Despite wind power’s going “mainstream,” original equipment manufacturers and end users struggle to pin down quality standards for ever-evolving wind turbine component technologies.
As more utilities embrace wind power, the U.S. wind turbine market has expanded tremendously over the years. It has proliferated into numerous facilities that specialize in the roughly 8,000 component parts that make up utility-scale wind turbines, including blades, rotors, generators, and other parts located inside the nacelle.
To drive the domestic wind industry’s growth, turbine makers have sought to scale up turbines to multi-megawatt power ratings, while also increasing cost-effectiveness. A telling statistic from the Department of Energy’s (DOE’s) National Renewable Energy Laboratory is that since 1999, the average turbine generating capacity has increased by 170%, to 1.94 MW. This has been achieved by advancements in composite materials, automation, and more efficient manufacturing processes.
Perhaps as important to the sector’s future is turbine reliability. Unplanned maintenance and component failure are concerns to both wind farm owners—because they directly affect revenue streams—and original equipment manufacturers (OEMs) (Figure 1), who often have to cover the costs of warrantied maintenance work.
As the Lawrence Berkeley National Laboratory notes, a wind turbine is a significant investment. The lab reported in 2011 that turbine transaction price quotes can range from as low as $900/kW to as high as $1,400/kW—meaning an average 2-MW turbine could cost between $1.8 million and $2.8 million, plus installation costs. Observed and projected potential failures of components of the gearboxes, generators, or blades alone could contribute about 3% to a wind project’s cost of energy. That’s a significant amount when you consider that a project’s civil work construction contributes an estimated 13%.
The Quality Question
From some perspectives, turbine quality concerns can be traced back to a lack of long-term renewable energy policies in the U.S. These have created an uncertain environment for turbine fabrication, prompting fluctuating volumes that lead to boom-and-bust planning scenarios and that do not support efficient, quality-ingrained manufacturing environments. Ramps in production sometimes mean hiring employees with limited experience in manufacturing components.
At the same time, because wind power’s rapid installation increases have spurred an accelerated evolution, some technologies, like those affecting blade materials, have changed quickly over the past few years. Some equipment currently in operation can easily be considered in the “infancy” category, say some stakeholders. The increased use of carbon in new blade designs, for example, makes certain processes complex, like inspection.
Other experts point out that the stochastic nature of wind power production contributes to challenges in standardizing reliability enhancements and reducing maintenance costs. That is, particular turbine models have different issues based on geography and operations.
Among several initiatives to address this challenge is the DOE’s Sandia National Laboratories’ Blade Reliability Collaborative, and its Continuous Reliability Enhancements for Wind (CREW) database and analysis program, which serves to characterize wind turbine and wind plant reliability performance issues by collecting data from thousands of wind turbines nationwide. CREW also provides a public benchmark that allows wind power generation owners and turbine makers to self-assess their performance against their peers.
Yet, during 2013, when CREW team members performed an internal reassessment of their effort to identify the benchmark’s value, any data gaps, and opportunities to be better aligned with the wind industry, they found that wind plant owners and operators have been developing in-house reliability-data collection tools with similar goals to the CREW project. It said in a September 2014 statement announcing this finding that the lab strongly supported a national reliability benchmark to assist the CREW team in their objectives of maximizing power performance yield; decreasing financial risk and uncertainty; and understanding reliability trends across turbine models (and components), geographical locations, and age.
Such a benchmark would be useful to end users of turbine equipment, many of whom say there is a dearth of failure/reliability data. “While trends show that [independent power producers] and utilities are developing their own projections of equipment failures, those analyses are in the early stages of development,” Michelle Arenson, director of wind energy development at Alliant Energy told POWER in June.
Meanwhile, OEMs continue to hold failure data as proprietary, which means there is a lack of transparency between user and OEM, she noted. Other factors that make quality issues difficult to understand include the lack of material certification, material origin, poor serial numbers, and a lack of lot tracking.
The relatively recent spotlight on quality management will only become more intense, Arenson predicted. She noted that the majority of turbines built in the U.S. were installed in the 2007 to 2009 time period, and that many of those turbines are coming off the original warranty periods, “so owners may be more in tune with failures than they were when the OEM had responsibility for major component failures.”
What OEMs Are Doing
So what are OEMs doing to improve performance? Although POWER contacted a number of major OEMs with this question, only one turbine maker responded.
“Over the past year we have seen an increased focus on quality in every area of the Wind Power Division,” said Klaus Hauschulte, who heads the Quality Management and Environmental Protection, Health Management and Safety arm for Siemens’ Wind Power and Renewable Division. “We collect key performance indicators (KPIs) monthly in all areas, including suppliers’ performance. The improvement we achieved with this is that we now have transparent reports on different levels. This helps to create awareness of quality.”
Experts recommend that it would be more prudent for turbine users to do their own research before the purchase of a major component. That may not always be practical, as wind typically represents a small percentage of most generating companies’ portfolios, and those companies do not have the long history of experience with the technology nor the industry contacts that could help in that effort.
For an example of what this might entail, one of the most effective ways of preventing a gearbox failure is to ensure that the design of the gearbox is appropriate for its usage environment. According to Bob Errichello, owner of gear consulting company GEARTECH and the primary author of the blog gearboxfailure.com, the procurement specification should include interface requirements, lubrication requirements, gearbox load information, minimum gear and bearing design margin, required calculation methodologies for gear and bearing life, material cleanliness, gear accuracy, required system life and reliability, minimum and maximum operating temperature, allowable maintenance intervals, acoustic noise requirements, and other such information.
ANSI/AGMA/AWEA 6006-A03 provides guidelines for design and specification of gearboxes for wind turbines that can also be used as a model for procurement specifications for other applications, said Errichello. If the application requires very high reliability, or if the consequence of failure is high, a user should commission a due diligence review of the design performed by an independent third-party expert, he suggested.
Bracing for the Blades
Blade failures often occur without warning—and they are costly and catastrophic. GE, for example, saw a series of six blade breaks (Figure 2) between mid-2012 and the end of 2013 on 1.6-MW turbines—either GE’s 1.6-100 model or 1.6-82.5 model—in Illinois, Michigan, and New York.
Though the company did not respond to repeated queries from POWER for this article, it was widely reported at the time that at the end of 2013, GE notified the owners of all 1,087 GE 1.6-100s installed around the world of the blade breaks. It later said it had identified a “suspect population” of roughly 1.5% of total blades in its 22,000 wind turbine fleet. Once it identified the root cause (some breaks were linked to a spar cap manufacturing anomaly, one to a disrupted cure in carbon spar, and another, to a lightning strike), it said it worked closely with customers to keep the turbines running reliably and safely. According to DNV GL, which has compiled blade failure rates from 10 GW of operating wind projects, between 1% and 3% of turbines in North America require blade replacements annually. Most of these replacements occur within the first 10 years of operation, with the highest failure rates usually occurring within the first five years.
DNV GL suggests that manufacturing defects are currently the leading cause of blade failure, specifically, when blades are not made to design specifications. Most utility-scale wind turbine blades are manufactured in several parts, which are then bonded together in a secondary assembly process, it notes.
But damage can also occur from lightning strikes, during transportation and handling, and due to operational factors such as incorrect pitch setpoints, incorrect shutdown sequencing, or failure to maintain yaw alignment during high winds.
OEMs can take a number of steps to reduce the risk of blade defects, DNV GL says, including certifying their quality systems to the ISO 9001 standard, ensuring proper day-to-day detailed execution of manufacturing and quality operations, and automating processes such as fabric placement and adhesive application.
The company acknowledges, however, that to minimize wind turbine component defects, much more research and development is needed. Echoing concerns from end users, it says that an enhanced knowledge base would allow for more “precise development” of quality acceptance criteria, and that increased clarity on turbine model–specific lightning protection system performance is needed.
Finally, the industry would vastly benefit from sharing information about blade failure investigations and root cause analyses, it says. (POWER would be happy to report on such findings.) ■
—Sonal Patel is a POWER associate editor