hero cooling tower fans
Asset monitoring / Cooling tower fans

Stay off the fan deck.
See the gearbox from the MCC.

Cooling tower fans sit on wet, windy decks with long drive shafts and right-angle gearboxes. Artesis follows motor, shaft, gearbox and fan condition from the electrical panel — nothing to mount in the plume.

62,572 kWhYearly energy effect of faults on two fans in the featured case
3 sitesSemiconductor, automotive and refinery
0 sensors in the towerCurrent and voltage measured in the MCC
01 / The challenge

Wet, windy
and hard to reach.

A cooling tower fan is driven through a long shaft and a right-angle gearbox, often 10 m or more above ground, in humid air and spray. Vibration sensors there corrode, and route rounds mean climbing onto the deck.

When a tower cell trips, condenser water warms and chillers or process units lose capacity. Because motor, shaft, gearbox and fan are one mechanical chain, their condition shows up in the motor current — readable from the MCC in the dry.

01

Connect at the motor control cabinet

e-MCM or a portable AMTPro test uses the existing current and voltage signals. Nobody needs to go onto the fan deck.

02

Learn each cell’s normal behaviour

Baselines reflect the fan’s real load, including VFD speed changes with the seasons.

03

Plan work before the hot season

Findings are classified as looseness, unbalance, gearbox or bearing, with the energy each fault costs.

02 / Featured case · Semiconductor

Two tower fans. Gearbox,
looseness and 62,572 kWh a year.

DOCUMENTED FIELD CASE
COOLING TOWER FAN 01 · 37 kW / CONDITION REPORT
Artesis AMTPro condition report for a cooling tower fan with loose foundation, unbalance/misalignment, transmission and bearing indicators in the caution band, and their energy effect

Four fault indicators in the caution band, each with its yearly energy effect.

COOLING TOWER FAN 01 / POWER SPECTRAL DENSITY
Artesis power spectral density of the cooling tower fan with looseness, misalignment and transmission/gear components marked

Looseness, misalignment and gear components above the learned envelope. Select to open full size.

2 fans

Faults found on both tested cells
Both 480 V, VFD-driven fans with gearboxes showed looseness and gearbox/transmission faults.

33,208kWh / year · fan 01 (37 kW)
29,364kWh / year · fan 08 (55 kW)
62,572kWh / year · both fans

Noisy fans, clear answers.

480 VNominal voltage
84 ANominal current
VFDGearbox drive

The situationCooling tower fans at a semiconductor plant were running with unusual noise and malfunction. AMTPro tests were run from the MCC on two cells.

The findingFan 01 showed loose foundation, unbalance/misalignment, transmission-element (gearbox) and bearing faults. Fan 08 showed loose foundation and transmission/gear faults.

The valueThe reports put the energy effect of the faults at 33,208 and 29,364 kWh per year — a clear case for correcting them before summer load.

  1. 1AMTPro tests from the MCC
  2. 2Gearbox and looseness faults on both fans
  3. 3Energy effect quantified per fault

Source: Artesis Case Study 2024, slides 6–8, and AMTPro reports for both fans (2023). Customer name withheld.

03 / More field cases

Gear mesh, loose bolts
and oversized motors.

3 INDUSTRIES
Artesis power spectral density of a second cooling tower fan with looseness and transmission/gear components marked
Semiconductor · Gearbox

The second cell: looseness and gear

On the 55 kW cell the spectrum showed a raised looseness floor and transmission/gear components. The report estimated the faults’ energy effect at 29,364 kWh per year.

Rating
480 V · VFD · gearbox
Energy effect
29,364 kWh / year
Artesis AMTPro power spectral density with a loose foundation pattern detected
Today’s AMTPro view: a loose foundation pattern detected in the PSD. Example from another asset.
Automotive · Looseness

A loose bolt, followed for four months

e-MCM first warned of looseness in October. A renewed alert in November started an inspection, and in February a loose attachment bolt was found and fixed.

Duty
Central cooling tower
Outcome
Loose bolt found and fixed
Refinery · Two 415 V cooling tower fans

Healthy — but oversized

AMTPro found no mechanical faults, but both motors ran at about 41% load. The reports advised a smaller, more efficient motor at the next replacement.

Cases are individual outcomes from Artesis field deployments between 2005 and 2023. All visuals are from Artesis software. Customer names are withheld.

04 / What we detect

Fault modes, signals
and evidence strength.

Evidence strength reflects how reliably a condition produces an observable electrical pattern in typical cooling tower fan installations — not a guarantee of detection.

ConditionWhat changes in the signalEvidenceField case
Drive train · mechanical
Gearbox and transmissionTransmission-element indicatorGoodFan 01 and 08 →
Looseness and foundationLooseness-related patternGoodFan 01, fan 08, automotive
Unbalance and misalignmentAlignment-related patternGoodFan 01, automotive
Bearing wearBearing-related pattern in the current spectrumGoodFan 01
Motor · electrical
Rotor barsRotor-related pattern in the current spectrumStrong—
Stator and insulationCurrent imbalance well above voltage imbalanceStrong—
Supply and connectionsVoltage imbalance, harmonics, per-phase RMSStrong—
Load · energy
Energy lost to faultskWh effect per detected faultStrongFan 01, fan 08
Oversized motorMotor load below ~60% of ratingStrongRefinery fans
Blade pitch or airflow changeLoad level versus learned baselineGood—
05 / Outside the envelope

Where electrical signatures are weaker.

We state the limits up front, so monitoring is set up where it adds most and paired with the right complementary checks.

Slow fan speeds

Fan-side components after a large gear reduction are small; gearbox and motor faults are clearer than blade faults.

Blade cracks and erosion

Early blade damage may not change load or balance enough to be seen.

Very low VFD speed

Below about 20 Hz in winter operation, confidence is reduced.

Gear oil condition

Oil quality is not seen directly; keep oil checks in the routine.

Observed patterns are interpreted by qualified personnel; they guide investigation rather than replace it. AMTPro spot tests need near-constant speed (±1% frequency) during capture.

06 / How you deploy it

Continuous where it matters.
Portable everywhere else.

Online · permanently installed

e-MCM

  • Critical tower cells and large fans
  • 24/7 learning, trending and alarms
  • Supports VFD applications
  • Results in OmniSight across sites
Portable · route-based

AMTPro

  • Seasonal checks across all cells
  • Spot tests from the MCC in minutes
  • Automatic condition report per test
  • Energy impact estimate in every report

Artesis Insight explains the finding.

Artesis Insight turns spectra and trends into plain-language explanations and next steps. Your team does not need to be an ESA specialist to act on a result.

Illustrative maintenance guidance

Check the gearbox and fan mounting.

Transmission and looseness indicators together suggest checking gearbox mounting, drive shaft couplings and fan hub bolts.

After action
Re-test at a similar fan speed and confirm the energy effect has fallen.

Example wording only; not a live alarm or a reproduced AI report.

07 / Applications

Across your cooling towers.

Induced-draft cooling tower fansForced-draft cooling tower fansGearbox-driven fansBelt-driven tower fansAir-cooled condenser fansChiller plant tower cells
Does anyone need to go onto the fan deck?

No. All measurements are taken in the motor control cabinet, in the dry.

Can you see the gearbox?

Gearbox and transmission faults show up as transmission-element components in the motor current, as both featured fans show. Very small fan-side effects behind a large reduction are harder to see.

Our fans run on VFDs. Is that a problem?

No. e-MCM supports VFD applications and AMTPro tests at the running speed. Confidence is reduced below roughly 20 Hz.

Can this help with energy?

Yes. Every AMTPro report estimates the energy effect of detected faults and flags oversized motors running below about 60% load.

What information is needed to start?

Motor nameplate data, gearbox or belt arrangement, fan speed range, drive type and the measurement points in the cabinet.

Bring your cooling towers into view.

Tell us about your tower cells and drives. We’ll show which fans to test first.

Talk to Artesis ↗

Contact Info

58 Thomas St, New York, 10013 USA

+1 443 315 2056

enquiry@artesis.com

Sitemap
Please enable JavaScript in your browser to complete this form.