Independent servicers of wind turbines are using empirical data to press home their case to the wind energy industry that they are a cost efficient way to improve performance and profitability. The pitch comes at a time when many wind plant owners are struggling to cope with underperforming wind projects, higher service costs and poor spare part availability.
There are a number of business models for wind plant service, many of which have been around for decades. Wind plant owners frequently agree contracts with turbine vendors for service through the initial warranty stage, and often pay for an extended service agreement once the warranty runs out. Other wind plant operators, particularly larger owners of big fleets, operate their own in-house service teams. Independent service providers (ISPs) fill in the gaps. But as more turbines are installed throughout the world, and vendor warranties on more and more turbines expire, proper operations and maintenance (O&M) often proves a challenge for turbine owners. New ISP companies are entering the market, aiming to meet the new demand. In some cases, they argue, they can provide better value than is being offered by the turbine suppliers that have cornered the market in many countries. These issues were debated at an operations and maintenance conference in Hamburg, Germany, this summer, hosted by Windpower Monthly. There was widespread agreement that, irrespective of who does the service, many plants could be performing better.
One of the best measures of the profitability of wind turbines is their availability factor, which is the percentage of time a wind plant is fully online and able to capture available winds. Because the wind resource is variable, it is critical that when good winds blow, the plants are able to run at their peak capacity. As a starting point, says Klaus Kruger of Voith Industrial Services, which provides O&M to a range of industries, the wind industry has overestimated the expected energy production of its wind stations, and projects often have not properly factored in the ongoing O&M costs to keep a plant running efficiently. Kruger says that availabilities in Europe range from below 95% at lower performing wind stations and average around 97% at the higher performing facilities. Windpower Monthly has reported extensively on the American fleet where low plant availability accounts for half of an approximate 10% gap between pre-construction estimates of power production and actual production (Windpower Monthly, January 2009). Even at the better 97% range, the wind industry should look for higher goals, says Kruger.
While some owners hesitate to spend more on maintenance or do not have the money for it, better maintenance can have a net saving to a business rather than a net cost. Voith's experience provides an example. The company took over O&M of a medium-sized commercial wind plant for an owner that had previously relied on the turbine vendor for service. The wind plant availability factor was 94.3% when Voith undertook an initial audit. After two years under Voith service, availability had increased to 97.6% and today it stands at 99%, which is rare for wind turbines. This translated to an eventual increase of plant income of 6% compared to before the maintenance took place.
Matthias Brandt, CEO of Deutsche Windtechnik AG, provides a similar example. His two-year-old company, which is based in Husum, Germany, specialises in the region's ubiquitous Vestas and NEG Micon - themselves acquired by Vestas in 2004.
Deutsche Windtechnik took over O&M from the turbine vendor on a ten-unit wind plant with 1.5 MW NEG Micon machines. The first four years of operation had been under the vendor's watch and at the handover the wind plant operated at 95.5% availability. After audits, new servicing and some minor parts replacements, Windtechnik increased availability to 97.1%. This represents an additional 600,000 kWh and additional revenue of EUR52,000 yearly. The turbine vendor's previous yearly maintenance costs had been EUR106,000, more than Windtechnik's annual charge of EUR85,000, leaving the wind plant owners EUR73,000 a year more profitable.
Brandt says: "On most turbines you will find lots of possibilities for improvements and retrofits." He echoes a point expressed by other wind service experts at the conference - that a key to cost-effective maintenance is having experienced staff who know all the particulars of specific turbine models and are able to turn a mass of data into usable and actionable information. Raw data is useless unless it is understood and applied, he says.
But despite the push from ISPs, the vendors, or original equipment manufacturers (OEMs), are not allowing ISPs to take more of the maintenance market share without a challenge. Axel Birk, director of global service for turbine supplier Repower Systems, believes that turbine vendors should stay closely involved in maintenance or provide full after-warranty service because they often know the turbines and subcomponents best and have better access to spare parts. "They can provide the best quality and technical competence," he says. "And they have a shortcut to the design department. There's a very close loop - it's a big advantage to using the services of the OEMs."
In Repower's case, it has an advantage on the spare parts business because when the company makes a certain model of turbine, it sets up a contractual agreement with the subcomponent suppliers that requires the suppliers to share detailed parts information with Repower. So the turbine vendor can react quickly and rapidly to provide needed spare parts to the field. "With all this information," says Birk, "we have a lot of reasons to have the OEM in the background as a supplier for wind farms."
As the machines age
The argument that the OEM knows the turbine best is pertinent for new machines but is less valid with older machines, says Strange Skriver, chief technical consultant of the Danish Wind Turbine Owners Association. Half of all wind service in Denmark is done by independent service companies. "The most knowledge of many machines is from the ISP even if the manufacturer is still alive," says Skriver, referring to the disappearance of some companies whose machines continue to turn over the years.
He believes that a spare parts strategy should begin early on, and developers should allocate money for delivery of some spares, particularly a gearbox or two, when they receive the turbines. This is a useful hedge against a gearbox failure and the potential downtime, as a replacement unit can take months to arrive.
Lack of spare parts is one of the critical issues facing the wind industry, says Voith's Kruger, who adds that the average lead time for replacement gearboxes last year was one and a half years. "This is something that the owners of a wind farm and the banks cannot live with and we have to look for ways compensate for that," he says. "The downtimes lead to reduced income for the owner and in some cases we have seen insolvencies of wind farms that were performing poorly." Kruger adds that reduced component lifecycle is also a key problem. There are examples of wind components breaking after two or three years, even though they were designed for a ten-year or higher life cycle. Part of the problem may be that some components are operating today under different conditions then they were in the past, says Kruger. The problems are especially clear for those joining the industry from more mature sectors.
"The wind industry today exhibits some early-phase characteristics, by which I mean failure rates in components and systems is too high," says Sverre Trollnes, who heads up the new energy division for international energy company StatoilHydro ASA. The company recently completed a study on failure rates at multiple wind farms. "It was astonishing," says Trollnes. "If this was in our other businesses like oil and gas we would not allow our investment to go ahead, I'm sure of it. So this really needs to be addressed. When short times between failures is combined with a poor operations and maintenance strategy then you have really poor performance, and that is what we see in some of the assets."
There was much discussion at the conference about moving away from solely corrective and scheduled maintenance to predictive and preventative maintenance. Certain procedures will always be done to a schedule and certain actions will always be taken to remedy a problem, but Trollnes believes that most of the operations and maintenance strategy in the wind industry is based on running plants until they fail.
About 70% of maintenance at most wind plants is either scheduled or corrective, he adds, which often fails to improve the performance of a wind plant or to identify problems that can lead to unplanned stops. "We should look at component failure from a new angle," says Trollnes. "We have decided that 80% of all stops should be planned and preventative in nature. It's a vast target from where we are today, but we know that it's possible because we've done it in other industries."
Trollnes points out that it is easy to blame the turbine vendor or the servicer but, in reality, the situation will need everyone in the wind industry to put their heads together to make needed improvements. It will also need a wholesale adoption of preventive and predictive maintenance technologies, such as systems that automatically and remotely monitor for changes in component vibrations. "It's a combination," says Trollnes. "You need expert OEMs, and you need ISPs - it's about how you put the team together."