Somewhat ill-defined, the term "operation and maintenance" covers a range of costs, including land or seabed rent, insurance, servicing and spare parts, electricity purchases from the grid, and administration costs.
Attention tends to focus on the servicing and spare parts element, which includes both scheduled and unscheduled maintenance, and typically accounts for 25-40% of O&M costs - that is 5-8% of electricity generation costs in the case of onshore wind.
O&M costs, according to a recent report from the US Energy Information Administration (EIA), amount to around $40/kW/year for onshore wind and $74/kW/year for offshore wind. The UK's Department of Energy and Climate Change quotes slightly lower figures at $30/kW/year and $57/kW/year, respectively. The Danish Energy Agency's (DEA) figures are slightly lower than those of US DoE and also suggest these costs will fall by around 30% - both onshore and offshore - by 2030.
A downward trend in servicing costs has been identified in a recent report by Bloomberg New Energy Finance (BNEF). Analysing data from 5GW of new US projects, the report shows that the cost of full-service contracts - which include scheduled and unscheduled maintenance costs, plus replacement parts - fell from around $40/kW/year in 2008 to around $25/kW/year in 2012.
BNEF also observed that the average length of full-service contracts has increased from 4.5 years in 2008 to 6.9 years in 2012 and average availability guarantees were 96.9%. The decreasing cost and increasing contract length suggest that turbine reliability is increasing.
At first sight, research from GL Garrad Hassan based on data from 8GW of wind farms, paints a more pessimistic picture. This shows that O&M costs increased, on average, from $21/kW/year in 2008 to $31/kW/year in 2011. In practice, these costs relate to wind farms that are already operational, whereas the BNEF data relate to new projects. These costs reflect the confidence of manufacturers that machines are now able to operate with a high level of reliability.
Most analyses of failures conclude that electrical and electronic systems are the most common causes of wind-turbine outages, although most of these faults can be rectified fairly quickly. High-profile generator and gearbox failures are less common, but take longer to fix. Increasingly sophisticated instrumentation is being used to monitor the performance of turbine components in order to provide warnings when key parameters, such as gearbox oil temperature or vibration levels, start to change. This raises an early alarm of incipient faults in the bearings, for example, or may indicate excessive wear on the gear teeth. Timely action or additional inspections might pre-empt costly failures.
This is a good example of condition monitoring, and similar techniques can be applied to rotor blades, main bearings, the pitch-control mechanism and other components. Many turbine manufacturers now offer in-built condition monitoring equipment and a number of specialised instrumentation suppliers are able to supply suitable sensors and/or software. Sophisticated software can also be used to quantify optimum stock levels for spare components and the best time to carry out repairs, reducing downtime of the turbine. It may be possible, for example, to delay repairs slightly to coincide with scheduled maintenance. Another factor to be taken into account is whether or not repairs need to be undertaken during windy seasons, or whether they can safely be delayed.