An open-source tool to assess the carbon footprint of research
Jérôme Mariette, Odile Blanchard, Olivier Berné, Tamara Ben Ari
AAn open-source tool to assess the carbon footprint ofresearch
J´erˆome Mariette ∗ Odile Blanchard † Olivier Bern´e ‡ Tamara Ben-Ari § . . .January 26, 2021 ———————————— Version 1 of the paper : author list to be completed ————————————
Abstract
Research institutions are bound to contribute to greenhouse gas emission (GHG)reduction efforts for several reasons. First, part of the scientific community’s researchdeals with climate change issues. Second, scientists contribute to students’ education :they must be consistent and role models. Third the literature on the carbon footprintof researchers points to the high level of some individual footprints. In a quest for con-sistency and role models, scientists, teams of scientists or universities have started toquantify their carbon footprints and debate on reduction options. Indeed, measuringthe carbon footprint of research activities requires tools designed to tackle its specificfeatures. In this paper, we present an open-source web application,
GES 1point5 , de-veloped by an interdisciplinary team of scientists from several research labs in France.
GES 1point5 is specifically designed to estimate the carbon footprint of research ac-tivities in France. It operates at the scale of research labs, i.e., laboratoires , which arethe social structures around which research is organized in France and the smallestdecision making entities in the French research system. The application allows French ∗ Universit´e de Toulouse, INRAE, UR MIAT, F-31320, Castanet-Tolosan, France † Univiversit´e Grenoble Alpes, CNRS, INRAE, Grenoble INP, GAEL, 38000 Grenoble, France ‡ Institut de Recherche en Astrophysique et Planetologie, Universit´e de Toulouse, CNRS, CNES, UPS,Toulouse, France, 9 Av. du colonel Roche, 31028 Toulouse Cedex 04, France § UMR 211, INRAE, AgroParisTech, Universit´e Paris-Saclay, F-78850 Thiverval-Grignon, France; CIRED,UMR 8568, F-94736 Nogent-sur-Marne, France a r X i v : . [ c s . C Y ] J a n esearch labs to compute their own carbon footprint along a standardized, open pro-tocol. The data collected in a rapidly growing network of labs will be used as part ofthe Labos 1point5 project to estimate France’s research carbon footprint. We expectthat an international adoption of
GES 1point5 (adapted to fit domestic specifics) couldcontribute to establishing a global understanding of the drivers of the research carbonfootprint worldwide and the levers to decrease it.
Availability and implementation:
GES 1point5 is available online for French re-search labs at http://labos1point5.org/ges-1point5 and source code can be down-loaded from the GitLab platform at https://framagit.org/labos1point5/l1p5-vuejs . Keywords: carbon footprint, research sector, WEB application
Introduction
When signing the Paris Agreement, France committed to drastically reducing its domesticgreenhouse gas (GHG) emissions within the next decades. All sectors are expected to con-tribute to this mitigation effort, including the public sector. Although it is expected thatthe research sector’s direct contribution to domestic GHG emissions is small, research insti-tutions are bound to contribute to these reduction efforts for several reasons. First, part ofthe scientific community’s research deals with climate change issues; their results point tothe need to mitigate and adapt to climate change. The whole scientific community cannotignore those results. Second, scientists contribute to students’ education and often standat the forefront of the discussions on climate change and its impacts on ecosystems andsocieties : they must be consistent and be role models. Third the literature on the researchsector’s GHG emissions points a high footprints of some scientists [Nature Astronomy, 2020,Spinellis and Louridas, 2013, Fox et al., 2009, Gr´emillet, 2008].At the international level, only a small number of studies report on estimates of thecarbon footprint of the research sector. These studies either focus on one single entity suchas a research lab or a university [G¨uereca et al., 2013, Wynes et al., 2019], on a specificevent such as a conference [Spinellis and Louridas, 2013, Desiere, 2016, Stroud and Feeley,2015, Kl¨ower et al., 2020], on a single source of GHG emissions such as air travel [Cierset al., 2019], or on a specific research project [Achten et al., 2013, Barret, 2020]. Whenmore comprehensive, these estimates relate only to one single year [G¨uereca et al., 2013].Furthermore, all of these studies rely on the development of ad-hoc methodologies relyingon a specific set of emission factors ( i.e. , the parameters used to convert activity levels intoGHG emissions). This point prevents from any comparison between carbon footprints andhighlights the need for a dedicated and standardized application.Several tools are actually freely available. Some are designed to estimate individual or1ouseholds’ GHG emissions and may be bounded to lucrative carbon offset schemes .The GHG Protocol offers a set of worksheets dedicated to industries, businesses, countriesor cities. The University of New Hampshire has developed SIMAP ® , a standardized soft-ware to address the specific needs of North American education institutions, colleges, anduniversities. In France, Bilan Carbone ® is a generic tool developed as a set of worksheetsand requires its users to pay for a licence and a training program. None of these tools arespecifically designed to estimate the carbon footprints of research activities.In this paper, we present a free and open source standardized web application, GES1point5 . It has been developed by an interdisciplinary team of engineers and researchers whowork in various research labs in France and interact in the non governmental organization
Labos 1point5 . The tool allows any research lab to (i) estimate the emissions relating tothe energy consumption and refrigerant gases of its buildings, and to its members’ travels(ii) easily highlight, via a graphic interface, the main GHG emissions drivers of a researchlab, and (iii) design emission reduction actions and evaluate their impact over time. Becausethey rely on a standardized protocol, completed GES 1point5 -based GHG inventories can becompared. This enables to make progress in our understanding of the drivers of the researchfootprint (e.g., disciplinary or geographic). At the national scale, the data collected willallow to estimate the carbon footprint of the French public research and thus support theexploration of evidence-based emission reduction strategies.The paper is organised as follows: we first describe a few specifics of the French researchsystem and the goals of
GES 1point5 ; we then describe
GES 1point5 methodology and itsimplementation; the fifth section provides an illustration of
GES1 point5 outputs, namely https://co2.myclimate.org/en/ https://offset.climateneutralnow.org/footprintcalc https://nosgestesclimat.fr/simulateur/bilan https://ghgprotocol.org/ https://labos1point5.org/ The French research system encompasses several types of institutions : national researchinstitutes (such as CNRS, INRAe, CEA, IRD, . . . ), semi-private research institutions (suchas CIRAD, Ifremer, etc.) and universities. These institutions take part in social structurescalled laboratoires (referred to as “research labs” in the following sections). Their financialcontributions to the operation of research labs may be multiple, e.g. they may pay thesalaries or stipends of its members, provide fixed assets such as buildings or infrastructures,pay for resources such as supplies, electricity, etc. A typical laboratoire benefits from theinvolvement of several institutions or universities. These research labs comprise betweentens and at most a few hundred members, and occupy one or several buildings. A laboratoire usually focuses on a specific scientific field. France counts over 1000 public research labsoverall.In this context, the implementation of GES 1point5 was initiated to provide the Frenchscientific community with a free and open source tool able to estimate the carbon footprintof a research lab using a standardized methodology.
GES 1point5 operates at the scale ofresearch labs for several reasons.
GES 1point5 is part of
Labos 1point5 , a large inter-disciplinary project in France that aims at fueling a bottom-up momentum by mobilizingmembers of the research community as opposed to decision-makers in the academic bureau-cracy. Research labs are the smallest human scale entities in the research system. They relyon independent and collective decision-making processes: experimental designs, as well asaccess to research facilities are decided and managed at the research lab scale. Consequently,a number of decisions relevant to the reduction of the carbon footprint can be made at the labos1point5.org GES 1point5 ismeant to be a decision support tool for mitigation actions and experiments. For example,to reduce emissions from professional travels, different mitigation decisions or policies maybe tested, such as an internal carbon tax or individual emission quotas.At a smaller scale,
GES 1point5 may also allow to estimate the carbon footprint of specificresearch projects or specific teams within the lab, that of PhD theses or even of conferences.At a larger scale, another objective with
GES 1point5 is to create a dataset which isa collection of the GHG footprints of a large number of research labs in France in a largearray of disciplines. Combining this dataset with statistical analyses, it will be possible toperform an extrapolation to assess the overall carbon footprint of French public research,and to describe the distribution of sources of emissions across regions and disciplines.Coupling the analyses performed at different scales with the experiences led by researchlabs will contribute to the construction of robust emission reduction scenarios at the globallevel and recommendations in terms of public research policy. GES 1point5 methodology
GES 1point5 complies with and refers to the French legislation [MEEM, 2016]. When es-timating the carbon footprint of an entity, it is necessary to precisely define its scope i.e. the GHGs considered and the emission sources included, as well as to mention the emissionfactors chosen. The next two subsections present
GES 1point5 features in these respects.4 .1 Scope
As required by the French legislation, the GHGs considered in
GES 1point5 are those of theKyoto Protocol. In terms of emission sources,
GES 1point5 takes into account those that arecommon and most often critical in research labs: buildings through energy consumption, airconditioning and refrigeration processes, daily commutes and professional travels due to theuse of cars, boats or planes to attend meetings or harvest field data.
An emission factor represents the amount of GHG emissions generated by a unit of activity.The emission factors included in
GES 1point5 mainly stem from the ADEME database that is the backbone of the French legislation. When an emission factor is missing fromthe ADEME database, the application refers to the most recent available studies. This isthe case for example for an electric bike [AVEM, 2015] or for an electric scooter [Arcadis,2019]. Furthermore, the GES 1point5 team has created customized emission factors totake into account specific research lab activities. For example, the application includes anemission factor for research campaigns at sea, based on ad-hoc estimates carried out by somelaboratories.
GES 1point5 also corrects some of the ADEME database inconsistencies. For example,in its current version , the ADEME database does not include manufacturing emissions inthe emission factor per kilometer of a gasoline car, whereas it does for hybrid and electriccars. As manufacturing emissions are significant when considering the total emissions of acar, these emissions are included in GES 1point5 for all types of cars, using factors thatwere provided in a previous version of the ADEME database. More generally,
GES 1point5 includes manufacturing emission factors for all vehicles when these emissions are significant, i.e. , cars, buses, coaches, trains, streetcars and subways. The application includes man- version 19.0 GES 1point5 has been implemented as a set of components using VueJS and django frameworks. The application uses input information that can be gathered reasonably easily(provided support is granted by the administration) and converts it into a GHG footprint.For a given number of emission sources, GES 1point5 converts GHG-emitting activity levelsinto CO e (carbon dioxyde equivalent) using emission factors as described in Section 3.2.From its welcome page (Figure 1), the application offers its users the opportunity toestimate GHG emissions anonymously or using an authenticated account. In the latter case, GES 1point5 allows to store input and output data relating to the research lab emissions.To gather the data required, the application provides a set of forms and routines brieflydescribed below. • General information : year of the GHG inventory ; number of lab members accord-ing to their position (e.g. research members, technicians, administrative staff, PhDstudents,post-doctoral fellows). • Data on buildings : floor area ; consumption of electricity, heat, and refrigerant gases; specifics related to the generation of electricity, when applicable (e.g. use of solar https://vuejs.org/ • Data on transportation modes operated by the lab : type (e.g. car, truck, aircraft) ;type of fuel ; power, distances travelled, number of hours of operation when applicable.Figure 2 presents the form dedicated to add a new vehicle and entering its energyconsumption. • Commutes : a standardized survey dedicated to collect lab members’ daily commutes isprovided by
GES 1point5 , using the
Framaform web application provided by framasoft ,a not-for-profitorganization that makes the safe handling of personal data overarching.The survey is to be sent to all lab members. Once the survey period is over, resultscan be exported from
Framaform and directly imported into
GES 1point5 . • Professional travel : the raw data are extracted from the information systems of thevarious research institutions that pay for the travels of the lab members. For each trip,they comprise information on the date, the departure and destination places (cities,countries), the travel modes, the travel purpose, the status of the lab member. Thesedata are imported as a .tsv file into
GES 1point5 , Table 1 defines the required format.
As previously mentioned,
GES 1point5 complies with the French legislation, which itselfabides by the GHG Protocol standard [WRI and WBCSD, 2004]. The application thusprovides the regulatory table covering scope 1 (direct emissions from owned or controlledsources), scope 2 (indirect emissions from the generation of purchased electricity, heatingand cooling), as well as some emission categories among all other indirect emissions of scope3. The resulting table, presented in Figure 3 can be downloaded in csv format.On top of displaying GHG emissions distribution within the three regulatory scopes,
GES 1point5 provides a user-friendly synthetic representation of the research lab’s carbon7igure 1: The
GES 1point5 welcome page offers its user the opportunity to estimate GHGemissions anonymously or using an authenticated account. The application menu, on theleft, allows to navigate between documentation pages and the different emission sourcesforms. 8igure 2: Form to add a new vehicle to the research lab’s inventory. The vehicle can beeither a car, a motorbike, a bike, a scooter, an aircraft or a boat. The form requires to definethe vehicle motorisation and its annual consumption.9able 1: Column description of the tsv file accepted by
GES 1point5 to import professionaltravels.
Column ID DescriptionTrip number One line per leg of the trip ; the trip number is asimple sequence number (from 1 to n) which allowsto gather all the legs of the same trip. If the returntrip is different from the outward trip, two lines mustbe used.Departure date A date in dd/mm/yyyy format.Departure city The departure city name. This field is used to ob-tain the departure city GPS coordinates using the geonames database.Departure country The departure country or its ISO3166 code.Destination city The destination city name. This field is used to ob-tain destination city GPS coordinates using the geon-ames database.Destination country The destination country or its ISO3166 code.Travel mode This field can take a value among [ ”Avion”, ”Train”,”Voiture”, ”Taxi”, ”Bus”, ”Tramway”, ”RER”,”M´etro”, ”Ferry” ].Number of people in the car In case of a trip in a car or a taxi, number of peoplein the vehicleOne way / return If the outward and return trips are identical, enter”OUI”, otherwise ”NON”.Travel purpose (optional) This optional field allows the application to per-form emissions statistics based on the travel pur-pose. This field can take a value among [ ”Etude ter-rain”, ”Colloque-Congr`es”, ”S´eminaire”, ”Enseigne-ment”, ”Collaboration”, ”Visite”, ”Administrationde la recherche”, ”Autre” ].Agent status (optional) This optional field allows the application to performemissions statistics based on agents’ statuses. Thisfield can take a value among [ ”Chercheur.e-EC”,”ITA”, ”Doc-Post doc”, ”Personne invit´ee” ]. footprint. It helps the user to analyse which emission sources most impact the overall carbonfootprint of the research lab, and to decide which actions to implement in order to mitigatethe lab’s GHG emissions. For example, in this representation, the buildings’ direct andindirect emissions generated from the heating and cooling systems are aggregated, whereasthey are split among the three scopes in the regulatory display. Similarly, the travel carbon10ootprint aggregates GHG emissions of vehicles, members’ daily commutes and professionaltravels. In the regulatory table, these emissions are reported in scopes 1 and 3. Someexamples of this operational representation of the carbon footprint are reported in Figure 4and 5. All the figures provided can be downloaded in multiple formats, such as png, jpeg,svg or pdf. This section provides and discusses results of the 2019 greenhouse gas inventory of a fictitiousresearch lab named Cogitamus. The lab comprises 80 members distributed as follows : 14researchers, 24 associate professors, 17 engineers or administrative staff and 25 PhD studentsor postdoctoral fellows. Cogitamus occupies one building shared with another lab and 60%of the total floor space, 3 300 m . In 2019, the building consumed 200 000 kWh PCI fromthe Toulouse Canceropole urban heating network, 120 000 kWh of electricity and 0.3 kg ofthe R32 cooling gas. The lab owns one diesel car which traveled 12 000 km in 2019. For afull reproducibility of the results presented in this section, the commuting survey results and the professional travel file used are freely available.Figures 3, 4 and 5 illustrate the GHG emissions inventory that complies with the GHGProtocol standard, the user-friendly carbon footprint representation, graphs excerpted from GES 1point5 , respectively. More specifically, figure 3 and 4 display the total emissions ofthe Cogitamus lab (98 032 ±
28 525 kg CO2e) from two different perspectives, i.e. ,theregulatory GHG inventory table and the detailed carbon footprint. Figure 4 (top) showsthat these emissions are mainly driven by professional travels (64%) followed by members’daily commutes (29%) and electricity (4%). Moreover, Figure 4 (bottom) illustrates thatemissions generated by professional travels are dominated by congress attendees’ travels andmore specifically by researchers’ travels to these congresses. https://cloud.le-pic.org/s/HaZWEae7ef22KMy https://cloud.le-pic.org/s/msGYms2w7kqCNqJ GES 1point5 topoint to the need to focus primarily on travel emissions for an efficient mitigation action,even if reducing the buildings’ emissions may be implemented in case of low-hanging fruits.Such results become even more valuable when analysing the changes in emissions over theyears and evaluating the efficiency of the mitigation actions implemented.
GES 1point5 is a first step in a larger endeavour to facilitate and inform emission reductionsfrom research activities in France. Estimating the carbon footprints is a crucial step tomitigate research labs’ GHG emissions as it brings clearer information compared to thatstemming from scopes 1, 2 and 3 of the GHG Protocol standard. The carbon footprintspoint to the relative weights of the main building blocks of the research labs’ GHG emissions.For example, if the buildings’ carbon footprint of a research lab is very low compared to itstravel carbon footprint, mitigation actions should mainly focus on the travel emissions -even if reducing the buildings’ emissions may be implemented in case of low-hanging fruits.Similarly, within the travel carbon footprint, if the emissions from the professional travelsovershadow the emissions from the staff and faculty commutes, mitigation actions shouldprimarily focus on professional travels - even if raising awareness on the emissions fromthe commutes and acting to reduce them should also be addressed.
GES 1point5 helpsto design and monitor mitigation actions in other respects, too. At the research lab level,while storing the GHG inventories year after year, it paves the way for analyses of the lab’semissions dynamics, define and implement action plans, monitor progress and define newaction plans in the long run. At the national level, statistical analyses can be carried outbased on the aggregated data-set of all labs, and help to design public policies towards themitigation of GHG emissions of the public research sector.This first version of
GES 1point5 presented in this paper focuses on the common emission12igure 3: Illustration of a regulatory table obtained with the
GES 1point5 tool for a ficti-tious research lab, presenting the emissions distributed among the three scopes of the GHGProtocol standard. The translation to English is given in Table 213igure 4: Illustration of the carbon footprint information provided as an output of
GES1point5 : table (upper panel) and of the pie chart (lower panel), showing the distribution ofemissions in kg CO e. Translation of the terms is provided in Table 3.14igure 5: Illustrative distribution of the professional travel carbon footprint provided by GES 1point5 according to travel purposes and agent status.sources that are often critical in research labs. However, other potentially important emis-sion sources need to be soon assessed, among which internal information technology (IT)emissions, and emissions linked to purchased goods.The estimation of the carbon footprint of IT systems internal to the research labs willbe developed in
GES 1point5 in future work. It will rely on the EcoDiag database thatprovides manufacturing and transportation emission factors for multiple types of IT devices.In this case, electricity consumption emissions will not be estimated as they are alreadyincluded in the emissions of the labs’ buildings .The estimation of the carbon footprint of purchased goods is complex as emission factorsdo not exist for all the categories of products that research labs can purchase. For example, itis common to use specific chemical solvents in biology labs, with only very little informationon the manufacturing and supply chains. Various methodologies are currently under studyand Labos 1point5 will thrive to consider the right trade-off between comprehensiveness andrepresentativeness.
GES 1point5 will later on be complemented by the estimation of the GHG emissions https://ecoinfo.cnrs.fr/ecodiag/ The research sector must reduce the carbon footprint of its activities along with all othersectors, in order for France to reach the goal of the Paris Agreement and, more indirectly, tosustain the bond between science and society. A first step naturally pertains to estimatingcurrent emissions level. In this context, a standardized tool is essential as it paves theway for comparisons of carbon footprints, statistical analyses, fruitful dialogues, coordinatedmitigation strategies and reporting of mitigation actions results in a consistent way.As an open-source software,
GES 1point5 is freely accessible to any research lab in theworld. Foreign research labs may have to adjust emission factors to their country when usingthe application, thus enabling comparisons between research labs worldwide.
GES 1point5 opens up a very broad range of international research perspectives and initiates a strategyto discuss on the best ways to reduce research emissions worldwide.
References
Nature Astronomy. The climate issue.
Nature Astronomy , 4:811, 2020. doi: 10.1038/s41550-020-01216-9.Diomidis Spinellis and Panos Louridas. The carbon footprint of conference papers.
PloSone , 8(6):e66508, 2013.Helen E Fox, Peter Kareiva, Brian Silliman, Jessica Hitt, David A Lytle, Benjamin SHalpern, Christine V Hawkes, Joshua Lawler, Maile Neel, Julian D Olden, Martin ASchlaepfer, Katherine Smith, and Heather Tallis. Why do we fly? ecologists’ sins of emis-sion.
Frontiers in Ecology and the Environment , 7(6):294–296, 2009. doi: https://doi.org/10.1890/09.WB.019. URL https://esajournals.onlinelibrary.wiley.com/doi/abs/10.1890/09.WB.019 .D Gr´emillet. Paradox of flying to meetings to protect the environment.
Nature Correspon-dence , 455:1175, 2008. doi: 10.1038/s41550-020-01216-9.Leonor Patricia G¨uereca, Nathalia Torres, and Adalberto Noyola. Carbon footprint as a basisfor a cleaner research institute in mexico.
Journal of Cleaner Production , 47:396–403, 2013.Seth Wynes, Simon D Donner, Steuart Tannason, and Noni Nabors. Academic air travel has16 limited influence on professional success.
Journal of Cleaner Production , 226:959–967,2019.Sam Desiere. The carbon footprint of academic conferences: Evidence from the 14th eaaecongress in slovenia.
EuroChoices , 15(2):56–61, 2016.James T Stroud and Kenneth J Feeley. Responsible academia: optimizing conference loca-tions to minimize greenhouse gas emissions.
Ecography , 38(4):402–404, 2015.Milan Kl¨ower, Debbie Hopkins, Myles Allen, and James Higham. An analysis of ways todecarbonize conference travel after covid-19, 2020.Joachim Ciers, Aleksandra Mandic, Laszlo Daniel Toth, and Giel Op’t Veld. Carbon footprintof academic air travel: A case study in switzerland.
Sustainability , 11(1):80, 2019.Wouter MJ Achten, Joana Almeida, and Bart Muys. Carbon footprint of science: More thanflying.
Ecological indicators , 34:352–355, 2013.Didier Barret. Estimating, monitoring and minimizing the travel footprint associated withthe development of the athena x-ray integral field unit.
Experimental Astronomy , pages1–34, 2020.Minist`ere de l’environnement de l’´energie et de la mer MEEM. M´ethode pour la r´ealisationdes bilans d’´emissions de ges, conform´ement `a l’article l.229-25 du code de l’environnement,2016.Association pour l’Avenir du V´ehicule Electro-Mobile AVEM. Est-ce que le v´ehicule ´electrique est polluant ? , 2015.Arcadis. Extrait de l’´etude portant sur l’impact environnemental des trottinettes ´electriques.´Etude de cas dans le contexte parisien. ,2019.World Resources Institute WRI and World Business Councilfor Sustainable DevelopmentWBCSD. The greenhouse gas protocol, a corporate accounting and reporting standard.revised edition, 2004.
A Appendix tables
GES 1point5 and presented inFigure 3. The three scopes correspond to those of the GHG protocol.
Scope 1
Scope 2
Scope 3
Carbon footprint Emissions in kg CO e Share of total fooprintCarbon footprint of thebuildings- Heating- Electricity- Refrigerant gasesTravel carbon footprint- Commutes- Vehicles- Professional travele Share of total fooprintCarbon footprint of thebuildings- Heating- Electricity- Refrigerant gasesTravel carbon footprint- Commutes- Vehicles- Professional travel