Latest / Red Talks / Episode 9: Predictive Maintenance: Rethinking Cable Reliability
Transcript
- 0:00SPEAKER_00: Uh what I see is a very simple recipe for success is make sure they're installed properly, protect them against extreme duty cycle, and let them go.
- 0:18SPEAKER_01: How do installation practices impact the long-term reliability of cables?
- 0:28SPEAKER_00: Installation practices are really the key to longevity because the vast majority of anomalies in cable systems exist from day one.
- 0:37SPEAKER_00: And the majority of those anomalies are installation issues.
- 0:42SPEAKER_00: Human contact with cable systems is by far the biggest risk factor to future failure.
- 0:49SPEAKER_00: So everything, in fact, we really want to focus on the quality of new installations to prevent future outages.
- 0:56SPEAKER_00: And uh there's many reasons for that, but one in particular is the fact that as an aged cable, say about five to ten years into service, it's ten times more likely to fail a second time if it's failed once.
- 1:08SPEAKER_00: So we really want to prevent all of that future escalation and exponential growth of failures by focusing on the beginning of the life of a cable system.
- 1:20SPEAKER_01: Can you discuss the role of environmental conditions in cable degradation?
- 1:27SPEAKER_00: Yeah, so environmental conditions, many people misunderstand how cable systems fail.
- 1:33SPEAKER_00: So when we're looking at uh low voltage cable systems, uh the predominant failure mechanism is just mechanical damage.
- 1:42SPEAKER_00: When we talk about medium and high voltage cable systems are which are more coaxial in structure, uh there's multiple layers of a ground layer on the outside and a conductor in the middle.
- 1:55SPEAKER_00: Uh once you get over about, well, say 5,000 volts and higher, 5 kV class and higher, the world changes and the failure mechanism is completely different.
- 2:05SPEAKER_00: So while at low voltage, say less than 1,000 volts, the primary failure mechanism is just mechanical damage and then uh leakage currents through uh conduction.
- 2:18SPEAKER_00: Um and if it's an aluminum conductor, it actually oxidizes the aluminum and it becomes more and more of an open circuit over time and eventually you have brownouts and eventually open circuit.
- 2:29SPEAKER_00: That's more on the utility side, mostly the industrial side, a lot of copper cable.
- 2:33SPEAKER_00: But when it comes to medium and high voltage cable systems, the the failure mechanism is associated with what we call partial discharge.
- 2:42SPEAKER_00: It's a micro erosion or micro arcing process that erodes insulation over time.
- 2:48SPEAKER_00: And it's not a continuous erosion process.
- 2:51SPEAKER_00: In the vast majority of cases for modern cable systems, it's an intermittent erosion process.
- 2:57SPEAKER_00: So it's not growing all the time, it only grows during uh extreme environmental, or you could call it environmental conditions, but it's not say the moisture in the ground or something like that.
- 3:10SPEAKER_00: It's actually uh the extreme voltage, so transients, microsecond to millisecond bursts of voltage uh that occur every day on power systems that turn the defect on and turn it off, turn on, turn off.
- 3:23SPEAKER_00: And over time, that erosion, that intermittent erosion, uh will eat through all the material and eventually cause partial discharge uh continuous at the operating voltage and then eventual failure.
- 3:37SPEAKER_01: Can you share examples where predictive maintenance has prevented potential failures?
- 3:45SPEAKER_00: Well, for sure.
- 3:46SPEAKER_00: Uh there's many great case studies that show that if we can identify where these defects are and protect the system against extreme operating conditions.
- 3:59SPEAKER_00: So we can we can extend the life two to three times, uh what most people would say 30 to 40 years.
- 4:06SPEAKER_00: The reality is cable systems can last longer than 100 years if, again, we remove those defects, primarily those issues that existed from day one.
- 4:16SPEAKER_00: And ideally, we would get those out of the system from day one.
- 4:19SPEAKER_00: However, if we didn't do that with an effective uh technology to scan the asset, we can do that scan later in life.
- 4:30SPEAKER_00: Uh that scan in particular has a detailed specification, what we call an offline 50 or 60 hertz partial discharge test with 5 PC sensitivity.
- 4:39SPEAKER_00: That is a direct comparison to the cable and accessory manufacturer's quality control standard that they use for all uh production and design testing.
- 4:51SPEAKER_00: So if we can repeat that process in the field on a meter by meter, foot by foot basis and find these issues that are substandard, and then once we remove those defects, protect the system against extreme conditions, primarily voltage, extreme voltage, which means you put surgery arrestors on the system, and then secondarily uh protect them against extreme loading, both of which we can control through design and operation practices.
- 5:19SPEAKER_01: What are the most common failure modes observed in high and medium voltage power cables?
- 5:26SPEAKER_00: Yeah, the most common uh failure modes for medium and high voltage cable systems, uh, as I mentioned earlier, are associated with partial discharge and they are virtually mechanical damage, usually at a very small level.
- 5:41SPEAKER_00: We're talking about a few mils, very uh small, sometimes difficult to see, but definitely still in the visual range of defects that occur in the joints and terminations.
- 5:54SPEAKER_00: It's about 1% of cable system segments, about 4% of splices, and 4% of terminations that get installed incorrectly from day one.
- 6:04SPEAKER_00: And if you look at statistically, just about 40% of new installs will have at least one substandard component, meaning the component no longer meets the minimum expected performance that where it as it left the factory.
- 6:17SPEAKER_00: So some of us have damaged it along the way, shipping, handling, installing.
- 6:23SPEAKER_00: And this is what is damaging cable systems.
- 6:26SPEAKER_00: Now, there are things that can happen over time to the cable, but the vast majority of the first failure in a system is caused most often by a defect that existed from day one.
- 6:40SPEAKER_00: Uh but there are, let's just say, phenomena that can happen over time, but those are usually only derived, or they're the the origin story of even defects that can grow over time most often have this origin story from day one.
- 6:58SPEAKER_00: So, really, that's what we want to focus on the existing defects uh that are caused by either manufacturing process or human interaction from day one.
- 7:12SPEAKER_01: Water trees one of those silent killers.
- 7:14SPEAKER_01: How does it form?
- 7:16SPEAKER_01: And how long can it go unnoticed?
- 7:20SPEAKER_00: Yeah, well, water trees are perhaps one of if someone studies cable at medium and high voltage cables, it's almost impossible not to see water trees mentioned in so-called degradation mechanisms.
- 7:36SPEAKER_00: And the reality is water doesn't fail cable.
- 7:39SPEAKER_00: Water trees don't fail cable.
- 7:41SPEAKER_00: Uh, it's a phenomenon called electrical treeing that actually fails the cable.
- 7:46SPEAKER_00: Water trees, the only way you get a water tree is if there's an existing anomaly.
- 7:52SPEAKER_00: See, water is a polar molecule and it's attracted to higher electric stress.
- 7:57SPEAKER_00: So if I have a manufacturing, a manufacturer's defect in the system that creates a little higher stress in a certain area, it'll draw the water molecule to it.
- 8:07SPEAKER_00: And then as the electric field is distributed throughout that insulation, it it draws a higher concentration of moisture to that area.
- 8:14SPEAKER_00: And when you have a wet, dry interface, so an area of high moisture and an area of low moisture, you get a higher stress in the drier part of the insulation.
- 8:27SPEAKER_00: And when that happens, if by chance a perfect storm scenario where you have this plume of moisture inside the cable insulation, originally caused by a stress enhancement from a manufacturing defect, when you have the stress, higher stress in that dry part, if there happens to be a secondary manufacturing defect, so perfect storm scenario, in that location, you can achieve stresses in that 10 to 12 kb per millimeter range, which is where you can create new defects.
- 8:57SPEAKER_00: So, under transient conditions, if you have this little dry area of the insulation that's taking all the stress, now you have the scenario, this perfect scenario to create a new defects called an electrical tree.
- 9:09SPEAKER_00: And the electrical tree will grow intermittently.
- 9:12SPEAKER_00: Uh, and it actually takes uh a very long time mostly uh to fail because again, it's only growing when a microsecond or millisecond transient occurs.
- 9:22SPEAKER_00: So, this whole idea of water treeing, there are trillions and trillions of water trees out there in every age cable that I've ever dissected, and I've dissected many, and we've um I've had the privilege of not only not only overseeing hundreds of thousands of cable assessments on new and aged aged systems, but also thousands of dissections.
- 9:47SPEAKER_00: Um, and I can tell you, I have yet to extract an aged cable that didn't have water treeing in it.
- 9:54SPEAKER_00: So, water trees are not the problem, and random moisture is not the problem.
- 9:59SPEAKER_00: It's concentrated moisture in this unique scenario, and the likelihood of that happening is actually quite rare.
- 10:07SPEAKER_00: And if we address the quality control process in the beginning of the life of the asset, these water trees generally are not an issue.
- 10:16SPEAKER_00: So uh it is part of the story, but like every good myth, there's some truth and a lot of confusion.
- 10:24SPEAKER_00: And as it relates, if we can very use this very simple recipe, scan these systems from day one, make sure that they're installed correctly, and then protect them against extreme duty cycle, and perhaps throw a couple more specifications into a cable systems design, such as shear bolt connectors to prevent poor connections and eliminating, say, constant force springs on the shields over a uh over a splice.
- 10:53SPEAKER_00: A couple things like this, if we can tune the specification up on the design side, commission it, make sure it doesn't have defects, fix all those issues, and then protect it against extreme duty cycle, there's no reason why these assets uh shouldn't last uh an 80, 90, 100 years.
- 11:12SPEAKER_00: And ultimately, uh what I see is a very simple recipe for success is make sure they're installed properly, protect them against extreme duty cycle, and let them go.
- 11:23SPEAKER_00: You shouldn't have to touch that asset at least for 20 years.
- 11:26SPEAKER_00: So this taking it out every couple years to see if it's still good is ridiculous.
- 11:30SPEAKER_00: Unless we have an extreme duty cycle scenario, which in many cases we know.
- 11:36SPEAKER_00: If there was an arrestor that failed, obviously the voltage was too high.
- 11:40SPEAKER_00: Now we probably should re-baseline that asset.
- 11:42SPEAKER_00: If you've overloaded the system, now it's time to take another look.
- 11:46SPEAKER_00: But those are clear indications of when we need to take another look.
- 11:50SPEAKER_00: And the idea of just continual looking at assets uh is really a waste of time.
- 11:58SPEAKER_01: Red Talks features voices driving safety, reliability, and innovation forward in the industrial world.
- 12:05SPEAKER_01: If your work challenges the status quo, we'd like to hear from you.
- 12:09SPEAKER_01: Learn more at redtalk.comslashabout.