---
title: "Do V-Belts Waste 5% of Motor Energy? Tracing the DOE Numbers"
slug: v-belt-energy-loss-doe-numbers
url: /topics/guides/v-belt-energy-loss-doe-numbers/index.md
publishedAt: "2026-11-05T21:06:21"
updatedAt: "2026-11-05T21:06:21"
excerpt: "The Department of Energy’s belt tip sheet says a V-belt drive can reach 95% efficiency or better when it is installed, and that the efficiency can fall by as much as 5% over time if the belt slips because nobody retensioned it. That is a conditional estimate of efficiency deterioration. It describes how far a […]"
authors:
  - Reliable Media
---

# Do V-Belts Waste 5% of Motor Energy? Tracing the DOE Numbers

By [Reliable Media](/authors/reliable-media/index.md)

The Department of Energy's belt tip sheet says a V-belt drive can reach 95% efficiency or better when it is installed, and that the efficiency can fall by as much as 5% over time if the belt slips because nobody retensioned it. That is a conditional estimate of efficiency deterioration. It describes how far a drive can slide if slippage develops.

A 5% belt figure can mean three different things: a drop in drive efficiency, the gap between two drive efficiencies, or a cut in electricity use after a belt change. The documents behind this claim use all three. I went through the three editions of the DOE tip sheet reviewed here, the fan study NREL ran for the General Services Administration, and Volumes 1 and 3 of DOE's 2021 and 2022 motor assessment to sort out which is which.

Here is the short version. DOE's own worked example put the electricity saving from replacing a V-belt with a synchronous belt at 5.1% in 2000 and 2005 and at 3.1% in 2012. NREL calculated normalized electrical savings of 1.6% to 20.1% for synchronous belts on one fan, depending on speed, and essentially no speed-normalized saving from retensioning belts on the other.

DOE's 2022 national model assumes a 2% improvement in overall motor-system efficiency. The 5% deterioration claim appears in all three tip-sheet editions reviewed here. The first two also calculate about 5% electricity savings in their worked example, and a separate 5% efficiency claim appears in Gates literature.

Why a maintenance man cares: DOE's 2021 assessment found belt tightening was the most common motor maintenance procedure among the categories it asked about, reported on 28% of industrial motors over two years. Before you can say what a belt change saves, you need to know the condition of the belt you are replacing. Most of the numbers below depend on that starting point.

## Where the 5% Appears

DOE has issued at least three editions of its belt tip sheet, and this article reviews those three. The Office of Industrial Technologies issued it in January 2000 as [Motor Tip Sheet #3](https://docs.nlr.gov/docs/fy00osti/27833.pdf). The Industrial Technologies Program reissued it in September 2005 as Motor Systems Tip Sheet #5, [reproduced](https://www1.eere.energy.gov/manufacturing/tech_assistance/pdfs/motor.pdf#page=81) in DOE's 2008 motor sourcebook. The Advanced Manufacturing Office revised it in November 2012 as [Motor Systems Tip Sheet #5](https://www1.eere.energy.gov/manufacturing/tech_assistance/pdfs/replace_vbelts_motor_systemts5.pdf), the edition at the DOE link cited here.

The 2000 and 2005 editions use the same sentences. A V-belt drive can have a peak efficiency of 95% to 98% at the time of installation. Efficiency deteriorates by as much as 5%, to a nominal 93%, over time if slippage occurs because the belt is not periodically re-tensioned.

Cogged belts, in those editions, run about 2% more efficient than standard V-belts. Synchronous belts offer about 98% and hold it across a wide load range. Neither early edition cites a source for any of those figures.

The 2012 revision keeps the 5% deterioration, drops the 93% nominal figure, and restates the installed efficiency as 95% or more. It adds two footnotes. The claim that about one-third of motors use belt drives goes to an article by a Greenheck Fan Corporation engineer in HPAC Engineering, July 2011. The 98% synchronous figure goes to a December 1997 Gates Rubber Company publication.

Those are the only references DOE has attached to these numbers in the three editions reviewed here. One supports the prevalence statement and one supports the synchronous-belt efficiency. Neither documents where the installed-efficiency range or the 5% deterioration figure was measured.

> DOE's deterioration estimate is conditional on slippage when the belt is not retensioned.

The table tracks the three editions. The 2005 sheet appears on pages 77 and 78 of the 2008 sourcebook (pages 81 and 82 of the PDF) with its September 2005 date and document number. NREL's publication record identifies the September 2005 version as superseded by the November 2012 edition.

## What DOE's Own Example Says

All three editions work the same machine: a continuously running 100 hp supply-air fan motor at 75% average load, consuming 527,000 kWh a year. The 2000 and 2005 editions replace a 93% efficient V-belt with a 98% synchronous belt. The formula is annual energy use multiplied by (1 minus η1/η2), and it returns 26,888 kWh a year, 5.1% of the fan's energy, worth $1,345 at the 5 cents per kWh DOE assumed then.

The 2012 edition keeps the fan and raises the assumed baseline belt efficiency to 95%, along with the electricity price. The same formula returns 16,130 kWh, 3.1% of the fan's energy, worth $1,290 at 8 cents.

The earlier example starts from the sheet's own figure for a belt that has slipped, so part of its 5.1% is the slippage allowance. The documents do not explain why DOE changed the baseline, and a revised calculation says nothing about how much of that difference a maintenance adjustment would recover in practice.

In DOE's examples, an assumed change from 93% to 98% is a five-percentage-point efficiency gain and a 5.1% electricity saving at the same useful output. From 95% to 98%, it is three percentage points and a 3.1% electricity saving. Other 5% figures may rest on different efficiencies or measurements; this arithmetic explains DOE's examples.

## What NREL Measured for GSA

The General Services Administration's Green Proving Ground program commissioned NREL to measure synchronous and cogged belts against the original V-belts on two fans at the Byron G. Rogers Federal Building in Denver, one constant-volume and one on a variable frequency drive. NREL's [full report](https://docs.nlr.gov/docs/fy14osti/61448.pdf) is dated February 2014, and GSA published a four-page [findings summary](https://www.gsa.gov/system/files/GPG012-Fan%20Belts-Findings-508R.pdf) the following month.

On the VFD fan, NREL ran the drive at ten frequencies from 15 to 60 Hz. Synchronous belts cut electrical input by 1.6% at 55 Hz and 2.3% at full tested speed of 60 Hz, rising to 20.1% at 15 Hz. Cogged V-belts cut it by 0.8% at 55 Hz, 1.2% at 60 Hz, and 9.3% at 15 Hz. Those are electrical savings normalized for fan speed and air density, calculated from a few minutes of measurements per point.

> The result at one fan speed has to be weighted by the hours the fan spends there.

A large percentage at low speed is a share of a small power draw, so annual savings depend on the duty cycle. NREL modeled the fan at 20 Hz for 20 hours a day and 60 Hz for four, and projected 1,953 kWh a year for synchronous belts and 976 kWh for cogged belts, about $164 and $82 at the $0.084 per kWh it assumed.

Modeled simple payback came to 3.87 years for synchronous belts and 0.49 years for cogged belts, with every belt assumed to last three years. Synchronous belts showed the lowest nine-year life-cycle cost under the energy, replacement, and maintenance assumptions, including one synchronous belt replacing four V-belts.

The constant-volume fan told a different story. Its original V-belts were loose. Retensioning changed measured input from 7,107 to 7,153 watts. After accounting for the small increase in fan speed, NREL calculated essentially no power saving.

Then it installed a synchronous belt. The report puts its potential saving at about 1.9% once the data were normalized for fan speed, but the fan sped up from 1,161.5 to 1,183 rpm and measured power rose to 7,369 watts, about 3.7% higher than with the loose belts.

NREL's conclusion was that the synchronous belt, as installed, increased the cost of running that fan. Cogged V-belts were never tested on the constant-volume fan; the report says installation logistics prevented it. GSA's four-page summary nonetheless describes cogged belts demonstrating savings on the CV fan, which the full report does not support.

The report still recommends cogged belts for constant-volume fans, citing the install and speed problems synchronous belts pose there, and cogged belts saved energy at every fixed speed tested on the VFD fan. It also warns that an incorrectly sized synchronous retrofit can raise fan speed and power demand. The earlier work it cites provides additional background.

That earlier work, as NREL reports it, was a University of Virginia facilities study of a constant-speed 40 hp HVAC fan. Replacing worn, tensioned V-belts with new V-belts cut energy demand by 2.9%, and cogged V-belts saved a further 3.3% over the new V-belts, 6.3% over the worn ones. I have not seen the Virginia document itself, so take those figures as NREL's account.

Two more things to know. GSA's summary credits Gates Corporation with providing the technology for the test bed. NREL lists the power logger's energy accuracy at plus or minus 1.5% but calls that an absolute error, and says Fluke's technical support confirmed that the error between repeated measurements on the same meter is negligible.

## How Many Motors Run on Belts

The tip sheets say about one-third of industrial and commercial motors use belt drives. The 2000 and 2005 editions gave no source, and the 2012 edition cites the Greenheck article. The better number comes from DOE's own field assessment.

Between September 2016 and March 2019, Lawrence Berkeley National Laboratory assessed 369 industrial and commercial facilities for DOE's [Motor System Market Assessment](https://escholarship.org/uc/item/42f631k3), published in 2021 and covering three-phase and DC motors of 1 hp and up. The sample-based estimates: 52.5 million motor systems consuming about 1,079 TWh a year, 29% of the U.S. grid load. Direct drive was the most common transmission in both sectors, at 34% of industrial motor systems and 44% of commercial ones.

In commercial buildings, 27% of motor systems run on belts, accounting for 27% of the sector's motor electricity. Where the assessors could tell the belt type, V-belts were on 12% of units and toothed belts on 4%. In industry, belts of some type were the second most common transmission after direct drive, with identified V-belts on 4% of motor systems and toothed belts on 1%. Many belts were recorded without a type.

The assessment also asked about maintenance. Among the categories surveyed, belt tightening was the most common procedure in both sectors: facility staff reported it for 28% of industrial motors and 18% of commercial motors over the previous two years. Rewinds, bearing replacements, and shaft alignments each ran at 7% or less. The survey recorded that belts were touched; it did not measure what tension they were left at.

## What DOE's 2022 Model Assumes

The assessment's [third volume](https://escholarship.org/uc/item/16x3d348), published in July 2022, estimated national savings from replacing V-belts with synchronous or notched belts. It cites a 3% improvement in maximum efficiency as achievable, then uses a conservative 2% improvement in overall motor-system efficiency for the modeled upgrade.

Applied to the motor systems the assessors identified as V-belt driven, that comes to 456 GWh a year in industry and 1,116 GWh in commercial buildings, 1,572 GWh in all, worth about $189 million a year.

LBNL rounds the total to 0.1% of motor-system electricity; against 1,079 TWh it is about 0.15%. Belts of unknown type were excluded, which the authors say makes the figure deliberately conservative, and the estimate carries the model's assumptions.

For scale, the same report put improved load control through VFDs and right-sizing at about 115,000 GWh and distribution-system improvements at more than 40,000 GWh. Those are modeled national opportunities, and they say nothing about what pays back first in a given plant.

The point for this article is narrower. LBNL's model did its arithmetic at an assumed 2% improvement in overall motor-system efficiency, a different quantity from DOE's conditional 5% deterioration figure. The model neither confirms nor refutes that figure; it shows what DOE-funded analysts chose to assume for a national estimate.

Terminology matters: the 2000 and 2005 sheets called notched V-belts "cogged." The 2012 sheet renamed them "notched" and listed "cogged" as another name for synchronous belts. LBNL's glossary follows the 2012 usage, and its savings tables label the measure "cogged V-belts" while meaning the synchronous-or-notched upgrade. When a page promises cogged-belt savings, find out which belt it means before you buy.

## Where the Manufacturers' 5% Fits

Gates Corporation's document "Energy Savings from Synchronous Belt Drives," the one GSA's summary cites for the manufacturer claim, states that a synchronous belt drive is on average 5% more efficient than a V-belt drive. It also says properly maintained V-belt drives can run at 95% to 98% at installation and deteriorate as much as 5% in operation, and that poorly maintained drives may be up to 10% less efficient.

That installation-and-deterioration sentence matches the DOE tip sheet wording almost word for word. The Gates document is undated and cites no test for the 5% average, though it includes a customer case study. It is a different publication from the 1997 Gates document DOE footnotes for the 98% synchronous figure.

The documents reviewed for this article do not establish where those shared sentences were first measured or which direction the wording was reused. Similar figures appear in DOE guidance and in manufacturer literature, and that is as far as the paper trail goes. A manufacturer's publication is a fair source for what the manufacturer claims. It is not independent confirmation of the claim, and a manufacturer supplying belts for a test does not by itself invalidate a laboratory's measurements.

So the manufacturer claim stays attributed to the manufacturer, and DOE's figures stay attributed to DOE. Synchronous drives are a good product in the right application. Price the saving against your own drive, under your own operating conditions, before you put a vendor's average into a business case.

## What This Means on Your Floor

The evidence supports a sequence, and the sequence starts with the drive you already have. Check tension, alignment, belt condition, and sheave condition before you evaluate a belt-type conversion. Correcting a defect may reduce losses, but the saving depends on the defect and the operating conditions. Here is how I would run it.

* **Tension on the manufacturer's procedure.** For conventional V-belts, my practice is to check tension at installation and again after 24 hours, as Ricky Smith laid out in [Mechanical Maintenance 101](https://reliamag.com/topics/articles/mechanical-maintenance-belt-chain-coupling), then on DOE's interval of every 3 to 6 months after break-in. The belt maker's run-in and tensioning procedure takes precedence where it differs. DOE's rule is the lowest tension that will not slip at peak load, since overtensioning can shorten bearing life.
* **Check the sheaves.** A belt riding in the bottom of a worn groove loses its wedging action and can slip. Gauge groove wear at every belt change and replace worn sheaves. NREL points to small-diameter sheaves as the largest opportunity for cogged belts, because bending losses rise as sheaves shrink, which is a selection issue separate from groove wear.
* **Align the sheaves.** The general guidance I use is 1/10 inch per foot of center distance for V-belts and 1/16 inch for positive-drive belts, with the manufacturer's limits taking precedence. Misalignment can cause wear and increase losses. Belt dust on the guard is one sign to investigate during a [visual inspection](https://reliamag.com/topics/articles/visual-inspections-predictive-maintenance-programs).
* **Replace belts as matched sets, per the manufacturer's specification.** A single new belt in a multi-belt drive can carry a disproportionate share of the load and fail early.
* **Measure real power, before and after.** Real power on an AC motor depends on voltage, current, and power factor, and power factor moves with load, so record kW with a power analyzer or log kWh over comparable periods; amps alone are a diagnostic. Record driven speed, output, and load conditions, and compare like with like before crediting the belt or building the [business case](https://reliamag.com/topics/articles/how-to-calculate-roi-in-maintenance-and-reliability-initiatives-effectively). On a VFD, set the measurement boundary and use instruments rated for it.
* **Match the belt to the application.** Synchronous belts can suit VFD fans and long-running drives, with sprockets sized to preserve driven speed. Notched V-belts can suit constant-speed fans and shock loads when correctly sized for the existing sheaves. Under the fan affinity laws, a 2% speed increase raises power about 6%, assuming unchanged air density, system resistance, and fan efficiency.

Chain drives have their own version of this story, which Ricky Smith covered in [Chain Drive Best Practices](https://reliamag.com/topics/articles/chain-drive-best-practices-for-plant-maintenance-teams).

> After accounting for fan speed, retensioning produced essentially no saving on this Denver fan.

## Run DOE's Formula With Your Own Numbers

DOE's method starts with annual energy use. Energy savings equal annual kWh multiplied by (1 minus efficiency before divided by efficiency after). Get annual kWh from a meter if you have one.

Otherwise, estimate annual kWh as rated motor output horsepower × 0.746 × average load fraction × run hours ÷ motor efficiency at that load. Enter the load fraction and efficiency as decimals. If you already know actual motor-shaft horsepower, omit the load fraction.

Here is an illustration with hypothetical drive efficiencies. A fan is driven by a motor rated at 50 hp, initially operating at 75% of rated shaft output for 6,000 hours a year. At an assumed motor efficiency of 94%, baseline electricity use is about 178,600 kWh.

Assume the original belt is 93% efficient. If retensioning raised it to 95%, the calculated saving would be about 3,760 kWh a year, leaving about 174,800 kWh. Replacing that belt with a 98% synchronous drive would then save a further 5,350 kWh.

The combined step, 93% to 98%, returns about 9,110 kWh. At 8 cents per kWh, the three results are roughly $300, $430, and $730 a year, and the retensioning step is about 41% of the combined total. This hypothetical sequence uses efficiency values appearing in DOE's examples, with useful output and motor efficiency held constant.

The retensioning gain is an illustration. DOE has neither measured nor promised that recovery.

Your drive's efficiencies will differ, and the formula assumes the load stays the same, which it will not on a centrifugal fan that speeds up. Substitute measured kWh and your own electricity rate, and keep the formula.

## What the DOE Numbers Can't Tell You

* None of the three DOE editions reviewed here cites a measurement for the 95% to 98% installed efficiency, the 5% deterioration, or the 2% notched-belt gain. The 2012 footnotes support the prevalence statement and the synchronous-belt figure only.
* The NREL study is two fans in one building, with minutes of data per operating point, a modeled duty cycle for the annual figures, and belts credited to a manufacturer. Its one retensioning comparison showed essentially no speed-normalized saving; one comparison cannot establish whether retensioning will save energy elsewhere.
* LBNL's 2% improvement in overall motor-system efficiency is a modeling assumption applied to the 4% of industrial and 12% of commercial motor systems identified as V-belt driven. Unidentified belts were excluded, and the result carries the model's assumptions.
* None of these sources measured how far belts in service have drifted from installed tension. The 28% belt-tightening rate tells you crews are touching belts. What tension they left behind is unrecorded.
* The 2012 sheet lists pulley size, pulley wear, alignment, transmitted torque, and belt sizing as factors in V-belt efficiency. An efficiency figure needs operating conditions attached before it can describe a particular drive.

## Frequently Asked Questions

### Do V-belts waste 5% of motor energy?

A 5% belt figure can describe a drop in drive efficiency, a gap between two drive efficiencies, or a cut in electricity use after a belt change, and those are different quantities. DOE's tip sheet says a V-belt drive can reach 95% or more at installation and lose as much as 5% to slippage if it is not retensioned. DOE's 2012 worked example saves 3.1% with a synchronous belt; the 2022 model assumes a 2% gain in overall motor-system efficiency.

### How efficient is a V-belt drive?

DOE's 2000 and 2005 tip sheets say V-belt drives can reach peak efficiencies of 95% to 98% at installation and deteriorate to a nominal 93% with slippage. The 2012 edition says 95% or more at installation and keeps the 5% deterioration allowance. The 2012 sheet lists pulley size, pulley wear, alignment, transmitted torque, and belt sizing as the factors that move it.

### How much energy do synchronous belts save compared with V-belts?

DOE's 2012 worked example shows 3.1% of a fan's annual electricity. NREL's measurements for GSA on a VFD fan ranged from 1.6% at 55 Hz to 20.1% at 15 Hz, with 2.3% at 60 Hz, and the modeled annual saving on that fan was 1,953 kWh. On a constant-volume fan the synchronous belt raised fan speed and increased consumption.

LBNL's 2022 national model assumes a 2% improvement in overall motor-system efficiency. Gates claims an average 5% improvement in drive efficiency.

### How often should V-belt tension be checked?

DOE recommends checking tension every 3 to 6 months after the break-in period, at the lowest tension that will not slip at peak load. For conventional V-belts I check immediately after installation and again after 24 hours of operation, then on DOE's interval, with the belt manufacturer's procedure taking precedence.

### What share of motors use belt drives?

DOE's tip sheets say about one-third, a figure the 2012 edition attributes to a fan manufacturer's article. LBNL's 2021 assessment of 369 facilities found 27% of commercial motor systems on belts and direct drive the most common transmission in both sectors. Identified V-belts were on 12% of commercial and 4% of industrial motor systems, with many belts recorded without a type.

### Should I replace V-belts with synchronous belts to save energy?

Check tension, alignment, belt condition, and sheave condition first, then measure real power on the drive. On VFD-driven fans, NREL's measurements support considering a conversion, with duty cycle and cost checked and sprockets selected to preserve the required speed. On constant-volume fans, GSA recommended notched V-belts in the existing sheaves, because a synchronous retrofit that raises fan speed can consume more than it saves.

## Sources

* U.S. Department of Energy, Advanced Manufacturing Office. [Replace V-Belts with Notched or Synchronous Belt Drives, Motor Systems Tip Sheet #5](https://www1.eere.energy.gov/manufacturing/tech_assistance/pdfs/replace_vbelts_motor_systemts5.pdf) (November 2012, DOE/GO-102012-3740).
* U.S. Department of Energy, Office of Industrial Technologies. [Replace V-Belts with Cogged or Synchronous Belt Drives, Motor Tip Sheet #3](https://docs.nlr.gov/docs/fy00osti/27833.pdf) (January 2000, DOE/GO-102000-0972, hosted by NREL; the September 2005 reissue, Motor Systems Tip Sheet #5, DOE/GO-102005-2060 and NREL/FS-840-37502, is linked in the text to its reproduction in Appendix B of DOE's September 2008 sourcebook Improving Motor and Drive System Performance, printed pages 77 to 78, PDF pages 81 to 82; NREL's [publication record](https://research-hub.nlr.gov/en/publications/replace-v-belts-with-cogged-or-synchronous-belt-drives-revision/) for the 2005 sheet notes its replacement by the November 2012 edition).
* National Renewable Energy Laboratory. [Synchronous and Cogged Fan Belt Performance Assessment](https://docs.nlr.gov/docs/fy14osti/61448.pdf) (Cutler, Dean, and Acosta, February 2014, prepared for the General Services Administration; two fans at the Byron G. Rogers Federal Building and U.S. Courthouse, Denver).
* U.S. General Services Administration, Green Proving Ground. [Fan Belts: Synchronous and Cogged, GPG-012 Findings](https://www.gsa.gov/system/files/GPG012-Fan%20Belts-Findings-508R.pdf) (March 2014 four-page summary of the NREL report; credits Gates Corporation with providing the test-bed technology).
* Lawrence Berkeley National Laboratory. [U.S. Industrial and Commercial Motor System Market Assessment Report Volume 1: Characteristics of the Installed Base](https://escholarship.org/uc/item/42f631k3) (January 2021, revised September 2021; 369 facilities assessed September 2016 to March 2019 for DOE's Advanced Manufacturing Office).
* Lawrence Berkeley National Laboratory. [U.S. Industrial and Commercial Motor System Market Assessment Report Volume 3: Energy Savings Opportunity](https://escholarship.org/uc/item/16x3d348) (July 2022; belt-upgrade assumptions and Table 19).
* Gates Corporation. [Energy Savings from Synchronous Belt Drives](https://www.abettercity.org/docs/sustainability/Energy%20Savings%20From%20Synchronous%20Belt%20Drivers.pdf) (undated manufacturer literature, mirrored copy hosted by A Better City; the document GSA's summary cites for the manufacturer claim of up to 5% savings; belt manufacturer with a commercial stake in the comparison).