Electrical Troubleshooting Quiz — Jan. 18, 2022

You work for an electrical services firm. Your coworker, Sandra, was sent on a particular call two days before her scheduled vacation, and this job has been transferred to you.

A multi-building manufacturing plant installed a distributed control system (DCS) for one operation that is in Building F. It is the only operation in that building, but there are multiple processes involved and they must be synchronized. Thus, the choice of a DCS to control it all.

The DCS seems to have a high frequency of control module failures, and the manufacturer is at a loss to remedy the problem. The DCS is installed per the manufacturer’s installation guide. The failures are causing missed shipping targets and missed quality targets.

Sandra met with you for a project handoff: “That was already a 10-hour day, and I had to shoehorn this project in. I didn’t have time to look at much. I met with the customer and listened, which is, of course, very important, but it is also time-consuming.”

After relating a few details, she said: “I set up a power analyzer at the service in Buildings C, D, and F. I think this is a power quality problem, so now you can go in and compare a couple of days of stored events and other data.” She concluded by noting there was a large pond out back and a building on the edge of it: “That is probably their main fire pump. I didn’t have time to check. If it is, a problem with it might be creating a voltage sag or something.”

What would be a good way to get to the bottom of things?

Grounding and bonding

Begin by looking at the installation instructions. These are normally written by people who don’t understand grounding. They see a ground connection as a point of vulnerability, as if transients can enter only that way. So, they instruct to create an “isolated” (not bonded back to the source) ground and often add in some recommendation or instruction to thread several rods together and drive the assembly down 40 or 50 feet.

None of the other building systems (e.g., power, communications) are bonded to this. Internally, bonding may be deliberately omitted for the same rationale. You can justify this setup with electrical theory if you (somehow) prove that Kirchhoff was wrong about parallel circuits. There’s also a small problem with 250.50, which is very clear about bonding all grounding electrodes together (the CMP apparently won’t argue with Kirchhoff).

And there is another reason this fails. The DCS is load equipment, not supply equipment — so it should not be grounded. It needs to be bonded to the equipment grounding jumper, which forms a path back to the main ground at the supply. Making this change puts the DCS in compliance with Article 250, Part V. Therefore, it eliminates dangerous differences of potential.

As a consequence of the isolated ground, you have the very thing the customer tried to avoid: In addition to a bunch of NEC violations, you have a conductive loop (ground loop) that collects and circulates undesired current. And you have no “drain” installed on the DCS. It’s sitting there with no way for undesired current to get back to the source except through the equipment.

Power analyzers

But this is not the whole story. Sandra was smart to set up those power analyzers. One thing you need to look for is a voltage spike that shows up in each of the three power analyzers. If you see this, then Sandra’s observation about the fire pump becomes important. Most transients are created on the load side, and these pumps typically start across the line.

If that transient got recorded, you have several options to present to this customer. If the fire marshal approves, you can install a soft starter on that pump. No doubt, restrictions will apply so that a soft-starter failure won’t mean the pump won’t start. But you could go through all of that only to find there are other large loads that create big transients. You will have to identify all large motors that start across the line before thinking about how to squelch those transients.

Before you think about it too much, see 12, 14, 16, and 18 for the types of SPDs. Make sure you understand that they go to different places in the distribution system.

If those power analyzers don’t show transients, you can rule out transients as something to address. Once the DCS is properly bonded, the main cause of module failure will no longer exist. That will have the biggest impact, so your company’s sales rep needs to sell this project first.

Let the customer know you need to go through those power analyzer files to see if they show other issues, such as voltage imbalance, high third harmonics, low power factor, or waveform distortion. If any of these issues exist, they will also have to be solved to prevent equipment damage and ensure equipment availability.