A review of empirical data of sustainability initiatives in university campus operations

Given the need to actively address the challenges of climate change, university leaders have a growing interest in reducing their campuses’ environmental impact. This article carries out a comprehensive literature review on the implemented actions and initiatives in university campuses reported in scientific publications. In addition, case studies carried out in universities are also reviewed, giving particular attention to the methods and tools used, targeting the current trends in sustainable campus scientific research. Key actions and initiatives were identified and categorized according to Energy, Buildings, Water, Waste, Transportation, Grounds, Air and Climate, and Food. Results show that the increase in energy generation on campus and the decrease of energy consumption in buildings are by far the leading policies adopted, however with limited dissemination of their impact. Moreover, there seems to be a tendency for countries with higher income economies to engage in initiatives that involve greater investment, such as the adoption of renewable energy systems or efficient buildings systems. The need to establish an integrated framework to disseminate and monitor the impact of key actions and their feasibility is suggested, in order to leverage strategic programs and actions, helping to optimize investments, and leading advances towards a sustainable university campus.


Introduction 1
Since the Brundtland Report and the establishment of the Sustainable Development (SD) 2 concept, governments and public institutions are aware of the responsibility in considering 3 environmental, economic and social sustainability in their activities.Higher Education 4 Institutions (HEIs) play a special and crucial role, due mainly to their inherent characteristics 5 and mission: a) as educational institutions, HEIs have the responsibility of preparing future 6 leaders and citizens to be more conscious and active in the dissemination of sustainable 7 principles; b) as owners of physical structures that consume energy and other resources, HEIs 8 have the opportunity to implement actions to decrease costs and impacts associated to campus 9 operations; c) as administrative structures, HEIs have to manage people from diverse socio-10 cultural backgrounds, finances and, still, seek an engagement between staff, academia and 11 community; and d) HEIs have the social responsibility of incorporating all these issues, acting 12 by example.13 Considerable work on the subject of sustainability has been done, taking an increasingly 14 important place in the lifespan of any HEI, either through governance or teaching models, 15 and/or through the management of the campus buildings.The number of HEI websites 16 exclusively dedicated to reporting sustainability practices have increased, providing 17 information to the general public on targets, planned initiatives and eventually on the current 18 status of execution.However, increasing evidence in literature indicates a substantial number 19 of failures in implementing sustainability initiatives (Mohammadalizadehkorde and Weaver, 20 2018), being the main reasons given by the HEIs themselves identified in literature.21 Apart from the information provided in websites, reports or declarations, it is important: i) to 5 -38 reviewed by Wright (2002) and Lozano et al. (2015Lozano et al. ( , 2013) ) are the first official expression of a 75 commitment to embrace environmental conservation.In 1990, the Talloires Declaration ("The 76 Talloires Declaration," 1990), placed emphasis on the need to enhance education and 77 environmental literacy, resource conservation and to involve stakeholders and community, 78 amongst others.Subsequent declarations have covered these topics as well.Wright (2002) 79 identified several emerging themes common to declarations and institutional policies, such as 80 physical operations, environmental literacy, curriculum and research, or public outreach.81 Cortese (2003) identified education, research, operations and outreach as an integrated system 82 of a sustainable university.However, Ceulemans et al. (2015) noted the complexity in 83 ascertaining what could be considered as an indicator of sustainability, mainly due to the 84 different objectives and interpretations that the various stakeholders could have.85 Conversely, the vagueness with which the commitments were addressed throughout the 86 different statements is reflected in the constant need, over time, for new and renewed 87 declarations, initiatives and commitments, raising the question whether they have actually had 88 any practical impact.As an example, after ten years following the Halifax Declaration's 89 Action Plan ("Halifax Declaration," 1992), the highest implementation rate of the proposed 90 initiatives was below 50 % (Wright, 2003).91 To overcome the HEIs lack of commitment to fulfill the purpose of the declarations, several 92 methodologies, models and tools have emerged.They have been reviewed in literature, acting 93 in the different stages of the sustainability process: a) implementation (Amaral et al., 2015;94 Testa et al., 2014); b) assessment (Berzosa et al., 2017;Fischer et al., 2015;Shriberg, 2002;95 Yarime and Tanaka, 2012); and c) reporting (Alonso-Almeida et al., 2015;Ceulemans et al., 96 2015;Lozano, 2011;Yarime and Tanaka, 2012).97 Despite the wide diversity of available tools, there is no current information on the status of 98 application of many of these tools.Currently, international rankings such as UI GreenMetric 99 (Lauder et al., 2015;Suwartha and Sari, 2013) or rating systems such as STARS (Fischer et 100 al., 2015;Lidstone et al., 2015;Yarime and Tanaka, 2012) have gained prominence, 101 becoming the most used in practice, since universities are able to audit, compare and 102 communicate their performance with each other and with stakeholders.103 Whatever the model used for implementing sustainability, top-down and bottom-up are 104 possible approaches.Nonetheless, both show weaknesses, as North and Ryan (2018) state that 105 top-down initiatives tend to fail when arriving at the community, and Ávila et al. (2017) 106 suggest that bottom-up initiatives may fail due to the lack of funding and support from the 107 administrative boards.Therefore, a mixed bottom-up and top-down approach (Ramísio et al., 108 2019), also indicated as in-between (Brinkhurst et al., 2011), is suggested.A great potential of 109 successful initiatives in the long term is reported, when carried out by faculty and staff, where 110 the operations experts, often belonging to sustainability offices or centers, are included.111

The contribution of operations dimension to the sustainable campus 112
Regardless of the type and functional program, buildings account for 30 % of total energy 113 consumption worldwide (International Energy Agency, 2018).In the European Union, 16 % 114 of non-residential buildings are universities and other educational institutions (European 115 Commision, 2013).Moreover, it is estimated that approximately 70 % of the life-cycle costs 116 of a building are incurred with operations, maintenance, utilities, and renovations (Carlson, 117 2012).For HEIs, this represents a huge slice of total expenditures; therefore, it is only natural 118 that major attention is given to reducing costs and the consumption of resources in the use 119

phase. 120
Overall, university campuses comprise a large amount of built-up areas with a substantial 121 number of users, involving complex and diverse activities on a continuous basis, if residences 122 are included.The occupancy profiles are so vast and the use of spaces so diverse that a 123 university campus resembles a community or a city district (Ávila et al., 2017;Gu et al., 124 2019).In fact, environmental concerns are quite similar to those verified in urban districts or 125 communities, involving not only greenhouse gas (GHG) emissions and the consumption of 126 resources such as energy, water, materials and food, but also the management of 127 transportation and waste production.128 Under the so-called operations umbrella, the environmental performance of buildings, 129 facilities and outdoor spaces may be improved with specific actions and initiatives that may 130 produce higher savings on a short to medium term basis.This may explain the attention that 131 the field of operations has been receiving in literature, when compared to other aspects such 132 as education or outreach (Yarime and Tanaka, 2012).133 The maintenance and management of facilities is frequently under the supervision of technical 134 departments that report to the administration or the rectorate and are not necessarily related to 135 faculty or research.Sustainability concerns have brought new challenges to these teams, since 136 these require an integrated approach, which include aspects typically dispersed and performed 137 independently by diverse staff members, as is the case of energy, waste, food or purchasing 138 areas.Therefore, sustainability offices play a crucial role, assuming diverse typologies 139 (Adomßent et al., 2019;Soini et al., 2018), where the operational aspects are predominant 140 (Filho et al., 2019).The main advantage of these structures is to gather qualified and 141 motivated people for the holistic implementation of actions, campaigns or projects, able of 142 engaging the academic community.As also pointed out by Filho et al. (2019), specific 143 campus operation issues may only be correctly addressed by HEI technicians.Consequently, 144 providing staff with training and/or guidelines is also an essential task in order to involve all 145 those concerned in the conservation of resources (Ferrão and Matos, 2017).Acting towards 146 reducing consumption and increasing the efficiency of university buildings is not only a 147 mission for the technical staff; it is a unique opportunity of working in a living laboratory 148 where actions may be planned, implemented, monitored and evaluated by professors and 149 students, as Shahidehpour and Clair (2012)  It is clear that the barriers to implementation are common to a variety of sustainability issues, 162 such as the lack of funding, lack of human and technological resources, lack of support from 163 administration, and resistance from staff, students or directors in moving forward.In this 164 sense, even initiatives within a technical area as is the case with operations, are strongly 165 influenced by internal social, organizational and economic policies and constraints.This 166 analysis is consistent with previous results based on other methods found in literature (Barth, 167 2013;Godemann et al., 2014;Hoover and Harder, 2015;Velazquez et al., 2005).Despite such 168 barriers, respondents also reported the support from management, funding and/or community 169 engagement as the main motivators to the successful implementation and prosecution of 170 sustainable university principles.171 The specific field of operations is mostly driven by financial incentives and regulatory 172 compliance, and usually obstructed by the lack of leadership support or by resource 173 constraints (Ralph and Stubbs, 2014).Several strategies may overcome these obstacles.174 Simple no-cost actions such as awareness campaigns or switching off equipment during 175 unoccupied periods are explored by Gul and Patidar (2015) and by Ferrão and Matos (2017).176 However one of the most cost-effective may be the paid-from-savings scheme, where funds 177 resulting from energy conservation measures are applied on financing further energy-related 178 projects (Faghihi et al., 2015), even when reluctance hinder its establishment (Maiorano and 179 Savan, 2015).180

Material and Methods 181
The literature review was carried out by focusing exclusively on scientific documents 182 published since 2010.These were collected by searching on Science Direct and Google 183 Scholar websites, with the keywords "sustainab", "university", and "campus".A total of 357 184 publications were retrieved, of which 250 were journal articles, 38 conference proceedings, 66 185 book chapters and 3 others.186 The articles were organized according to the field of sustainability in HEIs: Education, 187

Operations and Governance. 188
Within the collected publications, 120 were selected as they report actions and initiatives in 189 the area of Operations that were actually implemented, thus allowing ranking the most 190 common practices, as well as identifying their impact according to eight key subareas.This 191 terminology is based on the STARS rating scheme (AASHE, 2017) and can be briefly 192 Additionally, 112 articles presenting universities as case studies were also reviewed.These 216 focus on the development of methods and tools for implementing sustainability-related 217 actions, and even when not effectively executed, they help to understand the current trends on 218 sustainable campus scientific research.219 The complete and detailed description of the initiatives and of the case studies is accessible in 220 the Supplementary Material.A total of 424 initiatives were retrieved from the mentioned 120 221 articles, along with 201 case studies from the 112 articles, respectively.These are organized 222 according to the eight subareas related to campus operations and provide information on the 223 methodologies used and the results achieved, according to the available data in literature.224 4 Results and Discussion 225

Overview 227
In line with Velazquez et al. (2006), and according to Figure 1, results show that the highest 228 number of actions are associated with the energy paradigm.It is noticeable that Energy and 229 Buildings, the subareas that present the greatest number and diversity in initiatives, are 230 related.The Buildings subarea involves measures to reduce energy consumption and the 231 Energy subarea comprises actions to increase energy generation and distribution at the 232 campus level.This disparity may be justified by the larger and more visible savings in these 233 areas, which allow HEI decision-makers to expect a likely and tangible return on their 234 investment.In addition, there is a global awareness that energy is one of the major 235 contributors to environmental footprint, and acting on the reduction of energy consumption is 236 also linked to reducing embodied GHG emissions.237 In general, the focus on renewable sources for energy generation is the most substantial 238 initiative (12 % of the 424 initiatives), and summing its application to water and/or ambient 239 heating and to combined heat and power (CHP) systems, a significant expression is reached 240 (4 % and 2 %, respectively).This trend is in line with the transition from high-emission fossil 241 fuels to clean energy systems to meet the climate targets and the energy independency of 242 national policies.However, energy supply from renewable sources is reported as covering 243 rather a disparate percentage of the annual electricity demand -from 3,76 % (Kalkan et al., 244 2011), between 35 % and 40 % (Eggleston, 2015;Helling, 2018;Radhakrishnan and 245 Viswanathan, 2015) and above 80 % (Kobiski et al., 2015) to almost the overall electricity 246 demand (Walker and Mendler, 2017).Thus, results on its feasibility are not consensual.247 Kalkan et al. (2011) consider it is not a profitable investment, while Paudel and Sarper (2013) 248 consider it is, showing a payback period of 8 years.Supporting initiatives may allow 249 surpassing undesired results, such as the use of energy storage systems and the 250 implementation of microgrids.Machamint et al. (2018) concluded that the benefits of the 251 microgrid compensate the investment cost, and Shahidehpour and Clair (2012) and Washom 252 et al. (2013) showed that the combination of renewable generation and microgrid can supply 253 50 % and 92 % of campus electricity load, respectively.This illustrates how acting only on 254 the supply side may be reductive if omitting the demand side.Accordingly, Leal Filho et al. 255 (2019) concluded that a majority of HEIs are committed to energy efficiency and the 256 implementation of renewables, however these cover a small portion of energy demand.257 The environmental certification of university buildings, especially by LEED, is the second 258 most found initiative (5 %) although results are not unanimous.While Petratos and Damaskou 259 (2015) state that LEED-certified university buildings are designed to consume less 50 % than 260 other similar buildings, some studies show that these may consume more energy than non-261 certified ones (Agdas et al., 2015).Therefore, the use of this rating system may be justified 262 with the fact that majority of the participating HEIs are North American and listed on STARS 263 ranking, which foresees the LEED certification of university buildings -from the 26 North 264 American HEIs that have LEED certification, 19 are ranked at STARS (STARS, 2019).In 265 addition, the BREEAM-certified examples show how buildings designed to achieve a 266 certification do not necessarily demonstrate improved performance in the operational phase, 267 as there may be a gap between estimated and monitored consumptions (Forman et al., 2017;268 Gupta and Gregg, 2016).Inappropriate building management systems and erroneous 269 prediction of end-user energy behavior are some possible causes.270 The most consistent results are related to initiatives in the Buildings subarea, with either 271 passive design actions or active systems.Carlson (2012) highlights a reduction in energy use 272 of 70 % from passive building design, being 80 % more efficient than a conventional 273 building.More specifically, the improvement of efficient opaque envelopes through thermal 274 insulation shows a substantial reduction in building energy demand (Geng et al., 2013).275 Reductions in energy consumption for lighting and air conditioning systems are reported to 276 vary between 7,5 % (Escobedo et al., 2014) and 40 % (Opel et al., 2017), and reach up to 277 60 % (Jain and Pant, 2010).278 Other initiatives and subareas stand out: the treatment of wastewater (2 %), rainwater 279 harvesting to be used in the irrigation of green spaces (4 %), which Edwin et al. (2015) 280 demonstrated saved from 25 % to 30 % of water used on irrigation, while Walker and 281 Mendler (2017) indicated a percentage of 100 % of wastewater treated onsite; the waste 282 separation for recycling purposes (3 %), which increased separation and recycling rates (Geng 283 et al., 2013;Reidy et al., 2015); the adoption of a bicycling culture, namely through support 284 facilities, "pick-and-ride" schemes or for internal small cargo transportation to reduce the use 285 of motorized vehicles on campus (3 %); and the use of native plants in green open spaces 286 (2 %).287 The variation in results of each initiative is attributed to local specificities, but also to the type 288 of use, as occupant behavior significantly influences energy and resource consumption in 289 university buildings.Masoso and Grobler (2010) argue that more than half of the energy in a 290 Botswana university is consumed during non-working hours, as occupants leave lights and 291 equipment always on, becoming evident the importance of altering the use of energy services, 292 in particular users' behaviors and/or control automation.Although in a reduced number, some 293 initiatives seek to motivate students and staff in reducing resource consumption and waste 294 generation.The use of real time smart meter displays purposely designed for users control 295 (2 %) has shown a reduction in energy demand ranging from 6,4 % to 9 % (Boulton et al., 296 2017;Chiang et al., 2014;Sintov et al., 2016).In the subarea of Waste, several campaigns on 297 recycling and on the reduction of paper use (2 %) have revealed that about 74 % of academia 298 has changed their behavior (Cole and Fieselman, 2013), an increase of 10-12 % in the overall 299 campus recycling rates (Tangwanichagapong et al., 2017) and a reduction of up to 58 % on 300 paper use (Zen et al., 2016).In Grounds subarea, a study reported the involvement of all the 301 academic community in cleaning the campus' public open areas, promoting the discussion of 302 beneficial practices for the environment (de Castro and Jabbour, 2013).Despite all these 303 initiatives, Transportation is the subarea that obtains the greatest contribution from the 304 community.All of the initiatives (8 %) were related to the decrease in use of fossil fuel 305 vehicles.The use of automation for active systems is another reported strategy to deal with 306 detected energy waste.The use of controls for artificial lighting, systems setpoints, 307 temperatures and/or gas use during unoccupied periods represents 4 % of the initiatives; 308 Granderson et al. (2011) account for a reduction between 30 % to 35 % in energy demand, 309 and Coccolo et al. (2015) estimate about 13 % in heating demand due to changing air-310 conditioning setpoints.Integrated energy management systems are more comprehensive 311 solutions to control active systems (4 %), namely the HVAC schedules, lighting and 312 appliances, allowing it to be done remotely with the help of information and communications 313 technologies (Ferrão and Matos, 2017;Gomes et al., 2017).Reidy et al. (2015) and Gomes et 314 al. (2017) report a reduction of 20 % and up to 40 % in energy consumption, respectively.A 315 combination of approaches was described by Kettemann et al. (2017) in which a mobile 316 application with interactive maps allows all stakeholders to report any environmentally 317 relevant observation.318 Several initiatives are interrelated and show how adopting a holistic approach and/or working 319 under a global plan can be beneficial.For example, the use of native plants on campus 320 landscape, contributes not only to the CO 2 capture (Oyama et al., 2018;Sundarapandian et al., 321 2014) but also to the conservation of local biodiversity and the reduction of water use for 322 irrigation (Radhakrishnan and Viswanathan, 2015).The treatment of organic waste from 323 university restaurants is commonly used as fertilizer for the campuses' green spaces (de 324 Castro and Jabbour, 2013;Eatmon et al., 2015;Jain and Pant, 2010;Najad et al., 2018;325 Nandhivarman et al., 2015;Reidy et al., 2015).The "waste-to-energy" principle is also 326 conveyed, with examples of electricity or biogas generation from waste sources (Bauer, 2018;327 Helling, 2018;Nandhivarman et al., 2015;Tu et al., 2015).A significant reduction in 328 petroleum gas use (Nandhivarman et al., 2015) and CO 2 emissions (Tu et al., 2015) is 329 reported.This strategy draws up a good example of a circular economy, and may bring an 330 important contribution to decreasing HEIs environmental impact, by closing the loop on two 331 crucial actions -energy generation and waste treatment -as further supported by the 332 European Commission (Antoniou et al., 2019;Pan et al., 2015).333 In general, articles reporting practices and initiatives are not precise in describing the methods 334 used to apply the reported actions.This in turn denotes that limited information is available on 335 the methods and tools applied in real context.In addition, even though a significant amount of 336 initiatives is described, neither the impact nor the achievements are provided, making it 337 difficult to quantify their importance, especially in indicators as Grounds, Air and Climate or 338 Food.339

Framework of initiatives in national scenarios 340
Initiatives found in literature involve 106 HEIs dispersed over the world in 31 countries, 341 North America and Europe being the regions with the highest number of identified 342 institutions (see Figure 2).343 When comparing the total number of existing HEIs listed in Scimago (Scimago, 2018) and 344 the number of identified HEIs, these represent 3 % of a total of 3234.As may be observed in 345 Figure 3, the ratio between participating and total HEIs by country is notably low.As an 346 example, the USA is by far the country with the most reported publications and actions; 347 however, when compared to the national panorama, only 34 out of 432 institutions were 348 identified, which represents about 8 % of American HEIs. 349 This finding is corroborated by the conclusions presented by Lozano (2011) and Townsend 350 and Barrett (2015), who report the lack of commitment from HEIs in measuring and reporting 351 the progress of sustainability initiatives.Luxembourg must be seen as an exception, since the 352 19 -38 country only has one HEI, which justifies the percentage of 100 % of the university' 353 publishing initiatives.Yet, it is acknowledged that other studies and other countries may have 354 been left out of this review derived from the search method used.355 presents the highest energy values as well as the highest number of initiatives related to 366 Energy, particularly to renewable energy systems.Similarly, it is also the country with the 367 highest waste generation and the highest number of initiatives related to Waste, and the same 368 is noticed in the Transportation subarea.However, it is not possible to establish any other 369 association, as none of the remaining countries or initiatives seem to be related to the value of 370 energy and water consumption, or to waste generation or GHG emissions.On the contrary, 371 China carries higher values of waste generation and CO 2 emissions but a lower number of 372 initiatives on Waste and none on Air and Climate.In this sense, with the available data it is 373 difficult to establish a relation between the national scenarios for resources consumption, 374 waste generation and/or GHG emissions and the actions taken by HEIs.375 Figure 6 compares all of the identified initiatives with the gross domestic product (GDP) of 376 the countries where they were promoted (World Bank, 2017).The aim is to understand 377 whether the initiatives that required a large initial investment, namely those based on 378 technologies such as energy generation and distribution systems, or active buildings systems, 379 as lighting, HVAC or equipment, are associated to high-income countries.380 21 -38 Again, the USA shows a high GDP and the highest number of initiatives in Energy as well as 381 in other subareas.In fact, high-income countries such as the European member-states tend to 382 invest more on Energy and Buildings initiatives.However, there is no linear association.For 383 instance, China presents the second highest GDP value, but a low number of initiatives that 384 require a large initial investment.It is noticeable that the percentage of participating countries 385 GDP does not correlate to the percentage of initiatives; nonetheless, a trend is apparent 386 denoting a higher number of initiatives with higher income countries.In order to further 387 analyze this relationship, other factors should be explored, in particular the existence of local 388 or governmental incentives and/or financing programs for the adoption and implementation of 389 energy generation or efficiency measures that are available in some countries (ACEEE, 2018;390 Sustain, 2017).The example given by Drahein et al. (2019) providing information on the 391 inexistence of financing programs for energy or water efficiency in Brazil, could help to 392 understand the significant number of initiatives found in other areas that do not require initial 393 support, such as Waste.However, the lack of detailed information, either in the reviewed 394 articles or in web contents, inhibited the possibility of an accurate and transversal analysis, 395 leaving it open for further investigation.396 Some survey-based studies mention a relation between geographical distribution and 397 particular drivers and barriers.Ralph and Stubbs (2014)  provide.Also, some of these surveys showed that there has been a stronger interest in 404 bringing sustainability forward in HEIs in Europe than in other continents (Leal Filho et al., 405 2019a;Lozano et al., 2015).However, Blanco-Portela et al. (2018) show that the typology of 406 difficulties and stimuli to the implementation of sustainability in Latin American HEIs is 407 similar in all the inquired countries.Ávila et al. (2019) found that, even with different levels 408 of adoption maturity of innovation and sustainability in each continent, the same barriers are 409 found in all geographies.They concluded that developed countries are leading in 410 sustainability implementation, while developed ones were considered laggards, which is in 411 accordance with the trends observed in this work.412

Case studies on sustainable campus operations 413
Articles using universities as case studies show current concerns and advances in research in 414 this specific domain reflecting, in some cases, the attempt to tackle real challenges and 415 difficulties that universities face.However, there is no evidence that the conclusions of these 416 studies have any real execution.417  Regarding Energy subarea, literature has been converging to the studies of renewable energy 424 generation potential and the estimation of buildings energy consumption to better act on its 425 reduction.Photovoltaic systems are prominent, representing 38 % of the case studies on 426 renewables, however the combination of sources has created a growing interest (18 %), as 427 well as the methods for optimizing their management (Bonanno et al., 2012;Bracco et al., 428 2014;Dursun, 2012;Ghenai and Bettayeb, 2019;Park and Kwon, 2016).These may 429 contribute to a cost-effectiveness that Kalkan et al. (2011) and Kwan and Kwan (2011) stated 430 was not always possible when a single source is used -in these cases, solar.431 Simulation software is the tool most commonly employed to assess the potential and 432 feasibility of diverse renewable sources (20 % of Energy ), and their possible combination 433 and integration on microgrids, according to each campus conditions (Çetinbaş et al., 2019;434 Dursun, 2012;Mancini et al., 2017;Manni et al., 2017;Mewes et al., 2017;Mytafides et al., 435 2017;Park and Kwon, 2016).Learning algorithms are used as surrogate methods for 436 simulation (12 % of Energy), in order to produce robust estimations of energy consumption 437 (Hawkins et al., 2012;Jovanović et al., 2015;Yuan et al., 2018).438 Regarding the Buildings subarea, there is an attempt to improve current energy and thermal 439 performance of university buildings through the analysis of various retrofitting strategies.440 Again, simulation engines are the ones most employed in this area (44 %) to evaluate the 441 impact of improving thermal insulation of roofs and façades, the glazing type of windows 442 (Ascione et al., 2017;Manni et al., 2017;Mytafides et al., 2017;Zhou et al., 2019) or even the 443 replacement of the existing lighting system (Fonseca et al., 2018).Life Cycle Assessment 444 (LCA) is commonly used as well, when the objective implies a broader perspective and an 445 analysis of the environmental impacts of retrofitting measures along the lifespan of the 446 buildings (Huang et al., 2012;Tabatabaee and Weil, 2017). 447 In what concerns the other indicators, the scarcity of publications is notorious.Nevertheless, 448 LCA remains a procedure widely used in areas as diverse as Transportation, Waste or Air and 449 Climate (11 % of the total case studies).It is applied in the evaluation of management 450 strategies for parking on campus or the shift from private cars to public transport (Cruz et al., 451 2017); in the comparison of the environmental impact of using information in paper or 452 electronic (Ingwersen et al., 2012); or even to estimate the total CO 2 emissions of a university 453 campus, taking into account the direct emissions of facilities, the indirect emissions of 454 purchased electricity and others, namely from commuting or waste (Sangwan et al., 2018).soft modes of transportation, such as bicycling, are noteworthy (Çelebi et al., 2019;Peer, 458 2019;Ryu et al., 2019;Zhu et al., 2019).Regarding Air and Climate, the assessment of 459 environmental impacts of the whole campus, of which the development of methodologies to 460 estimate GHG emissions and mitigation policies (Leach et al., 2013;Medina and Belcena, 461 2018;Sangwan et al., 2018;Williamson, 2012) is highlighted.Also, broader studies were 462 found approaching a more general concept of sustainability on campus; the proposal of a 463 Green Campus model based on Multi-Criteria Decision Analysis for operations indicators 464 (Ribeiro et al., 2017); the proposal of a holistic tool covering all indicators, acting as an open 465 system by allocating a database with international implementation experiences (Baletic et al., 466 2017).467 The developed methodologies and tools are further clarified and thoroughly described in 468 research that uses the universities as case studies, either to develop exercises or to plan future 469 actions.However, in these cases it is not clear when a university is simply used as a case to 470 experiment a methodology, to develop an isolated exercise, or whether it is part of a wider 471 sustainability plan.Nevertheless, theoretical exercises as those carried out by of Gul and 472 Patidar (2015) or Costa et al. (2019) demonstrate their importance in raising awareness to 473 improve management decisions and policies against actual conditions.474

Conclusions 475
Regardless of the vast amount and scope of initiatives, this work provides supporting 476 information in the identification of strategies and opportunities for institutions to improve 477 environmental sustainability, with tested results on real case scenarios.478 By outlining the actions and institutions, not only a better understanding of the intervention 479 areas and their success is provided, but also the barriers to their implementation, disclosing 480 the impact of possible technical or local context reasons, in addition to those already Having specialized teams, namely sustainability offices, has been a valuable contribution to 484 advancing the cooperation and alignment on decisions and actions.Assuming the campus as a 485 living laboratory may represent a significant contribution to training in a sustainability-486 learning environment, to stimulate scientific research in this field, and also to foster the 487 adoption of more sustainable behaviors in the future.488 The reported initiatives and HEIs are strongly diverse and dispersed worldwide, and do not 489 present relevant relations between national indicators for resources consumption, waste 490 production, emissions generation and the actions taken.However, some trends were 491 identified, being perceivable that HEIs from upper middle-and higher-income countries tend 492 to implement more sustainable initiatives.The results are also highly variable between 493 universities, due to the specificities of each campus, culture, climate or policies.494 The field of operations is the most endorsed in literature, particularly in the area of decreasing 495 energy consumption in buildings and increasing the use of renewable energy on campus, both 496 in practical situations and in case studies.A small number of studies focused on subareas such 497 as Grounds, Air and Climate, and Food for implemented initiatives, and Waste and Water for 498 case studies were found.Moreover, the limited results and the lack of a connection between 499 initiative and impact can hinder reaching definite conclusions on the efficacy of implementing 500 the proposed initiatives.Regardless of the methodology to be adopted, a sustainability culture 501 reflected in an integrated strategy seems to produce better results, rather than implementing 502 isolated actions, as demonstrated by the greater impact of the studies presenting a 503 combination, either in one or in various subareas.504 This analysis also suggests that the successful implementation of sustainable initiatives in 505 HEIs is strongly influenced by internal social and governance restraints even when dealing 506 with a technical component as campus operations.In this sense, this work provides the basis 507 for follow-up mixed-methods studies.Questioning participating countries in relation to local, 508 national, social and economic aspects would provide useful insights into better understanding 509 specific differences and similarities and, thus, support the choice of the best initiatives for 510 each case.In order to contribute to a sustainable university, campus operation initiatives must 511 bring social and economic benefits -and, as literature has shown, to be outreached.To 512 overcome barriers, due to financial difficulties or even resistance to change, investments need 513 to be optimized and effective, in order to show that it is worth investing in strategic actions 514 that bring numerous benefits.An approach based on a ranking of measures with some 515 criterion -investment, payback, energy payback time, consumptions reduction, etc. -could 516 bring positive insights to support decision-making.517 Nevertheless, the exhaustive qualitative analysis that was carried out raises the need for a 518 future quantitative approach, as well as an investigation into the feasibility of the 519 implemented actions that may contribute to the development of comprehensive frameworks 520 able to push forward the sustainable campus principles and practices.521 In this sense, more research is needed -or at least, more empirical information with greater 522 and better dissemination of plans and their results -in order to produce more robust findings, 523 capable of being generalized and eventually inspiring for other universities.524 comprising mostly the deployment of energy generation systems from 194 renewables and respective distribution and storage; 195 Buildings: initiatives that act on the energy performance of buildings, being related to 196 active systems or passive design; 197 Waste: initiatives related with reducing solid waste production, by reducing, recycling 198 and/or reusing actions.Diverse types of waste are considered, such as food waste, 199 consumable materials or hazardous waste; 200 Water: initiatives related with water management and treatment, the second most 201 consumed resource on campuses; 202 Transportation: initiatives that promote sustainable transportation systems serving 203 campuses and their community, namely by decreasing the prevalence of fossil fuel 204 vehicles or by proposing alternative means of commuting; 205 Grounds: initiatives on campuses public and open spaces, as the management of 206 sustainable landscape and healthy ecosystems; 207 Air and Climate: initiatives related with the reduction of GHG and other pollutant 208 emissions and improvement of the air quality, either indoor as outdoor.In addition, 209 initiatives to counteract climate change on a wider perspective of environmental 210 footprint; 211 Food: initiatives related with commitment with sustainable food systems, which intend to 212 mitigate environmental and social impacts of industrial food production, by privileging 213 organic ingredients and/or local producers, also reducing pollution associated to 214 transportation.215

Figure 1 .
Figure 1.Number of initiatives distributed by operations subareas.

Figure 2 .
Figure 2.World weighted distribution of the HEI involved in the initiatives found on literature.

Figure 3 .
Figure 3. Association between the number of initiatives and institutions by country, and ratio of the participating HEIs and the total number of HEIs by country.
highlighted different English and 398 Australian national contexts and specific governmental requirements, which justifies the 399 slight variances found in HEI motivations.Molthan-Hill et al. (2019) found a geographical 400 pattern in HEIs regarding the importance given to climate change.Salvia et al. (2019) 401 investigated the extent to which the various approaches to SDGs are related to local contexts, 402 and noticed a possible relation between local challenges and the areas of interest that SDGs 403

Figure 7
Figure 7 displays the distribution of published case studies by each operations subarea.Unlike 418

Figure 7 .
Figure 7. Number of case studies distributed by the operations subareas.Energy and Buildings remain the most studied indicators (both expressing 37 %) and, 421

Table 1 .
Drivers and barriers to the sustainability implementation.