Power transformer monitoring device

[kelechiorbiter]

Hello all,

I am new to this forum and I have found it a very exciting and resourceful place.

A little background. I am an electrical engineer and I live and work in Nigeria. We have frequent outages in my country and sometimes it takes weeks before the utilities even know the status of a transformer. This results in people being in darkness for weeks and sometimes months. I want to help solve this problem

I am working on a project that involves real-time monitoring of power transformers. The parameters I am interested in are the operating status of the transformer (on/off), temperature of the transformer, the percentage loading of the transformer and sent the results real-time on-line. I am looking at sending it via SIM module(GPRS/EDGE/HSPDA/4G) and I want two sim modules for redundancy.

Can I get any help from this forum?

I look towards having a very engaging and intensive discussion here that will result to an actual physical device.

Thanks.

 

[BradtheRad]

Although you know to exercise caution getting close to a utility transformer, it's wise to make a project that is easy to remove at a moment's notice. Because if the electric company finds out you touched their equipment then they are likely to say you were the one who made it go bad.

operating status of the transformer (on/off)

It may be sufficient if your sensor is a coil of wire (inductor), placed close to a wire carrying high voltage and/or high Amperage.
Align your coil parallel to the other wire. It may pick up AC flux at a distance of a few inches (or even feet). Amplify its readings. Theoretically you'll have greatest sensitivity if your detector is resonant at AC mains frequency.

temperature of the transformer

Radio Shack carried a temperature sensor 271-110 (two dollars). Place it directly in contact with the transformer. To get proper comparison with ambient air, you should put a second sensor a few inches away. Make a comparison circuit which tells you how much hotter is the transformer than ambient air.

If you're willing to experiment you can make a temperature sensor from a diode or transistor. Squeeze it between your fingers. In just a few seconds it causes slight changes in the component's conductivity.

 

[kelechiorbiter]

BradtheRad, thanks for the reply and keep it coming.

Do not worry about utility. I am the utility. It a project I told the superiors that it can be solved so we can improve the community and I have been asked to go ahead.

I suggest we go very practical by listing components, procedures and methods to build the device.

I will acknowledge everybody who contributed when we succeed and maybe some of us can meet up in Nigeria for some drinks at the end of it all.

Thanks BradtheRad, let's keep up the discussion.

 

[betwixt]

Assuming you have some money allocated to the task, I would suggest a simple SMS system using a cell network SIM adapter along with a temperature sensor and a simple circuit to monitor the transformer input and output voltages. I suggest monitoring both voltages as it will help to isolate a fault as being before or inside the transformer. You will need a backup battery so it can report a fault when the line has no power and a small MCU to coordinate things. Total cost I would guess around $US 100 and the parts are easily obtained.

You may be able to monitor the voltage drop inside the transformer to assess the load on it instead of using a current measuring system, it would make the design a lot simpler and you are checking input and output voltages anyway.

Is this a single or three phase transformer, or are both used?

Brian.

 

[kelechiorbiter]

Thanks for the very helpful reply @betwixt.

Yes I already have money allocated for this project.

The transformers are all three phase transformers.

I have worked in the power industry for 10 years however when it comes to developing such a device I do not experience assembling it. Can i get recommended sensors, part lists and recommended vendors? Methods and procedures for assembling them into a single device will also help.

A simple SMS will work fine for a start.

Thanks. Please keep the replies coming.

 

[KlausST]

Hi,

What is the high voktage of the transformer and what the low voltage?
What frequency, 1 phase or three phase?

Do you know what are the most frequent errors that cause outages and how to solve them?
What about failure against earth gnd?

Klaus

 

[kelechiorbiter]

@KlausST,

33/0.415KV
11/0.415KV
33/11KV

All are three phase transformers

We are not looking at identifying the errors at this point, we just want the transformers to talk to us.

We need the transformers talking to us so we can rspond faster and correct the problems instead of leaving people in darkness for weeks.

Thanks a lot.

 

[BradtheRad]

operating status of the transformer (on/off)

Another sensing method is to leave a disconnected input on a logic gate, or fet/mosfet, or op amp, or high-gain transistor. I have seen 60 Hz square-ish pulses coming from the output of such arrangements. Sometimes it does not appear until I touch the input pin. Or other times it is sufficient to attach a short length of wire.

These easy methods detect mains hum in house wiring several feet away. So I think it should be simple to detect thousands of volts the same way.

This method disregards the normal caution not to leave input pins floating unconnected. Nevertheless it could be a good idea to add safeguards against static charge reaching sensitive devices.

 

[KlausST]

Hi,

This method disregards the normal caution not to leave input pins floating unconnected. Nevertheless it could be a good idea to add safeguards against static charge reaching sensitive devices.

I strongly recommend this, especially in the given environmemt and for the given application.

A better solution:
A CMOS input OPAMP, inverting circuit, feedbacked with a zener.
Additionally high input resistor with high voltage rating and two diodes anti parallel to GND as overvoltage protection.

With this you may even touch live wires without the circuit to be destroyed.

But at first one should decide if checking high side voltage is needed at all, maybe the presence of low side voltage is enough information.

Klaus

 

[betwixt]

Open circuit high-impedance inputs will sense the presence of voltage by capacitive coupling but not indicate whether an open circuit has occurred. If placed close to an open ended cable for example it would show a signal, even though nothing useful could be done with it. In remote locations there is also a very high risk of damage from lightning strikes as the likes. I would definitely use a hard connection and protection devices for reliability of measurement and the measuring circuit.

Being three phase introduces its own problems of course, the average pick up from a capacitive probe would be zero if everything was OK and also zero if there was a complete failure. Each phase has to be monitored individually. The best method is by current transformer as it not only demonstrates voltage is present but that is has a path to flow through but it can also be done with a simple software routine that monitors the phase of the three outputs relative to each other so that a break in any one or two shows as a phase shift error in the voltage between them. I have seen simple voltage monitors on three phase systems that show no fault even with one phase missing because of reflected voltage from a three phase motor load, a phase detector would have shown the missing phase was 'out of step'.

Brian.

 

[fourtytwo]

Hi I read this thread with interest and I guess total non-invasiveness is a requirement (no opening up the transformer).

For load estimation I would go for temperature rise using two sensors one shielded (from sun/wind) one in free air near the transformer bottom and the other glued to the case at the top. As you are the utility you have the transformer specification so you know the efficiency and cooling coefficients.

If you want to measure the voltage of an MV circuit then a standard mains transformer (that could also power the monitor circuit) would do nicely but if this is a HV distribution transformer then you need thr standard utility grade capacitive dividers.

 

[KlausST]

Hi,

I suggest monitoring both voltages as it will help to isolate a fault as being before or inside the transformer.

I have been working in the power distribution field. 20kV 3 wires / 400V 4 wires (including earted neutral), 50 Hz, 3 phase
Over the years I can not remind that I ever have seen a distribution transformer fail in a way that the high voltage was present and the low voltage was missing. --> three phase open winding fail either primary or secondary.

***
Are the HV cables below or above ground?
And the low voltage cables?

***
Sensing the HV voltages and currents is problematic ... mainly because of installation when power is applied and because of relability and safety over the years.
I'd rather think about a current measurement as a "clip on" device with the use of a HV rated tool (like those rods to operate the HV switches or to change the HV fuses). Solar and battery powered. Wireless data transfer (RF, IR, plastic fiber..)

Trasmission rate could be rather slow: an overview over the past quarter of an hour... and additional ones when an exceptional event was detected)

But for now this may be too detailed.
An overview with all the expected functions, some sketches, photos, timings, interfaces.....is how I'd start.

Klaus

 

[kelechiorbiter]

Thanks for the replies.

All parameters will be picked from the LV side for thus first project.

I will suggest we break this project into its constituent components and discuss it in details. I will need to source part lists from US,Europe or Asia as I do not have suppliers here.

Can we all start from the temperature sensor.

Function- To read the transformer oil temperature.

Required- Recommended temperature sensor consideration affordability, suitability in oil and effectiveness.

When we are done brainstorming on the temperature sensor we proceed to other required sensors then we can look at the circuit design.

Thanks to you all helping in solving a problem a world away.

- - - Updated - - -

The LV cables are generally above the ground. We do not need to sense the HV voltages and currents, I do not need it. I am thinking about the device as a clip on as a whole.

Battery powered.

SIM module for communication (2G/3G/EDGE/HSPDA/4G/4G LTE) that is what we use here in Nigeria.

And we can start with SMS functionality and then for later upgrades add sending data over the internet.

 

[kelechiorbiter]

@klausST I would like to know whether I can share some journals on this forum. Is it allowed?

 

[kelechiorbiter]

Thanks all, I do not know whether you can see this attachment.

Can we discuss this so we can move with more focus.

 

[KlausST]

Hi,

I also prefer starting with the temperature sensor.

If possible I'd avoid to put the sensor inside the transformer, because it may be hard to install without switching off the transformer. Adiitionally I see an added source of failure: (metallic) dirt, humidity, water may come inside the transformer, the sensor wiring my influence isolation situation inside the transformer.

Thus please check if a measurement at the outside of the transformer gives satisfying results.

Even here an isolation (maybe with optocouplers) may improve reliability. Especially in case of an error where unexpected high currents may flow (through earthing) ... and shift potentials.

Klaus

- - - Updated - - -

Hi,

@klausST I would like to know whether I can share some journals on this forum. Is it allowed?

It's not allowed to post copyrighted material.
But you may post a link to such documents.

Non copyrighted material may be uploaded.

Klaus

 

[kelechiorbiter]

We are i different time zones so I may reply after several hours.

Just coming back from work and I made some research on the temperature sensors.

There are contact and non-contact temperature sensors.

Since I am looking for a non-invasive solution, I will be going for non-contact temperature sensor.

Infra-red maybe.

Can I get more replies? Thanks

 

[betwixt]

Infra red will make it very complicated. A 'contact' sensor is one you can glue or clip to the outside surface of the transformer without any invasive action at all.

Brian.

 

[BradtheRad]

I once tried to get an IR sensor to respond to a hot stove element. I thought it should work because infrared is similar to heat. I had the notion of making one of those newfangled thermometers that you point at your ear and it displays the temperature.

I turned on the stovetop and it started heating up but the IR sensor did nothing. It was not until the element became red that I saw a reading from the sensor. It told me I wouldn't get far with this project.

There are storebought infrared sensors which detect a warm-blooded creature moving in the dark. These might yield a workable concept which can detect a warm transformer.

 

[c_mitra]

First you should plan a power source: Use a small coil that can inductively couple to the transformer magnetic field and produce a few volts; that voltage can be rectified and amplified and made to charge a battery so that it will work even when the power goes off.

A few turns of copper wire pasted on the outer body (which is iron sheet) should be sufficient. This will be at triple of the mains frequency and you will have little problem to rectify and amplify to feed the few mw of power needed for the electronics.

The temp sensor can also be included in this block. You will also need current and voltage transformers that can be transmitted together with a time stamp and transformer ID.

You may process the results locally (recommended) and transmit the final data.

If the power cable is underground you can still get decent reading from the ripples (current) but voltage readings will be difficult (can you try the low voltage end of the transformer?)