Getting experts to agree on most things is challenging at best, and often, downright impossible. Getting transformer experts to agree might even be more challenging. Especially when it relates to transformer failures.
Despite the fact that transformers have been around for nearly a century with varying opinions on transformer design and maintenance, most experts agree that not all transformer failures are alike. And also, that the life of the insulation or paper is the primary limiting factor towards the life of the transformer.
What are the risk factors when it comes to transformer life or the lack thereof?
Condition-based risk is simply the fact that as your transformer’s condition gets worse, the greater the risk of unplanned transformer failure.
On the face of it all, that seems obvious enough. Too often, the condition is not something in which we have paid all that much attention. While transformer oil testing has been around for many years, amazingly enough, a large number of companies test only because their insurance providers require it.
What do you do with the tests? What conditions do oil tests uncover?
A simple Karl Fisher test indicates the amount of moisture in the transformer’s oil, and from that, the determination of the amount of moisture in the paper. Moisture is a leading cause of aging and potential transformer failure once the moisture builds up to the point when a wet transformer is rapidly ailing and at-risk. Gasses, acid and heat are all additional factors affecting the condition of the unit. If these potential problems are not addressed, the condition will deteriorate more rapidly.
To this day, “Do Nothing” is the biggest competitor for most testing and maintenance companies.
While most companies plan their end-of-life strategies for equipment replacement, few plan for transformer end-of-life in the same way. Consider that many manufacturers of specialty units are months and even years out on delivery and that transportation and installation times are unpredictable. Then, it is no wonder why so many people are caught in a bind when a critical unit goes down. In one case study, we estimated that the potential down time for a company with 17 locations in the U.S. could be reduced from a possible six to twelve weeks to two days. Consider the cost of that downtime.
Even worse, the original plan was to replace as needed, rather than to plan the end-of-life process across all locations with a “rolling spare” solution. This would save the company hundreds of thousands of dollars in replacement costs alone, not to mention the loss of production saved.
When it comes to catastrophic transformer failures, there are risks that cannot be calculated monetarily. An oil spill from a failed unit may well be a hazardous chemical spill resulting in fines, delays and EPA-type scrutinizing — something that could have been avoided.
In 2012 a transformer failure in Florida led to the death of two people in the vicinity. What’s even more unfortunate is that it might have been avoided had the risks been considered. In a case still pending in the courts, the fire from the transformer failure resulted in the loss of a complete production line. What was a $250K transformer loss is now also a $19M court case trying to affix blame. Failure risks are seldom considered when it comes to developing operational and capital budgets. They usually lead to heads rolling once the corporate leaders realize that some body along the way dropped the ball. Be rest-assured, the insurance industry and risk management people are taking a much harder look at failure risks today.
Finally, what does the transformer power? If it is a critical application that would lead to a highly valued production line being shut down, how can you monetize that risk? Recently, in a conference call with a specialty metals processing company, it was discovered that a shutdown of its main furnace unit in any of its 17 plants would lead to a $1M per day loss of production. I can assure you that these critical units are being tested, maintained and monitored at a level commensurate with their application risk.
Sadly, our industry has become more dis-integrated than ever before when it comes to supporting Intelligent Transformer Management. Between manufacturers, testing service, contractors and suppliers, we have taken an “Our Way is Best” approach.
Why would manufacturers want to extend the life of transformers when they are in the business of selling transformers? Ask most people: What is the single best testing method for condition monitoring to avoid transformer failures? The answer will likely depend on whether they provide oil testing or electrical testing. What is the best way to eliminate moisture in a wet transformer? It depends on whether the person you question owns an oil processing rig or on-line dryer. And the customer is always served more poorly when an industry is serving in a dis-integrated manner.
The best approach is an integrated one where electrical and chemical testing is supplemented by good mechanical inspections. What’s more, standards for both testing and preventative maintenance are established in advance. Amazingly, this is the exception and not the rule when it comes to transformer life extension. So what should an integrated approach look like?
First, it should start with an Impact Assessment.
What should the corporate standards be? We can attest to the fact that few companies highly energy-dependent have developed a Corporate Standards program for transformer reliability. There are testing standards developed by IEEE and companies like ours. We believe that our MaxLife philosophy is the best way to determine standards but would gladly accept any “good” standard. Adopting corporate standards are better than allowing each plant location to determine its own based on how much money it wants to save on this year’s maintenance budget.
In the same way, one should pre-determine the “Best Practices” for maintenance standards. When is acid too high? When is moisture too high? What happened as a result of the latest spike in gassing? What should you do about it?
We believe it’s best to standardize, test and execute a maintenance plan according to those standards. And for those critical units, is there a corporate plan in place to share that risk, reduce the cost of replacement and make informed long-term capital plans? Why not? The cost of not planning is much greater than the cost of planning.
At the site level, the criticality of each unit, and the availability of a backup plan, should be a part of the overall corporate plan to avoid transformer failures. One site with the ability to share a load over a like unit even overloading for a period, may be the most cost-effective solution for that site. But a site with only one unit, and not load-sharing capability, has a different need. And finally, the age and condition of each unit should be considered when developing a standardized testing and maintenance program.
We have previously mentioned the condition risk that relates directly to a condition assessment of individual units. Transformers made over the last decade may have the same “rating” as units made as long ago as the ‘70s. We can assure you that they are not the same as it relates to the standards. Tolerances have become very tight with less copper, steel, insulation and oil being used today, due in part to better design modeling capability.
Transformers were overbuilt in the past and are being built to spec today. The result is that the condition of a 1976-built 4mVa unit versus a 2012 4mVa unit may give us a false sense of peace of mind. That 1976 unit is still running and the 2012 unit with the same condition factors just failed. It is happening more frequently today. This means that today it take much more brainpower to develop a good set of standards to avoid unplanned outages.
Why test? What tests should be standard?
When it comes to oil analysis, the chart above is a good indicator of what tests can tell you and which ones need to be run. Of course, not all units need a CriticalPack of tests, but the chart is a good guideline. Most experts also agree that oil testing can give you the best chance of discovering the condition of your units, with about 75 percent accuracy. The more accuracy you need, the more you come to rely on additional testing like IR and electrical testing. But in most instances this 75 percent is a good basis for developing a good testing and maintenance program.
The cost and difficulty in obtaining additional electrical testing may not be worth the additional accuracy since shutting down a unit for electrical testing is not usually convenient or even possible for most critical units. If the oil testing indicates a serious enough condition, the decision for further testing can be made.
Where do you house all this data and trending analysis?
You would be amazed at how often test results are filed away. They are never correlated or trended over time, or worse yet, never used to develop a maintenance program. We have Asset Management programs for every conceivable piece of production equipment, but let the file cabinet be the Asset Management process for transformers. A good testing program is only as good as the access to the data. Access must be simple, clear and reliable to prevent transformer failures.
How much simpler can it be than “Acceptable,” “Questionable” (test again and or monitor) or “Unacceptable” (do something)? The question becomes: Whose standards are you going to use? A Transformer Dashboard is the simplest, most robust and cost-effective way to manage these critical productive assets. Whether the fleet is reviewed on a site-by-site basis or across all facilities as a fleet, a powerful and effective tool in Intelligent Transformer Management is what you do with the data. Dashboards are used everywhere, yet our industry has been one of the last to develop and apply the tool. Just as the fact that not all testing companies are alike, not all spreadsheets are the Dashboard.
Finally, what are some of the common misconceptions about transformer failures?
We have addressed the most common and potentially disastrous misconceptions throughout the paper. In the long run the biggest misconception is that transformers need not be maintained. Unlike rotating equipment, they have no moving parts unless equipped with a Load Tap Changer. The truth is: A well-maintained transformer could last as long as 50 years, if not longer with no transformer failures. Given the industry standard of a 17-year average life expectancy, a half-century of reliable, uninterrupted and low-cost power seems like a great return on a little Intelligent Transformer Management.
This article was previously published in the Reliable Plant 2013 Conference Proceedings.
By Alan Ross, SD Myers, Inc.
Don’t miss insightful case studies from these industry leaders…