A wind farm has no fuel cost, which leads people to assume it has almost no operating cost. It does not. Operations and maintenance is a significant and continuing line item over a twenty five year life, and how well it is managed separates projects that meet their financial model from projects that quietly underperform it for two decades.
The three kinds of work
Scheduled maintenance
Each turbine receives planned service at defined intervals, typically once or twice a year. The visit covers lubrication, oil and filter changes, bolt torque checks on critical connections, inspection of blades, brakes, pitch and yaw systems, electrical connections, safety equipment and the climbing and lifting apparatus inside the tower.
Scheduled work is the cheapest form of maintenance per hour and the easiest to plan, since it can be timed for low wind periods when the production lost costs least. Good operators schedule around the wind rather than around the calendar.
Unscheduled repairs
Faults happen, and the cost distribution is extremely uneven. Many faults are resets and minor component swaps, taking a technician a few hours. A small number are major component exchanges of gearbox, generator, main bearing or blade, each requiring a large crane, specialist personnel and a long outage. Those rare events dominate the lifetime repair budget, which is why the reliability discussion in our note on gearbox failures in wind turbines and what causes them matters commercially rather than only technically.
Condition based intervention
This is the work that exists to prevent the second category. Vibration monitoring, oil analysis, debris counting, thermography, blade inspection by drone and continuous review of the supervisory data all aim to catch developing faults while they are still small. A bearing replaced uptower during a planned low wind window costs a fraction of the gearbox exchange it prevents.
Reading the data the turbines already produce
Every modern turbine reports extensively through its supervisory control system. Power output against wind speed, rotor speed, pitch angle, temperatures at a dozen points, alarm and fault logs, and operating hours in each state. Most of this is recorded and a surprising amount of it is never examined.
The single most useful analysis is the power curve. Plotting measured output against measured wind speed and comparing it with the manufacturer’s warranted curve reveals underperformance that no alarm reports. A turbine producing consistently below its curve in a particular wind speed band has something wrong, and the shape of the deviation narrows down what.
Blade degradation from leading edge erosion shows as a loss concentrated in mid range winds. Pitch miscalibration shifts the whole curve. Yaw misalignment, which is invisible and common, costs a steady percentage across all conditions, as discussed in our note on pitch and yaw systems, how a turbine faces the wind. Comparing turbines within the same farm against each other is powerful, because they share the same weather and a consistent outlier is a machine problem rather than a wind problem.
Blades, the asset people inspect least
Blades are the largest, most expensive and least accessible components, and they degrade continuously. Leading edge erosion from rain, hail, dust and insects roughens the aerodynamic surface and reduces energy capture measurably before any structural concern arises. Lightning strikes damage tips and the conduction system. Bond lines between shells can begin to separate. Small cracks propagate.
Drone photography has transformed this, allowing a full external survey of every blade in a fraction of the time rope access takes, at a cost low enough to repeat annually. Repairing erosion or a small crack early is a straightforward operation. Letting either develop means a blade replacement, which means a crane.
Contracting and what the arrangements actually cover
Turbine suppliers offer full service agreements covering scheduled work, repairs and often an availability guarantee, which transfers most risk to the supplier at a predictable annual cost. Independent service providers compete on price and flexibility. Larger owners sometimes self perform, building their own technician capability, which gives the most control and requires real scale to justify.
Whatever the route, a few provisions decide whether the contract is worth what it costs. How availability is defined and what is excluded from it, since definitions vary considerably and exclusions can be extensive. Whether major components are included or carve outs apply. What response times are guaranteed and what happens when they are missed. Who owns the operating data, which matters enormously for the analysis described above and is frequently overlooked. And how crane mobilisation is handled, because that is where the large costs sit.
The things that are easy to forget
Access roads need maintaining, and in regions with hard winters that includes snow clearance so that a crane can reach a turbine when it must. The electrical collection system, substation and transformers need their own inspection programme, and a transformer failure takes out a whole string rather than one machine. Met masts and instruments need calibration, since every analysis above depends on their accuracy. Safety equipment, lifts and fall arrest systems require certification. Spares holdings need managing, because lead times on large components can be long.
And somebody has to own the whole picture. The most common failure in wind farm operations is not technical but organisational, where data is collected, alarms are acknowledged, and no one is responsible for noticing that turbine seven has been producing four percent below its neighbours since the spring. Public research on operations and reliability is published through the National Renewable Energy Laboratory.
Frequently asked questions
How often are turbines serviced?
Typically once or twice a year for scheduled maintenance, with the timing ideally chosen to fall in low wind periods so that lost production costs less.
What drives the lifetime repair budget?
A small number of major component exchanges requiring cranes, rather than the many minor faults. That is why condition monitoring aimed at preventing them pays for itself.
What is the most useful analysis of turbine data?
Comparing the measured power curve with the warranted curve, and comparing turbines within the same farm against each other. Both reveal underperformance that triggers no alarm.
Why do blades need regular inspection?
Because leading edge erosion reduces energy capture well before it becomes a structural issue, and small cracks or bond separations are far cheaper to repair early than to let develop.
What should be checked in a service contract?
How availability is defined and excluded, whether major components are covered, guaranteed response times, how crane mobilisation is handled, and who owns the operating data.