ESA for Compressors: Bearings, Valves, Rotor Bars and Surge — No Machine Sensors

Why compressors are not just “utilities”

Air, gas, process, refrigeration, pipeline and HVAC compressors sit at the intersection of reliability and energy cost. Compressed air can represent ~10% of site electricity — sometimes 30% or more.

For reliability managers the business case shifts:

• Downtime risk: one compressor trip cascades into process instability.
• Energy risk: inefficiency is often the earliest sign of mechanical or process degradation (contamination, valve leak, lubrication issues). kW drift is a leading indicator, not just a cost line.
• Staffing reality: fewer expert diagnosticians for more assets. Automated diagnostics with clear actions matter.

Artesis key idea: continuous monitoring and automatic diagnosis without machine-mounted sensors — because motor electrical signals already carry information about the whole drive train.

---

What electrical signature analysis (ESA) is

ESA is an ISO-recognised approach to online condition monitoring of three-phase induction motors from electrical measurements.

ISO states that variations in motor electrical characteristics can come from motor faults or driven equipment problems. ESA is therefore a system-level health indicator from motor to working machine.

Artesis implements ESA with a model-based (“digital twin”) approach:

Product	Role
e-MCM	Permanent online monitor
AMT Pro	Portable tester from MCC
Artesis IoT / OmniSight	Cloud dashboard, trends, energy
e-PCM	Generators and prime movers

---

Compressor fault map: what ESA sees

ESA works best when the failure mode changes average load/torque, torque pulsation, slip dynamics, or electrical modulation (sidebands).

Bearing degradation and lubrication

Bearing defects change friction and add cyclic load. Early signs: subtle torque pulsation and efficiency drift; later: stronger modulation and harmonics.

Misalignment, looseness, coupling issues

In vibration these link to 1× and harmonics. In ESA: stator current modulation (sidebands around electrical components). ISO 20958 lists coupling misalignment and driven equipment faults as MCSA-detectable.

Broken rotor bars and air-gap eccentricity

A core ESA case. ISO 20958 gives sideband relation: f_k = (1 ± 2s) · f_s. Compressors often run heavily loaded — rotor integrity is critical.

---

Failure physics ESA can “see”

Reciprocating compressors: cyclic load and valve leak

Strong cyclic torque is natural. Motor current components sensitive to valve leak, belt slack, intercooler leak.

Centrifugal compressors: surge/stall dynamics

Surge involves low-frequency flow/pressure oscillations (~1–10 Hz), modulating motor torque — visible as LF content in current/power envelopes and sidebands.

---

ESA math: theory to practice

Sampling: Nyquist f_N = f_s/2; record length T = N/f_s; FFT resolution Δf = 1/T.

Rule: surge and slow load need longer records; harmonic content needs adequate f_s.

Modulation: AM creates sidebands at f_c ± f_m. For compressors f_c is line frequency (50/60 Hz); f_m is shaft speed, compressor cycle, or surge frequency.

---

Artesis on compressor drives

Platform	Best for	Output
e-MCM	Critical drives 24/7	Traffic-light status, broad fault coverage
AMT Pro	Commissioning, routes	Auto report with fault + actions
OmniSight	Fleet view	Cloud trends and diagnostics
e-PCM	Turbine/generator drives	Model-based generator monitoring

e-MCM installs at the MCC with CTs (and VTs if needed) — ideal for hazardous or inaccessible assets.

---

Fault–signature map

Fault	ESA mechanism	Signatures
Broken rotor bars	Electrical asymmetry	Sidebands at (1 ± 2s)·f_s
Misalignment/coupling	Torque pulsation	Sidebands at 50/60 Hz ± shaft
Recip valve leak	Cyclic load change	Modulation index rise; f_c ± cycle
Centrifugal surge	LF load oscillation	LF rise 1–10 Hz; surge sidebands
Fouling/efficiency loss	Mean load + PF drift	kW, PF, THD trend

---

Rollout playbook

Step 1 — Criticality: process compressors, instrument air, refrigeration → e-MCM; secondary assets → AMT Pro routes.

Step 2 — Operating modes before baseline: fixed/variable speed; steady vs. load/unload; recycle valve; anti-surge zone.

Step 3 — Data rules: stable capture — e-MCM requires load variation <45% over 6 s. Do not mix unlike modes in one baseline bucket.

Step 4 — Alarms: Watch (plan inspection) → Examine (schedule intervention) → Red (reduce load, inspect now).

---

Sustainability and lifecycle

PdM reduces lifecycle cost by preventing major failures and catching energy waste from defects and process issues. High consumption and low efficiency are detectable process problems via PF and power monitoring.

---

Case templates

Case 1 — Valve leak and belt slack (recip): rising kW; cycle-sensitive current features; check valves, intercooler, belt tension at outage.

Case 2 — Surge approach (centrifugal): LF surge bands in current envelope; verify anti-surge, instrumentation, fouling on compressor map.

Case 3 — Rotor bar defect: subtle current change; track sidebands vs. load; plan rotor inspection if trend continues.

---

Limits and verification

VFDs can add components mistaken for faults — Artesis uses operating-mode learning. Rapid load changes during capture blur spectra. Some faults (localised bearings) may be clearer on vibration or oil — verify with process data, targeted vibration, alignment checks, post-maintenance trend.

---

Summary

Criterion	ESA (e-MCM/AMT Pro)
Machine sensors	Not required
Real-time monitoring	Continuous
Fault coverage	Electrical + mechanical + process
Early energy loss	Yes — kW, PF, THD
Hazardous zones	Yes — MCC install
Standards	ISO 20958
Auto diagnosis	Yes — traffic light + guidance

ESA turns motor supply cables into sensors for the whole drive — rotor to working machine — without extra machine instrumentation.

---

Learn more

• Artesis e-MCM — continuous compressor drive diagnostics
• Artesis AMT Pro — route testing from MCC
• Motor monitoring portfolio

Contact KEG TRK for a free compressor reliability programme assessment.