More than 1,200MW of wind power capacity was installed in the United States in the third quarter of 2011 alone, bringing new installations through the first three quarters of the year to 3,360MW, according to the American Wind Energy Association (AWEA). The U.S. wind industry now totals 43,461MW of cumulative wind capacity through the end of September 2011. Of this total, more than 35% of all new generating capacity has been installed within the past four years.

"Within the international standards, the prescribed design life is 20 years," says Mike Derby, land-based R&D team lead (acting), Wind and Water Power Program, Energy Efficiency and Renewable Energy, U.S. Department of Energy (DOE). "Certainly, there are some that are older than that."

According to Derby, the first wind turbines, ranging from 40kW to 100kW, were installed in the early 1980s. "Modern large, utility-scale wind turbines for commercial production of electric power are considerably younger than those turbines," he says. "The latest generation started being deployed in the mid-2000s."

Annual operations and maintenance (O&M) costs for wind turbines, which include insurance, regular maintenance, repair, spare parts, and administration, are estimated between 3% to 5% of the total cost of installation (click here to see Fig. 1). A different estimate indicates that costs equal 20% to 25% of the cost per kilowatt hour produced over the 20-year lifetime of the turbine. If the turbine is still relatively new, the costs are estimated to reach between 10% to 15%; however, this percentage may increase to at least 20% to 35% by the end of the turbine’s life cycle. Costs for repair and spare parts, in particular, increase as the turbine ages. According to the "2010 Wind Technologies Market Report," released in June 2011 by the DOE and the Lawrence Berkeley National Laboratory, "Despite limited data availability, it appears that projects installed more recently have, on average, incurred lower O&M costs than older projects in their first several years of operation, and that O&M costs increase as projects age."

Performance Data

Many of the same processes used to reduce O&M costs in industrial plants apply to wind power projects. However, the variable environments of both inland and off-shore wind turbine locations, as well as the height of the turbines themselves, pose unique challenges. On October 31, Sandia National Laboratories released the first findings from its Continuous Reliability Enhancement for Wind (CREW) database. Although the data is considered "directional" rather than "actionable" because of its small sample size — current data covers three seasons and 58,000 turbine-days of data of wind turbines at or above 1MW in size — the dataset provides a "useful initial view of the U.S. fleet’s operational and reliability performance."

Among the goals of the project are to benchmark reliability performance, identify technology improvement opportunities, and enable O&M cost reductions. "The data from the CREW database is expected to go to the owner/operators primarily," says Derby. "It’s the same model that’s used for the gas turbine industry and other power generation."

According to Derby, plants report their reliability statistics in return for specialized reports about how they’re doing relative to other generators.

Notwithstanding 19,000 turbine-days of unknown time, attributable to the pilot status of the program and associated beta time, the wind farms reported overall performance in line with expectations. Operational availability, which includes a generating factor of 78.5% plus a reserve shutdown of 16.3%, totals 94.8%, and capacity factor equals 33.4% (click here to see Fig. 2). On a national basis, according to the 2010 Wind Technologies Market Report, the average operating capacity, or capacity factor, for wind projects has grown steadily over the past 30 years to reach a 34% capacity factor in 2010.

Reported downtime includes unscheduled maintenance at 3% and forced outage/unavailability at 1.8%. Scheduled maintenance accounts for 0.5%. "There are really two categories — planned maintenance and unplanned events," says Derby.

Newly installed wind turbines fall under manufacturer warranties, so the manufacturer specifies required service and maintenance. In addition, next year, AWEA will publish its Operations and Maintenance Recommended Practices Manual, which will be distributed to its members and made available for sale in its online store.

Generally, the entire wind system, including the tower, storage devices, and wiring, should be inspected at least once a year. Routine maintenance also includes: changing the gear oil, coolants, seals, brake pads, and filters; greasing the bearings; adjusting sensors and actuators; and visually inspecting the blades, tower, and electrical connections. Although the turbines have a life expectancy of 20 years, many of their components do not. Preventive maintenance includes the replacement or refurbishment of these components.

Unplanned stoppages requiring maintenance, according to Derby, occur under fault conditions. "If you have some kind of a failure in the system — for instance, a bearing failure or maybe a gear tooth breaks — that’s an example of unplanned maintenance," he explains.

In addition, condition monitoring is emerging in the wind industry. This method, which uses high-tech multiple sensors in the turbine, massive data collection, and sophisticated analysis to track the condition of individual turbine components, is already being used in more mature energy generation industries, such as oil and gas, coal, and nuclear. It has also been used in industrial applications in the aircraft, military, and processing sectors. In recent years, more sophisticated online monitoring systems have been introduced to wind turbines, the most common of which are vibration monitors and fluid contamination monitors. "Condition monitoring is typically done in large power plants where you only have one system serving an entire 100MW plant versus a 100MW wind system that would require 50 such systems that have to be monitored in order to determine if the turbine components are operating properly," says John Dunlop, senior technical programs manager, AWEA.

Studies by the Electric Power Research Institute (EPRI) indicate a 47% reduction in overall maintenance costs when using predictive maintenance techniques. When deciding to use any of these condition monitoring systems, however, an end-user should consider the cost of installing and commissioning these systems versus the usefulness of the information.

According to Andy Milburn, gear and bearing consultant, Milburn Engineering, Inc., Seattle, data collection can lead to data overload. "There are so many sensors on the gearboxes and the turbines themselves that owners and operators are getting too much information," he says. "There’s so much information that they’re ignoring it. It’s hard for them to pick out what’s important and what’s not. There’s good information in there, but it’s not being disseminated in a way that people can handle."

Inside the Box

Within the wind energy system, certain components may require more critical O&M care than others. Individual components are known to be more prone to failure, are essential to turbine operation, or are more expensive or time-consuming to repair. Although there are variations among components by manufacturer, configuration, and operating environment, certain items, such as generators, power converters, and gearboxes, have been singled out as failure prone by particular studies.

Excluding "other," the CREW findings list the top three contributors to turbine unavailability as rotor/blades, electric generator, and balance of plant (BOP), as shown in Fig. 3 (click here to see Fig. 3). Surprisingly, in this study, the gearbox failures, which, in other studies, are estimated to cause 30% of wind farm O&M costs, don’t rate in the top five.

A separate study conducted by the German Wind Energy Measurement Programme tracked the performance of around 1,500 wind turbines in Germany for 10 years (1997 to 2006). The German study’s findings pointed to electrical equipment as the cause of most stoppages, with approximately 5.5 incidents every 10 machine-years (click here to see Fig. 4). However, these problems are resolved fairly quickly, and the turbines are back online in less than two days. In the same study, gearboxes account for only about 1.5 incidents every 10 machine-years, but the outage time logs in at more than six days.

A smaller but more recent database from the agricultural commission of Schleswig Holstein, Germany, which tracked 5,800 turbine-years, indicates similar failure rates for the various components, but significantly longer outage periods per failure. In the case of gearboxes, the average outage period was 14 days.

To address the perceived problems with gearbox reliability, in 2009, the DOE launched its Gearbox Reliability Collaborative (GRC). The department designed and conducted field and dynamometer tests to evaluate, validate, and develop gearbox analytical tools and models. The effort also established a database of gearbox failures. "We engineered a gearbox so that the DOE owned the intellectual property for that and instrumented it so we could try to understand what was happening to the inside of a gearbox as it was being loaded," says Derby.

Although the newer gearboxes don’t yet match the 20-year lifespan of turbines in general, there have been some technological advances made that improve their reliability. However, there are still reasons for gearbox failures specific to wind turbines (see The Gearbox Problem).

"They’re doing better than they were in the past, but the longevity at this point is unknown for the newer turbines," says Milburn, who has been working with gears for 35 years, the last 10 of which have been focused on gearboxes in wind turbines. "Right now, bearings are the biggest problem. In the old days, they had lots of gear problems, but those have mostly been solved. Now, they’re more application problems. The bearings are being exposed to loads and conditions that no one has seen before."

According to Milburn, the most common problem with gearboxes in wind turbines is bearing inner race axial cracks (Photo 1). Problems with micropitting, surface temper (Photo 2), and inclusions on the gear tooth are also still occurring on megawatt-size turbines.

As an example of engineering away a problem, recent turbine designs have eliminated the gearbox by using a low-speed, large-diameter synchronous generator. However, results of studies of the performance of these direct-drive turbines are mixed. The Dutch Offshore Wind Energy Concepts (DOWEC) study compared projected reliability for six design strategies for 5MW machines and found that the failure rates will increase by approximately 20% for innovative variable-speed and direct-drive designs. Northern Power Systems, in its design study for an innovative direct drive wind turbine, concluded that unscheduled maintenance costs will decrease by 60% with their design because of the elimination of the gearbox.

Help Wanted

Until all failures in wind systems can be engineered away, there will be a need for wind turbine technicians. There are no technical or physical barriers for wind to supply at least 20% of U.S. electricity by 2030, according to the DOE; therefore, the organization estimates there will be 80,000 O&M jobs by that year. As the latest generation of wind turbines is launched, the skills required of maintenance technicians will increase in scope, particularly in the areas of diagnostic, control, and power electronic systems.

Currently, the North American Board of Certified Energy Professionals (NABCEP) does not offer certification on systems larger than 100kW. "It evaluates the skills of individual workers and typically has been focused primarily on small turbine installations," says Dunlop.

However, many additional programs are available. Most turbine manufacturers, for example, offer comprehensive training for their own technicians as well for the site owner’s personnel. In addition, AWEA directs a program that offers its Seal of Approval to technical training programs. "Our approach, with the cooperation from the training institutions across America, looks at the programs themselves," says Dunlop. "The Seal of Approval program is based on skills that were extracted from our operators. We pushed them pretty hard to provide us with the minimum skills that they would expect a graduate of a technician training program to have. It addresses all of the aspects of how those skills are taught, in the classroom as well as the hands-on portion."

Medford, Ore.-based wind systems service provider UpWind Solutions offers its 170 technicians a combination of manufacturer-provided and independent training. "We’ve got technicians that have been through the electrical training programs," says Guy Dees, director of operations at the firm’s Sweetwater, Texas-based operations office. "Other technicians have gone through technology schools — one-year or associate’s degrees."

In addition, the firm employs veteran technicians, with equal or more experience than Dees, who has been involved with wind energy since 1998 and with UpWind Solutions since 2008.

The company provides full-service O&M for utility-scale wind projects. "To make the projects work, especially sites in remote areas, we bring on promising individuals from local areas and provide them with OEM training or different types of industry training that’s available out there," says Dees. "Our technicians work on everything from BOP to major component replacement and everything in between."

Therefore, most of the company’s projects are long term. "We work on day-to-day maintenance activities, preventive maintenance, and troubleshooting," says Dees. "Our technicians travel across the United States and do everything from catching up on maintenance schedules to large-scale retrofits. It just depends on the location and the work warranted."

These long-term contracts make the work "recession proof," according to Dees. Despite the recent lull in new installation, particularly in 2010, UpWind Solutions remains busy. "Our business is not tied to development or construction of new turbines," says Dees. "Our business is all about operations, and we’re still operating the turbines that were installed during the heyday."


SIDEBAR:The Gearbox Problem

According to Andy Milburn, gear and bearing consultant, Milburn Engineering, Inc., Seattle, gearboxes in wind turbines are prone to failures stemming from their application in wind energy systems. The following are the most common problems:

Highly variable load and speed: Wind is an intermittent energy source. It alternates between gusting and still. Therefore, the load that these gearboxes are trying to transmit is a lot more variable than it is in a plant operation. "There are some applications where the load in plants is pretty variable," says Milburn. "But I think wind turbines are probably the worst of the lot."

Low gearbox safety factors: The drive system in wind turbines is designed to be compact. "They try to make things as small as they can, so that means the safety factors that they’re using are low compared to a typical industrial application where they don’t have to be too concerned about weight," says Milburn.

Flexible foundation: Typically, a plant gearbox and motor are mounted on a large concrete foundation or a steel structure that’s bolted to a concrete foundation. "Up in the nacelle, we don’t have that luxury," says Milburn. "The nacelle is flexing and the rotor itself is causing lots of loads in the whole structure. This causes misalignment between the generator and the gearbox."

Periods of no rotation: The gearboxes operate only around 30% of the time, so they are often idle. "Most rotating machinery doesn’t like to sit at rest," says Milburn. "They run into problems with lubrication then."

Operating as a speed increaser: Wind system gearboxes operate as a speed increaser instead of a speed reducer. "Most gearbox applications use a high-speed motor that drives something that’s turning slower," says Milburn. "In this case, you’re taking the blade that’s rotating slowly, and you’re increasing the speed up to the generator speed. So that has some affect on things in terms of lubrication."

Extreme operating environment: The turbines have to operate in extremely cold or extremely hot settings. Although the gearbox is in a nacelle and protected from rain it can still be subjected to extreme temperatures.

High operating temperature: Manufacturers are resistant to adding large radiators to wind turbines. "They allow these gearboxes to run pretty hot, and that means the oil viscosity gets low," says Milburn. "When they’re rotating slowly, you don’t get a thick oil film between bearings and gears, so you get metal-to-metal contact — and that’s a problem."

Design details: Design flaws may cause failures, even if they occur slowly over time.