Article by David Clark, condition monitoring specialist, Bachmann Electronics, www.bachmann.info
Condition monitoring, now more than ever, plays a huge role in the reliability of wind assets. With the number of turbines coming out of warranty and the Production Tax Credit soon to die, how do you keep an asset available, productive, and reliable as it ages? The perception of the effectiveness of condition monitoring systems (CMS) has been tainted recently by a number of events. All were avoidable, but they point to a larger problem. The lack of vibration-analysis education, and application knowledge missing from the wind industry’s maintenance culture.
Halloween is over, but a lot of scary things still happen on wind farms. In the last several weeks, for instance, there has been a rash of horror stories regarding people installing incorrectly and misusing condition monitoring in wind. Companies and people that are the source of these tales are doing more harm than good to the wind industry. Here’s the scary part: all of these are avoidable. My intent here is two-fold: To shame those who are making such bad calls – for reasons that should be obvious – and to inform their victims that there is recourse. Let’s take a midnight walk through the graveyard of recent and true CMS horror stories.
Scary story: A CMS company installs a system that fails to detect a known issue.
Why it’s scary: It was discovered that the CMS company was using an incorrect sensor. Two common vibration sensors are often used in the wind industry. One is a slow-speed sensor, rated for shaft rotations down to about 12 rpm. The other is rated for use above 60 rpm. Guess what happens when you use a 60-rpm sensor on something turning at 16 rpm? Right. You miss issues and a large part of the relevant vibration signal.
As a result, the owner initially dismissed CMS wholesale. Not the analyst, not the CMS hardware or software, but the condition monitoring system. Sadly, this happens quite often. This is topic for vibration 101 and it is astonishing that a CMS company would install an incorrect sensor at the particular location, but it happens often. The situation has been reported twice in the last few weeks.
A similar version of the story has a CMS company telling an owner that vibration sensors are not appropriate for the main bearing because they cannot produce data. Having installed several thousand, I can tell you this is completely untrue.
Scary story: A CMS company tells a customer that months of data must be taken to develop trends before any meaningful information can be derived.
Why it’s scary: How many readings from a tire gauge are needed to tell tire pressure? One, or a few hundred? It’s astonishing to me that people claim that this is the case when clearly many readings are not required. This is a danger sign that indicates the analyst is not familiar with vibration, or the application, or both. If it was true, there would not be ISO vibration levels for alarms, but subjective trends based upon the analyst’s mood.
Another example is a test on a single tower in which a university instructor, a CMS manufacturer, and an analyst unfamiliar with the application, all examined the same data from a problem tower. All three provided different opinions of what was wrong in the nacelle. Only one was right. Vibration analysis is not this subjective. The variables are the analyst, the application knowledge, and that person’s experience.
Scary story: A CMS company installs a system and receives many false alarms after months of taking data.
Why it’s scary: Adjustments are common for new equipment , of course, but not wholesale “alarm as you learn” or a less fancy term…“guessing”. The situation suggests someone is unfamiliar with the application and appropriate methods to even acquire meaningful CMS data. There are several methods to avoid false alarms, most of which include training and application knowledge.
Scary story: Incorrect analyses.
Why it’s scary: There is a history of incorrect calls in wind because it is indeed a difficult application for condition monitoring. More than a decade ago, a CMS company told a turbine manufacturer that a quantity of its gearboxes would fail soon after taking readings. Two years later, when the instrumented turbines were all still running well, the manufacturer dismissed CMS as a whole. Not just the analyst and not the CMS hardware and software, but CMS entirely. This single event set back the validity of CMS a decade. That’s why these recent reports of mistakes are problematic.
A few more cases of incorrect calls cost the customer a crane call-out on a large part that did not need replacing. The only problem was that the inexperience cost the owner a crane call-out. Similar to the alarm-as-you-learn school, the expensive guess benefits no one, except the crane company.
When looking at suspected vibration readings, it was clear that the amplitudes and harmonics did not equate to severity. Why? Because there are methods and measurements used to qualify meaningful data. The analyst made an incorrect assessment of the severity and his customer paid the price. When methods and measurements are not in place, the result is a completely avoidable incorrect call. And that’s scary.
Spooky cell service
Scary story: CMS companies claim 3G wireless is used to transfer data from wind turbines to data-storage servers.
Why it’s scary: Does your cell phone work on all wind farms? If it does not, how can you reliably expect to get CMS data? The company making the claim had no installations to its credit, just theory. If they had installs, they would not recommend this.
Scary story: A CMS company uses non-gear-lubricant rated cables for permanent installations.
Why it’s scary: When gear lube touches these cables, it eventually eats through the insulating jacket. If you don’t mind replacing every cable in three to five years on every tower, then this isn’t scary. If a CMS company recommends a cable that is not environmentally rated, run.
Scary story: At a recent exhibition, I saw six new companies claiming that they install condition monitoring systems for wind turbines.
Why it’s scary: These companies were not installing CMS last year. In fact, I spoke with three of them and none had a single install in wind, and had none had ever been up a tower. So how does a turbine owner know who is legitimate? A few suggestions:
- A few variables are the analyst, the application knowledge, and that person’s experience. So look for those that have two things: number of installations and years of experience. That list is short.
- Some are claiming they have thousands of installations. Ask if these are permanent and installations that they have actually performed. Documented installs helps to keep everyone honest.
- Also, ask how many people on staff are certified vibration analysts. While this is no guarantee that they have wind experience or even experience with your model, at least they understand vibration.
While a vibration certification is no guarantee, neither is the certification course. Most certification courses in vibration are heavily weighted towards handheld portable vibration analysis. None of these are used in most wind cases due to application, safety, or incomparable data between conditions, and operational consistency. Certification courses also weigh heavy on axis and phase which again have little use in the wind industry.
There are many analysts but few with wind experience. In fact, a colleague was a Level 3 certified analyst with 20 years experience. He was looking at wind vibration data and seeing things he had never seen before. Also, I know of respected analysts with experience on exactly one version wind turbine.
A happy ending
CMS is capable of detecting and predicting failures and correctable conditions on all aspects of a wind-turbine drive trains. This is why CMS is used in most all mines, paper mills, refineries, power plants, and steel mills in the U.S. and has been the standard for almost 40 years.
Remember, total verified installs, number of monitored turbines, and years of experience helps to keep horror stories from happening. It also helps to keep goblins from testing you and your assets. You’ve been warned. WPE