Wind turbines are theoretically designed to operate for 20 years and the fatigue life of main components is calculated on that basis. The theory, however, is not being borne out in practice, as those who operate and maintain wind plant are well aware of. "With hindsight it has to be said that not all the calculations have proved to be right," says Strange Skriver, chief technical consultant for the Danish association of wind turbine owners. "We all know about components that have broken down well before the 20 years have passed, components that are designed to last for the entire life of the turbine."
The problem, according to Skriver, is that the haste to get new and bigger products on the market ahead of competitors means that a finished wind turbine has yet to be launched. Turbines are sold before all the components are sufficiently tested. "When parts break, the manufacturer develops a better solution which may be applied to all turbines of that type free of charge, or offered for sale as a turbine upgrade," Skriver told a gathering of Danish wind turbine owners in November. In effect, wind turbines are being put into commercial operation without being fully developed, he says. New models are instead perfected through a series of retrofit programs during their first five to seven years of operation. "So far, when a prototype is being tested, the next generation turbine is reaching the last stages of its design and the next again is being planned. That means that operational experience with one turbine has not been built into the next without a time lag," says Skriver. "Development has gone faster than the industry could keep up with."
In the early days of the modern wind industry, little was know about wind turbine loads. To be on the safe side, the Danish national laboratory, which had the job of issuing type certificates for wind turbine series before they could qualify for government subsidies, applied the same wind loading criteria to turbine rotors as it used for buildings. "That resulted in some rather over dimensioned machines," says Skriver. "When wind turbines have broken down, it's because specific components have been under dimensioned. After the component has been redesigned and retrofitted, the turbine can operate again."
Retrofits can produce good results, he points out, at least on older components that had a tendency to be over dimensioned. As an example he cites the failure of Alternegy 7.5 metre blades installed on hundreds of 55 kW turbines in Europe and California. The cause of the failures was poor quality manufacturing, particularly at the blade root. By testing root stiffness it was possible to retrofit poor blades before they were thrown off the rotor. "Time revealed that even though the quality of the blade root was really poor, the blades managed to operate for 15-20 years after a modest repair. Not many believed that would be the case at the time," says Skriver.
The example shows that a big enough safety margin had been built into the early blades, he continues. Even when the quality was poor, once retrofitted, the blades run to the end of their design lives. Skriver points out that the following generation of 11-13 metre blades used in large numbers on turbines with rated capacities of 150-225 kW were much lighter than the early 7.5 metre models. But manufacturing quality was good enough and the blades have not failed.
Much effort has gone into making today's far larger blades as light as possible, says Skriver. Lighter blades means lighter loads on the entire structure, reducing material costs. "We have seen several examples with bigger blades where relatively small deviations in production give fatal results, with cracks in the blades causing the blade to break. That reveals that safety margins are not as big in today's newer turbines as they were in the older models," he says.
"For larger wind turbines the demand for spare blades can also be expected to be larger," adds Skriver. "I have argued that safety margins have fallen as blades have grown, but in reality we don't yet know if that is the case for blades in the 19-25 metre range. But the turbines are taller and the probability of lightning hits on blade tips is much bigger. Even though these blades have effective lightning conductors, we experience the occasional total failure following a lightning strike. Access to spare blades in this size range will be important in future." Vestas, says Skriver, has retained the moulds to all its blade models and produces blades for stockpiling.
Beyond the well documented series failures of gearboxes in wind turbines with rated capacities between 600 kW and 1 MW (Windpower Monthly, November 2005), Skriver says gearbox damage is being recorded on larger wind turbines too, though the problems are no longer related to the use of spherical bearings, since use of these has been dropped. He does not, however, expect to see a repeat of the widespread gearbox failures seen previously, though by far the majority of problems still result from bearing failure.
"We have seen many failures of bearings in gearboxes. Nobody has been able to demonstrate why and no calculations can explain why they fail as early as they do. That shows that gearboxes suffer loads or forces or momentary forces that we don't know about and therefore haven't taken account of in calculating bearings. Vital information is still missing on loads on gearboxes in wind turbines. As long as we lack knowledge of the loads, we will see gearboxes failing."
The price difference between fitting a new gearbox or a refurbished one is not large, adds Skriver, but the earlier damage is discovered, the cheaper it is to repair. If replacement of all the bearings is all that is required -- and this can be done by a local workshop -- the cost of a repair can be half that of a new gearbox.
Operational losses can be kept to a minimum by installing a replacement gearbox when the failed unit is taken down for repair. For this reason, Skriver recommends building up stocks of refurbished gearboxes. "Manufacturers have not shown any interest in building up stocks of spare gearboxes in this way. They would rather sell a new gearbox or a refurbished one." He points out that small third-party service companies are better than large manufacturers at managing retrofits and individual solutions.
"Vestas has talked about establishing a workshop for refurbishing gearboxes for a long time so that it can be part of this sector, but so far it's all been talk. Now it looks as if Vestas will co-operate with an external company in Denmark on renovating gearboxes. We can only hope that will lead to better and cheaper solutions from Vestas in future."
Most large components, such as gearboxes, generators, bearings, electric motors, yaw drives, cooling systems, hydraulic components, switch gear, and so on, can be bought through the original wind turbine equipment manufacturer or in some cases directly from the sub-supplier. These standard parts will continue to be available, believes Skriver, but that is not the case for specialised components.
"The large specialised components made specifically for wind turbines are control systems and blades. Today we have problems getting hold of spare parts for particular control systems and we have seen prices which are hard to understand." In some situations the answer has been to install a complete new control system. "It naturally costs a lot to replace an entire control system, but if this is the only option for keeping the wind turbine running satisfactorily -- and it can pay for itself in less than a year -- it becomes a very interesting solution."
There is no certainty, however, that the required part will be on the shelf when it is needed. "We have seen very long delivery times on large bearings and gearboxes in particular. Delivery times of 14-18 months are now the norm. There is nothing to indicate these will get shorter," says Skriver. They may get longer. "We've seen several cases of wind turbines at a standstill for many months as a result of gearbox and main bearing failure. For these manufacturers, stockpiling spare gearboxes would be a good investment."
In Denmark, service is largely carried out by the original equipment manufacturer, but Skriver welcomes the first award of an ISO:9001 certificate to an independent Danish wind turbine service company. Several more Danish companies are in the process of being certified. Skriver says it is a welcome development that will lead to a fitting number of qualified service companies and the benefits of price competition.
"The argument for sticking with the manufacturer, despite the higher price, is that he knows the turbine better. That is true for most turbine types. But the older the turbines get, the less truth there is in that argument. With hindsight, wind turbines are not further developed once they are more than seven or eight years old. From that point, knowledge of the turbine rests entirely in the manufacturer's service division. The older the turbine is, the fewer people in the service division that will have knowledge of it."
Economics decide whether a major component retrofit or replacement is worthwhile. "Just as long as a repair can pay for itself within a reasonable timeframe, it's worth keeping the wind turbine running. With a major renovation, such as a new gearbox, the extended operating period would justify a very large investment," says Skriver. Wind turbines could turn a profit well past a 20 year working life, he believes.