control panel ghosts 1

[amazing azza]

Hello friends, have you ever had ghost-like symptoms in your controls panels? What do I mean by that? Here are some examples:

- Indicator lights (LED, 220VAC) flickering when contactors or even regular relays are activated/deactivated.

- Latching relays (for motor control) coming undone in the middle of an otherwise steady pumping run. Just turn off by themselves while steadily pumping.

- Relays driven by external sensors (ie. there is a DC loop going out to a sensor relay, which in turn activates a coil in a panel-local relay, which then participates in the control circuit as an NO/NC contact) activate suddenly in the middle of a pumping run.

Some details about my panel:
-pumps are driven by inverters under PID control (if that matters)
-most of control circuits are 220 VAC, but a few a 24 VDC
-there are a lot of relays and buttons and there are lots of wires to/from in the cable ducts
-I tried my best to separate control and power, but in some places they do intersect (for example near the contactors)
-DC and AC controls share ducts as there is simply no space to separate them
-Analog signals and wires going out to sensors use shielding and are grounded at the panel-side only
-There is only one ground at the panel and it is shared between instruments and motors (I have not yet measured the resistance to ground at the panel. Would it be worth checking?)
-DC relays use diodes on the coils, AC relays and contactors do not
-I have a surge suppressor on the phase that powers the control circuits

Have any of you ever dealt with such symptoms? What could be causing them? How can one go about isolating and diagnosing this behavior?

 

[waross]

First try to see if the issue is process related or power supply related.
Process related, or both:

-pumps are driven by inverters under PID control (if that matters)

Possibly; Have you calculated or considered forbidden frequencies?
You may be getting reflected pressure transients at forbidden frequencies that are shutting you down on high pressure.
Any fairly rapid valve movements may be generating pressure transients that are causing high pressure transients.

- Indicator lights (LED, 220VAC) flickering when contactors or even regular relays are activated/deactivated.

220 Volts? Are you at 60 Hz? Most of your devices are probably rated for the standard voltage of 240 Volts @ 60 Hz. If you are running at 220 Volts, then line voltage dips can be more of an issue.
Also your control transformer may be undersized, you may have some loose/bad connections, or both.

-There is only one ground at the panel and it is shared between instruments and motors (I have not yet measured the resistance to ground at the panel. Would it be worth checking?)

Sometimes more important than ground resistance is the resistance from the ground bus to the grounded circuit conductors or neutral bus.

How can one go about isolating and diagnosing this behavior?

Identify the devices that may be sending a shutdown signal to the control scheme and look for unexpected operation.
Check the actual voltage at the panel and the voltage rating of your devices.
Investigate the quality of the supply power. Is it subject to dips and/or transients?
Check your system for reflected high pressure pulses due to possible forbidden frequency/speed operation.
Check for anything that may be causing an abrupt change in the flow rate. Although water hammer is generally associated with abrupt stoppage of the flow an abrupt reduction in the flow rate may also cause water hammer or pressure transients.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[amazing azza]

Hi Bill, thank you so much for replying.

I have not investigated the forbidden frequencies. But the operation during cut-out is stable pumping at 3.5 barg. There are no swings in pressure prior to the contactor latch coming undone.

I am at 220 VAC (line to neutral), 50Hz.

 

[waross]

Hi Azza. For some strange reason I thought that you were located in the US.

I am at 220 VAC (line to neutral), 50Hz.

220 VAC should be good at 50 Hz.
Anecdote:
The first time I saw an issue with a pump running at a forbidden frequency was a 60 HP pump pumping straight into a fairly long discharge pipe followed by a 90 degree elbow.
The pump was VFD driven under PID control. The first indication of trouble was when a piece of the pump case about the size of my hand was blown out of the pump. The first thought was a faulty casting. The pump was replaced and a warranty claim prepared. Then the replacement pump was destroyed in the same way. That was when we started to realize that some frequencies/speeds should be avoided. You may be getting transient pressure peaks that are tripping the pressure switches but are not apparent on the instrumentation.
You may be having power issues.
Are you able to monitor the incoming supply for disturbances?
The flickering lamps may be an indication of an undersized control power transformer or of poor connections in the wiring.
Does the panel look to have been wired in a workmanlike manner?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[danw2]

What signal/event unlatches the latching relay in the motor control circuit?

 

[HCBFlash]

Yes I have. Many/most have been due to "wiring problems" particularly terminations of conductors at terminal strips (particularly where "combs" are used across multiple terminals) or multiconductor cable connectors in the main control panels/MCCs. Don't quickly rule that out.

.

(Me,,,wrong? ...aw, just fine-tuning my sarcasm!)

 

[amazing azza]

Does the panel look to have been wired in a workmanlike manner?

Hehe, I tried my best Power quality seems ok, but I will keep an eye on it.

What signal/event unlatches the latching relay in the motor control circuit?

Usually an NC pushbutton (stop)... But there is also an NC interlock (a low water pump lockout). Hmm, the latter seems like it could be the cause, as it is one of the "relays driven by external sensors", which are themselves subject to flaky behavior. I have not thought of that one before. Thank you!

Many/most have been due to "wiring problems" particularly terminations of conductors at terminal strips

Indeed... There is a lot of wiring. Some of it has been redone... a couple of times And some terminals use "combs" or equivalents. Sigh... as much as I hate doing it, I think you're right - need to go over all the connections one more time. Any tips on doing connection checks?

I will post a picture of the inside of the cabinet for you guys tomorrow.

 

[danw2]

Check what the failsafe setting, and/or the "loss-of-echo" setting is on the ultrasonic level sensor. It could be dropping to 4mA on a loss of echo and tripping the motor relay. If

 

[roydm]

Coils like solenoids or starter contactors can cause large spikes on inputs particularly where they are run in a common cable.
A 120 VAC solenoid turning off can create a spike in adjacent 120VAC input that will cause the logic to flip.

The way we got around this is to put MOVs (Metal Oxide Varistors) in parallel with the coil to absorb the spike.
For a 120 VAC coil something like a 150V MOV would do.

 

[HCBFlash]

Tips?
- a Yankee screwdriver works very well; a little speed with good feel
- a long-nosed pliers, and a bent-nose-long-nosed pliers.
- a dental inspection mirror
- a small bright flashlight; 1xAA or 1xAAA
- remember, it's conductive, firm connections you're going for NOT tight screws (yeah, we all know that, but.....)

.

(Me,,,wrong? ...aw, just fine-tuning my sarcasm!)

 

[amazing azza]

Sorry about the delay, here's the pics I promised previously.

Would there be a point to trying looking around with a scope? Or is this not something commonly done on controls panels?

 

[waross]

Well you're looking for a signal or lack thereof that may be transient. There is a chance that one of the field wiring runs is picking up a lot of noise that at times exceeds a thresh hold and causes a trip, so I can't rule out checking inputs with a scope.
Looking at the power to the flickering lights may be a good idea. Look also at any control signals to any led drivers if there are any such circuits.
Personally I would follow the following protocol:
1. Start by looking at the latching relay that is dropping out; It may be a faulty relay.
2. Look at the control circuit and identify and list all devices that may send a signal to the latching relay to drop it out.
3. Scope the input to the drop-out coil of the latching relay to see if there is any noise on the signal line.
4. Check each device on the list generated at step 2 for physical damage or wear.
5. I suspect that you may have one or more alarm relays, driven by a 4-20 ma signal. If so:
a) Look at the 4-20 ma signal coming into the relay with the scope. You can monitor the voltage across the relay. That voltage will be dependent on the current, no need to open the wiring.
If you have a problem with process pressure variations this should show up in the 4-20 ma signal. Note: the scope will show pulsating pressures but you may miss one time transients.
b) Go back and check the engineering on the loop controlling the alarm relay;
i) Add up the input impedances of all devices in the loop and make sure that the maximum allowable resistance does not exceed the maximum allowable resistance for the supply voltage.
eg. With a 24 Volt DC power supply feeding a 4-20 ma loop, the maximum resistance allowable is 24V/20 ma = 1200 Ohms.
Add up the listed resistances of all the devices and verify that you are not getting close to the 1200 Ohm maximum loop resistance.
Anecdote: I once had a problem with an alarm relay failing to operate under alarm conditions. The relay was intended to initiate a safety shutdown when the loop current reached about 19 ma. The relay was off spec and the input resistance was greater than specified. As a result, the loop resistance was too high and the maximum current that the power supply could generate into that loop was about 18 ma. The relay several times failed to respond to alarm conditions with resulting equipment damage.
6. Look at your process for conditions that may develop a transient pressure pulse.
A pump discharging into a fairly long straight discharge line may develop pressure pulses if operated at a forbidden speed. Fast acting valves and rapid changes in the flow rate may develop pressure pulses. (Think Water hammer).
If possible ramp up the speed of the pump while checking the output pressure with a fast responding pressure indicator.
Good luck
Yours
Bill

Bill
--------------------
"Why not the best?"
Jimmy Carter