Plug Load: What It Is, Why It Keeps Growing and How to Cut It

Plug load is the electricity used by equipment that plugs into an ordinary wall socket. In an office, that means computers, monitors, printers, copiers, phone chargers, desk fans, space heaters, coffee machines and refrigerators. At home, it means televisions, game consoles, routers, set-top boxes, smart speakers, kitchen appliances and the dozens of small adapters that charge everything else.

For most of the twentieth century, building energy efficiency focused on the big, built-in systems: heating, cooling, ventilation, lighting and water heating. Those systems are designed by engineers, covered by building codes and controlled by facility managers. Plug loads are different. They are bought by individual occupants, brought in after a building is finished, rarely metered separately and mostly controlled by whoever happens to be sitting nearby. As the big systems became more efficient, plug loads grew to become one of the largest and fastest-growing parts of building energy use.

This article explains how plug loads are defined, how large they are, how they behave, how they are measured, how they have been regulated, and what households and organizations can do to reduce them. It includes worked examples (clearly labeled as hypothetical) to show how the arithmetic works.

Definitions

The basic definition

Wikipedia, drawing on a 2006 paper by Nordman and McWhinney of Lawrence Berkeley National Laboratory (LBNL), defines plug load as the energy used by products that are powered by an ordinary AC plug. The term generally excludes energy used by the major building end uses: heating, ventilation and air conditioning (HVAC), lighting and water heating.

Related terms

Many terms overlap with "plug load", and building codes rarely use the phrase itself. The most common are:

The ASHRAE and LEED definitions of process energy list similar categories: office and general equipment, computers, elevators and escalators, kitchen cooking and refrigeration, laundry, lighting exempt from the lighting power allowance, and other uses. Not all of these plug into a wall socket. Elevators, for example, are hard-wired. That is why researchers such as NREL use the combined term "plug and process loads", which captures everything outside the major building systems.

Standby power, idle load and "vampire" power

Several terms describe energy used when equipment isn't doing anything useful:

• Standby power is the electricity an appliance uses while switched "off" but still in a low-power state, for example to keep a clock running, listen for a remote control or show a display. It is also called vampire power, phantom load or ghost load.
• Idle load is a broader term used by the Natural Resources Defense Council (NRDC) for all electricity used by devices that are always on but not actively in use, including devices that never switch off, such as routers, and devices left in standby.
• Sleep mode is a reduced-power state that a device enters automatically after a period of inactivity, from which it can wake quickly.

How large are plug loads?

Commercial buildings

Estimates of the share of commercial building energy taken by plug and process loads depend heavily on what is counted and what base is used.

The numbers are different because they measure different things. The 47% figure covers all plug and process loads, including hard-wired process equipment, while the 19% figure is a narrower office-equipment measure for one building type from 2003 data. They should not be compared as if they were a trend.

A growing share

What all sources agree on is that plug loads are growing in relative importance:

• In its Annual Energy Outlook 2000, published in 1999, the U.S. Energy Information Administration (EIA) projected that office equipment would be the fastest-growing commercial end use between 1998 and 2020.
• The NREL guide cites U.S. Department of Energy projections that, by 2030, commercial building energy use would rise by 24% while plug and process load energy use would rise by 49%.
• In June 2025, the EIA reported that computing (which in its definition covers data center servers, desktop and laptop computers and monitors in commercial spaces) used an estimated 8% of U.S. commercial sector electricity in 2024. In its Annual Energy Outlook 2025 reference case, that share rises to 20% by 2050, making computing the fastest-growing end use in commercial buildings and potentially the largest, ahead of lighting, space cooling and ventilation.

Two forces drive this growth. First, as HVAC and lighting become more efficient, whatever is left takes a larger share. Second, the number of plugged-in devices keeps rising.

Homes

In homes, the headline numbers concern standby and idle loads:

• A 1998 LBNL study estimated standby power at about 5% of U.S. residential electricity, costing more than $3 billion a year.
• LBNL researcher Alan Meier has estimated that standby power accounted for up to 10% of household electricity use in some cases before efficiency rules took effect. A French study found about 7%, and some estimates have reached 13%.
• In 2015, NRDC published Home Idle Load: Devices Wasting Huge Amounts of Electricity When Not in Active Use, produced with Home Energy Analytics and the Stanford Sustainable Systems Lab. Using smart-meter data from about 70,000 homes in northern California, along with detailed audits of 10 homes, it found that always-on idle loads accounted for nearly 23% of electricity use in California homes on average. The audited homes had an average of 65 devices, and about two-thirds of them drew more than 1 watt while in always-on mode. NRDC extrapolated the results nationally, estimating that idle loads cost U.S. households about $19 billion a year, or roughly $165 per household at average national electricity rates, and up to $440 a year at some utilities' top-tier rates.

The NRDC estimate is based on California data extrapolated nationally, and it includes devices that legitimately need to stay on, such as refrigerators' control boards, security systems and routers. It is best read as the size of the "always on" slice, not as the amount that could realistically be saved.

How plug loads behave

Power states

Most modern electronic devices move between several power states:

Savings from sleep modes vary greatly. A 2006 study by Carrie Webber and colleagues, cited by Wikipedia, found that sleep-mode savings ranged from about 55% for desktop computers to about 94% for CRT monitors. That variability is one reason plug load estimates are uncertain.

Time matters more than power

Office equipment spends most of its time unused. LBNL's 2001 field study noted that in an office open from 9 to 5, office equipment is used for less than 25% of the hours in a week. The rest of the time, its energy use depends on whether it has been switched off, put to sleep or left on.

The same study, Field Surveys of Office Equipment Operating Patterns (LBNL-46930), carried out 11 after-hours walk-throughs of offices in the San Francisco and Washington, D.C. areas. It found that only 44% of computers, 32% of monitors and 25% of printers were turned off at night. It estimated that power management was working on about 56% of monitors and enabled on about 96% of printers.

Nameplate power is misleading

The label on a device shows its rated (nameplate) power, which is a maximum for electrical safety, not a measure of typical use. ASHRAE research by Hosni and Beck (2009), cited by Wikipedia, found that average consumption can be as low as 10 to 15% of the nameplate value. Designers who size cooling systems or estimate energy use from nameplate ratings will overestimate considerably.

Plug loads make heat

Almost all the electricity used by plug loads ends up as heat in the room. In summer, or in buildings that need cooling year-round, that heat adds to the cooling load, so the building pays twice: once for the plug load and again to remove its heat. In winter, plug-load heat offsets some heating demand. As Wikipedia notes, using electricity this way is a less efficient form of heating than a heat pump, but the offset is real if the electricity is being used anyway. The net effect depends on climate, building type and heating system. Reducing plug loads almost always cuts electricity use overall, but in heating-dominated buildings some of the saving is offset by additional heating.

A short history of plug load efficiency

ENERGY STAR (1992)

The U.S. Environmental Protection Agency launched the ENERGY STAR program in 1992, and the first products labeled were computers and monitors, followed by printers in 1993. According to EPA material, the first computer specification required a sleep mode of 30 watts or less, which represented roughly a 70% saving compared with normal use. Federal purchasing rules later required agencies to buy ENERGY STAR-compliant computers, monitors and printers. A 2000 study in Energy Policy by Carrie Webber, Richard Brown and Jonathan Koomey found that ENERGY STAR-labeled computers, fax machines, scanners and printers achieved savings of more than 50% compared with standard equipment.

The One Watt Initiative (1999)

In 1999, the International Energy Agency (IEA) adopted a target proposed by Alan Meier of LBNL: that standby power for new appliances should be one watt or less. LBNL's research argued that nearly all standby functions could run on one watt or less at little or no extra cost to manufacturers. The IEA set a goal of 1 watt by 2010 and 0.5 watts by 2013, and estimated that the initiative would cut carbon dioxide emissions in OECD countries by 50 million tons by 2010.

Government action (2000s)

Building codes start to cover plugs (2010s)

Because plug loads were long seen as the occupant's business rather than the building designer's, building codes largely ignored them. That began to change:

• ASHRAE Standard 90.1-2010 introduced automatic receptacle control, requiring that 50% of 125-volt, 15- and 20-amp receptacles in private offices, open offices and computer classrooms be controlled by an automatic device, such as a schedule or an occupancy sensor. Plug-in power strips do not count; the receptacle itself must be controlled.
• California's Title 24 (2013 edition) required controlled receptacles in rooms such as private offices, open offices, reception areas, conference rooms, kitchenettes in office areas and copy rooms. Later editions have continued to develop these requirements.

These rules don't limit how much equipment occupants plug in, but they ensure that a share of outlets can be switched off automatically when nobody is there.

The equipment itself has improved

Wikipedia notes that while total plug load energy is rising, the stock of equipment is becoming more efficient. Factors include low-power LCD monitors replacing CRTs, better sleep modes and the shift from desktop computers to laptops. LCD and LED displays use far less power than the CRT monitors they replaced, and laptops are designed to run on batteries, so they are inherently frugal. The catch is that there are many more devices than there used to be, and new categories, from smart speakers to data center servers, keep appearing.

Plug loads in different kinds of buildings

The mix of plug loads, and the best ways to tackle them, varies by building type. The descriptions below are general patterns rather than measured averages.

Server rooms and data center space

Small server rooms and network closets inside ordinary buildings deserve special mention, because they run continuously and need cooling around the clock. The EIA has reported that its 2018 Commercial Buildings Energy Consumption Survey found data center rooms most commonly in healthcare and large office buildings. In its 2025 analysis, the EIA projected that by 2050 up to 7% of all U.S. commercial floorspace may need extra energy to meet data center demand, spread across most building types, and noted that data center equipment also raises ventilation and cooling needs because of the heat it produces. Moving on-site servers to efficient shared facilities, consolidating underused servers and raising server-room temperature setpoints within manufacturers' limits are common ways organizations reduce this load.

Why plug loads dominate very efficient buildings

In a conventional building, heating, cooling and lighting dominate the energy bill, so plug loads seem minor. In a highly efficient building, with a tight envelope, efficient heat pumps and LED lighting with good controls, the picture reverses: the building systems use so little energy that plug loads become the largest remaining category. That is why designers of zero-energy and very low-energy buildings pay close attention to plug loads, often setting equipment budgets for tenants, specifying laptops and efficient displays, and installing controlled outlets. A building cannot reach net zero energy if its occupants' equipment runs around the clock.

Plug loads in energy modeling

Energy models used to design buildings and demonstrate code compliance must assume some level of plug load. Because the real load depends on future tenants, modelers usually use standard assumptions expressed in watts per square foot or square meter, along with schedules that describe when equipment is on. If those assumptions are too high, cooling systems are oversized; if they are too low, the building uses more energy in practice than predicted. The nameplate problem described above is one reason early models often overestimated plug loads. Measured data from existing buildings, where available, gives more realistic inputs. Plug loads are also one of the most common reasons why a building's actual energy use differs from its design prediction, a mismatch often called the "performance gap".

Measuring plug loads

Why measurement is hard

Plug loads are scattered across hundreds of outlets, change constantly as equipment is replaced, and are not usually metered separately from lighting and other circuits. Typical approaches include:

• Inventory: walking through a building and listing every plugged-in device, its type and, where possible, its power rating and power-management settings.
• Spot measurement: using a plug-in power meter to measure a sample of devices in different states.
• Logging: leaving plug meters in place for days or weeks to capture patterns over time.
• Circuit-level metering: measuring whole circuits or panels that serve mainly plug loads.
• Whole-building interval data: looking at night-time and weekend consumption from a smart meter or building management system as a proxy for always-on loads.

Accuracy at low power

Measuring very small loads accurately is difficult. The international standard for measuring standby power, IEC 62301, requires an uncertainty of no more than 2% for measurements of 0.5 watts or more, and 0.01 watts for measurements below that. Inexpensive plug-in meters, which cost around US$10, are often inaccurate at low power, and non-sinusoidal current waveforms can distort their readings. They are good for finding large idle loads but should not be trusted for fractions of a watt.

A simple conversion to remember

A device that draws power continuously uses:

watts × 8,760 hours per year ÷ 1,000 = kilowatt-hours (kWh) per year

So a continuous 1-watt load uses 8.76 kWh a year. Wikipedia's article on standby power rounds this to about 9 kWh, or roughly US$1 a year at average U.S. rates.

Worked examples (hypothetical)

The following examples use invented but realistic numbers to show how the arithmetic works. Electricity prices vary widely by country, utility and tariff; substitute your own rate.

Example 1: The cost of one always-on watt

Assume electricity costs $0.17 per kWh.

• 1 watt × 8,760 hours = 8,760 watt-hours = 8.76 kWh per year.
• 8.76 kWh × $0.17 = about $1.49 per year.

Example 2: A household's idle load

Suppose a household's smart meter shows that its lowest overnight draw, with everyone asleep and nothing actively in use, is 40 watts more than it needs to be, because of game consoles, set-top boxes, audio equipment and chargers.

• 40 W × 8,760 h = 350,400 Wh = 350.4 kWh per year.
• 350.4 kWh × $0.17 = about $59.57 per year.

If the household cut that excess idle load in half, it would save about 175 kWh and roughly $30 a year. That isn't life-changing, but it costs little to achieve and repeats every year.

Example 3: Office workstations left on after hours

An office has 100 workstations. Each workstation (computer, monitor and accessories) draws 30 watts when left on but unused. The office is occupied 50 hours a week, leaving 118 unoccupied hours (168 − 50).

• Per workstation: 30 W × 118 h = 3,540 Wh = 3.54 kWh per week.
• Per year: 3.54 kWh × 52 weeks = 184.08 kWh.
• For 100 workstations: 18,408 kWh per year.
• At a commercial rate of $0.13 per kWh: 18,408 × $0.13 = about $2,393 per year.

If schedule-based controls cut that after-hours waste by 26%, the average workstation reduction found in GSA's study (discussed below), the saving would be about 4,786 kWh, or roughly $622 a year. If equipment were reliably put to sleep or switched off, the saving would be larger. In buildings that need cooling, there is an additional saving from removing the heat these devices produce.

Example 4: Sizing using nameplate ratings

A designer adds up the nameplate ratings of equipment in an office zone and gets 10 kW. If actual average consumption is 10–15% of nameplate, as the ASHRAE research cited above found possible, the real average load could be closer to 1–1.5 kW. Sizing cooling equipment on 10 kW would mean installing far more capacity than needed, at higher cost and lower efficiency.

How to reduce plug loads

Strategies fall into five groups, roughly in order of how reliable they are.

1. Buy efficient equipment

The cheapest kilowatt-hour is the one never used. When equipment is replaced:

• Choose ENERGY STAR-certified products where the category exists, or products meeting equivalent efficiency labels elsewhere.
• Prefer laptops to desktops where they meet users' needs.
• Choose monitors of an appropriate size and brightness.
• Check standby and network-standby power ratings for TVs, set-top boxes and other electronics.
• Consider the total cost of ownership, including electricity, not just the purchase price.

2. Remove, consolidate and share

Many buildings contain equipment that doesn't need to exist:

• Replace personal desktop printers with shared networked printers.
• Remove personal space heaters, mini-fridges and coffee makers where shared alternatives are available and building policy allows.
• Consolidate underused kitchen appliances.
• Retire old equipment that is still plugged in but no longer used.

3. Use power management

• Enable sleep settings on computers and monitors, with sensible delays.
• Make sure printers and copiers have power-saving modes enabled.
• Turn off wake-on-LAN and similar network-wake features where IT departments don't need them, or manage them centrally.
• Use operating-system power management centrally across an organization rather than relying on each user.
• Remember that an animated screensaver keeps the display and computer working. Sleep mode, not a screensaver, saves energy.

4. Install controls

• Advanced power strips switch off connected devices on a schedule, when an occupancy sensor detects no one is present, or when a "master" device such as a computer is turned off.
• Timers and smart plugs switch off equipment such as water coolers, coffee machines and vending machines outside business hours, where safe and appropriate.
• Automatic receptacle control, as required by ASHRAE 90.1-2010 and California's Title 24 in certain spaces, switches whole outlets off on a schedule or when spaces are empty.

The best-known U.S. field test of these controls was carried out by the General Services Administration (GSA), which tested advanced power strips in eight government buildings in the Mid-Atlantic region and reported results in 2012. Schedule-based control was the most effective strategy. It reduced workstation plug loads by an average of 26%, with savings of nearly 50% in kitchens and printer rooms, where equipment had previously run around the clock. A later GSA summary reported annual savings of more than $200,000 across the deployment, with a payback of about 1.7 years overall and 2.6 years for workstations.

Controls have limits. Some devices, such as network equipment, medical devices, refrigerators and anything that must download updates overnight, should not be switched off automatically. Occupancy sensors can switch off equipment people are using if they sit still. Poorly configured controls frustrate users, who then bypass them.

5. Change behavior

Behavior campaigns, such as asking staff to switch off monitors at the end of the day, can work but tend to fade over time. The LBNL surveys showing that most monitors and printers were left on overnight suggest that relying on behavior alone produces mixed results. Behavior works best when combined with defaults and controls that make the efficient choice automatic.

A plug load action plan for organizations

Steps for households

• Check your home's lowest overnight draw on your smart meter or utility app, if available. That number is your always-on load.
• Walk around with a plug-in meter and measure devices in standby, especially older electronics, game consoles, set-top boxes, audio systems and computer peripherals.
• Use switched power strips or smart plugs for entertainment systems and home office equipment that doesn't need to stay on.
• Check settings on game consoles and smart TVs for "instant on" or quick-start modes that keep them drawing power.
• Unplug or switch off chargers and adapters you don't use. Modern ones draw very little, but older ones can draw more.
• Keep routers, refrigerators, freezers, medical equipment and security systems on.

Running a simple plug load walk-through

A walk-through is the cheapest way to start, and it needs nothing more than a notebook, a flashlight and, ideally, a plug-in power meter. The most revealing time to do it is outside working hours, which is how the LBNL researchers ran their surveys.

1. Pick a time when the building should be quiet, such as an evening or weekend.
2. Walk every space, including kitchens, copy rooms, meeting rooms, reception, storerooms and server closets.
3. Record every device that is on, noting its type, whether it is active, idle, asleep or in standby, and whether anybody needs it on.
4. Look for the usual suspects: monitors left on, printers and copiers awake, coffee machines and water coolers heating or chilling all night, meeting-room displays left on, chargers and space heaters under desks, vending machines with lit fronts.
5. Measure a sample with the plug meter, focusing on equipment that appears frequently or looks power-hungry.
6. Estimate the annual waste with the watts × hours calculation shown above.
7. Rank the opportunities by size and ease, starting with equipment that runs all night for no reason.
8. Repeat after changes to confirm the improvement held.

What to check when buying equipment

Product specifications often hide the numbers that matter most for plug load. When comparing models, look for:

Default settings deserve special emphasis. Many users never change them, so a product that ships with sleep enabled and a short delay will save energy in practice, while one that ships with power management switched off may not, however efficient it is in theory.

Limits and complications

• Savings estimates are uncertain. Equipment varies widely, even within one product category. Wikipedia's article on standby power notes that CRT monitors, for example, were measured at anywhere from 1.6 watts to 74.5 watts in standby.
• Some loads need to stay on. Network equipment, servers, fridges, medical devices, alarms and controls cannot simply be switched off.
• Rebound and growth. As each device becomes more efficient, more devices appear. Total plug load energy can rise even as individual products improve.
• Heating interaction. In heating-dominated buildings, some savings are offset by extra heating.
• Data centers and AI. The EIA's projections for computing assume no additional growth from accelerating AI expansion, so actual growth in computing loads could be faster than its reference case.
• Split incentives. In leased buildings, tenants often control plug loads while landlords pay for cooling, or vice versa, which weakens the incentive to act.

Common misconceptions

"Plug loads are too small to matter." Individually, many devices draw little power. Together, they account for a large and growing share of commercial building energy, and NRDC's study found always-on idle loads made up nearly 23% of electricity use in the California homes it studied.

"Turning a computer off and on uses more energy than leaving it on." For modern computers, the brief power draw at start-up is small compared with leaving the machine on for hours. Sleep and off modes save energy.

"Screensavers save energy." They were designed to prevent image burn-in on old CRT screens. An animated screensaver keeps both the computer and the display working. Sleep mode is what saves energy.

"Switched-off means zero." Many devices still draw standby power when "off". Only unplugging, or switching off at a switched outlet or power strip, guarantees zero.

"The nameplate tells you how much energy a device uses." Nameplate power is a maximum rating. Average use can be a small fraction of it.

"Unplugging phone chargers will slash my bill." Modern chargers in regulated markets draw very little when nothing is connected. The larger idle loads usually come from entertainment systems, older electronics, game consoles in quick-start modes and always-on appliances.

Frequently asked questions

What counts as a plug load?

Any device that is powered through a standard wall socket. Building energy analysts usually exclude heating, ventilation, air conditioning, lighting and water heating, even if some of these devices plug in.

Are plug loads the same as standby power?

No. Standby power is part of plug load: the energy used when a plugged-in device is "off" but still drawing power. Plug load also includes energy used when devices are actively working.

How much do plug loads cost in a typical office?

It depends on the building, its equipment and how it is used, so there is no reliable single figure. NREL estimates that plug and process loads account for 47% of U.S. commercial building energy, though that includes hard-wired process loads. A building's own meter data is the best guide.

What's the easiest way to cut plug loads in an office?

Find equipment that runs 24 hours a day without needing to, such as printers, kitchen appliances, displays and vending machines, and put it on schedules. GSA's study found the largest savings, nearly 50%, in kitchens and printer rooms.

Do power strips save energy?

A standard power strip only saves energy if someone switches it off. Advanced power strips with timers, occupancy sensors or master-controlled outlets switch off automatically and were found effective in GSA's study.

Are building codes regulating plug loads?

Partly. ASHRAE 90.1-2010 and California's Title 24 require automatic control of a share of receptacles in certain office spaces. They don't limit what equipment occupants bring, but they allow outlets to be switched off when spaces are empty.

What's the One Watt Initiative?

A target adopted by the International Energy Agency in 1999 that new appliances should use one watt or less in standby. It was followed by laws and purchasing rules including the U.S. Executive Order 13221 in 2001 and the EU's Regulation 1275/2008.

How do I measure my home's plug load?

Look at your lowest overnight power draw on a smart meter display or utility app; that is roughly your always-on load. Then use a plug-in power meter to measure individual devices. Remember that cheap meters can be inaccurate below about a watt.

Is computing really going to become the biggest commercial electricity use?

In the EIA's Annual Energy Outlook 2025 reference case, computing grows from about 8% of commercial electricity in 2024 to 20% in 2050, and could overtake lighting, space cooling and ventilation. That projection excludes any extra growth from accelerating AI adoption, so the actual figure could be higher. It is a projection, not a certainty.

Should I unplug everything when I go on vacation?

Unplugging entertainment systems, computers and small kitchen appliances before a long trip is an easy way to cut idle load to zero while you're away. Leave refrigerators, freezers, security systems and anything else that must keep running plugged in, and check that devices such as routers or thermostats aren't needed for remote monitoring.

Summary

Plug load is the electricity used by devices that plug into ordinary outlets. Once a minor concern, it has become one of the largest and fastest-growing parts of building energy use, as heating, cooling and lighting have become more efficient and the number of devices has multiplied. Reliable research shows that much of this energy is used when nobody is using the equipment: only 44% of computers and 25% of printers were turned off at night in LBNL's 2001 office surveys, and NRDC found always-on idle loads made up nearly a quarter of electricity use in the California homes it studied.

Efficiency programs and rules have made a real difference, from ENERGY STAR's first computer specifications in 1992 to the IEA's One Watt Initiative, the EU's standby limits and building codes requiring automatic receptacle control. The most effective actions for organizations and households are buying efficient equipment, removing what isn't needed, enabling power management, scheduling equipment that runs around the clock, and measuring the results.

Sources

• Plug load – Wikipedia
• Standby power – Wikipedia
• One Watt Initiative – Wikipedia
• Assessing and Reducing Plug and Process Loads in Office Buildings (NREL/BR-5500-76994) – OSTI
• Field Surveys of Office Equipment Operating Patterns (LBNL-46930) – OSTI
• Pulling the plug on standby power – Lawrence Berkeley National Laboratory
• Electricity use for commercial computing could surpass space cooling, ventilation – U.S. EIA Today in Energy
• Always-On Devices May Cost Americans $19 Billion Annually – NRDC
• Report: Always-on devices waste $19B in energy annually – Utility Dive
• GSA studies showcase emerging technologies aimed at reducing federal buildings' energy costs – FMLink
• Plug Load Control (Green Proving Ground findings) – U.S. GSA
• Energy Star – Wikipedia
• Automatic Receptacle Control to Meet ASHRAE 90.1-2010 and California Title 24 – NEMA