1. September 2026
News

Wind Turbine Condition Monitoring: A Practical Guide

Short answer: wind turbine condition monitoring is the continuous measurement of vibration, temperature, oil and load data from turbine components, so that developing faults are found while they are still repairable rather than after a failure.

A standard condition monitoring system covers the mechanical drivetrain: main bearings, gearbox, generator bearings and, on newer machines, blades and tower. The frequency converter sits outside that scope on most turbines, even though it is one of the highest contributors to unplanned downtime in the nacelle.

ConverterTec condition monitoring covers the frequency converter itself, using converter data that a drivetrain CMS does not collect.

What is wind turbine condition monitoring?

A wind turbine condition monitoring system, usually shortened to CMS, is a permanently installed measurement chain: sensors on rotating and load bearing components, a data acquisition unit in the nacelle, and condition monitoring software that turns raw signals into a health assessment. It runs while the turbine produces power, which is what separates it from periodic inspection. As wind power has grown into a core part of renewable energy portfolios, that continuous view has moved from an option on prototype machines to a normal part of asset management.

The purpose is narrow and practical. A CMS does not tell you that a component is broken, because at that point the turbine has already stopped and the SCADA system has already raised an alarm. It tells you that a component has started to change, weeks or months earlier, so maintenance planning can happen on your schedule rather than on the weather’s. For turbine operators, that single shift is where the reliability gain and the reduction in maintenance costs actually come from.

That distinction matters commercially. An unplanned gearbox exchange in winter, with a crane mobilisation and no parts on site, costs a multiple of the same exchange planned in advance for a low wind week. The condition monitoring data is what buys the planning time. This is the same logic covered in more depth in our comparison of predictive maintenance and preventive maintenance.

What components of a wind turbine can be monitored?

Coverage varies by turbine platform and by how the wind farm was specified at purchase. The table below sets out what is typically monitored, with what technique, and what each channel actually detects.

Component	Technique	Typical sensor	What it detects early
Main bearings	Vibration analysis	Low frequency accelerometer	Raceway spalling, misalignment, lubrication starvation
Gearbox	Vibration plus oil analysis	Accelerometer, oil debris sensor	Gear tooth wear, bearing faults, metal particles in gearbox oil
Generator bearings	Vibration analysis	High frequency accelerometer	Bearing degradation, imbalance, electrical erosion
Blades and pitch system	Structural health monitoring	Strain gauges, acoustic sensors	Cracks, imbalance, pitch system drift
Tower and foundation	Structural health monitoring	Strain gauge, inclinometer	Tower sway, stress and strain trends, foundation movement
Yaw system	Vibration and torque	Accelerometer, torque measurement	Yaw gear wear, brake pad degradation
Frequency converter	Electrical and thermal monitoring	Converter internal measurement	Temperature drift, cooling loss, capacitor and IGBT ageing

The first six rows are what a conventional CMS covers. The last row is the one that is usually missing, and it is the reason this guide exists.

What types of sensors and techniques are used?

Vibration analysis

Vibration is the backbone of drivetrain monitoring. Accelerometers mounted on bearing housings capture vibration signals across a wide frequency band, from below 1 Hz on the low speed shaft to several kHz at the generator end. Signal processing, usually an FFT together with envelope analysis, separates a specific bearing defect frequency from the general noise of a machine that changes rotational speed constantly.

A typical drivetrain installation uses between six and twelve vibration sensors. Fewer than six and you cannot separate a gearbox fault from a generator fault; more than twelve and the additional channels rarely pay for themselves.

Oil and particle monitoring

Gearbox oil carries the wear history of the whole gear set. Inline oil debris sensors count ferrous metal particles in real-time, which gives early detection of gear and bearing wear before the vibration signature changes. Oil condition sensors add viscosity, water content and temperature data.

Temperature and SCADA data analytics

Every turbine already produces a large amount of SCADA data: wind speed, power, rotational speed, temperatures at dozens of points, and alarm history. Data analytics applied to SCADA data can flag a component drifting away from its own normal behaviour without any additional hardware. This is the cheapest form of condition monitoring available, and on many wind farms it is underused. It is also the basis of most digital twin work in the energy sector.

Structural health monitoring

Strain gauges, fibre optic sensors and acoustic emission sensors measure stress and strain in blades, wind towers and foundations. Structural health monitoring is standard on offshore wind turbine foundations, where inspection access is expensive, and increasingly common on long blade onshore machines.

Why is the converter the gap in most condition monitoring programs?

Ask an operator what their CMS covers and the answer is almost always mechanical: gearbox, bearings, sometimes blades. That reflects how the market developed. Condition monitoring solutions grew out of rotating machinery vibration analysis in heavy industry, arrived in wind energy with the first large gearboxes, and were shaped by what turbine OEMs offered as a factory option at the point of sale. The frequency converter is not a rotating machine, so it never entered that scope.

The failure statistics do not follow that logic. Gearbox failures are the most expensive single event, which is why they attract the monitoring budget, but power electronics and control systems generate a far higher number of stoppages per turbine per year. Each individual converter fault is cheaper than a gearbox exchange. Multiplied across a fleet and across a year, the lost production is comparable, and it arrives as a stream of short unplanned outages that are difficult to plan around.

The practical consequenceA wind farm can have a fully instrumented drivetrain, a mature condition monitoring program, an asset management dashboard, and still be blind to the ageing of the component that trips the turbine most often.

There is a second reason. Converter health is visible in data the converter itself already produces, not in data a bolt-on accelerometer can capture. Reading it requires knowing what the control software logs, what the thresholds mean on that specific platform, and what a slow drift in one temperature channel implies about a cooling circuit. That is converter engineering knowledge, not vibration analysis knowledge.

How does converter condition monitoring work?

ConverterTec condition monitoring is applied to the frequency converter. It is not a whole turbine CMS and does not replace one. It uses the measurement points already present inside the converter, combined with analytics built on 30+ years of design and repair data from the platforms we engineer.

The channels that carry the most information are consistent across platforms:

Temperature data across power stacks, cooling circuits and cabinet air, trended against wind speed and load rather than read as absolute values.
DC link voltage behaviour, which changes measurably as capacitor banks age long before a protection trip occurs.
Cooling system performance: flow, coolant temperature and fan behaviour, where a slow loss of cooling capacity is the single most common precursor to a power stack failure. How those loops degrade is set out in our guide to water cooling in wind turbine frequency converters.
Switching and protection event history, where a rising count of self clearing events is a reliable warning that a module is approaching its limit.
Control system logs and fault code patterns, which distinguish a genuine developing fault from false alarms caused by grid disturbances. Reading those patterns is covered in our guide to frequency converter fault codes and troubleshooting.

Interpretation is the part that determines whether a condition monitoring program produces informed decisions or noise. A converter that reports a slightly elevated stack temperature on a hot afternoon at full power is behaving normally. The same reading at partial load in October is not. Our analysis tools, including DataVisualizer and CTSAnalytics, are provided through your regional contact together with the monitoring service, so that the data is read against the platform it came from.

When monitoring identifies a developing issue, the response path is already in place: spare parts from stock, workshop repair and refurbishment, on-site field service, or where the platform is reaching the end of its economic life, an upgrade or retrofit.

Are condition monitoring systems compatible with older wind turbines?

Yes, and this is where the economics are strongest. A turbine commissioned in 2005 has passed the point where component ageing is theoretical. It is also the machine least likely to have been delivered with a CMS, because condition monitoring was an option rather than a default at the time.

For the mechanical drivetrain, retrofitting a CMS means installing sensors, cabling and a data acquisition unit, which is a real capital cost. For the converter, much less is required. The measurement points already exist, because the control system needs them to run the machine. What is added is the extraction and interpretation of that monitoring data.

This applies across the converter platforms we support, including current ConverterTec systems and the platforms delivered under the former SEG, Woodward and IDS names, as well as units manufactured by Efacec under contract. If you are unsure which of those applies to your fleet, our article on which company services SEG, Woodward and IDS converters today explains the relationship between those names.

How do you build a business case for condition monitoring?

Most business cases fail because they are built on avoided catastrophic failure, which is speculative and easy for a finance department to discount. The stronger case is built on four effects that can be quantified from your own O&M records.

Planned instead of unplanned work. Compare the actual cost of your last three unplanned converter interventions with the cost of the same work carried out in a scheduled window.
Fewer call outs. Early detection converts a site visit for diagnosis plus a second visit with parts into one visit with the right part already loaded.
Derated operation instead of stopping. Some developing faults allow a turbine to keep producing at reduced output until a planned window. That option only exists if you know about the fault.
Combined interventions. If three turbines on a site each need attention within a quarter, knowing that in advance allows one mobilisation instead of three.

A useful budgeting note: monitoring is often easier to approve as an operational cost than as capital expenditure. Converter condition monitoring is typically structured on a per turbine per year basis, which keeps it inside the O&M budget line and avoids a CAPEX approval cycle. It is frequently combined with a service contract, so that monitoring and the response to what it finds sit with the same team.

ConverterTec supports more than 24,500 installed converters representing over 45 GW of capacity, for 237 customers, under an ISO 9001 certified quality system. That installed base is what the analytics are calibrated against: the comparison for your converter is not a generic model but several thousand units of the same platform. That fleet history is what makes the approach cost-effective at wind farm scale, because the forecast of remaining life for a given component is drawn from thousands of units rather than from one site, which turns monitoring from a line of operational costs into a lever on long-term reliability.

Frequently asked questions

What is the difference between a SCADA system and a condition monitoring system?

A SCADA system runs and protects the turbine. It reports the state of the machine now and stops it when a limit is exceeded. A condition monitoring system exists to find trends before any limit is reached. SCADA data can be used for condition monitoring through data analytics, but a SCADA system on its own is a control and alarm tool, not a diagnostic one.

Does condition monitoring cover the frequency converter?

A standard drivetrain CMS does not. It is built around vibration sensors on rotating components such as main bearings, the gearbox and generator bearings. Converter condition monitoring is a separate discipline that uses electrical and thermal measurements from inside the converter, and ConverterTec provides it for the converter specifically.

How many vibration sensors does a wind turbine need?

A workable drivetrain installation uses roughly six to twelve accelerometers, covering the main bearing in radial and axial directions, the low speed, intermediate and high speed gearbox stages, and both generator bearings. Below six channels it becomes difficult to locate a fault rather than simply detect one.

Can condition monitoring be retrofitted to an older wind turbine?

Yes. Drivetrain CMS retrofits require new sensors and cabling and are a capital project. Converter condition monitoring generally does not, because the converter control system already measures the relevant temperatures, voltages and events. What is added is access to that data and the analysis behind it.

What is structural health monitoring and how does it differ from condition monitoring?

Structural health monitoring measures stress and strain in load bearing structures such as blades, wind towers and offshore foundations, usually with strain gauges or fibre optic sensors. Condition monitoring in the narrower sense tracks the health of machinery components. Many wind farm programmes run both, with structural health monitoring focused on remaining useful life of the structure.

Which converter platforms does ConverterTec monitor?

Condition monitoring is available for the converter platforms we engineer and support, including current ConverterTec systems and those previously delivered under the SEG, Woodward and IDS names, together with units built by Efacec under contract. These platforms are installed in turbine series including Nordex Beta, Gamma and Delta, Senvion MM, MD, 3XM, 5M and 6M, Suzlon S9.x and S11.x, DeWind D4 and D6, Fuhrlander MD and FL2500, and Ecotecnia and Alstom ECO machines.

Key takeaways

Condition monitoring buys planning time. Its value is the difference between a scheduled intervention and an unplanned one, not the prevention of every failure.
A conventional CMS covers the mechanical drivetrain. Gearbox, main bearings and generator bearings are well served by vibration analysis; blades and towers by structural health monitoring.
The frequency converter is the common blind spot. It causes more stoppages per turbine per year than the gearbox, and it is rarely inside the monitoring scope.
Converter monitoring uses data the converter already produces. That makes it unusually straightforward to apply to older turbines with no new hardware in the nacelle.
The business case is built on planned work, fewer call outs, derated running and combined interventions, all of which can be quantified from existing O&M records.

To discuss condition monitoring for the converters in your fleet, see our condition monitoring service or contact Mark Rush at ConverterTec USA in Loveland, Colorado on +1 970 227 9597.

More Media

18. August 2026
News

Do you have any questions?

Get in touch now!