Connections +

Focus on Maintenance & Testing: Maximizing Energy Savings From Your UPS

A high efficiency UPS system can provide substantial savings on energy costs -- and offer a host of other benefits.

February 1, 2002  

Print this page

While energy efficiency may have had little or no weight in past purchasing decisions, the cost savings resulting from today’s high efficiency uninterruptible power supply (UPS) systems are so significant they can no longer be ignored.

Energy savings gained by using a high efficiency UPS, versus a conventional UPS, often equal the value of the UPS in as little as three to five years. Even a slight advantage in energy efficiency can save you thousands of dollars in a few years with a small UPS system — or hundreds of thousands of dollars with a large UPS system.


The energy efficiency of a UPS can be expressed as the difference between the amount of energy that goes into a UPS versus the amount of useful energy that comes out of the UPS to power your loads. In all UPS systems, some amount of energy is lost as heat when it passes through the internal components (transformers, rectifiers, inverters, etc.).

Just how much energy is lost between the input and output can be significant when you consider how much the wasted energy costs. Energy efficiency advantages of as little as one per cent between one UPS and another can translate into tens of thousands of dollars in savings per year, depending on the size of the UPS.

Diagram 1 shows a typical UPS. While the UPS loads use 90 kW of power, the UPS demands 100 kW of power. This means that 10 kW of power is being lost (as heat) as it goes through the UPS and renders it only 90 per cent efficient. Since the lost power is being dissipated as heat, the air conditioner will have to use more energy to cool the facility.

While 10 kW of lost energy may not seem like a lot of power, UPS loads operate continuously, 365 days a year. 10 kW of lost energy now equates to 87,600 kWh of power each year. At a typical utility rate of $0.10 / kWh, this equates to $8,760 dollars (all figures in U.S dollars) of energy wasted by the UPS, in addition to $2,600 in extra energy costs just to cool the heat rejected by the UPS. This scenario would result in a total of $11,360 in wasted energy.

No matter what UPS system you select, there will be some energy lost between the utility and the output, but high-efficiency UPS systems can dramatically limit the energy loss, resulting in substantial cost savings.

Consider the previous scenario. What if the same UPS was 95 per cent efficient (as opposed to 90 per cent efficient) and lost only five kW of energy as heat? The difference in energy savings would be more than $5,500 per year. Considering that a typical 100 kW UPS costs approximately $25,000, the energy savings would pay for the UPS in five years.


When comparing the energy efficiency of UPS systems, you may find that published efficiency specifications from UPS vendors seem very similar. This experience can leave you wondering whether there are any differences among UPS manufacturers. The only way to ensure that a vendor is giving you straight facts about UPS efficiency is to demand a witness test on your specific UPS — prior to shipment from the factory.

Getting an efficiency test is like test-driving a car to measure fuel efficiency. Just as a car will get radically different mileage when driving on a highway versus a windy mountain road, UPS efficiency can also change with the type of load it powers.

Today almost all UPSs power 100 per cent non-linear loads (computers, servers, motors and electronic equipment). When most manufacturers test their UPS systems, they use linear loads, which are not accurate representative of customers’ actual loads. As UPS efficiency will often be much higher when powering linear loads, customers may not be shown the ‘true’ energy consumption of a system when it is installed at their site. The only way to be sure of a system’s true efficiency is by insisting that a manufacturer demonstrate efficiency with non-linear loads that are representative of real-world environments.

A second factor to keep in mind when testing UPS efficiency is the power level at which the efficiency is measured. Just as a car obtains its best mileage when operating at approximately 60 mph, most UPS systems will typically be most efficient when operating at a 50 to 100 per cent load level. In the real world, most UPS systems operate at 25 to 60 per cent of their nominal load. To determine accurate efficiency, a system should demonstrate efficiency at loads below 25 to 50 per cent, where most UPSs will likely be operating — especially if operating in redundant configurations.

By testing your UPS under actual expected load levels, and simulating an actual load profile in your facility, you will know the real efficiency of your UPS.


Switching losses in the inverter and transformers is the primary energy waster in a UPS. To minimize switching losses, a digitally based power management system will optimize the switching characteristics of the UPS inverter for specific load types and load levels by precisely controlling every pulse of the switching cycle. This results in the most efficient switching patterns with the least losses, and outperforms older style systems with fixed switching patterns.

The power management system operates by creating the output waveform from hundreds of small, tightly controlled pulses. The inverter output is constantly compared to a computer generated theoretical sine wave, and corrective micro frequency adjustments are made by the pulses to keep the output waveform to within a few per cent of a perfect sine wave. By actively contouring the output to the load conditions, the need for power hungry output filters is eliminated, and you can maintain a higher efficiency and deliver a perfect sinewave output (please see Diagrams 2 and 3 below).

The inverter output waveform is created from hundreds of precisely controlled pulses with the switching constantly optimized for the load level and profile. This Digital Power Quality Management technology also keeps the output waveform within +/-1 per cent of a perfect sinewave.

Other energy saving features include high efficiency transformers that meet or exceed national standards for high-energy efficiency (as required for the Energy Star certification). High efficiency transformers can offer a two- to three-per-cent efficiency advantage over a generic lower efficiency transformer.


To calculate how much money you can save by using a high efficiency UPS, estimate your actual load and utility rate category. Your efficiency advantage will vary depending on your UPS load, but will often range from two to four per cent (please see “Efficiency Advantage” chart on p. 24).

While reliability is the number one factor in selecting a system, price is often the next deciding factor. Small differences in operating costs between UPS systems of varying efficiencies often result in savings greater than the purchase price.

To ensure savings, a vigilant comparison is needed. Evaluate UPSs at their anticipated operating load levels and on non-linear loads in the manufacturers test area. Do not rely on paper specifications. Some manufacturers claim very high efficiencies based on proprietary topologies, which, upon closer investigation, turn out to be line interactive UPS systems (feeding utility power straight to the critical load) and are unsuitable for high-end applications. Even with a two per cent efficiency difference between systems, the savings can be extremely lucrative — and well worth the time it takes to investigate.

Alan Katz is a Senior Product Manager with MGE UPS Systems in Costa Mesa, CA. Alan’s background is in the power conversion and UPS business with a focus on data center critical power systems. In his current position, Alan specifies designs for new UPS and critical power system products and evaluates emerging technologies for integration in conventional applications.

Table 1: Efficiency Advantage

Energy Rate Load 5.0% 4.0% 3.0% 2.0%
(S/kWh) (kW) 1 yr. 5 yr. 1 yr. 5 yr. 1 yr. 5 yr. 1 yr. 5 yr.
$0.15 100 $10,209 $80,909 $8,077 $48,186 $5,991 $35,742 $3,951 $23,569
500 $51,046 $304,544 $40,383 $240,928 $29,954 $178,710 $19,753 $117,845
1000 $102,092 $609,087 $80,766 $481,856 $59,909 $357,420 $39,420 $235,690
2500 $255,230 $1,522,718 $201,915 $1,204,639 $149,772 $893,551 $98,763 $589,226
$0.10 100 $6,806 $40,606 $5,384 $32,124 $3,994 $23,828 $2,634 $15,713
500 $34,031 $203,029 $26,922 $160,619 $19,970 $119,140 $13,168 $78,563
1000 $68,061 $406,058 $53,844 $321,237 $39,939 $238,280 $26,337 $157,127
2500 $170,153 $1,015,146 $134,610 $803,093 $99,848 $595,701 $65,842 $392,817
$0.05 100 $3,403 $20,303 $2,692 $16,082 $1,997 $11,914 $1,317 $7,858
500 $17,015 $101,515 $13,461 $80,309 $9,985 $59,570 $6,584 $39,282
1000 $34,031 $203,029 $26,922 $160,619 $19,970 $119,140 $13,168 $78,563
2500 $85,077 $507,573 $67,305 $401,546 $49,924 $297,850 $32,921 $196,409

Print this page