Understanding Baseline Data

Devices

Portable balancer & Vibration analyzer Balanset-1A

€1,975.00 + VAT (if applicable)

Parts

Vibration sensor

€90.00 + VAT (if applicable)

Parts

Optical Sensor (Laser Tachometer)

€124.00 + VAT (if applicable)

Devices

Balanset-4

€6,803.00 + VAT (if applicable)

Parts

Magnetic Stand Insize-60-kgf

€46.00 + VAT (if applicable)

Parts

Reflective tape

€10.00 + VAT (if applicable)

Devices

Dynamic balancer “Balanset-1A” OEM

€1,735.00 + VAT (if applicable)

Baseline data is the complete set of reference measurements, signatures, and operating parameters captured from a machine while it is in known-good condition — the yardstick against which every future reading is judged in a condition monitoring programme. It is closely related to the broader concept of the baseline, but where “baseline” names the idea, baseline data is the tangible archive: the vibration spectra, time waveforms, overall levels, phase readings, process variables, and documentation that together define a healthy machine’s signature. Good baseline data is an investment that pays back across the whole life of the asset, because almost every diagnostic decision is ultimately a comparison against it.

1. Definition: What Baseline Data Really Captures

Comprehensive baseline data goes far beyond a single overall amplitude number. A useful baseline is a layered snapshot — vibration in several forms, the operating conditions that produced it, and enough written context to reproduce the measurement later. Without that context a reading is just a number; with it, the same reading becomes evidence of how the machine behaves when nothing is wrong.

The reason this matters is simple: machine vibration is never zero, and “normal” is different for every machine. A pump that always runs at 2.1 mm/s RMS is healthy; an identical pump that historically ran at 0.8 mm/s and has climbed to 2.1 mm/s is sounding an alarm. Only a baseline lets you tell those two situations apart, which is why establishing one is the first task in any serious predictive maintenance effort.

2. Components of a Comprehensive Baseline

Vibration measurements

The core of the baseline is a structured set of vibration readings taken at every measurement point and in every direction (horizontal, vertical, axial):

• Overall amplitude values: RMS velocity (mm/s or in/s) is the most common metric, with peak velocity or displacement recorded for low-speed equipment and peak acceleration for bearing-defect detection. Capture both filtered and unfiltered values.
• Frequency spectra: FFT spectra at each point, ideally over more than one frequency range (for example 0–1 kHz for machine condition and 0–10 kHz for bearings), at a resolution fine enough to separate the key running-speed orders, and stored as data files rather than only as pictures.
• Time waveforms: several seconds of the raw signal versus time, which reveal the character of the vibration — pure sinusoid, impacting, or modulation — that a spectrum alone can hide.
• Specialised measurements: the envelope spectrum for bearing condition, orbit plots for critical machines, Bode plots from any startup or coastdown, and phase at the key orders (1×, 2×, and so on).

Operating parameters

Vibration only means something in the context of how the machine was running. Record the actual operating speed in RPM, the load or output (power, flow, pressure), the relevant process conditions, the bearing temperatures, and the power draw. A baseline taken at 60 % load is not comparable to a later reading taken at full load unless you know both.

Documentation

• Equipment data: make, model, serial number, and nameplate specifications.
• Measurement setup: sensor types and locations, mounting method, and instrument settings, so the geometry can be repeated exactly.
• Date and personnel: when it was taken and by whom.
• Conditions: operating state, any recent maintenance, and free-text observations.
• Photos: the measurement locations and the general condition of the machine.

3. Storing and Managing Baseline Data

A baseline that cannot be found when a problem appears is worthless, so storage and structure matter as much as the data itself.

• Organisation: a hierarchical layout (plant → area → equipment → measurement point), consistent naming, cross-referencing to the equipment database, and version control whenever a baseline is updated.
• Formats: keep the native instrument files for full re-analysis, plus portable copies (CSV, PDF), spectrum and waveform images, and the headline values pushed into the trending database.
• Accessibility: centralised storage on a network drive, cloud, or CMMS; quick retrieval for side-by-side comparison; access control to prevent accidental deletion; and regular backups.

4. Using Baseline Data in Analysis

The baseline is not an archive to admire — it is an active reference used in three everyday tasks.

• Trend analysis: plot current values against the baseline over time, calculate the rate of change, extrapolate toward the alarm limit, and watch for accelerating (non-linear) growth that signals a fault entering its final stage. The free Remaining Life from Vibration Trend calculator turns such a trend into an estimated time-to-limit.
• Fault diagnosis: overlay the current spectrum on the baseline spectrum. New peaks mean new faults; taller existing peaks mean a known fault is progressing; a changed pattern suggests the failure mechanism itself has shifted.
• Alarm setting: relative alarms expressed as multiples of baseline (for example, Alert at 2× and Alarm at 4× baseline), absolute alarms drawn from a standard but sanity-checked against baseline, or adaptive alarms that move with operating conditions using the baseline as their anchor. The ISO 20816-1 zone system (the successor to ISO 10816) pairs naturally with this approach.

5. Quality Assurance for the Baseline

A baseline taken from a machine that already has a hidden fault will silently mask that fault forever, so the data must be validated before it is trusted.

• Repeatability: repeated measurements should agree within roughly 10–15 %; wider scatter points to mounting or setup problems.
• Reasonableness: compare the levels to similar machines and to published norms such as the vibration severity bands.
• Completeness: confirm every required parameter is present.
• Operating conditions: verify the machine was in steady-state, normal operation.
• Peer review: an experienced analyst should review the data before archiving, confirming no obvious unbalance, misalignment, or bearing fault is baked into the “healthy” reference.

6. Capturing the Baseline in the Field

For most machines the baseline is gathered on site, at the machine’s own operating speed, with a portable instrument rather than on a test rig. A two-channel analyser such as the Balanset-1A records the overall levels, FFT spectra, time waveforms, and 1× amplitude and phase at each point in a single pass, capturing the true running state — including foundation, thermal, and load effects that a laboratory measurement would never see. Just as importantly, if that first survey reveals the machine is already out of balance, the same instrument performs field balancing there and then, so the baseline you archive is of a genuinely healthy machine.

7. Legal, Contractual, and System Integration Roles

Baseline data also has a life outside diagnostics. At commissioning it is frequently part of the acceptance test, documenting that a new machine met its contractual vibration limits and providing a warranty reference point. It becomes a legal record of the machine’s condition at a given date — useful for insurance, liability, and later failure analysis — and forms the foundation of the maintenance history.

Inside a CMMS or condition-monitoring platform, the baseline is linked to the equipment record so the software can compare new data automatically, generate alarms from baseline deviations, trigger work orders, overlay spectra for visual review, and report exceptions without manual effort. In short, baseline data is the foundation of every effective monitoring programme: time invested in capturing a complete, validated, high-quality reference while the machine is healthy is what makes all subsequent trending, diagnosis, and early warning possible — and is ultimately what delivers the return that justifies a predictive maintenance strategy.

← Back to Main Index