Is there a way to monitor the position of open blade outdoor 15kV air switches? 1

[bdn2004]

Attached is a picture of some horizontally mounted, 15kV outdoor air switches that we are going to replace. What we would like to do is bring the status position (open/closed) of each of these switches into a SCADA system. I'm looking for an idea on a relatively easy, safe way to do this, if there is one?

The switches are gang operated by a mechanical lever. At present the position of the 3-phase group of switches is monitored by limit switches that monitor the position of the of the lever as it is rotated. The issue with that however is that sometimes one of the 3-phases of the air switch can get stuck - which gives a false reading.

The problem with mounting a proximity switch on this high structure is that we would be introducing low voltage at high elevation and surrounded by open 15kV. This would be a maintenance issue long term.

Any ideas?

 

[PHovnanian]

Who operates the lever? When the operator opens/closes the switch group, make it a part of the procedure to look up and verify the position of the phases.

Given the cluttered bus and structure between an operator on the ground and the switch blades, the addition of some colored reflective tape on the switch blades might make spotting the switch positions easier from the ground (particularly at night).

If these switches are motor operated and you want to avoid installing monitoring suitable for the system voltages, I'd install a CCTV so the system operators can examine the switch positions after a command is issued. The use of reflective tape/labels on each blade would still be a good idea.

This is stretching the state of the art (a bit), but it may be possible to find some computer vision s/w that can read the camera images and provide status inputs to SCADA. QRC codes on each switch?

 

[DTR2011]

A common method for this is Aux switches mounted at the operator handle, at the base of the switch. It sounds like this is what you have already and is problematic in your case. Likely a new switch will also have this available and this should be considered as well as other options.

Have you looked into having a fiber optic signal for the proximity sensor?

Do you have 3 phase VT's available (both sides of switch)? If you do, you can check for the loss of voltage.

 

[rcw retired EE]

My experience with this style of switch in a similar facility (aluminum plant) was the center insulator breaks free of its pin and the arm stays close when the limit switch indicates its open. We always had to visually check the switch arm position.

Most cost effective is probably web cams and lighting to remotely check them visually. This picture did a good job in daylight, good weather.

Fiber optic limit switches might be able to sense the blade position and have enough voltage isolation. But there are mounting challenges and cost of one sensor per phase adds up fast. Plus how reliable would an optic sensor be long term in that outdoor environment?

 

[bdn2004]

Visually checking the switches is the procedure. Reflective tape is a good idea, but I don't think its a real issue not being able to see the switch position. As you can tell the switches are pretty old.

I like the camera ideas, but the point is to put this into a SCADA open/close - not just to check on the position remotely...not sure how this is done without creative programming. And there are no 3-phase VT's on the load side of these switches. The switches are not motor operated.

I am going to investigate the fiber optic proximity switch idea with the Vendor. On the surface that seems like the best solution - likely relatively inexpensive, safe and something we can specify to be built with the new switch. Thanks!

 

[PHovnanian]

I like the camera ideas, but the point is to put this into a SCADA open/close - not just to check on the position remotely...not sure how this is done without creative programming.

The creative programming has already been done. Have your IT people look at some computer vision products. The reflective tape might still be a good idea. It will give the CV app some targets to look for.

Given that CV apps will have some error rate, I wouldn't trust them to provide more than a SCADA alarm for a handle/blade disagree.

 

[racookpe1978]

OK.

So try this idea - but remember, I'm NOT a switchyard nor SCADA expert, so critique carefully the following concept.

Put a highly reflective surface on the bottom of the switch "arm" so incidence light is reflected back down towards the observer. Arrange 4x light sensors (lasers ?) aimed "up" so #1 confirms positively to SCADA that the switch is "closed and locked" (switch is in the "fully shut" position) by reflecting light back down to the receiver ONLY if the switch is completely closed. Assume this is 90 degrees. Maybe 95 degrees.

#2 light is aimed so that as soon as the switch is "no longer completely shut" it reflects light back down to the receiver. Assume this is at an 85 degrees point.

#3 light is aimed upwards so that as soon as the switch arm rotates "far enough to be open but not locked open, but isn't really safe yet" This might be at the 10 or 15 degree position from fully open.

#4 light is only to confirm that the switch is in the "fully open" position. 0 degrees. Maybe at -5 degrees. As soon as the switch begins to close, this light should trip "off" to SCADA.

You'd only have to work out the logic in the SCADA display to show fully-shut (#1 = "tripped", #2 and #3 and #4 "off"), opening-but-not-fully-open (#1 "off" AND #2 OR #3 "tripped" or #4 "tripped"), and fully-open states (#4 "tripped" AND #3 "off" AND #2 "off" AND #1 "off").

The reflective surface is on the bottom of the switch, so it should not get too dirty to be effective. The switches (light transmitters and receivers) are on the ground and away from the high voltage potential, and are not themselves powered or threatened by static voltages or shocks. They can't lose "calibration" since they are mounted on unistruts or a single metal plate.

 

[davidbeach]

In other words, you're asking for something that isn't really practical. The best that seems to work is a basic position indicator that will indicate Open or Not Open. Not Open with current through the switch is generally a reasonable indication of closed. Open with current through the switch is an even better indicator of a failed switch mechanism. Proving a switch closed that can't be visually verified (if it was closed and hasn't been operated it is highly likely to still be closed) is very difficult without passing current through the switch and then verifying that it isn't heating (IR scan) more than a similar switch when known to be closed.

 

[bass1mj]

As per one of the previous replies, use voltage sensing to confirm switch position. Easy to integrate into SCADA.

 

[davidbeach]

Voltage sensing can work if the only source is always on one side of the switch. If there can be sources on both sides of the switch then voltage sensing tells nothing.

 

[waross]

It seems that the issue is not switch position, the limit switch on the operating handle provides that. The issue is detecting a switch with a broken pivot that is still closed but which may have moved enough from the normally closed position to give an open signal. If you focus on and resolve this issue, it will probably be easy to extract normal operating data. This may be solved with some variation of racookpe1978's suggestion. Infared co-axial relays are common an economical. That is an infa-red source and an infa-red receiver in a common enclosure which give a switch closure when the infa-red beam is not returned from a reflective target.
You may consider using optics to detect a broken pivot pin.

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

 

[bdn2004]

Davidbeach is right, there is always voltage present on each side of these switches. If you look at the photo, the source side is on the outside right, the load side of the switch goes down to the transformer - but is also connected to the load side of another set of switches that forms the other source. One set is either open or closed but never both closed.

The infrared ideas, which I'm thinking you mean something similar to a garage door opener...seem worth investigating except the transformer is sitting directly under these switches. We don't want guys up there either if we can help it.

 

[stevenal]

http://www.seecoswitch.com/sensorcs.htm

Here's something that can verify current through the switch to use Davidbeach's logic above. Does the switch pickup current when closed? Can currents be monitored elsewhere for change?
The cap and pin insulators are famous for separating as rcwilson spoke of.

 

[stevenal]

http://www.lindsey-usa.com/catalogs/CVMI/Lindsey-MulticoreSensor.pdf