Keeping Wind Turbines Spinning

 

Sep 18 - Power Engineering

By Blankinship, Steve

Maintenance routines for wind farms differ greatly compared with those for fossil, hydro and nuclear plants Today's typical wind farm consists of several hundred relatively thin towers about 35 stories tall, each supporting more than 65 tons of sophisticated mechanical and electrical equipment inside a nacelle on top of the structure. That same turbine, typically with a rated output of 1.5 MW, has a blade span longer than a football field, meaning that the blade tips travel at speeds of around 200 MPH under typical wind conditions.

Not surprisingly, torque-related stresses on wind turbine gearbox components are unlike those of almost any other technology. Add to that wind's inherent speed fluctuations and the frequent onslaught of rain, snow, hail, dust and the elements in general and maintenance becomes a major aspect of operating wind generating assets.

While rising natural gas prices, combined with a presumed extension of the 1.9 cent/kWh production tax credit for wind generation, should keep wind competitive with other forms of generation, wind's intermittency means that keeping turbines producing at their maximum potential is also critical to holding prices close to other forms of power production. Wind turbine maintenance costs are typically less than those for conventional forms of power generation. This, combined with the fact that wind generation uses no fuel or water, has helped wind's bottom-line economics.

Scheduled wind turbine maintenance is usually conducted two to three times a year, requiring 12 to 18 hours of downtime for each outage. Generally, only a few turbines in a wind farm are down at one time. Usually the only time an entire facility is brought off- line intentionally is for substation maintenance, which typically takes about 12 hours and occurs twice a year during low production periods.

Wind farm maintenance scheduling differs dramatically from that of fossil or nuclear plants, which are typically taken offline for maintenance during periods of low demand for power. Wind farms are often serviced when their availability is lowest due to unfavorable wind conditions. This can often be in a period of high electricity demand, too, since winds in many places tend to be light during the summer.

"We try to avoid maintenance when a plant is in ? peak production period," said Todd Brogna, regional operations manager, Pacific North West for enXco Service Corp., a division of Electricite de France and one of the largest wind turbine maintenance firms in the United States. Firms such as enXco have not found it beneficial to try to perform all the maintenance on a wind turbine at one time, a departure from practice at other types of power plants. Maintenance firms also schedule work for the morning hours when the wind is low, hoping to be finished by the time afternoon winds pick up.

A firm like enXco deploys six to eight people on a routine maintenance project. The bulk of its maintenance personnel are technicians trained to perform both electrical and mechanical work. A few electronics specialists are skilled at troubleshooting. The maintenance team usually also includes a high-voltage specialist. A crew expects to need about three days to complete scheduled maintenance on each turbine.

But while scheduled wind turbine maintenance costs are relatively low, unscheduled maintenance can be another story. Original equipment manufacturers (OEMs) shouldered much of the cost associated with unscheduled maintenance and repair as part of their maintenance/warranty agreements. But original warrantees usually expire after two years, after which the asset owner is responsible for keeping things working and for fulfilling power agreement terms with the plant's energy customers. And while unscheduled maintenance costs have not deterred wind energy's growth, they can have a negative affect on a wind farm owner's bottom line.

Gearbox: Achilles Heel?

Generator and gearbox rebuilds are generally a wind facility's two most costly maintenance items, while other maintenance issues are highly specific to particular manufacturers, models and even site conditions, said Dave Luck, director of business development for enXco Service. "Unlike gearboxes on gas turbines, the variable loads on a wind turbine are an issue," he said. In addition to taking routine lubrication oil samples, enXco's technical services group performs root cause analysis of identified issues, evaluating explanations for failures by the original equipment manufacturers. "Root cause analysis and proactive questioning of OEMs' latest ideas pay dividends," said Luck.

As wind turbines have grown larger, gearbox reliability has suffered more than perhaps any other maintenance item. Variable loads that wind places on them are sometimes extremely difficult to predict. As the wind turbines get bigger, so do their design challenges. Blades on larger machines produce massive torque loads through the typical three-stage gearbox used in these big turbines. In an attempt to meet the increased torque requirement, manufacturers have developed huge, costly ring gears and bearings. When these components fail-often due to torque-related stress- replacement components often are expensive, not to mention difficult and time-consuming, to replace.

Transparency Needed

According to a 2007 report by the National Renewable Energy Laboratory (NREL), the wind energy industry has experienced high gearbox failure rates almost from its inception. Early wind turbine designs were dogged by fundamental gearbox design errors compounded by what NREL said was consistent underestimation of the operating loads. Over time, however, wind turbine manufacturers, gear designers, bearing manufacturers, consultants and lubrication engineers have improved load prediction, design, fabrication and operation.

But because gearboxes are one of a wind turbine system's most expensive components, higher-than-expected failure rates add to the cost of wind energy. The NREL authors said future uncertainty of gearbox life expectancy is contributing to wind turbine price escalation.

The majority of wind turbine gearbox failures appear to start in the bearings, NREL said. These failures occur in spite of the fact that most gearboxes have been designed and developed using the best bearing-design practices available. Therefore, an early focus for researchers was discovering weaknesses in wind turbine gearbox bearing applications and deficiencies in the design process.

A major factor contributing to the problem's complexity, said NREL, is that much of the bearing design-life assessment process is often proprietary to bearing manufacturers. Gearbox designers, working with bearing manufacturers, initially select the bearing for a particular location and determine the specifications for rating. The bearing manufacturer then conducts a fatigue life rating analysis to determine if the correct bearing has been selected for the specific application and location. Generally, some degree of faith is required to accept the outcome of this analysis, which typically is done with little transparency. Even though bearing manufacturers claim they are adhering to international rating standards, many manufacturers use internally developed design codes, said NREL. These codes have the potential to introduce significant differences that can affect actual calculated bearing life without revealing the details to customers.

What's more, since bearing manufacturers generally do not have broad or intimate knowledge of gearbox system loads and responses that may be contributing to unpredicted bearing behavior beyond the bearing mounting location (such as housing deformations), they may not be well qualified to make valid root-cause analyses on their own.

The NREL report said the wind industry has reached a point where design practices for gearboxes do not result in sufficient life and institutional barriers are hindering forward progress. As a result, NREL initiated the Gearbox Reliability Collaborative aimed at providing a fresh approach toward building better gearboxes. The initiative combines resources of key members of the supply chain to investigate design-level root causes of field problems and solutions that will lead to higher gearbox reliability.

Check the oil

Lubrication is another element of wind turbine operation and maintenance that differs from fossil and nuclear power plant O&M. In a typical wind turbine, a large supply of lubricating oil is placed into the gearbox. The lubrication system contains filters for the oil and, depending on the turbine design, lubricant is either pumped through the system or is gravity fed. Gearboxes on the generally smaller-sized turbines installed in the mid-1980s hold about 10 gallons of oil or less. Newer, larger machines might hold as much as 60 gallons.

Brogna said one school of thought holds that if the unit is designed correctly, lubrication should not be an issue. A second school of thought, he said, argues that lubricants need to be adapted and improved to meet the unique needs of wind turbines.

Access

Manufacturers also consider human factors in new turbine designs. Through years of maintenance and repair, OEMs have learned the importance of making generators, rotors, bearings and high-speed gears easy to remove and replace. No less critical is making components easy to access. New designs allow ample working space for technicians, an especially important factor for rotor hub access. Making the rotor hub accessible from the inside-and eliminating the need for maintenance crews to work outside the tower-allows technicians to better attend to the hub during routine maintenance. The addition of visual inspection ports for gears has also proven helpful. And because climbing to the top of a wind tower may become more difficult for technicians as they age, elevators are now being included on most new generation turbine towers.

Predictive maintenance programs, for decades standard tools for the fossil and nuclear sectors, are also finding their way into wind generation now that individual units are large enough to make such proactive procedures meaningful. Predictive maintenance is generally defined as maintenance that emphasizes early failure prediction using non-destructive techniques such as vibration analysis, thermography and oil analysis.

"When one of these modern generation machines goes down, there is now enough lost revenue to justify additional instrumentation and analysis of data," said enXco's Luck. "Those of us who have spent a lot of time in the traditional power segments see their value and are starting to recommend using predictive failure analysis to asset owners."

Workforce Issues

The wind industry faces the same workforce issues common in every other sector of the power sector. There are simply not enough qualified technicians to meet the demands of the rapidly expanding wind energy sector.

"Rapid growth of the wind industry put considerable stress on providing qualified personnel," said Luck. His company currently has about 225 wind technicians working full time.

The gas turbine sector is expected to be a valuable source for wind industry expertise. Edward Lowe, who for years was a key executive responsible for GE Energy's gas turbine business, is now involved with GE Wind and said that GE Wind is able to pull a great deal of expertise from the gas turbine side of the business and apply it to wind. That's because the basic design for gas turbines and wind turbines are very similar.

One of the more common approaches is to establish technical training programs in regions where large scale wind forms are already operating. Major wind technology training programs are underway in collaboration with technical schools and junior colleges in California and Texas.

Supply Chain

Another constraint to maintaining wind facilities is a critical lack of supply chain infrastructure. Establishing a strong and consistent supply chain for new construction and replacement parts can be a "chicken or egg" dilemma. That's because a reasonable assumption is that a critical mass of wind farms is necessary before a strong supply chain infrastructure can be established. Without that supply chain, however, wind development (along with the ability to repair and maintain turbines) may be stymied. Uncertainty about the future of the 1.9 cent/ kW production tax credit for wind (which has historically been renewed at two-year intervals and almost always is accompanied by debate and political wrangling) also has a chilling effect on decisions by wind turbine manufacturers and developers to invest in supply chain infrastructure.

One commonly reoccurring theme discussed at the American Wind Energy Conference in Houston last spring was supply chain issues that hindered both development of new wind farms and ongoing maintenance. A new study released by Cambridge, Mass.-based Emerging Energy Research entitled Global Wind Turbine Markets and Strategies, 2008-2020 addressed the topic.

The report stated that hundreds of parts comprise a wind turbine, and all components require adaptation to turbine models and form part of the critical path to assembly and shipment. But some components have been deemed more strategic than others by turbine OEMs. These components may be manufactured in-house so that their proprietary nature can be better protected.

"Positioning on the supply chain poses a key strategic question for component suppliers as they address increasingly sophisticated demand for larger turbine models," said Stephanie Aldock, Emerging Energy Research marketing manager. "A key issue for manufacturers is the set of criteria turbine OEMs consider to supply a component in- house or to outsource." The seven key components identified in the report have several pros and cons, although in general all OEMs across the industry rely to a degree on outsourcing.

Life Expectancy

One critical difference between wind turbines and traditional power plant turbines has been longevity. Because the wind industry is generally less than 25 years old, turbine longevity remains a question that can only be answered as time passes.

Most steam generating plants were built with a projected life of 30 years, said Luck. Life extension programs mean that some of those plants are capable of operating far longer than that. Experience with wind turbines supports the idea that repowering can make economic sense well before the 20-year life expectancy is reached.

By Steve Blankinship, Associate Editor

Copyright PennWell Corporation Aug 2008

(c) 2008 Power Engineering. Provided by ProQuest LLC. All rights Reserved.