Review of Computer Energy Consumption and Potential Savings

Summary

After lighting, computers and monitors have the highest energy consumption in office environments. Studies have shown that power management of computers and monitors can significantly reduce their energy consumption, saving hundreds or thousands of dollars a year on electricity costs.

The energy consumption of computers and monitors is determined by the amount of energy they require to operate and how they are used. While the energy requirements of a device make an important contribution to its overall energy consumption, the key to reducing energy consumption is changing how devices are used.

Approximately half of all office computers are left on overnight and on weekends. Evenings and weekends account for 75% of the week, so ensuring computers are turned off at night dramatically reduces their energy consumption. Further savings are made by ensuring computers enter low power mode when they are idle during the day.

Power management is a way of ensuring computers and monitors are turned off when not required and in low power mode during idle periods.

Manual power management, which relies on education users to turn off their computers, can achieve impressive results with ongoing education and reinforcement. Alternatively, automatic power management relies on software, or built in energy saving features. Theoretically, automatic power management can achieve 100% power management, with all computers turned off when not required and in low power mode when idle.

Acronyms and abbreviations

CRT

Cathode ray tube

kWh

Kilowatt hours

LCD

Liquid crystal display

PM

Power management (indicates the device enters low power mode after a period of idle time)

W

Watts

Introduction

Computers and monitors account for 40%-60% of the energy used by office equipment. Their energy consumption is second only to office lighting (Picklum et al: 1999, Roth et al: 2002).

Reducing the energy consumption of computers and monitors is simple. A power managed computer consumes less than half the energy of a computer without power management,[1] and depending on how your computers are used, power management can reduce the annual energy consumption of your computers and monitors by 80%.[2]

The average computer and monitor use 30% of their energy while idle and 40% of their energy outside business hours (Kawamoto et al: 2004). Power management reduces the energy consumed by computers and monitors while they are not in use. This represents a clear opportunity for saving money on energy costs.

The energy consumption of computers and monitors is influenced by two factors:

1. The energy required to run the device, or the power draw;

2. How and when the device is used, that is, its usage pattern.

This document reviews studies on computer energy consumption and usage patterns. All of these studies indicate that effective power management significantly reduces the energy consumption of computers and monitors.

We begin by looking at the energy requirements and usage patterns of computers and monitors. This is followed by estimates of the annual energy consumption of computers and monitors. Finally, we discuss the potential savings that power management can achieve.

Background

This document focuses on research results for desktop computers and their monitors in commercial environments. Although some of the studies include data for server and mainframe computers, we have not included them here, as they tend to have different power management requirements to desktop computers. We have included a discussion on laptop computers for comparison with desktop computers, and because anecdotal evidence suggests they are replacing desktop computers in some office environments.[3]

The research sites documented in the studies are predominantly public sector offices, but sites also include schools, universities, and large and small private sector offices. We have not differentiated between different types of research sites in this document, although some studies suggest computer usage patterns may differ between the public and private sector.[4]

Terminology

We have adopted the terms ‘active’, ’low power’, and ‘off’ to describe the different power states of computers and monitors. These terms are used by a number of the studies. ‘Active’ refers to when the computer is turned on ready for use, but not necessarily being used. ‘Low power’ encompasses commonly used terms such as ‘sleep’ or ‘suspend’. Where a device has multiple low power settings, ’low power’ refers to the lowest of these settings. ‘Off’ describes a device that is turned off but still connected to mains power.

We also make a differentiation between energy requirements and energy consumption. The energy requirement of a device refers to how much power the device requires to operate at a given moment. The energy consumption of a device refers to how much electricity the device uses over a period of time. Energy consumption is a function of the energy requirement of a device and how it is used.

Energy consumption of computers and monitors

The energy consumption of computers and monitors is determined by their energy requirements and their usage pattern. This section looks at both of these factors and how their interaction influences annual energy consumption.

Energy requirements of computers and monitors

Computers and monitors vary in the amount of energy they require.

The energy efficiency of computers and monitors has improved over time, which reduces their energy requirements. However, at the same time increased processing power in computers, and higher resolution monitors have increased energy requirements. As a result, newer computers and monitors tend to use more energy when they are active, but less energy in low power mode than older computers and monitors. Newer computers are also more likely to have power management features than older computers.

The energy requirements of computers and monitors are indicated by their power draw. The power draw, measured in watts (W), is an indication of how much energy a device requires at any given moment.

Computers

There is huge variation in energy required by different computers. The average computer requires between 36W and 250W when it is active, and between 1W and 27W in low power mode.

Computers, like many other electrical devices, draw power even when they are turned off. The average computer draws between 1.5W and 3W when it is turned off but plugged into a mains socket.

A number of studies have looked at the power draw of computers, with different results. Table 1 summarises these results.

Table 1: Energy requirements of computers

The difference in the energy requirements of newer and older computers is highlighted by two studies: one by Roberson et al (2002), and the other by Kawamoto et al (2001).

The study by Roberson et al (2002) looked at computers manufactured between July 2000 and October 2001. The study by Kawamoto et al (2001), which was done around the same time, looked at existing computers in office sites As such these computers were a mix of older models manufactured before 2001.

These studies show that, on average, newer computers use 70W when active and 9W in low power mode (Roberson et al: 2002), whereas older computers use 55W when active and 25W in low power mode (Kawamoto et al: 2001).[5]

Figure 1: Energy requirements of old and new computers (source: Roberson et al: 2002, Kawamoto et al: 2001)

Monitors

Monitors, like computers, vary in the amount of energy they require. The average CRT monitor requires between 66W and 135W when active, between 0W and 19W in low power mode, and between 0W and 5W when switched off.

Table 2 summarises the results of studies looking at the power draw of different monitors.

Newer monitors are more energy efficient than older monitors. Two factors contribute to the increased energy efficiency of monitors:

1. New CRT monitors require less energy than older CRT monitors of equivalent size;

2. LCD monitors require much less energy than CRT monitors.

The two studies by Roberson et al (2002), and Kawamoto et al (2001) highlight the difference in energy requirements between newer CRT monitors and older CRT monitors.

Table 2: Energy requirements of monitors

The study by Roberson et al (2002) looked at monitors manufactured between July 2000 and October 2001. The study by Kawamoto et al (2001) looked at existing monitors in office sites, as such these monitors were a mix of older models manufactured before 2001.

These studies show that, for 17 inch CRT monitors, the newer models require 61W when active, 2W in low power mode, and 1W when turned off (Roberson et al: 2002). In contrast, the older models require 85W when active, 5W in low power mode, and 0.5W when turned off (Kawamoto et al: 2001).

Figure 2: Energy requirements of old and new CRT monitors (source: Roberson et al: 2002, Kawamoto et al: 2001)

LCD monitors require a fraction of the energy of CRT monitors. The average LCD monitor requires 15W when active, 1.5W in low power mode, and 0.5W when turned off (Kawamoto et al: 2004).

Figure 3: Energy requirements of CRT and LCD monitors (source: Kawamoto et al: 2004)

Anecdotal evidence suggests that although offices may update their computers regularly, they do not update their monitors as regularly. On average, only 17% of office monitors are LCD (Webber et al: 2006).

The size of a monitor makes a significant contribution to its energy requirement.

On average, CRT monitors require 0.53W/inch², and LCD monitors require 0.23W/inch². However, the research suggests that difference in energy consumption between CRT monitors and LCD monitors decreases as size increases (Roberson et al: 2002).[6]

Conclusions

As expected, all the studies indicate that computers and monitors use much more power when they are active compared to when they are in low power mode or turned off. This is true for both older and newer models.

Computers and monitors vary significantly in their power requirements. Computers and monitors with a lower power draw can make a contribution to reducing energy costs. It appears the most significant savings would be achieved by replacing CRT monitors with LCD monitors.

Usage patterns of computers and monitors

The power draw of computers and monitors is only one factor affecting total energy consumption. The energy efficiency of a computer makes a small contribution to its total energy consumption. The way in which a computer is used is a far more significant factor in determining the total energy consumption.

An energy efficient computer that is always on consumes more energy than a less energy efficient computer that is regularly turned off.

Looking at typical usage patterns of computers and monitors provides a clearer picture of how their total energy consumption can be reduced than simply looking at energy requirements.

Usage patterns are typically separated into two periods: nighttime usage, which usually includes weekends, and daytime usage.

Nighttime usage

On average, only half of all computers and monitors are turned off at night. In some cases computers are left on overnight to allow back-ups or maintenance to occur, but in other cases there is no obvious reason for computers to be left on.

The study by Webber et al (2001) found, on average, 44% of computers were turned off at night, 3% were in low power mode, and the remainder, approximately 54% of computers, were active.

This study found huge variability in turn off rates. At some research sites only 2% of computers were turned off, while at other research sites 91% of computers were turned off.[7]

A more recent study by Webber et al (2006) found, on average, only 36% of computers were turned off at night, 4% of computers were in low power mode, and the remainder, 60%, were active.

Other studies have found a much higher proportion of computers are turned off at night.

A study by Kawamoto et al (2001) found 65% of computers were turned off at night and the remaining 35% were active.

A study by Nordman et al (2000) also found, on average, 65% of computers were turned off at night, 10% were in low power mode, and 25% were active. This study compiled the data from 17 previous studies conducted between 1990 and 2000. Like other studies, this study also showed huge variability in turn off rates, which ranged from 0%- 91% of computers turned off at night.

Figure 4: Nighttime power state of computers

The picture for monitors is slightly different. The studies have shown a much lower proportion of monitors are active at night.

The earlier study by Webber et al (2001) found 32% of monitors were turned off, 38% were in low power mode, and 30% were active.

The recent study by Webber et al (2006) showed similar results. For CRT monitors, 32% were turned off, 49% were in low power mode, and 19% were active. For LCD monitors, 18% were turned off, 61% were in low power mode, and 21% were active.

As with computers, the study by Nordman et al (2000) found a much higher proportion of monitors are turned off at night. They found 65% were turned off, 20% were in low power mode, and 15% were active.

Figure 5: Nighttime power state of monitors

A surprising number of computers and monitors are left on at night. Significantly, most of the computers and approximately half of the monitors are active, not in low power mode.

As the studies of power draw indicate, computers and monitors consume much more energy when they are active than they do in low power mode. Reducing the time computers and monitors spend unnecessarily in active mode will reduce their energy consumption.

In a 9 to 5 office, computers are used for less than 25% of the week (Webber et al: 2001). Ensuring computers are turned off at night reduces their running time by 75%. This represents an opportunity to make significant savings on energy consumption and electricity costs.

Daytime usage

The way in which computers are used during the day is also important. Of the 25% of the week the average computer is in use, a significant proportion is likely to be idle time.

A study by Kawamoto et al (2004) found that in the average office, a computer is used for 6.9 hours a day. Of those 6.9 hours it is in active use for 3 hours, and idle for the remaining 3.9 hours.

A computer which is actively used 3 hours a day, 5 days a week, is only in use 9% of the week.

The study by Mungwititkul and Mohanty (1997) and the study by Nordman (1999) both found that on average, computers are active for 9% of the year.

Ensuring computers are in low power mode during these idle periods further reduces their energy consumption.

Effect of idle time delay

It is unrealistic to assume that a computer is put into low power mode as soon as it becomes idle. Power managed computers usually enter low power mode after a specified delay. The length of the delay affects how much of the idle time the computer spends in active mode and how much of the idle time it spends in low power mode.

The study by Kawamoto et al (2004) found that the average computer is idle for 3.9 hours a day. If a computer goes into low power mode after 5 minutes of idle time it will spend 76% of those 3.9 hours in low power mode. If a computer goes into low power mode after 30 minutes of idle time it will spend just 34% of those 3.9 hours in low power mode.

The best length for the delay period will be determined by how the computer is used. Someone who spends a lot of time reading on screen will need a longer delay period than someone who spends most of their time typing.

Table 3: Effect of idle time delay on power state (source: Kawamoto et al: 2004)

Figure 6: Effect of idle time delay on power state (source: Kawamoto et al: 2004)

Conclusions

Turning computers and monitors off at night and putting them into low power mode when idle represents a significant saving in energy consumption. Almost half of the computers studied are active 100% of the time. Turning them off at night would reduce the time they spend active to 25% of the time. Turning them off at night and putting them into low power mode when idle would further reduce the time they spend active to less than 10% of the time.

Figure 7: Percentage of time in active mode for different usage patterns (source: Webber et al: 2001, Kawamoto et al: 2004)

Laptop computers: Energy requirements and usage patterns

Many of the energy efficient features in desktop computers were originally developed for laptop computers to prolong the time they could run off the battery.

Laptop computers consume a fraction of the energy of desktop computers. This is due to a number of factors:

• They draw less power;
• They go into low power mode more quickly than desktop computer to preserve battery power;
• They are more frequently turned off and unplugged than desktop computers.

Power draw

The study by Kawamoto et al (2001) found that laptops require between 12W and 22W when active, between 1.5W and 6W in low power, and between 1.5W and 2W when turned off if the battery is fully charged.

This study calculated that on average laptops require 15W when active, compared to 55W for active desktops, 3W in low power mode, compared to 25W for desktops in low power mode, and 2W when turned off, compared to 1.5W for desktops which are off.

Another study (Roberson et al: 2002) found similar results; laptop computers require 19W when active, and 3W in low power mode, whereas desktop computers use 70W when active and 9W in low power.

Figure 8: Energy requirements of laptop and desktop computers (source: Kawamoto et al: 2001)

Few studies consider the energy required to charge a laptop computer. The study by Nordman (1999) suggests that a laptop which is charging draws 12-24W when plugged in but turned off.

Usage pattern

There has been little detailed research done on the usage patterns of laptop computers. In part, this is because the methods used to determine usage patterns of desktop computers, such as night audits, do not give a complete picture for laptop computers. The portable nature of laptop computers means people have often taken them home when the audits take place. There was no data in the studies on how a ‘work’ laptop is used by an employee at home.

However, the research that has been done shows some important differences in the use laptop computers and desktop computers.

On average laptop computers spend more time turned off and in low power mode than desktop computers.

The study by Kawamoto et al (2004) found that 97% of laptop computers were turned off at night, compared to 82% of desktop computers turned of at night.[8]

The study by Roberson et al (2002) found that most laptop computers had light sleep mode. This mode draws less energy than active mode, but more energy than low power mode. Laptop computers require 11W in light sleep.

Anecdotal evidence suggests that laptop computers enter light sleep after a very short period of idle time before entering low power mode after longer periods of idle time. The purpose of light sleep is to prolong battery power when the laptop is not plugged into a mains socket. This feature also makes laptops much more energy efficient when they are plugged in.

A study by Roth et al (2002) shows that the average annual energy consumption of a laptop computer is less than 15% of the annual energy consumption of a desktop computer. A large proportion of this saving is due to the different usage patterns of laptop and desktop computers. Whereas desktop computers spend most of the time in active mode, laptop computers spend most of the time in low power mode. They assume that desktops and laptops spend the same amount of time turned off.

Figure 9: Typical usage pattern of laptop and desktop computers (source: Roth et al: 2002)

Annual energy consumption of computers and monitors

Annual energy consumption ranges from 52kWh to 482kWh per year for computers, and from 22kWh to 754kWh per year for monitors.

The annual energy consumption of the average office computer and monitor is influenced by two factors:[9]

1. Different energy requirements of individual computers;

2. Different usage patterns and user behaviour.

Over the course of a year, these differences can have a huge effect on the total energy consumption.

Table 4 and Table 5 summarise the results of five studies that measured the annual energy consumption of computers and monitors.

Table 4: Annual energy consumption of computers

Explaining variation in annual energy consumption.

The annual energy consumption of computers reported by the studies ranges from 52kWh per year to 482kWh per year. How do different studies achieve such different results?

The studies can be divided into two groups: those that calculate one annual energy consumption figure, which is an average of all computers in all power states, and those that calculate different annual energy consumption figures for different usage patterns.

Two of the studies (Roth et al: 2002, Kawamoto et al: 2001) calculate the annual energy consumption as an average of computers in all power states (that is, active, low power, and off). The study by Roth et al (2002) assumes 25% of computers have power management features. The study by Kawamoto et al (2001) also assumes that 25% of computers have power management features, and that 65% of computers are turned off at night. Both of these studies assume the same power draw for computers, so the difference between their results is explained by the different usage patterns they have used.

The other three studies (Webber et al: 2006, Mungwititkul and Mohanty: 1997, Nordman: 1999) calculate the annual energy consumption for computers in different power states.

The study by Webber et al (2006) calculates annual energy consumption of computers with the same power draw, but four different usage patterns: always on, turned off at night, always on with power management features, and turned off at night with power management features.[11] This study highlights the dramatic effect that different usage patterns can have on total energy consumption.

The annual energy consumption found by the other two studies results from differing usage patterns and power draw in each of the states they describe.

The study by Mungwititkul and Mohanty (1997) uses different power draws for computers that have energy saving features, and those that do not have energy saving features.

This study (Mungwititkul and Mohanty: 1997) gives one typical usage patterns for office computers. It does not provide individual usage patterns for the three states described: no energy saving features present, energy saving features present and enabled, energy saving features present, but disabled. Computers with energy saving features have a lower power draw than those without, and only those computers with energy saving features enabled enter low power mode.

The study by Nordman (1999) is similar in that it compares the annual energy consumption of computers and monitors that comply with the Energy Star[12] scheme, with those that do not comply with the scheme. To comply with the Energy Star scheme, computers and monitors must have a low power mode and their power draw must be below a specified level. As such, Energy Star compliant computers are likely to have a lower average power draw than non-compliant computers and the average usage pattern of Energy Star compliant computers will include a low power mode.

Comparing the research results

The study by Mungwititkul and Mohanty (1997) indicates that computers are almost always turned off at night. As such their ’no energy saving’ usage pattern is similar to the ‘off at night’ usage pattern in the study by Webber et al (2006). If you consider the lower power draw of the study by Mungwititkul and Mohanty (1997), the annual energy consumption for these states, 80kWh (Mungwititkul and Mohanty: 1997) and 100kWh (Webber et al 2006), seems comparable.

Similarly, the ’energy saving enabled’ usage pattern (Mungwititkul and Mohanty: 1997) and the ’turned off at night with power management’ usage pattern (Webber et al 2006) in these two studies are comparable. They give annual energy consumption estimates of 52kWh and 86kWh respectively. Once again, these studies use a different power draw for the computers. The study by Mungwititkul and Mohanty (1997) uses a lower power draw for this usage pattern (36W) than for the ’no energy saving’ pattern (48W) described above. In contrast, the study by Webber et al (2006) uses a power draw of 55W for all usage patterns.

We can also compare the two studies that give average annual energy consumption (Roth et al: 2002, Kawamoto et al: 2001) with the study by Webber et al (2006). These three studies all use the same power draw in their calculations.

The average annual energy consumption of the four usage patterns described by Webber et al (2006) is 231kWh. This average energy consumption assumes equal proportions of computers in each of the four states described. This average falls between the higher annual energy consumption of 297kWh found by Roth et al (2002), and the lower annual energy consumption of 213kWh found by Kawamoto et al (2001).

The higher average energy consumption in the study by Roth et al (2002) may be explained by the higher proportion of computers in active mode, while the lower average energy consumption in the study by Kawamoto et al (2006) may be explained by the higher proportion of computers turned off at night.

Monitors

Monitors show a much greater variation in annual energy consumption than computers, ranging from 22kWh to 754kWh per year.

The factors that cause variation in the annual energy consumption of monitors are the same as those that cause variation in computers discussed above.

In addition, there is also a large difference between the average power consumption of LCD monitors and the average power consumption of CRT monitors. LCD monitors require much less energy.

Table 5: Annual energy consumption of monitors

Most of the studies looked at CRT monitors, largely because CRT monitors are still much more common in office environments than LCD monitors. The study by Webber et al (2006) highlights the difference in annual energy consumption between CRT monitors and LCD monitors.

Figure 10: Annual energy consumption of CRT and LCD monitors (source: Webber et al: 2006)

The lowest annual energy consumption is achieved by a computer with LCD monitor that is turned off at night and power managed. The highest annual energy consumption is achieved by a computer with CRT monitor that is always on.[13] The latter will consume more than ten times the amount of energy every year.

Energy consumption of Mac computers

Although some of the research includes Mac computers in their sample, they rarely report the results separately. We did not find any research that focused solely on Mac computers. As such, there are no average values for the energy requirements of Mac computers. Some studies did report energy measurements for individual computers.[14]

Table 6 summarises the results of these studies.

Table 6: Energy requirements of older Mac computers

Many of the newer model Mac computers combine the computer and monitor in a single unit. These are referred to as ‘integrated computer systems’ in the research literature.[15]

As the computer and monitor are combined, their energy requirements cannot be measured separately. However, it is quite common for the monitor to enter low power while the computer remains active. Measurements by Bluejay (2006) indicate that this saves 35W.[16]

When you consider independent LCD monitors save 18W-51W in low power mode (Roberson et al: 2002), a saving of 35W when only the monitor is in low power mode seems reasonable.

Based on the environment specifications provided by Apple, current model Macs are more energy efficient than older models. Most current model Macs are also more energy efficient than the average results reported by the studies.

Table 7 sets out the energy requirements of current model Mac computers. Table 8 estimates the annual energy consumption of current model Mac computers for different usage patterns.

Table 7: Energy requirements of current model Mac computers (source: Apple Product Environmental Specifications)

Table 8: Annual energy consumption of current model Mac computers[18]

With the exception of the 24" iMac and the powerful, high end Mac Pro, the annual energy consumption of current model Mac computers compares favourably with the results reported by the studies for average computers.

Indirect energy use

In addition to the considerable direct contribution computers and monitors make to the energy consumption of an office, they also make an indirect contribution to energy consumption.

The study by Roth et al (2002) found that office equipment increases the load on air conditioning by 0.2-0.5kW per kilowatt of office equipment power draw.

So, for every kilowatt of electricity your computers use, your air conditioning system uses an extra 0.2-0.5kW to keep the office cool. Reducing the electricity consumption of computers reduces the heat they produce, and thereby reduces the burden on your air conditioning system.