Digitisation and the Environment

This article assesses the extent to which digitisation impacts on the environment in the creation, use and storage of its outputs.

Digitisation is defined as the conversion of analogue information into digital information.  Or, more specifically, the representation of an arts, cultural heritage or educational object, image, event or document in digital form, with the express purpose of providing a resource that can be re-used in learning, teaching and research.  The digitisation of the world's cultural heritage and other educational resources has grown considerably in the recent past.  Significant sums of public money have been invested in digitisation, and museums, galleries, archives and libraries have placed high priority in providing access to their collections and holdings in the formats expected by today’s increasingly digitally-connected learners, teachers and researchers.  

Among the positive impacts of digitisation are the reduction in handling and better conservation of fragile material, improved access to information and resources for more people, increased flexibility over where and when resources are used, the emergence of new avenues and methods for learning, teaching and research and, ultimately, a better understanding of the world around us.   Yet, the rise in digitisation, and in Information  and Communication Technology (ICT) use generally, puts added pressure on administrators to provide the equipment, tools, on-going management and storage necessary for the creation, archiving and use of digital resources.  Digitisation therefore results in a rise in carbon use at a time when national and international governmental bodies are setting targets for carbon reductions.

Between 2008 and 2009 the size of the all the digital files created and replicated in the world was estimated to grow by 62 percent.  By the year 2020, this digital universe is predicted to grow by a further 44 times to the unimaginable size of 1.2 zetabytes (1.2 trillion gigabytes).  (1) Greenhouse gas emissions from data storage centres are anticipated to triple by 2020, (2) and 1.4 Giga-tonnes Carbon Dioxide Equivalent (GtCO₂e) is expected to be used to fuel the manufacture and use of ICTs by that time. (3) Concerns, therefore, over the environmental impact of ICT are justifiable.

However, the fact that ICT's carbon use is rising while simultaneous pressure is being applied to reduce carbon dependency throughout the economy is not necessarily an incongruity.  On the contrary, Smart2020, a report published in 2008 by the Climate Group, identified a potential reduction of 7.8 GtCO₂e in global carbon emissions enabled by increasing the use of ICT by 2020. (4) This reduction therefore equates to some five times the total ICT-related carbon emissions predicted by that time.  Thus increasing the use of ICT and therefore raising its carbon emissions has the potential to enable net carbon reduction in the economy as a whole.

ICT can produce these carbon reductions primarily outside its own sector through the enablement of better energy efficiency in buildings, motor systems, electricity production, transport and logistics.   The Smart2020 report also highlights the potential of the ICT sector to at least negate rising carbon emissions in its own industry through employing more environmentally efficient practices and technological innovations, such as in equipment manufacture, storage systems and uses of ICT.  ICT also has the potential to facilitate what is known as dematerialisation.   That is, the gradual replacement of carbon intensive physical products and activities, like article and travel, with low carbon electronic alternatives like e-documents, tele-communications and home-working.

The article will begin with an overview of Government and sector targets and policy when addressing climate concerns.  It will then examine the rise in digitisation over the past decade or so, highlighting its educational benefits and also look for evidence of its contribution to greenhouse gas emissions.  The article will go on to assess the means by which digitisation can minimise its carbon emissions though more efficient working practices, greener systems architectures and use of technologies like cloud computing.  Finally the article will look it detail at the evidence for dematerialisation enabled by digitisation in activities like learning, travel and article use.

1.  Carbon reduction policy and activity in the Education and Cultural Heritage sectors

Current environmental legislation is a useful indicator of climate concerns in the Education and Cultural Heritage sectors.  In December 2008, European Union (EU) member states reached agreement over an energy and climate change package which included a target to cut greenhouse gas emissions across the EU by 20 percent by 2020, and generate one-fifth of the EU's power from renewable sources by that time. (1) Through the introduction of the Climate Change Act of 2008 the United Kingdom Government set an unilateral target to reduce carbon emissions by 80 percent by 2050 compared to 1990 levels, and by at least 34 percent by 2020.

UK Government departments have been responding to these targets.  When focussing on digitisation the two departments that have relevance to the Higher Education (HE), Further Education (FE) and Cultural Heritage sectors are the department for Business and Innovation (BIS), which has responsibility for Business and Further, Higher and Adult Education, and the Department for Culture, Media and Sport (DCMS).  

BIS produced its Sustainable Development Action plan for 2010-2011 in which it sets out plans for developing sustainable business practices and use of resources in response to climate change.  In its plan, the BIStasks the Higher Education Funding Council for England (HEFCE) to produce achievable carbon reduction targets for HE, and an accompanying strategy for achieving them. (2) Following a programme of consultancy with the sector in 2009, HEFCE published its Carbon reduction target and strategy for higher education in England. (3) This article set out the framework within which HE could reduce its carbon emissions by 80 percent by 2050, in line with the UK wide government target.  As well as providing a supporting mechanism in which HE can meet it target commitments, HEFCE has also linked some future capital funding allocations for HE to adherence to carbon reduction plans.

The DCMS which has responsibility for Museums, Libraries, Galleries, Archives, Arts Organisations and the Built and Historic Environments, published its Sustainable Development Strategy in 2007 and followed this up with a Climate Change Plan in 2010.  The Climate Change Plan sets out steps to embed climate change considerations into the DCMS’s own departments and estate as well as in the Non Departmental Public Bodies (NDPB's) that fall under its control, which include some national museums, archives and libraries.  Unlike HEFCE, the DCMS has so far not set explicit carbon reduction targets in specific areas, and until now has been involved in stakeholder engagement through a series of carbon usage awareness activities.

Despite the lack of specific reduction targets from the DCMS, the Culture and Heritage sector is acting to reduce its emissions.  The Museums Libraries and Archives (MLA) and the National Archives are to publish Environmental Standards for Cultural Heritage Collections, through the British Standards Instituteby Autumn 2011.  The MLA has also fundeda 'Green Tourism Business Scheme' which provides some small museums with environmental assessments and carbon reduction action plans.  Both national and regional museums are actively seeking to reduce emissions, for example, the Victoria and Albert (V&A) Museum has published figures that show a 20 percent carbon reduction in the museum's operations when compared to 2005 figures. (4) The National Archives recently published figures highlighting that it has reduced its carbon footprint by ten percent when compared to 2008/9 figures.(5) There is therefore a consensus, stemming from Brussels and national government and cascading through  the Education and Cultural Heritage sectors themselves, on the need for action to reduce carbon emissions and to ensure more environmentally efficient working practices.  

2.  The rise in digitisation and the rising ICT carbon footprint

These top down carbon reduction targets and sector activities are taking place at a time when the carbon footprint of ICT is rising.  The creation and use of digital materials in learning, teaching and research has grown considerably in recent years.   A report published in 2005 by the Consortium of University Research Libraries (CURL) estimated that some £130 million of UK public money had been invested in digitisation in research libraries and archives during the previous ten year period.(6) In addition to this direct public funding for digitisation, a similar rise in smaller scale locally funded initiatives is evident, and could equate to the  funding provided directly through central government. (7) This is likely make the total figure invested in research library and archive digitisation closer to £250 million.

The Cultural Heritage sector has also funded large scale digitisation. The sector received an early digitisation investment of £50 million in 2001 by the New Opportunities Fund (NOF, now renamed the Big Lottery), and museums have subsequently been striving to digitise entire collections to provide online access and to aid preservation and curation.  For example, The V&A can now can provide online access to one million works.  Investment in digitisation does not seem to be abating.  The French government has announced plans for an almost £700 million investment in digitisation, partly as a repost to the Google Books initiative which has to date digitised over 12 million books.  A recently published vision for future digitisation calls for a higher priority to be given to the complete digitisation of all European cultural heritage at an estimated cost of €100 billion. (8)

The rise in the availability of digital resources is also evident in the drive to foster easier access to education across the world in an initiative known as Open Educational Resources (OER).  OER has received hundreds of millions of US dollars funding from organisations such as the Hewlett and Mellon Foundations and UNESCO, since 2001.   In the UK, HEFCE has provided over £10 million since 2009 to fund an OER programme to kick-start the notion of sharing digital educational resources across UK education.  The aim of OER is to make it possible to share materials, tools and media, primarily via the Internet, for teaching and learning.  These resources are either free from copyright restrictions or licensed for anyone to use and re-purpose for educational means.  OER is seen by its exponents as a potential step towards a solution to world educational inequalities and to radically improve access to education for developing and developed countries alike. (9)

Examples abound of innovative and collaborative uses of digital resources to support the claim that digitisation can democratise and further knowledge by improving accesses to resources for learning, teaching and research.  For example the notion of 'citizen science' is becoming increasingly common, whereby digitised research material is provided to interested enthusiasts who supply the expertise to enhance it.  Galaxy Zoo is a good example, where some one million high resolution images of previously unseen galaxies have been classified by over 300,000 volunteers. (10)  These volunteers have been instrumental in providing the background data required by researchers investigating our galaxies and the universe.  Another project called Old Weather has also taken the citizen science model and asked volunteers to transcribe data from late nineteenth and early twentieth century naval records that have recently been digitised by a collaborative group working under the Atmospheric Circulation Reconstructions of the Earth (ACRE) initiative.(11) This work will enable better understanding of historical weather patterns that will inform future research on climate change today.  

The creation of digital resources can be truly transformative for entire cultures.  The almost forgotten music and lore associated with the historic Bantu kingdoms of Uganda is a case in point.  This music has been revived recently partly as a result of the digitisation of the Klaus Wachsmann collection, a unique record of this music recorded just after the second world war.  These digital recordings are for the first time being accessed by young Ugandan musicians who are now learning the music of their past.(12) These are only a few of the plethora of examples of how digitisation can engage us in our past history and culture, provide new avenues and methods for research and reshape our understanding of the world around us.

Digitisation can also provide opportunities for new economic activity through the development of technologies and processes for the creation and use of digital materials – such as scanners, cameras, PCs, and information management processes, as well as monetary gain for some types of pay on demand materials.  Digitisation thereforehelps create jobs in the wider economy and can provide funds directly for some cultural and educational institutions.  Indeed it seems that the Education and Cultural Heritage sectors' future use of ICT in learning, teaching and research, and their ability to provide openly available digital resources, will further affect their ability to generate future income from government, foundation and local institutional funding.   Providing digital access to collections, archives and learning and research materials promotes institutions, enhances the possibility of external funding and attracts visitors and students who provide consumer demand for the products of digitisation, and the flexibility in learning and teaching that digitisation can provide.

The educational benefits of digitisation and its link to future revenue generation can only result in more digitisation in the education and cultural heritage sectors.  And as research, learning and teaching become more ICT dependant the pressure on the managers of ICT to provide adequate data storage and infrastructure will increase.  Technological advances and potentially more complex use of resource intensive image and video formats and services will add to this pressure, as will the sheer volume of predicted data.

The environmental question then is the extent to which digitisation contributes to the world's carbon footprint and is the carbon cost of digitisation justifiable?  Digitisation's contribution to carbon emissions could be usefully measured by assessing the size of its outputs in terms of bits and bytes and by obtaining concrete data on usage figures.  However, this information is difficult to obtain, as a recent attempt to compile a statistical assessment of European digitisation activity proved.(13)  This project uncovered a lack of a standard means of describing digitisation - for example, outputs are recorded as files, images, pages and collections across various scenarios – which results in a profound difficulty in measuring like-for-like digitisation activity.  This is compounded by a lack of co-ordinated digitisation planning. Only one third of participating institutions  had formal digitisation plans, and, consequently, most institutions do not maintain rigorous information about their available digital resources from which statistical analysis can be gleaned.

Moreover, problems seem to exist when trying to gather empirical evidence on the use of resources.  The CURL report mentioned above, and others, have tried to gauge use satisfaction with digitisation.(14)   However, institutions generally do not maintain data on resource access figures on any great scale, and at the point of resource creation expert user-needs analysis seems to be lacking.  Digitisers often cite users as being central to their digitisation decision making, however the evidence shows that users' views can often be assumed rather than expressly solicited.(15) A more explicit link between funding for digitisation and the use of digital resources therefore needs to be found.   The lack of which makes compiling meaningful statistical data on the size and extent of digitised resources difficult, and presents a significant problem when trying to asses digitisation's environmental implications.    However, while there is little quantative research currently on the environmental impact of digitisation, there are undoubted repercussions to the environment in its increase.  Generic ICT-related carbon emissions, to which digitisation contributes, are more closely monitored and provide another means of assessment.

ICT accounts currently for two percent of global CO₂ emissions, and under BAU growth estimates, is expected to treble by 2020 to 1.4 GtCO₂e.(16) Figures show that the use of ICT in UK HE and FE currently generates over 500,000 tons of CO₂ annually.  Personal computing is the main area of ICT-related energy consumption in UK Universities and Colleges, accounting for an estimated 40 to 50 percent of the total, while printing accounts for between 10 and 16 percent.  The remainder is taken up by server, network and data storage facilities.  Using figures scaled up from those gathered from the University of Sheffield, a recent study estimated that HE and FE utilises nearly 1,470,000 computers, 250,000 printers and 240,000 servers, and consumes nearly one million mega watt-hours of electricity annually with a cost of approximately £116 million.(17) These figures account for the environmental impact of operating ICT services. On the other hand, the impact on the environment of manufacturing the hardware must be taken into account.  The carbon generated from the manufacture and materials of PCs, peripherals and other devices accounts for one quarter of the total ICT carbon footprint.

When assessing the particular issues associated with digitisation, it is the need for data storage that stands out. Data centres are expected to grow faster than any other aspect of ICT.   In 2006, the United States (US) Department of the Environment estimated that data centres in the US used 61 billion kWh of electricity, representing 1.5 percent of all US electricity use, the equivalent to the amount of electricity used by about six million US homes.(18)   The Smart2020 report estimates that if data centre growth continues in line with demand, the world will require some 122 million servers by 2020, up from a total of 18 million required servers in 2008.  The global carbon footprint of data centres was measured at 76MtCO₂ in 2002, and this is expected to treble by 2020.(19)

3. Dematerialisation in learning, teaching and research

A key mitigator in the rising use of ICT-related carbon emissions is the expected potential to reduce  dependance on carbon intensive physical products and activities. More specifically in terms of creating and using digitised resources, evidence for dematerialisation can be measured in in e-learning, travel, and article use.

3.1 E-Learning

Digitisation is one of the key enablers in any e-learning activity. Digitised objects and learning resources can offer a means of reducing the carbon footprint of HE through its implications for more home-working, less face-to-face teaching, reduced use of campus resources and the availability of electronic course documentation as opposed to printed handouts.  A study carried out by the Open University which looked at measures such as heating, travel, computer use and printing, found that distance learning courses on average consume nearly 90 percent less energy and emit 85 percent less CO₂ than traditional campus-based courses.(20)   However the same study also found that more traditional e-learning courses that utilised a blend of electronic and traditional teaching methods resulted in a much lower 12 percent reduction when compared to courses that used no electronic materials or tools.  Indeed, another study carried out by University of Leicester in 2009 points to even lower differences in carbon use between courses carried out with face to face teaching and others using more electronic components.(21)

While the Leicester survey was carried out on a comparatively small scale looking at a limited number of students and their travel requirements, it does highlight the potential difficulties there can be in reducing carbon emissions through more e-learning.  There are many contributing factors that affect the extent to which e-learning reduces the use of carbon, such as: where students and teachers live, travel methods and travel frequency, printing, computer use and heating requirements.  Moreover, e-learning has such a wide variety of implementations that it seems there is an equally wide scope for making carbon reductions.  In some instances, such as distance learning courses, environmental benefits can be significant, whereas courses that have a blend of campus based teaching alongside electronic components will tend to have less of an impact.  It seems therefore that more research is required in this area, providing detailed studies of travel, computer, heating and printing needs across a wide variety of learning scenarios, before one can say categorically that e-learning can in itself enable dematerialisation that results in significant carbon reductions.

3.2  Travel

As mentioned above e-learning in a blended or distance context can have implications on travel and its associated emissions. Furthermore, travel can be reduced through the increased use of electronic methods of communication – most notably in using Video Conferencing (VC) technologies to replace physical meetings.  The extent to which VC is used currently in the education and cultural sectors is difficult to ascertain due to a lack of available research.  However one recent study published by the European Telecommunications Network Operators' Association states that by replacing 30 per cent of business meetings with VC, over 30 million tonnes of CO₂ emissions could be saved.(22)   Statistics produced by the Janet Video Conferencing Service (JVC), a JISC funded service which held over 20,000 video conferences in 2007/8, indicated comparatively high use.(23) However this figure did not take into account the fact that users are often the same people and institutions returning for multiple visits, and therefore cannot be used as an indicator of the breadth of usage across HE and FE.   A small survey of HE carried out by SustainIT as part of their research for JISC, pointed to 43 percent of respondents never having used VC while nine percent were regular users.(24)

Another driver for reductions in travel could be availability of digital resources online which negate the need for users to travel to see the real thing.  However hard evidence for this is scant.  Museum access figures could be one indicator – the theory being that as collections go online access figures correspondingly drop.  The reality however is that museum access figures have generally risen in line with a rise in their collections being available on-line.(25) While this rise in visitor numbers coincides with the move to free entry to museums in the late 1990s, and many other factors impact on reasons for visiting a museum, it potentially raises questions about the notion that the availability of a digital surrogate of a resource equals less travel.  Indeed it is likely that digitisation acts more like a promotional tool for the museum, and the availability of a museum object online in many instances will encourage people to travel to see the real thing.  An interesting future study could begin to unravel this question, however currently this evidence is lacking.

3.3  Article use and printing

As noted above, printing accounts for at least 10-16 percent of ICT related electricity consumption in education. And reprographics departments that once would have had control of printing and copying are increasingly viewed as non-essential with most members of staff and students now having access to networked laser and ink-jet printers.(26) However, rather than digital resources reducing article use it would seem that there has been a rise in printing and article use alongside a simultaneous rise in the use of ICT.  A study by the printer manufacturer Lexmark concludes that the average UK worker prints 38 pages a day, with an average 29 percent of these wasteful. (27) A recent study into printing in HE found that respondents printed an average of 240 sheets of article per week (48 per day) with some printing up to 4,000 sheets per week.(28) It is not known how many of these could be avoided.  The same survey further estimates that the HE sector consumes 4,250 million sheets of article (approximately 21,250 tonnes of article) annually. (29)  The article used in printing has a relatively large environmental footprint, with the energy used to make one sheet of article amounting to more than is used to actually print it.(30)

So evidence would suggest that digital resources do not seem to negate the need for users to retain their need for hard copy, and therefore printing continues to grow.  It would seem that a shift in attitude amongst users is required to fully realise potential carbon reductions through reduced article consumption.  In their report on the environmental impact of distance learning and campus based courses, Roy et al. observed that there can be tendency amongst students todownload and print a high proportion of web based learning materials for ease of portability and study, particularly at the beginning of courses. (31) Change in printing behaviour is likely to happen as more people consume their printed material via e-books and tablet PCs and less on the printed page, and more institutions implement article saving and awareness raising to stop non-essential printing. However, at this time, increased ICT use in learning, teaching and research does not equate to less use of article.  Indeed, evidence would suggest the reverse is true. 

4  Reducing the carbon emissions generated through digitisation and ICT

Having looked at the emissions generated through the operation and manufacture of digitisation services, balanced against the ways in which the employment of digitisation can potentially reduce the amount of green house gases emitted through traditional carbon intensive products and activities, we will now look at how green house emissions from digitisation can be minimised through responsible working methods and best practice.

4.1 Server Virtualisation

It is clear that digitisation increases the volume of data stored in data centres and on servers. One method of reducing the potential for increases in carbon emissions in data centres is to deploy a virtualisation strategy to facilitate a more environmentally efficient use of servers.  Research has shown that  servers typically run at 10-15 percent of their capacity. (32) This is due to the fact that a single server is often dedicated to one task, which is easier for administrators to identify and resolve problems.  However this architecture can also result in wasted energy used in powering and cooling servers for relatively little output.  This problem is compounded by the fact that as data centres utilise more servers they begin to take up a lot of physical space and larger data centres become more energy intensive to run and cool. 

Virtualisation can facilitate the conversion of one physical server into multiple virtual machines.  Each virtual server acts like a unique device, capable of running its own operating system which enables one server to undertake multiple tasks and run at closer to its optimal capacity while not restricting the identification and isolation of technical problems.   This makes it possible to consolidate down to fewer servers and maintain the same level of operation, and therefore virtualisation can result in data centres which are smaller, give off less heat and use less energy.

4.2 Dispersed / Cloud Computing

Another means of enabling a more environmentally beneficial architecture for data storage is to utilise some form of dispersed computing.  That is, a shift from the computer user storing files and using software on their own device, or within the confines of an institutional network, to a model that uses the Internet to undertake a wider number of activities and tasks.  For example, a user can store images, access email, create documents, use social networks and stream video by using software and storage accessed via the Internet, as opposed to their own machine or local network.  Dispersed computing is increasingly being referred to as cloud computing.

Some expected benefits for end users of cloud computing are: cost reductions, as services can be paid for incrementally at the point of use; better access to software as new software is updated centrally and not at a local level; and also flexibility and mobility as individuals are no longer tied to their local institutional networks.  For those responsible for managing and providing digitisation services the added benefit of cloud computing could be realised in better access to data storage, paid for on demand, and away from the institution itself.   The institution is therefore relieved of the task of stocking, maintaining and cooling its own data centres.  

However, environmentally, there is some debate as to the value of the cloud.  Currently cloud computing is responsible for between one and two percent of the world’s electricity use and Greenpeace predicts that greenhouse gasses associated with it will triple by 2020. (33)  Another environmental drawback associated with cloud computing is that the institution could simply be displacing its carbon use from one location to another thus resulting in no net carbon reduction globally.

Furthermore, evidence suggests that until now cloud data centres have been, by and large, environmentally in-efficient.  A Greenpeace comparison of seven significant cloud data centres in the US showed that only one retrieved more than 50 percent of its energy from renewable sources. Furthermore, the largest data centres looked at in the study, built by Apple and Microsoft, utilised only 3.8 percent and 1.1 percent of renewable energy respectively. (34) That said, a considered and well planned use of the cloud, that takes into account these factors can potentially result in a net reduction in the carbon footprint of digitisation. 

4.3 Thin Architecture

The production of digital resources, particularly image and video rich data, tend to require the use of high powered desktop PCs with large processing and file storage capability.  One method of reducing the carbon footprint of this type of digitisation within institutions would be to explore the potential suitability of what is known as a thin architecture.  Thin architecture has the potential to minimise the processing and storage capability of PCs by using centralised facilities to host commonly used software and data.  These facilities could be located within the institution itself, or indeed, outsourced to the cloud.  The result of this shift from thick to thin architecture is that individual PCs require less data storage capability and processing power, which in turn means that they can be smaller and lighter and require less energy to run and cool.   This can lead to a lower environmental impact in production, a better use of space, as software is updated centrally, less disposals, and an overall reduction in energy requirements.  Estimates show that by 2020, 74 percent of all PCs in use will be laptops as opposed to desktops.  And while thin architecture can result in increased processing loads at the centre, more network energy consumption and may not be suitable in all digitisation scenarios, it can lead to net carbon savings and enable more efficient and targeted management of the sectors’ ICT energy consumption. 

5.  Conclusion

Policies, directives and guidelines as well as initiatives by the Education and Cultural Heritage sectors themselves indicate that there is broad agreement on the need to reduce carbon emissions.   However there is also the potential for a rise in emissions brought about by the demand for more  ICT.  These two trends are not contradictory.  Increasing the use of ICT can enable significant net carbon reductions throughout the economy.

Moreover, increasing digitisation and the availability of digital resources has unarguable benefit for education.  Resources are more easily accessible to more people, there is increased flexibility as to when and where researchers, learners and teachers use resources, new avenues and methodologies for research and teaching can be realised, institutions can promote themselves through their resources and expertise, attract and retain students and staff, and potentially increase funding.   These benefits may be enough in themselves to justify the environmental cost of digitisation.  However, in the context government driven carbon reduction targets and the will of the sectors themselves to reduce their emissions, digitisers and those responsible for digitisation services have an obligation to ensure their efforts are as environmentally efficient as possible.

It is extremely difficult to assess digitisation's environmental efficiency.  The extent to which digitisation and its growing demand for data storage contributes to the burgeoning digital universe is simply unknown.  Attempts to assess digitisation's size and use have uncovered a worrying lack of planning for digitisation at an institutional level, and a lack of expert user analysis at the resource creation stage.  This could result in environmental inefficiency and leaves digitisation open to accusations of wasted resources.  More research is required into the environmental cost of digitisation, and greater emphasis amongst those responsible for providing funds and services for digitisation should be placed on planning and the strategic development of digital materials.

As we have seen, various methods of reducing the environmental impact of digitisation can be deployed – such as more efficient running of servers, a careful use of cloud computing and the deployment of thin architectures.  These methods for reducing the carbon emissions associated with digitisation seem eminently achievable and are already underway at many institutions.  However, a key finding of this article is that digitisation and the use of digital resources do not automatically lead to the dematerialisation of carbon intensive physical products.  In the case of e-learning there are a range of potential implementations of ICT ranging from courses carried out entirely online to those that use a small proportion of digital components, and the available research currently points to an equally wide range of carbon reduction capability. 

Furthermore where dematerialisation is thought to be more obvious: in less printing and use of article and in less travel, the available evidence indicates that printing and article use in HE and FE is on the increase, and technologies like VC which have potential to reduce travel requirements are, as yet, to realise their full potential.   To some extent changes in user behaviour over time will compensate for this.  For example, as users become increasingly content to consume digital materials on screen as opposed to on article.  However as things stand currently, it seems that more could be done to increase potential for dematerialisation.

Thus digitisation remains vital across education and beyond, but it could do more to meet its environmental obligations.  And it must, in order to continue its role in providing significant drivers for educational enhancement and opportunity across the world, while also contributing to a truly sustainable future.