Electrical Reliability

You may or may not have heard about electronic marshalling. It's a new development with the DeltaV automation system. Emerson's DeltaV team has developed a new whitepaper, Electronic Marshalling Robustness, to describe this technology and its impact on overall system availability. If you're not familiar with DCS marshalling panels to see the traditional method of wire marshalling, here's a Google Image search showing pictures.

The basic purpose of a marshalling cabinet is to collect all the wires coming in from devices throughout the process, and then cross wire them over to the automation system's I/O terminations. Historically, this has been done to separate the installation efforts of field wiring, from control system wiring. These activities often occur at different stages in a project. With electronic marshalling, this cross-wiring process is eliminated.

The whitepaper describes two new components, characterization modules (CHARMs) and CHARM I/O cards (CIOCs). Each CHARM is a single-channel of I/O for the various I/O types and voltage options including digital input (DI), digital output (DO), HART analog input (AI), HART analog output (AO). Each CHARM has current limiting circuitry to prevent wiring faults such as short circuits and ground faults. Each CHARM is designed to fail open to isolate any voltage or current problems from neighboring CHARMs.

For the field loop with which an individual CHARM connects, it provides signal conditioning, analog to digital (A/D) conversion, loop power, line fault protection, field wiring disconnect capability, and a HART modem (for AI and AO CHARMs). Since CHARMs provide single channel integrity, their theoretical mean time to fail (MTTF) is higher than an I/O card that has multiple I/O channels per card. The whitepaper shows a multiplier of increased reliability ranging from 7.53x for an AO HART CHARM to 3.95x for a DI 24V dry contact CHARM.

Another important reliability measure, mean time between failure (MTBF) which is the average amount of time a system will run between failures. The measure is based upon the hardware components and their respective MTTFs. The whitepaper shares a comparison of similar systems done with classic I/O versus electronic marshalling I/O. It concludes that the electronic marshalling approach has higher availability based on the MTTF differential.

The individual CHARMS connect into redundant CHARM I/O cards, which in turn connect to the redundant network where the DeltaV controller resides in a traditional star network topology. They are segregated through firewalled switches. The whitepaper shows this topology with a diagram.

There's more including how CHARMs are physically keyed to prevent different types being accidently inserted in a maintenance situation and how these can be reset if different types of I/O are required over time.

Overall, the whitepaper provides a technical understanding of how availability is increased to augment other papers on why this electronic marshalling approach can improve project economics.

MP3 | iTunes

Update: A friend I've known for many years reminded me of the Electronic Marshalling Overview whitepaper as good background on this technology. There's also a YouTube video which is worth at least a thousand words.

Storms blew through Austin the week before last knocking out power to our building. It reminded me of the criticality of reliable power to conduct business in this information age. If you're responsible for maintaining the power distribution facility at your plant, you know how critical it is to avoid unplanned power outages.

I caught up with Emerson's Cliff Kirby, a manager in the Electrical Reliability Services business. Cliff described some advances to their partial discharge (PD) testing and monitoring services that have expanded to include switchgear.

Now, for those not steeped in an electrical engineering background, partial discharges are small electrical sparks that occur within a cable's insulation, or on the surface of the insulation of medium and high voltage electrical equipment. These sparks result in the electrical breakdown of a small portion of the insulation surface or in an air pocket within the insulation. Over time, these partial discharges will erode the insulation and can result in a complete breakdown.

According to the National Fire Protection Association (NFPA 70B), the leading cause of electrical failures is insulation breakdown. The National Electrical Code (NEC) states that these partial discharges are the first indication of insulation deterioration. Research from the Recommended Practice for the Design of Reliable Industrial & Commercial Power Systems (IEEE Gold Book), Table 36, indicates that cables, switchgear, and transformers suffer the greatest losses from insulation failure.

Cliff indicated that it's impractical for most process manufacturers to de-energize their power system to perform testing. He noted that partial discharge testing and monitoring service could be performed online (while the electrical equipment is energized.) There are a lot of factors to consider in performing partial discharge testing including whether it's practical to de-energize the system, the age of your electrical assets, the material composition of the insulation, weather conditions, etc. Online testing can range from the use of a simple handheld detection unit, to periodic testing with special capacitive or high-frequency current transformers (HFCT) sensors, to continuous monitoring for the most critical or hard-to-access electrical assets.

As implied by the term "offline," this type of partial discharge testing can only be performed when the system is de-energized-such as when the installation is new and has not been released to operation or during a plant turnaround.

In the United States, more than ten industry standards provide information about field-testing medium voltage cables and components. The IEEE 400-2001 standard defines six tests and their advantages and disadvantages. One of the well-established tests, DC Hipot testing, may cause damage and premature failure in certain types of cables including EPR and XLPE cables.

In addition to online and offline PD testing, Cliff listed some other tests that the electrical reliability team performs: Ultrasonics, Tan Delta (dissipation factor testing), and Very Low Frequency (VLF) testing. Every plant has unique circumstances and selecting the right mix of test methods and technologies helps provide the diagnostic information necessary to analyze the data to spot problems well before an unplanned shutdown occurs.

These tests are usually incorporated into an ongoing program to help improve the maintenance plan and length of service of the electrical assets.

MP3 | iTunes

Update: I updated the links above for the Electrical Reliability and Partial Discharge links to the new Emerson website location.

In my days as a young automation engineer putting in power, control, and safety systems on offshore oil and gas platforms, I had the "opportunity" to see improperly terminated motor leads burn up during a startup after power was supplied to them from a variable speed drive. So, it was with great interest that I listened to a presentation during the Emerson Exchange by Wally Vahlstrom and Cliff Kirby in Emerson's Electrical Reliability Services organization. I could have used their expertise back then to prevent that sinking feeling I had when I smelled smoke.

Their presentation, Early Detection of MV Cable Problems Improves Overall System Reliability, described how failures can occur and steps to diagnose impending failures.

Cable terminations and splices are the area where most prone to deterioration and failure since these are typically assembled by hand. The junctions account for 80 percent of the failures. Typical problems include nicked insulation, incorrectly connected or no drain wire, physical abuse, and environmental contamination-all of which can produce partial discharge (PD). Also, voltage transients caused by lightning and other sources, and manufacturing defects can create reliability problems.

Cables themselves can also fail caused by many things including manufacturing defects, damage caused by installation or physical abuse, metallic shield corrosion, water migration, and even cable test methods like DC Hi-Pot methods which can damage older cables. Typically, the cable will pass the test, but fail after AC power is reapplied after some period. There are many suspected reasons for this but one may be that 'space charges' develop in the insulation during application of the DC test voltage.

Wally discussed a form of deterioration known as water trees found in extruded dielectric cables. These trees are water-filled micro channels that develop in the insulation of cables operating in a wet environment. The patterns that form resemble trees that have lost their leaves. Water trees can continue to grow under operating voltage until they bridge the insulation. This often leads to cable failure.

Cliff discussed some of the US standards and guides for testing cables in the field. IEEE 400 warns against testing the cable using DC Hi-Pot methods on older medium voltage cables, especially in wet environments because it accelerates failure. Other test methods described by the IEEE 400 standard include AC Hi-Pot, Partial Discharge, Very Low Frequency (VLF), Dissipation Factor (Tan delta) and Oscillating Wave (OSW).

The Electrical Reliability Services team uses on-line partial discharge detection methods to test the reliability of the cable system. It is the only test of the ones mentioned that can be performed while the cable is energized and in service. This testing method is a non-destructive, non-invasive predictive maintenance tool that assesses aging cables. This test is also used to test for workmanship in new cable installations, given the 80% failures occurring around the handmade terminations and splices. Ah yes, this is what triggered my memory of those smoking motor terminations!

A spectrum analyzer, RF analyzer, and U-shaped sensor are used to identify partial discharge. This testing can see about 500 feet each way down a cable.

Cliff showed some installations with corrosion in other areas outside of the cables including medium voltage switchgear. Typically, this is caused by non-operational space heaters in the switchgear. These space heaters prevent condensation that causes this corrosion to occur.

Cliff recommends a site assessment be done which can be performed over time. What to assess should be based on criticality, past failure rates, and environmental conditions to prioritize how and where the partial discharge testing is done.

Update: I've removed the picture and associated text for the picture I incorrectly attributed to IEEE.