9% student response rate and 30.4%-33.2% staff response rate Scope 3 Solid waste generation High-estimation based on UP Cebu data (survey) and proxy per capita data from an Indonesian university [21] and IPCC recommended values [23] Scope 3 Wastewater generation Medium-based on UP Cebu data (survey) and IPCC recommended values [23] Intergovernmental Panel on Climate Change (IPPC) method for solid waste disposal [22]. Regarding wastewater, the university does not have a treatment facility such that most of its wastewater is released into the city sewer system. Emission from the disposal of untreated water was calculated following the IPCC method for wastewater discharge [23]. The values used in the emission calculation are tabulated in table 7 of the supplementary material.

Data uncertainty
An assessment of the data quality was performed, following a qualitative approach reported by Olatayo et al [24]. Data uncertainty can be handled using qualitative discussions of the results' reliability-qualitative approaches, which express our confidence in the results, are exempt from the formal characterisation of the uncertainty of the input data [25]. The evaluation was based on the data sources, ranging from low to very high uncertainty (table 3). As such: low uncertainty features data that is directly available such as data collected directly from statistics, surveys and direct measurements; medium uncertainty is for indirectly available data collected from proxy statistics; high uncertainty estimations are projections based on factual data; and, very high uncertainty corresponds to data computed by balancing input and output values [24].

Total carbon footprint
The total CO 2 emissions of UP Cebu in 2019 were estimated to be 1420.7 tCO 2 e, which is a 6.6% decrease from the 2018 emissions (table 4). In 2020, the total CO 2 emissions were estimated to be 555.8 tCO 2 e, equivalent to a 60.9% decrease from the 2019 emissions. A similar study in a United Kingdom university found that even within just the 3 month lockdown period, the carbon footprint had already decreased by 30% [26]. The inventory results of UP Cebu imply that the carbon footprint changed drastically during the pandemic. Figure 1 illustrates that the pre-pandemic direct emissions (scope 1) were significantly low but then increased to more than double in 2020 when the pandemic occurred. Indirect emissions (scope 2 and scope 3) did not significantly differ between 2018 and 2019 but decreased significantly in 2020. A detailed breakdown of inventory results per scope is presented in the following sections. Per capita emissions in UP Cebu for 2019 and 2020 were estimated to be 0.9 tCO 2 e and 0.3 tCO 2 e, respectively. Both were below the national average, equivalent to 1.4 tCO 2 e and 1.2 tCO 2 e in 2019 and 2020, respectively [27]. From the qualitative estimation of the data uncertainty degree, higher uncertainties for the scope 3 results are related to the use of projected data from a relatively low-response rate survey and estimated data set. Whereas the low uncertainties for the results of scopes 1 and 2 are due to the data being available and retrieved directly from UP Cebu records.

Emission from fuel consumption and reduction strategies
The fuel consumption of UP Cebu tripled in 2020, primarily from gasoline use. Consequently, the CO 2 emissions from this activity increased by 305.8% when the 2019 and 2020 inventories are compared. This was primary due to increased vehicle usage when the university provided vehicle service for the staff during the lockdown. The vehicle service continued until the end of 2020, even after easing the lockdown restrictions.   The change in gasoline consumption was greater because the primary vehicle used for the service was mainly gasoline-powered. Although the vehicle service had tremendously increased carbon footprint in scope 1, this setup contributed to decreased staff-mobility emissions. This coincides with a study at McGill University, which showed a sharp decline in GHG emissions from transportation, of which the remaining emissions were from vehicle use by limited on-site workers and students at the beginning of the pandemic [15]. The two critical factors influencing scope 1 emissions are the type and amount of fuel consumed [28,29], so the decarbonisation strategy should ideally focus on these areas. A report related to Philippine biofuels elucidated that the current 2% minimum biodiesel blend could reduce GHG emission by 1.2%-1.3%, while if the biodiesel blend is increased to 20%, GHG emission could reduce by 12.6%-12.9% [30]. In UP Cebu, the 1.2%-1.3% GHG reduction would be equivalent to 0.0-0.1 tCO 2 e, while the 12.6%-12.9% would be equivalent to 0.4-0.9 tCO 2 e. Figure 3 illustrates the monthly variation in electricity consumption by UP Cebu pre and during the pandemic. The total electricity consumption dropped by 43.2% from the start of the lockdown and has been continuously under 25 000 kWh until the end of 2020. The average monthly CO 2 emission in the Lahug campus before the pandemic was 28.5 tCO 2 e but was reduced by 38.2% (10.9 tCO 2 e) during the pandemic. For the SRP campus, the average monthly emission was 1.4 9 tCO 2 e before the pandemic and was later reduced by 48.8% (0.7 tCO 2 e) during the pandemic. Due to online education, this reduced emissions from decreased electricity consumption has also been observed among Chinese universities [31]. However, emissions from this category have consistently been the second-largest contributor to the carbon footprint of UP Cebu. Additionally, a degree of reduction below 50% implies that, despite the shift to online classes and  work-from-home setups, the energy consumption of buildings was not significantly reduced. A study on the COVID-19 pandemic and energy use at the University of Almera revealed that this could be due to the varying energy performances of buildings with different structures [14]. It was also reported that the energy consumption of buildings, whether occupied or unoccupied, such as during lockdown, had a significant impact on the environment [13].

Emission from electricity consumption and reduction strategies
To reduce carbon in this category, reducing electricity consumption is essential. A combined strategy tested for the University of Arizona, which reduced annual consumption by 43 437 kWh, included maximising daylight, using high-efficiency light bulbs, and using daylight sensors [32]. Similarly, energy reduction strategies of Osaka University-which include but are not limited to LED lighting, installation of motion sensors, and upgrade of air conditioners-resulted in a 22% decrease in energy use per unit floor [33]. As in the case of UP Cebu, these strategies seem ideal for universities in the tropics where the sun shines the most. Unplugging or converting all electronic device appliances to achieve high efficiency minimises energy use [14]. In UP Cebu, this includes, but is not limited to, computers, television, fans, air conditioners, and laboratory equipment. Furthermore, an energy management program that considers the contributions of renewable energy sources should also be in place. At University Malaya, implementing an energy management system resulted in an emission reduction of over 60% over three years, with a reduction of 90% in the first year [34]. These mitigations would still be applicable in the post-COVID-19 world when physical classes and normal office operations would resume.

Emission from travel and reduction strategies
Emissions from official travel accounted for 6.5% of total emissions in 2019 and 3.6% in 2020. This agrees with the global trend where official travel shares 5% on average of the university carbon emission [35]. Figure 4 shows that local and international travel increased from 2018 to 2019 then drastically decreased in 2020 which was expected due to the closure of borders. All international trips taken in 2019 and 2020 were long-haul flights. Emissions from long-haul flights were 42.0 tCO 2 e in 2019 and 10.5 tCO 2 e in 2020.
Official travel is a crucial activity among globalised and research-oriented universities [36,37] in building academic collaborations, albeit impacting largely on a university's mobility emissions. Figure 5 shows the international travel destinations of UP Cebu staff for official business purposes such as attendance at conferences, meetings, and workshops. From 2018 to March 2020, 107 trips were made internationally, 58 of which were conducted in 2019. Most of the trips were in Asia, with Taiwan being the most common destination. Consequently, Asian trips produced the bulk of international travel emissions which is equivalent to 40.4 tCO 2 e or 49.2% of the total. Trips to countries in Europe, North America, and Oceania make up 20.5%, 23.3%, and 6.9% respectively, of the total international travel-related emissions. A similar study at the University of Montreal, which traced the direction of official travel, found that academic mobility was mainly within proximity of the university [38]. However, regarding the average emission per trip, trips to Asia produce less carbon footprint than farther trips such as to Europe. Assuming the same travel frequency, we compared two scenarios where travel destinations differ (table 5). It was revealed that if business travel is limited within Asia, UP Cebu could reduce its travel-related emission by 39.6%. Three effective strategies to minimise international conference emissions include: choosing a more central venue, prioritising virtual participation, and shifting from annual to biennial events [39]. Although there could be other underlying reasons for each travel, university personnel are highly encouraged to choose a more accessible location and rethink their modes of participation. UP Cebu staff should consider virtual participation in conferences and local meetings whenever possible. In addition to reducing emissions, virtual academic events could also promote diversity, as the distance between countries and continents no longer reaches limits [40].

Emission from mobility and reduction strategies
Student and staff mobility contributed to 50.8% (721.1 tCO 2 e) of UP Cebu's carbon emissions in 2019 and 39.2% (217.8 tCO 2 e) in 2020. In terms of global trends, the share of mobility in the carbon emissions of universities is 45% of total emissions [41]. This indicates that emissions from students and staff mobility before the pandemic (57.3%) were above the global average. In terms of individual emissions, the case of UP Cebu at 4.7 tCO 2 e per student is below half that of Bina Nusantara University in Indonesia with similar modes of transportation at 11 tCO 2 e per student [10]. In particular, student mobility has a disproportionate share of mobility emissions, owing to the wide student-to-staff ratio [7]. Although emissions from mobility decreased during the pandemic, the share of mobility was still high, owing to the return of students and staff to their hometowns, particularly those living far from Cebu City. This campus proximity factor influences commuting behaviour and the types of transportation used, affecting carbon emissions [10]. Figure 6 shows the proportion of students and staff using public and private transportation during 2019 and 2020. About 47%-49% of the students mainly used public jeepneys for transportation, >25% walked or cycled to and from UP Cebu, and the rest used other public and private transportation modes, such as buses  and motorcycles. Meanwhile, 50%-51% of the staff used private vehicles, <10% of the staff walked to and from UP Cebu, and the remaining 40%-41% used public transportation. Between 2019 and 2020, private and public transportation of staff decreased approximately by 1% owing to the university's vehicle service during the lockdown. There is also the possibility that some are discouraged from using public transportation because of the associated high risks of transmission and implementation of social distancing [42].
Since the beginning of the inventory in 2018, student mobility has consistently been the most carbon-intensive activity in UP Cebu. When we examined how students' choice of transportation has changed, we found that fewer students used private and public transportation in 2019 than in 2018, eventually leading to a slight decrease in footprint. More students decided to live either in the dormitory inside the campus or in rental apartments nearby, so students who walked/cycled increased from 24.2% in 2018 to 30.3% in 2019. This trend continued in 2020 when 41.6% of the students walked/cycled to the university.
The implementation of the lockdown led the students and staff to return to their hometowns, as captured in the mobility survey. Most students (53.4%) stayed on Cebu island during the lockdown, 6.9% moved to Negros Oriental, 5.2% moved to Bohol, and <4% moved to other provinces in Visayas and Mindanao (figure 7). There was less movement among the staff, with 84.6% staying in Cebu, whereas the rest moved to provinces in the Visayas and Mindanao areas ( figure 8).
Through the provision of vehicle service, UP Cebu indirectly mitigated its emissions from staff mobility. A previous study pointed out that such a system and other similar approaches like carpooling could not only effectively reduce carbon dioxide by 5% and nitrogen oxides by 7% but also ensure safety and improve road traffic conditions [43]. Regarding the mitigation of emissions from student mobility, the carbon footprint produced by movement at the beginning and end of the semester is inevitable, mainly because UP Cebu caters to students from diverse backgrounds and origins. However, UP Cebu can extend the same vehicle service to students when face-to-face classes return. On an individual level, students' environmental awareness can be enhanced if they are to calculate their carbon footprint and make necessary lifestyle changes [36].

Emission from waste and reduction strategies
The total carbon footprints of solid waste and wastewater disposal in 2019 and 2020 are listed in table 2. Because these two activities are related to the population physically present at the university, we adjusted the calculation of emissions based on the population before and during the pandemic (table 1). About 0.1 tCO 2 e/day was emitted from solid waste disposal prior to the COVID-19 pandemic and was later reduced by 95.1% during the COVID-19 pandemic equating to 0.01 tCO 2 e/day (figure 9). Regarding wastewater  disposal, 0.5 tCO 2 e/day was emitted pre-pandemic and was later reduced by 94.8% during the COVID-19 pandemic, equating to 0.2 tCO 2 e/day. Owing to the absence of students on campus and the decreased number of staff reporting in person, waste production and disposal have decreased significantly.
Twenty percent of the solid waste generated at the university was composted, whereas the remaining 80% was sent to landfills for disposal and possible third-party recycling. This emphasises the strong potential for recycling within school premises, a viable method for reducing the waste sent to landfills thereby minimising the corresponding GHG emissions. Recycling strengthens sustainability practices on campus and can also be a source of income, promoting a circular economy [44]. Solid waste characterisation is the first step in prompting such changes and improving the university's waste management system, which can help identify opportunities for recycling and composting. Knowing the composition and sources of waste will facilitate the development of an integrated solid waste management plan that includes waste generation reduction strategies with proper segregation, collection, and treatment activities, which will reduce or eliminate the need for landfills [45,46].
At UP Cebu, wastewater from toilets and lavatories is collected and treated on-site using septic tanks which have been reported to be old. Therefore, there is a high probability of seepage into the ground. Regarding the options to reduce emissions, there is a need to upgrade the existing wastewater treatment system on campus, such as expanding the capacity and introducing a small-scale treatment in situ. An anaerobic digestion system can also receive the organic fractions of the generated solid waste, to capture and recover the emitted methane [45]. In addition, because wastewater treatment can also directly contribute to reducing solid waste [47], after removing contaminants, the resulting sludge can be used for landscaping and soil amendment instead of landfilling [48,49]. Lastly, the upward trend of the overall contribution of the emissions from waste indicates the need for in-depth future studies, which rely on the direct measurement and country-specific variables. In doing so, the associated high uncertainty of the results can be reduced and the impact of the potential waste composition change can be considered. For instance, to examine if using and discarding personal protective equipment such as masks changed the amount and characteristics of the emissions.

Conclusions
The total CO 2 emissions of UP Cebu during business-as-usual years were similar, at more than 1400 tCO 2 e. In 2020, the total CO 2 emissions were estimated to be 555.8 tCO 2 e, equivalent to a 60.9% decrease from the 2019 emissions. The pandemic and shift to online learning caused a significant decrease in emissions per activity, except for fuel-related emissions which increased by 305.8% due to the provision of a transportation service especially during the lockdown. As face-to-face classes are expected to resume in mid-2022, UP Cebu must consider mitigation strategies to curb its emissions. Besides the specific and obvious emission strategies mentioned in the paper, UP Cebu could also consider larger decarbonisation projects. For example, building a wastewater treatment facility is ideal for reducing emissions from wastewater. In the case of fuel-related emission, UP Cebu can produce its own biofuel, possibly resulting in carbon neutrality in this category. The study's methods and results could be used by other higher education institutions (HEIs) as a model in calculating for their own carbon footprint. For example, in the Philippines, the study method is beneficial because there is a general lack of know-how in GHG scoping and the choice of appropriate emission factor is not yet clearly sorted at the national level. However, this study proved that despite such limitations, a GHG inventory is still possible using US EPA and/or IPPC default values. Additionally, the results of this study highlights which activities are most or least carbon-intensive, and which are most or least likely affected by the pandemic-related changes. Quantifying carbon emission is already a significant step towards sustainable goals and practices among HEIs and other similar organisations for the new normal. For future studies, we recommend increasing the sample size of the mobility survey to improve the representation of the entire university population. Due to high data uncertainty in some carbon-emitting activities, we recommend direct data collection whenever possible to improve estimate accuracy. Lastly, connected with the previous point, to clarify the magnitude of influence each data set has on the results, and to improve the recommendations for emission reductions, future studies would benefit from applying a sensitivity analysis.

Data availability statement
The data generated and/or analysed during the current study are not publicly available for legal/ethical reasons but are available from the corresponding author on reasonable request.