Wind turbine blades are a major structural element of a wind turbine blade. Wind turbine blades have near aerospace quality demands at commodity prices; often two orders of magnitude less cost than a comparable aerospace structure. Blade failures are currently as the second most critical concern for wind turbine reliability. Early blade failures typically occur at manufacturing defects. There is a need to understand how to quantify, disposition, and mitigate manufacturing defects to protect the current wind turbine fleet, and for the future. This report is an overview of the needs, approaches, and strategies for addressing the effect of defects in wind turbine blades. The overall goal is to provide the wind turbine industry with a hierarchical procedure for addressing blade manufacturing defects relative to wind turbine reliability.
Excerpt: A. Reliability Issues
Reliability is an inherent characteristic of a system, therefore it should be considered a design parameter and its assessment should be performed as an integral part of the design process. In many low cost composite manufacturing instances reliability is performed as a subjective and qualitative step rather than incorporating a quantifiable reliability metric. In order to achieve a truly reliable system, quantitative terms must be allied. For simplicity, the reliability of a wind turbine can be divided into two components; design/manufacturing and in‐life evaluation. The first step in developing improved reliability is to establish which components of the system are prone to failure. According to a study performed by Institut für Solare Energieversorgungstechnik (ISET) on historical wind farm operating data, blades and rotors result in approximately 12% of a wind turbine down time (Figure 6) (7).
In 2008, a preliminary survey of wind turbine farm operators was performed by Sandia National Laboratory. Data collected from five wind farms showed that approximately 7% of all wind turbines blades had to be replaced. If one blade failed on a turbine at a time and each turbine has three blades, it can be assumed that approximately 20% of all turbines experienced a blade failure resulting in down-time. The operators reported that the leading causes of failure where manufacturing defects and lighting strikes (7). However, based on discussions with a number of major wind turbine OEMs and blade manufactures this number is probably higher (8).
Manufacturing quality is a critical issue for improved reliability. As recently as February 2010 Suzlon Energy Ltd., the world’s fifth leading wind turbine manufacturer, announced a retrofit program to resolve blade cracking issues discovered during the operations of some of its S88 turbines in the United States. The six‐month retrofit program will be carried out by maintaining a rolling stock of temporary replacement blades to minimize the downtime for operational turbines. The cost of the retrofit program is estimated to be $25 million (9). Problems such as these are not exclusive to the wind energy industry. Growing use of composites in the aerospace industry, for example, has led to similar problems. In August of 2009 a Boeing supplier halted manufacturing the barreled pieces of the 787's mid‐section because of a problem in the manufacturing process. This manufacturing problem resulted in microscopic wrinkles in structural stringer supports along the length of the airframe. Boeing had to develop a patch in order to repair the existing plane sections. Subsequently newer sections will be made utilizing a different manufacturing process (10).
Large blades are likely to use the heaviest possible reinforcing fabrics or prepreg ply thickness to achieve manufacturing efficiency. Increases in fabric weight—and therefore thickness—may affect basic in‐plane properties, delamination, and problems associated with ply drops where the thickness is tapered as confirmed by previous MSUCG research (11). Moreover, thick composite laminates have an increased likelihood of hidden flaws and multiple flaws being grouped in the same local area. A number of production‐related flaws may occur in larger structures which are more easily avoided in smaller structures, and rarely appear in test coupons. Typical of these are fabric joints and overlaps where individual rolls of fabric terminate, and flaws in fabric where individual strands terminate during production of the fabric. Other factors which are more likely in larger blades include fiber waviness, large scale porosity, large resin rich areas, and resin cure variations through the thickness (5).
It has been suggested that a flaw in a 50 meter structure is just as harmful as a flaw in a 5 meter structure (12). Moreover, utilizing a Weibull Distribution to compare the strength of a large structure to the strength of a small structure shows that larger structures have a greater probability of a critical flaw. This conclusion is valid when the distribution has the form (Vlarge/Vsmall)1/m, where V is the volume and m is the Weibull parameter (13).