Understanding Online Monitoring
Online monitoring — also called permanent monitoring, continuous monitoring, or installed-system monitoring — is a condition monitoring approach built on permanently mounted sensors and instruments that gather vibration, temperature and other data automatically, either continuously or at frequent intervals from seconds to hours, with no human intervention. Where route-based collection relies on a technician walking up to each machine on a schedule, an online system delivers round-the-clock surveillance, real-time alarming and automatic trending for the machinery that matters most.
It is the premium tier of condition monitoring, reserved for high-value or critical assets where the cost of permanent sensors and infrastructure is easily justified by early fault detection, immediate alarming, and protection against catastrophic failure. In exchange for that investment it returns maximum protection and the richest data set of any monitoring strategy.
1. System Components
An online system is a chain running from the shaft to the control room. Three layers make it up.
Permanently installed sensors
- Accelerometers at each bearing — often two or three per bearing.
- Proximity probes in X–Y pairs for shaft displacement and orbit analysis on fluid-film machines.
- Temperature sensors (RTDs or thermocouples) on bearings and windings.
- A tachometer or keyphasor for speed and phase reference, all cabled back to central hardware.
Monitoring hardware
- A multi-channel data-acquisition system with signal conditioning and processing.
- Local processing and alarming so protection works even if the network drops.
- Network connectivity, typically rack-mounted in the control room or a field enclosure.
Software and networking
- Real-time display and trending, with automatic alarm management.
- A data historian for long-term archiving and later analysis.
- Integration with plant systems and remote-access capability.
2. Advantages
The benefits cluster into four areas, all flowing from the fact that the sensors never leave the machine.
- Immediate fault detection: problems are seen as they develop rather than at the next route visit, giving the shortest possible time from onset to detection and the longest lead time for action — true early warning.
- Continuous protection: 24/7 watch with automatic shutdown on critical vibration, protecting the asset, adjacent equipment and personnel from catastrophic damage.
- Rich data: high-frequency measurement captures transient events and a detailed fault-progression history, enabling sophisticated analysis impossible from sparse samples.
- Operational gains: no manual collection, perfectly consistent measurement conditions, trending without human effort, and automated reporting.
3. When to Use Online Monitoring
The investment is warranted when one or more of the following apply.
Critical equipment
- Failure causes production loss above roughly $10,000 per hour, or the asset is safety-critical such as a turbine or compressor.
- The process is hazardous (toxic or flammable), or the machine is a single point of failure with no installed spare — classic critical machinery.
High-value assets
- Equipment worth more than about $500,000, or with prohibitive replacement cost or lead time, so that monitoring is a small fraction of the asset value.
Inaccessible or hazardous locations
- Machinery that is difficult, dangerous or remote to reach for routine periodic checks, or sited in continuous-process and automated areas.
Regulatory requirements
- Standards and mandates such as API 670 for turbomachinery protection, nuclear regulations, or industry-specific codes; the broader monitoring framework is set out in ISO 13374.
4. Implementation Considerations
A permanent system is itself an asset to be specified, budgeted and maintained.
- Cost: sensors run roughly $500–2,000 each, hardware from $5,000 to $50,000+ depending on channel count, and software from $5,000 to $100,000+ by capability; with wiring and commissioning, a typical machine lands at $10,000–100,000.
- Integration: DCS/PLC links for shutdown, CMMS linkage for automatic work orders, the supporting network infrastructure, and the cybersecurity that connecting protection systems now demands.
- Maintenance: the system needs upkeep too — sensor calibration, software updates, and spares for the monitoring hardware itself.
5. Technologies and Features
Modern online platforms offer several measurement modes and a deep analysis toolkit.
Measurement modes
- Continuous: real-time signal processing of every channel.
- Snapshot: periodic detailed captures every few minutes.
- Alarm-triggered: a full detailed capture the moment a limit is crossed.
- Transient capture: special start-up and coast-down recordings through critical speeds.
Analysis capabilities
- Overall level trending, FFT spectral analysis, and envelope analysis for early bearing defects.
- Orbit analysis where proximity probes are fitted, plus automated fault-detection algorithms.
Alarming
- Multi-level limits — alert, alarm, danger, trip — each a distinct threshold escalating from a warning through to a trip level.
- Email and SMS notification, automatic shutdown capability, and full alarm acknowledgment and logging.
6. ROI, Justification and Where Portable Tools Fit
The economics are usually compelling: preventing a single catastrophic failure often pays for the entire system. Typical payback is two to five years on critical equipment, and some facilities recover the cost in under a year. The value drivers are prevented failures, eliminated unplanned downtime, maintenance optimised to actual need, extended equipment life from timely intervention, and reduced spare-parts inventory — in short, the backbone of predictive maintenance for critical assets.
Online monitoring is not, however, an all-or-nothing choice. Most plants reserve permanent systems for their handful of critical machines and cover the rest with portable instruments, and the two are complementary. When an online system flags rising 1× vibration on a fan or pump, an engineer still has to diagnose and correct it — and a portable two-channel analyser such as the Balanset-1A can confirm the diagnosis on the spot, balance the rotor in its own bearings at operating speed, and verify the residual unbalance before the machine is handed back. The permanent system watches; the portable tool fixes.
