When it comes to environmental sustainability, the information technology industry's greatest strength is also its greatest weakness. The flip side of our industry's relentless technological progress is rapid product obsolescence and ever-rising piles of electronic waste (e-waste). Put simply, Moore's Law, the driving dynamic behind IT innovation for nearly five decades, is fundamentally not green, and perhaps not even sustainable – at least, not yet.

We have been making this point since our first Position Paper on 'Green IT' back in early 2008. But the results of our latest sustainability research project have re-confirmed that, while the IT industry appears on track to be energy-neutral in terms of its net energy consumption and savings, there is still no credible scenario for safely managing the global production and disposal of literally billions of personal computers, mobile phones and other electronic devices. Yet in recent years, energy savings have been the overwhelming focus of the Green IT community – this needs to be rebalanced to give the e-waste problem the attention it deserves.

Research Background

Our sustainability research has always stressed the need to take a holistic environmental approach that accounts for the production, consumption, application and disposal of IT products. Back in 2008, we learned from the available literature that all of the world's data centres, PCs and networks accounted for just 2 percent of total world energy consumption (and just 1.3 percent of greenhouse gas emissions). Based on a variety of sources, we also estimated that manufacturing and distributing these same hi-tech products (while much less discussed) required a roughly equivalent 2 percent of share of world energy.

Therefore, from a holistic energy perspective, if the application of IT could save just 5 percent of the remaining 96 percent of world energy usage, it could offset its own current consumption and production requirements, since (.05 x .96) > (.02 + .02). Many would argue that these savings have already occurred. But, as the energy used in producing and consuming IT continues to grow considerably faster than energy usage in the wider economy, we think that within five years, a 10 percent saving may well be needed to maintain this type of energy-neutral position.

To better understand the energy-saving benefits of IT and the means and likelihood of achieving these savings targets, in 2009 we commissioned and sponsored a special issue of the prestigious Journal of Industrial Ecology to look specifically at Environmental Applications of Information & Communications Technology (EAICT). This journal is now available online at http://www.lef.csc.com/projects/80, and printed copies have been sent to LEF clients. We are grateful to the EAICT project editors, Eric Masanet from Lawrence Berkeley National Laboratory and H. Scott Matthews from Carnegie Mellon University, for assembling a world-class set of important peer-reviewed articles, focused on three main areas of potential IT-enabled energy savings:

1. Optimization– the use of embedded IT to make products, machines, buildings, water systems and other activities smarter and more energy-efficient.
2. Behaviour - the use of IT to facilitate awareness, sharing, recycling, re-use, measurement, control, and the modelling of complex behavioural dynamics.
3. Dematerialization - saving energy by using IT to replace physical goods and activities (such as stores, offices, travel and music delivery) with more efficient digital alternatives.

We strongly encourage clients interested in environmental issues to read through these extremely well done journal papers, which contain a vast amount of data, insight and supporting documentation. In this commentary, we will summarize some of the key themes of the EAICT project. Our main message is that while progress will be slow and there is no single dominant environmental application for IT, the evidence suggests that our modest energy-neutrality target of 5-10 percent savings across the broader economy is clearly achievable, but major gains beyond this will likely take time to develop.

This is why we believe that from a sustainability perspective, the top IT industry priority in the near term should be better handling of the pollutants and toxic substances associated with IT production and disposal, especially the latter. To see why, let's look at the three classes of IT benefits in more detail:

• Optimization. For many years, we have heard about smart products, smart buildings, smart cities and even smart planets; but how real and measurable are the benefits? The EAICT project features papers in four areas: the use of semiconductors in the overall economy, smart equipment in small and medium-sized businesses, smart irrigation and water-use systems, and intelligent residential energy auditing. While there is real potential in each area and there are many stories in the marketplace about major savings, the actual results in these cases were mixed. The research shows that smart products and equipment can have relatively short payback periods, but the benefits are often not much greater than would be realized by just keeping existing equipment well maintained. In contrast, the economics of smart irrigation systems are generally not attractive outside of high-cost water markets such as the American southwest. Similarly, given today's technology, few households are willing to put in the time and effort needed to accurately monitor their residential electricity usage. In short, the overall Smart Planet optimization picture is that of a long journey, with significant chicken-and-egg and learning-curve challenges.

• Behaviour. While most behavioural change research focuses on the individual, the four EAICT project papers in this area assessed more systematic possibilities: environmental metrics, industrial ecology, mapping/visualization and complex systems modelling. Again, the results were mixed. By scientific community standards, the environmental field has been slow to adopt modern Web 2.0 tools, and thus a great deal of knowledge is still stuck in various silos, many of them off-line. Ideally, open, semantic web-based technologies would allow sustainability information to flow freely between companies, industries and nations, but this is still rare. This makes it hard for much-hoped-for developments such as industrial symbiosis (one industry's waste is another's input) to emerge. Visual mapping software (showing, for example, how the type and intensity of energy usage varies geographically) can be a powerful policy tool, but modelling the net impact of policy changes in complex environments such as Tokyo is a daunting challenge. At an individual level, the emergence of improved guides, information and peer pressure is reshaping behaviour, but not as fast as desirable. As with optimization, the overall rate of behavioural change is slow and uneven.

• Dematerialization. This is where big gains can be quickly realized. (Of course, within the computer industry, we tend to use the word 'virtualization'. But as virtualization is used in so many different – and often ambiguous – ways, 'dematerialization' is the more precise term.) The EAICT journal features a fascinating analysis of traditional and digital music delivery alternatives. But even in this most familiar and straightforward of net benefit examples, hard data can be elusive. For example, while the overall finding is that digital music delivery is 40 to 80 percent more energy-efficient than the traditional CD retail store, most of the savings come from eliminating the energy used by the customer in driving to the store. If customers also make other stops on the trip, the savings are much less clear. The savings would also shrink substantially if the quality of the digital music and album artwork – and thus the size of the required audio and graphics files – were raised to standard CD quality, but this won't happen until our underlying network and storage technologies are more capable.

  Yet despite these measurement ambiguities, we think dematerialization will have an increasingly strong impact. Consider the way smart phones are already replacing so many physical products – cameras, music players, calendars, tape recorders, compasses, maps, books, newspapers, watches, alarms and even PCs – or the way cloud computing is enabling virtual environments to replace their physical counterparts – data centres, office space, meeting rooms, hotels, stores, malls, theatres and so on. The lessons from the EAICT project suggest that while the accumulated first-, second- and third-order effects of these shifts will probably never be precisely measurable, they should be enough to credibly get us to our 10 percent energy savings target, particularly when combined with the optimization and behavioural gains described above. This means that the overall IT energy savings/consumption balance is likely to be positive over time.

Pages 1 2 3 Next