is produced by the non-profit Center for Energy & Climate Solutions
About the Site
Proven Results
How-To Guide
Internet & Energy
Law & Policy
Media Contact
Corporate Commitments
Profits & Productivity
A source for everyone interested in energy, business and the environment  
" invaluable guide to a more helpful future..."
    -Walter Cronkite


Analysis Debunks Myth that Net is Energy Hog,
Says Supply Troubles Would Likely be Even Worse Without it

The proliferation of Internet and Internet-based technology throughout the U.S. has been widely credited for tremendous productivity gains that propel both profits and growth. Less noticed, however, are the dramatic savings in energy use and the resulting pollution that are being generated by this New Economy technology - both as a function of broader productivity gains and direct efforts to reduce energy costs using better information technology. Indeed, many people believe incorrectly that the Internet is increasing our overall energy use.

Based on our extensive analysis over the past year, the Internet appears to be dramatically reducing the amount of energy America needs to propel its economy. Going forward, we strongly believe that Internet technology will unlock unprecedented savings of energy and pollution emissions. I am especially heartened by dramatic new data - data that gets stronger with each passing month - indicating that the fundamental relationship between energy use and economic growth in the United States has been changed permanently by the spread of New Economy technology to every corner of our lives.

A Fundamental Change Unfolds: A New Energy Economy

The story begins with a few simple, but truly amazing facts. Since 1996 - a period that corresponds with the tremendous growth of the Internet and e-Commerce - the nation experienced remarkable economic growth, driven to a significant extent by information technology (IT) industries. Overall productivity of the economy appears to have increased substantially as well, also driven by the IT sector. What is startling is that the nation's overall productivity gains have been accompanied by an equally impressive gain in energy productivity.

In the immediate pre-Internet era (1992-1996), GDP growth averaged 3.2 percent a year, while total energy demand grew 2.4 percent a year. In the Internet era (1996-2000), GDP growth is averaging more than 4 percent a year, while energy demand is growing only 1 percent a year. Demand for electricity, which some say is going up due to demand by the Internet itself, follows a similar trajectory.

This is a remarkable change - higher GDP growth and lower energy growth. From 1996 through 2000 U.S. energy intensity - the energy consumed per dollar of output - improved at the unprecedented rate of 3 percent per year, or triple the rate of the preceding decade.

From the point of view of greenhouse emissions, the immediate pre-Internet era saw 2 percent annual rises in carbon dioxide emissions, while the Internet era has seen rises of slightly over 1 percent. In 1998, U.S. emissions of greenhouse gases grew just 0.2 percent, the smallest rise since 1991, when the economy was in the throes of recession.

Analysis by EPA and the Argonne National Laboratory suggests one third to one half of the recent improvements in energy intensity are "structural" - when economic growth shifts to sectors of the economy like information technology that are less energy intensive than others. The remaining one-half to two-thirds of the improvement comes from energy efficiency improvements throughout the system as a whole, as businesses practices change in ways that reduce energy use relative to output. For example, a factory might use more efficient motors on its assembly line or better lighting in its buildings.

According to our findings, the Internet economy itself seems to be generating both structural and efficiency gains. If companies put their stores on the Internet, rather than constructing new retail buildings, which would represent an Internet structural, gain. If that same company used the Internet to more effectively manage its existing supply chain, it would be an efficiency gain.

Internet Technology Cuts Energy Use in New, Old Economy

Clearly, both sorts of activities are taking place, with major energy implications. In business-to-consumer e-commerce, for instance, a warehouse holds far more product per square foot than a retail store, and uses far less energy per square foot. We calculated the ratio of building energy per book sold in traditional bookstores versus on-line retailer to be 16-to-1. And contrary to what most people think, Internet shopping uses less energy to get a package to your house: Shipping 10 pounds of packages by overnight air - the most energy-intensive delivery -uses 40 percent less fuel than driving roundtrip to the mall. Shipping by truck saves 90 percent.

Business-to-business e-commerce, estimated at 5 to 10 times the size of business-to-consumer trade, may yield even bigger savings. As traditional manufacturing and commercial companies put their supply chain on the Internet, and reduce inventories, overproduction, unnecessary capital purchases, paper transactions, mistaken orders, and the like, they achieve greater output with less energy consumption.

Analysts at Ernst & Young, for example, estimate that collaborative planning systems between manufacturers and suppliers could reduce inventories by $250 to $350 billion across the economy, roughly 25 to 35 percent of finished goods stock. IBM says its e-commerce solutions are delivering inventory savings as high as 50 percent for some of their customers.

This is more important than you might think, because the energy used to create and transport the raw materials that a company uses may vastly exceed energy they use directly. For instance, Interface Flooring Systems calculates this "embodied energy" in raw materials for its carpet tile outstrips the energy needed to manufacture it by a factor of twelve. That means a 4 percent cut in wasted product could save the equivalent of fully half the energy used in manufacturing.

The resulting impact on energy use and global warming pollution would be dramatic.

By 2007, business-to-consumer and business-to-business e-commerce together could avoid the need for 1.5 billion square feet of retail space and up to 1 billion square feet of warehouse space. Internet technology may also eliminate as much as 2 billion square feet of commercial office space - the equivalent of almost 450 Sears Towers - along with all the lighting, heating and cooling that goes with it.

Energy savings just from the operations and maintenance of these "un-buildings" total 53 billion kilowatt hours per year, about 13 percent of total electricity growth projected under business-as-usual scenarios. That equals the output of 21 average power plants, plus 67 billion cubic feet of natural gas. Expressed in terms of the global warming issue, this Internet "un-building" scenario would prevent the release of 35 million metric tons of greenhouse gases.

Avoided construction of all those buildings saves the equivalent of 10 more power plants worth of energy, and another 40 million metric tons of greenhouse pollution. By 2010, e-materialization of paper, construction, and other activities could reduce U.S. industrial energy and greenhouse emissions by more than 1.5 percent.

What About Energy Use BY the Internet?

As to the important question whether the Internet itself is consuming vast amounts of electricity, the facts simply - and irrefutably - fail to support such a conclusion. To begin with, the rate at which U.S. electricity demand is growing has slowed since the start of the Internet boom.

The pre-internet era saw electricity demand rise 2.9 percent per year. Since 1996, electricity demand has risen only 2.2 percent per year. And this has all occurred in spite of higher GDP growth since 1995, hotter summers (1998 was the hottest summer in four decades in terms of cooling-degree days; 1999 was the second hottest summer), and less support by utilities for demand-side management, all of which would normally lead to higher growth in electricity demand. We suspect this has much to do with the trends already discussed here.

These figures run counter to a argument that is frequently - and incorrectly - repeated in the media, that the Internet has become a major energy consumer because it supposedly requires a great deal of electricity to run the computers and other hardware powering the Internet economy.

This hypothesis can often be traced back to analysts Peter Huber and Mark Mills. A closer examination, however, indicates that Mills and Huber have significantly overestimated the energy consumption of most critical pieces of equipment.

Scientists at Lawrence Berkeley National Laboratory (LBNL) examined in detail the numbers underlying a Mills and Huber analysis, and found that the estimates of the electricity used by the Internet were high by a factor of eight. Major overestimates were found in every category, including their calculations of energy used by major dot-com companies, by the nation's web servers, by telephone central offices, by Internet routers and local networks, and by business and home PCs.

Mills and Huber assumed, for instance, that a "typical computer and its peripherals require about 1,000 watts of power." In fact, the average PC and monitor use about 150 watts of power; this dips to 50 watts or less in energy-saving mode. Laptop computers, a key growth segment, are particularly low energy users, with some using under 30 watts. Moreover, computers are getting more energy-efficient because of steady improvements in technology driven in part by the growing market for portable equipment. New flat screens typically use about a quarter of the energy of traditional video display terminals with cathode ray tubes.

These basic mistakes are reflected in their conclusions. Mills and Huber claim that from 1996 to 1997, the increase in electricity consumed by all computers used for the Internet constituted more than 1.5 percent of all U.S. electricity consumed that year. Yet total electricity consumption for all purposes grew slightly less than 1.4 percent during that period, which would imply that electricity growth for everything else equaled zero - despite economic growth 4.5 percent. While we believe that the Internet reduces energy intensity, we don't believe it has quite that dramatic an effect.

But mathematical and data errors are only part of the problem. Indeed, we believe Mills and Huber have the entire Internet energy story almost completely backwards:

One of the reasons why energy intensity declined so slowly from 1987 through 1996 is likely that businesses in particular purchased a great many computers and other IT equipment that consume electricity, yet generated little accompanying productivity gains to offset that increased energy use. But Internet changed all that, unleashing a storm of new productivity in every sector of the economy. By then, of course, most desks already had computer. The added energy needed to shift PCs from traditional uses to the Internet is modest compared to its overall benefit.

A Few Unknowns About the Internet & Energy Use

There are aspects of the Internet that will probably entail more energy use, such as greater small-package delivery by truck. These cases may not, however, result in a net increase in energy use; relatively efficient package delivery by truck may replace at least some relatively inefficient personal driving to malls, supermarkets, bookstores and the like - particularly if most of the packages are delivered by the Post Office, which already visits virtually every home in the country on a daily basis.

The great unknown in this regard is whether or not a significant fraction of Americans will change their driving habits over the next few years once it is possible to make a critical mass of cyber-trips on the Internet. That is, will the Internet be the mall of the 21st Century? We suspect the Internet economy will be no worse than neutral in the transportation sector, but could well have a large positive impact. Already, in the last two and a half years, the growth rate in vehicle miles traveled (VMT) has slowed, and the VMT to GDP ratio has dropped dramatically.

Computers and the Internet may well lead to more home electricity consumption. This is part of a long-standing trend, as homes have for some time been getting bigger and more stocked with electronic equipment. But the question is, if people spend more time on the Internet, what are they spending less time doing?

Some will be watching television less; others reading newspapers less; some may be printing individual items of interest to them rather than receiving entire printed catalogs or directories in the mail; others will be working at home rather than in an office building; and, potentially, some may be not be driving to work or to malls as often as before. These are all activities that would normally consume a great deal of energy.

Changes in Energy Technology Meet Changes in Information Technology

The application of New Economy information technologies to traditional energy-use technologies has resulted in quantum improvements even in two classical sectors that are responsible for most electricity consumption: lighting and electric motors. The result is more energy savings in parts of the economy not traditionally considered "high-tech."

We have seen steady advances in solid-state electronic ballasts for running fluorescent lamps, which not only save considerable energy compared to magnetic ballasts, but also eliminate annoying flicker and hum. The new ballasts can be run with highly sophisticated, low-cost controls that automatically dim the lights to offset daylight in the room. These lamps can also be controlled even at the desktop by remote controls or through a PC. Greater control over the workplace environment in general, and lighting in particular, has been linked to productivity increases.

Similarly, computer-controlled adjustable speed drives for motors can simultaneously reduce energy consumption and improve process control, achieving significant direct cost savings as well as productivity gains. Even boilers and hot water heaters can cut energy consumption 25 percent or more through the installation of microprocessor-based controllers.

Digital energy management control systems (EMCS) can continuously gather data about what is taking place in a building and how its equipment is operating, feeding it into a central computer used to control building systems and optimize energy performance. Energy experts at Texas A&M have shown in two dozen Texas buildings that using such an approach can cut energy use 25 percent with an 18-month payback in buildings that have already received on upgrade with the latest energy-saving equipment.

Increasingly, such technologies will operate over the Internet itself. We know of one major energy service company pursuing the installation of digital EMCS's in the buildings they manage, so they can operate them over the Internet very efficiently and at low cost. A similar arrangement is already operating in Singapore.

Many utilities have begun exploring Internet-based home energy management systems, which would give individual homeowners more control and feedback over their home energy use, or the ability to have an outside energy company or expert software system optimize their energy consumption. Early trials of remote controlled home energy management systems suggest the savings in energy bills could be as high as 10 percent.

Spreading the Gospel: Rousing Corporate America to the Energy Challenge

As Fortune magazine noted in 1998, "only a third of U.S. manufacturers are seriously scrutinizing energy usage, where savings in five areas can move billions to the bottom line." Thanks to low energy prices and the benefits of energy efficiency investments in the 1970s, energy in mid-1980s became a much lower fraction of the cost of doing business. Naturally, companies reduced investments in energy-saving technologies. During the downsizings of the early 1990s, corporate energy staffs were often sharply reduced or eliminated entirely.

As a result, most companies have lacked both the motivation and the management expertise to improve energy performance for most of this decade. Many companies, including some of our largest and most energy intensive, have been making investments in energy-savings technologies only if they paid for themselves within about a year.

There are exceptions. Some companies, including IBM, Johnson & Johnson and Polaroid, have instituted corporate wide policies to adopt energy-saving technologies. They have been able to sustain steady improvements in their corporate energy intensity of 4 percent per year and 3 percent per year respectively throughout the 1990s.

Though virtually every company could do what IBM and J&J have done, they are still the exceptions. But energy outsourcing - another New Energy Economy trend - is starting to change this. Soon it may revolutionize corporate energy efficiency investments.

Because most companies typically consider energy issues as secondary to core business concerns, they typically pursue only simplest, most obvious solutions, which means investments in energy-efficient equipment only with a payback of a year or so. To an outside contractor, energy is the core business. That means they have more expertise and longer investment horizons that allow them solid returns on energy investments with five- to seven-year paybacks (or sometimes as long ten years).

Some companies have turned over their entire power supply needs to outside contractors. In March 1999, Ocean Spray announced a $100 million deal with the energy services division of Enron, a major natural gas and utility company based in Houston. Enron will use its own capital to improve lighting, heating, cooling and motors and to invest in cogeneration (the simultaneous generation of electricity and steam onsite, which is highly efficient). Ocean Spray will save millions of dollars in energy costs, have more reliable power and cut pollution, without putting up any of its own capital. In September 1999, Owens Corning, the fiberglass insulation manufacturer, announced a similar $1 billion deal with Enron.

Many other energy service companies are taking a similar approach. Some, like Sempra Energy Solutions, have even gone so far as to finance, build, own and manage the entire energy system of a customer. Substantial investments in such outsourcing deals are relatively recent phenomena. But I believe these deals will grow very rapidly in the next few years, and are likely to ultimately achieve savings well beyond that achieved by utility demand-side management (DSM) programs, which have scaled back dramatically with the onset of utility restructuring.

This is especially true for two reasons. First, traditional DSM often focused on retrofitting individual electricity-using components, whereas outsourcing encourages a whole systems approach to efficiency covering all fuels, an approach that can achieve deeper savings at lower cost. Second, traditional DSM did not encourage much cogeneration, which the outsourcing deals do. And cogeneration combined with energy efficiency can cut the energy consumption of a building or factory by 40 percent or more in a period of just a few years.

An Optimistic Prognosis

In conclusion, we find great cause for optimism over the prospects for reducing energy demand and energy-related pollution while maintaining a strong and vibrant economy. Indeed, it is that very vibrancy that has improved this prognosis substantially in recent years. And we challenge those pessimists who consider the Internet a problem, rather than a solution, to rethink their interpretation. With or without them, the New Economy is changing the way America uses energy; in concert with sound climate policies, we can count on the Internet revolution to help us protect and preserve our environment as well.