Understanding Warning Levels in Vibration Monitoring

Құрылғылар

Портативті теңгергіш және діріл анализатор Balanset-1A

€1,975.00 + VAT (if applicable)

Бөлшектер

Діріл сенсоры

€90.00 + VAT (if applicable)

Бөлшектер

Оптикалық сенсор (лазерлік тахометр)

€124.00 + VAT (if applicable)

Құрылғылар

Баланс-4

€6,803.00 + VAT (if applicable)

Бөлшектер

Магниттік тұғырдың өлшемі-60 кгс

€46.00 + VAT (if applicable)

Бөлшектер

Шағылыстырғыш таспа

€10.00 + VAT (if applicable)

Құрылғылар

«Balanset-1A» OEM динамикалық балансы

€1,735.00 + VAT (if applicable)

A warning level is an intermediate alarm threshold in a multi-level vibration monitoring system, sitting between the first low-level alert and the critical danger level. When vibration rises past the warning level it signals a confirmed, progressing problem that calls for planned maintenance — typically within one to four weeks. Conceptually the warning level aligns with the Zone C boundary of ISO 20816 (the modern successor to ISO 10816): unsatisfactory for prolonged continuous operation and requiring corrective action, yet not so severe that the machine must be stopped at once.

Warning levels are the primary action trigger in a predictive maintenance programme. Crossing one means the window for orderly intervention is open: the fault is real and developing, but enough time remains for scheduling, parts procurement and execution during planned downtime. Getting this threshold right is what turns raw condition monitoring data into cost-effective decisions.

1. Position in the Alarm Hierarchy

The warning level only makes sense as one rung of a graded ladder of thresholds, each prompting a different response.

The multi-tier system

1. Normal: below all thresholds — healthy operation against a known baseline.
2. Alert / caution: the first threshold — a possible problem worth investigating.
3. Warning: a confirmed problem; plan maintenance (this level).
4. Danger / critical: a severe condition demanding urgent action.
5. Trip / shutdown: the trip level that forces an emergency stop.

The alert level catches early hints; the warning level confirms and times the repair; the threshold structure as a whole protects the asset.

Typical values

• Baseline-referenced: warning ≈ 4× the baseline vibration recorded when the machine was known to be healthy.
• Standards-based: warning set at the ISO 20816 Zone C/D boundary for the machine class.
• Absolute: roughly 7–11 mm/s RMS velocity for general machinery, depending on size and mounting.
• Equipment-specific: tuned to machine type, size and speed rather than taken from a single rule.

2. Why the Warning Level Matters

A confirmed problem

A warning exceedance should represent a genuine, repeatable change — not a transient or a measurement error. The fault is confirmed through trending across successive readings, it requires corrective action, and simply continuing to watch it is no longer sufficient.

A planning window

The defining feature of the warning level is the time it buys. The response can be orderly rather than reactive: it is not an emergency, but neither is the window indefinite. A typical span of one to four weeks allows maintenance to be scheduled and parts to be ordered without disrupting production.

Economic optimisation

A well-placed warning level is early enough to head off secondary damage — preventing a minor bearing fault from wrecking a shaft — yet late enough to extract useful life from the component before intervening. It strikes the balance between premature, wasteful maintenance and the risk of running to failure, marking the optimal repair timing.

3. Setting Warning Levels

Several factors shape where the warning level should sit for a given machine.

Equipment criticality

• Critical machinery: lower, tighter warning levels to force earlier intervention on high-priority equipment.
• Non-critical machinery: higher, looser levels are acceptable where a failure has limited consequences.
• Safety impact: lower levels wherever a failure carries safety consequences.

Failure-mode speed

• Slow deterioration: a higher warning level is acceptable because weeks of lead time remain.
• Rapid progression: a lower level is needed when only days separate detection from failure.
• Evidence base: set from historical failure data for that equipment type and its dominant defect severities.

Maintenance scheduling

• The level must allow enough time to plan and schedule an outage.
• Account for typical spare-part procurement lead times.
• Respect production scheduling constraints.
• The warning should trigger with adequate planning time still in hand.

4. Responding to a Warning Exceedance

Immediate actions

1. Confirm the alarm: verify it is not a false alarm or measurement error before acting.
2. Detailed analysis: run a spectral (FFT) analysis to identify the specific fault behind the rise.
3. Severity assessment: judge how fast the vibration is climbing.
4. Generate a work order: initiate the maintenance-planning process.
5. Increase monitoring: step up to weekly or daily measurements to track progression.

Where the spectrum points to mechanical unbalance as the cause, the diagnosis can flow straight into a correction: a portable two-channel instrument such as the Балансет-1А can both confirm the elevated 1× level on the machine in place and balance the rotor on the same visit, often clearing the warning without a teardown.

Planning actions

• Determine the required repairs and parts.
• Procure the necessary spares.
• Schedule the outage window.
• Assign resources — personnel and tools.
• Prepare clear work instructions.

Execution timeline

• Non-critical equipment: the next convenient outage, around 2–4 weeks out.
• Important equipment: a dedicated outage planned within 1–2 weeks.
• Critical equipment: expedited scheduling, from days to a week.
• Adjust: tighten or relax the timeline according to the vibration progression rate.

5. Documentation, Optimisation and Refinement

Records to keep

Document every alarm setpoint for each machine, the rationale behind the chosen levels, the date each was established, any subsequent changes, and the approval and review process. Maintain alarm event logs too: when the warning level was exceeded, the vibration value and trend at the time, the investigation findings, the actions taken, and the resolution with its results. Good records feed directly into the machine’s diagnostic report history.

Tuning the levels

• False-alarm tracking: count alarms that did not lead to a found defect; aim below a 10% false-alarm rate, and relax levels upward if they trip too often.
• Missed-failure analysis: if a failure occurred before the warning level was reached, the level was too lenient — lower the threshold or increase monitoring frequency to catch the next one earlier.
• Continuous refinement: review annually or after any significant event, fold in operational experience, and update the levels after equipment modifications, keeping a documented trail of every change.

Warning levels are the action thresholds in a condition monitoring programme that trigger planned maintenance. Set correctly between the first alert and the critical condition, they provide the optimal intervention point — problems confirmed, but time still in hand for an orderly response — delivering the planned, cost-effective maintenance that is the whole promise of predictive maintenance.

---

← Back to Main Index