How to Select a Light Sensor for a Switchgear

maiza; is the light sensor to light the equipment room at night? perry

no..the light sensor i mean here is used for catching a light..it need to be applied at switchgear as a part of protection device..

Maiza are you thinking of some kind of light detector to catch a warning light on an existing protection relay device? There are two separate issued involved here. P&C equipment must met very stringent criteria in order to be approved for the intended use. ANY modification in whatsoever form must meet a similar approval scheme. Secondly, depending on where this P&C equipment is mounted you may have issues with inductive pick up from the energized power cables or even the control wiring. You can't expect to install and use ordinary consumer type equipment in a P&C environment. It could malfunction and give erroneous indications. It becomes debatable whether it is worse to signal a fault that doesn't exist or fail to reliably detect a true fault. If this is a school project what criteria il be applied to judge the suitability of your project?

dear elnav..

i'm sorry i dont understand what is P&C..however, i really appreciate your comment.you are right, it is a light detector..is light sensor is not similar with light detector?..there is no criteria given. thats why i need some guide on how i can judge where there it can work as protection device or not..in my opinion, one criteria needed is the light detector distance in sensing the light..

Hello Maiza, P&C is the short term description used in the industry to denote Protection and Control which is apparently the focus of your school assignment. Yes a light sensor and light detector is essentially the same. However final usage is often the determnining factor is choice of words and language. A light sensor usually refer to a photocell as used in an automatic light used for area security. It comes on at night and turns off in the daytime.

A light detector usually refers to a photosensitive semi conductor used as a switch in some manner inside automatic equipment. These are also found in opto couplers. Your reference to "protection device" suggests this is your application. But what sort of protection are we talking about?

Your name suggest you may be located in some part of the world other than North America. I do not know what regulations prevail in your juristictions. However some common sense itme are universal in applications. Inside any switch yard you must be aware of inductive pick-up as a possible source of interference. Your circuit must be designed to be immune to the effects of inductive pick up of stray voltages and harmful current flow. Is this intended as protection for people working in the switch gear area? Or is this some kind of detector of light (signals) that invokes electrical system protection. If the latter, it is much more critical to avoid false triggering from spurious signals induced by inductive coupling. When dealing with electrical system protection you must be aware of whatcould happen if the detector circuit is triggered at the wrong time or fails to trip when it should.

I won't do the work for you but I can point you in the right direction as to what general principles to look for etc.

I worked as a P&C engineering tech for four years. Previously I had worked as the Quality Assurance manager for a SCADA equipment manufacturer. During my employment as a P&C engineering technician we built a 2 megawatt transformer station and my partner and I hand built a complete SCADA control system from spares because the manufacture was unable to produce their system on schedule. Since the transformer station was intended to be remotely operated via a SCADA system it was imparative we install something.

My partner and I also spent our time installing SCADA gear on something like 160 remotely operated switch gears of 27.6 kW distribution lines. We used opto couplers to isolate circuits from inductive pick-up and from passing on current flow across long signal lines.

Because of the real risk of doing severe damage if a circuit malfunctions, fail safe design must be used in all protection applications..

Hello elnav,

GA to you for careful, considered advice!

Take care and happy new year..........

dear elnav

wow its really complicated..huhu..there are some electrical term i cant understand.. but i'll search for it ..tq very muchyour comment is really interesting..

actually i just do a MATLAB simulation..nothing involves with hardware..from the papers i've read through, most of them associate the current detector with light detector in providing a signal to trip open CB in a switchgear in response to fault..current detector i always heard during my class discussion. therefore, i have some idea on how to choose it according to need,on its working principle. so this make me undestand faster when i read those papers.. however, not like light detector..

Thank you very2 much..i really enjoy your comment very much..

Maiza wrote: actually i just do a MATLAB simulation..nothing involves with hardware..from the papers i've read through, most of them associate the current detector with light detector in providing a signal to trip open CB in a switchgear in response to fault.

REPLY

Hmmm I have mixed feelings about simulations. Its only as good as the parameters input as the situation conditions. The real world has a nasty habit of introducing unforseen events you would never think of with a computer simulation.

For your project look into things like ground loop issues. and opto coupler isolation of data circuits. Consider also how data from transducers is transmitted to physical meters and detector circuits. Think about when and why you would use 0-1 ma current loops and when to use 4-20 ma current loops. What is the function of your light and why is a detector needed to indicate th light is on or off? Why not just hard wire it with an auxilliary set of contacts. Can stray light from an extraneous source caue false triggering?

I will give you an example of real world situation no one ever figured could or would happen. Our transformer station had sixteen out feeders of 600 Amps each at 27.6kV Three of them blew simultaneously, as near as we could determiine from the data logger. The protective relaying indicated 1500 amps fault current on one of these three feeders. The other feeder showed slightly less, and even less on the third. Question: What could cause a simultaneous fault across three independent feeders.

In fact the third was only about 900 pr so amps. Barely 50% over current. These feeders left the transformer station undergorund and ran a distance of 500 meters before going up a riser pole to overhead distribution lines. The reason for the underground section was to dip under the High tension 220kV feeders that supplied the transformer station from the nuclear generation stations that fed the whole area. The riser was the first accesible point. My partner and I found the insulators and a surge arrestor and cross arms blown to smithereens. Shards and splinters lay on the ground below the pole. Operations had already isolated the whole section and back fed the customers affected by the outage. No indication of any fault on the cables or lightning arrestors downline. In other worsds it looked like a phantom fault.

It took two days to unravel the most probable cause and effect. The only clue was some bits of feathers I saw on the ground. Most probable cause was a large predator bird, possibly a big hawk we had seen in the area. We theorized it was landing on the cross arm on the riser pole and got itself across two phases. The plasma arc as it shorted out the lines caused enough of a surge to cause the other feeder to trip due to the inducted fault current. Incidentally some time later I found the cables in the cable tray had ripped loose all the cable fasteners in the steel racks. These were the cables from the actual CB to the underground ducts outdoors. The cable racks were one story below the switchgear and we only inspected the area once per month.

Each set of ties could withstand a few hundred pounds tension. Something like a hundred sets of ties broke. Evidently the fault current created a magnetic field strong enough to cause the feeder cables to whip and flex under the influence of the magnetic field created by the fault curent.

Imagine if you can, how such strong pulses can induce apparent fault currents in adjacent cables in a sub station or switch yard. Then try to visualize these induced stray currents circulating in other fault sensing circuits and leading to the protective relaying. Our engineers eventually calculated different parameters into our protective system to prevent a recurrence.

Perhaps this kind of thing goes beyond the bare minimum your instructor is expecting but if you do it right, you will get extra marks for getting it right and better.

Good luck

elnav

What is the function of your light and why is a detector needed to indicate th light is on or off?

maiza

based on what i understand, the fault esp arcing fault will produce a light..therefore, the arcing light will be the input to trip open the CB. the arcing fault will be sense by the light detector then it will go through A/D converter to produce a digital signal that will be fed to the CB..in other word, the light detector here is used to sense the arcing light in order to provide a tripping signal to CB..(but it only true for an enclosed switchgear..)

thank you very much..your comments are really helpfull

It sounds like you have cause and effect reversed. How will you know where an arcing fault is going to occur? Your description suggest you must pre position the light detector at the point at which an arc wil appear.

The only place where you can accurately predict the occurence of an arc is right at the switch gear contacts. BUT.... the arc will not materialize until after the CB contact is already being tripped and move to open the circuit. So the arc developss after the fault; not as a pre-cursor.

If the intent of the exercis is to detect the presence of a fault which will cause a CB trip, you need to look at some other characteristic. Back to the drawing board.

If your enclosed switch gear is SF6 or similar, containing arc quenching gas, how do you intend to see a quenched arc? The bottles are sealed and the arc is going to be proportinal to current amplitude? How will you deal with that? Better look for other fault characteristics.

In any case, consider also the latent response time of your fault detector circuit.

How many milliseconds will the overall circuit introduce into the feed back loop.

Remember that any physical, mechanically operated contacts will have a finite period of time in which it reacts. You need to know this time as well as part of your protective circuit. How do you propose to establish that time interval?

You have some more homework to do. Have fun!

elnav

It sounds like you have cause and effect reversed. How will you know where an arcing fault is going to occur?

maiza

hehehe you are right..i've introduce my arcing and determine which CB shold trip..i've group the light detector based on 4groups..each groups have many light detector n been differentiate by the wiring and coding..the busbar single line diagram connection i've found in one manual book on switchgear..

I agree that the arc develops after the faults..but that is for the overcurrent fault phenomena..for the arcing fault, it is the arc that cause fault..is it right or i just got wrong understanding?

about the time, i've found in 1 paper state that the fastset tripping time is 3cycle..if 1cycle 60Hz, 3cycle (0.05s)..for the mechanical/contactor i've choose 80ms based on relay manual book..i've calculate the time and predicted that it will operate about 0.13s..

ermm about the time interval, is it between the main and backup CB..if it is, i've assume 4s based on one talk that i've attend..

Thank you very2 much..

Hello maiza:

You know we are not allowed to help with school work?

I did not say this right? But it all depends how far away your light source is. Most lights and light reflector/receivers are just a few cm apart.

If you can get some 'Fresnel' glass it could help with the light gathered. You can make your own by running a glass cutter back and forth across the surface and fix this to your receptor/reflector. You can also try and use glass-paper/sand-paper on plain glass. The Fresnel with be very small but it may be helpful?

I cannot help any more. But try google and light sensor.

Thank you for joining CR4. All of us here are non paid advisors and you can write in any time. A 'Member' is given slightly more respect than a non member. Simply because they have taken the time to fill out the forms and join!

It is a brilliant site, enjoy it. And, thank you for your honest start in saying you are at school!

Take care and good luck.................

I'm sorry if it is wrong..for me i just want to get some advise..really sorry then..this site is really helpfull..thank again

Hello maiza:

Hey, no problem! Good luck in whatever you choose to do!

LEDs are reversed biased when exposed to external light source. Try using aa red LED and a laser pointer. Most laser pointers operate well at 2.5V. Most LEDs will handle 12V reverse bias. I use 5V. Computer power supplies work well for this type of project.