Monitoring a diesel water pump.

[Nigel Goodwin]

OK, I'm after a little info, if anyone has had any dealings with diesel powered water pumps?.

This is the information, as far as we know it so far.

We've been contacted to see if we can provide a solution for one of our water industry customers, they have a number of remote (Scotland, so VERY remote ) water pumping stations which in case of an electricity failure (or main pump failure) automatically switch to a diesel powered backup pump, and the water company want notifying if and when this happens. Currently they have no notification, and the first time they find out is when whatever the pump is supplying stops, after the generator has run out of fuel - which is never good on diesel engines.

Obviously, we would need to use a GMS modem to send a txt or an email, and I'd probably do it (as I often do) by connecting to a web server that runs a PHP script, that stores the information in a MYSQL database (so there's a record of what's happened), and sends one or more emails, to addresses stored in the same MYSQL system. This allows easy changes to the system, without needing to go out on site, and update the transmitters.

That's all fine, the issue is how to detect the diesel pump has fired up?.

We've no specs on the pumps, know nothing about them, and are currently waiting for then to arrange a site visit (or multiple site visits) to see what is there, and what we can use for detection.

Presumably electrical power will be no issue, as there's mains (until it goes off), and that could be used to keep the GMS batteries charged, or it could be powered from the batteries which are 'presumably' used to start the diesel pump.

One obvious (and easy) solution, would be to send a message when the mains electricity dies, but that doesn't tell them that the diesel pump is working, which I suspect is important to them?.

My thoughts are:

1) Detect electric power has gone off - send message informing of this.

2) Diesel pump has started running - send message confirming this.

3) Electricity has come back on, diesel pump has stopped, send message about this.

4) Electricity is still off, diesel pump has stopped (out of fuel?), send message about this.

So does anyone know, would such a diesel water pump have some easy way of detecting if it's running, like for example, do they have a light that shows they are running?.

I'm presuming the pumps are fairly large?.

 

[Diver300]

If the diesel pumps are in the 1 - 1000 kW range, they will almost certainly have a 12/24 V system with an alternator. Alternator voltage will tell you if they are running, and the frequency of the AC side of the rectifiers on the alternator would be an absolute indication that they are turning.

It's a well-proven method. Diesel cars made by Ford in the 1990s had an AC terminal on the alternator to run the rev counter.

There may well be a light that says the alternator is working or not - what used to be called the ignition warning light. It's also possible to use the oil pressure warning light.

Also, just about all diesel engines have a lot of fuel recirculating. A flow switch in the right place could work.

 

[alec_t]

Even if there is no alternator there must be a starter motor, so that running would provide one signal. A vibration sensor on the pump could provide another signal.

 

[Ian Rogers]

When we serviced the diesel backup generators if South Africa.
We attached a negative sense to the oil light switch. if it was high, it was running ( ie the oil light was off )
So as you already said, but its better that way because if the oil light looses pressure, problem or not running, the system needed to know.

 

[Nigel Goodwin]

Can't say I'd though of an oil pressure light, certainly an option though, as is a possible AC output from an alternator - be nice if it had some kind of socket we plug in to

 

[rjenkinsgb]

You could use a low cost vibration sensor on the PCB and bolt the whole thing to the engine / pump frame somewhere it can sense the running vibration of the diesel?

eg.

https://www.mouser.co.uk/datasheet/2/418/8/ENG_DS_LDT_with_Crimps_A1-3005079.pdf

 

[Nigel Goodwin]

You could use a low cost vibration sensor on the PCB and bolt the whole thing to the engine / pump frame somewhere it can sense the running vibration of the diesel?

eg.

https://www.mouser.co.uk/datasheet/2/418/8/ENG_DS_LDT_with_Crimps_A1-3005079.pdf

Ooh

Thinking outside the box.

 

[For The Popcorn]

The diesel engine may be air-start....

One sure way to tell if it's running, or even attempting to start unsuccessfully, would be to mount an optical sensor pointing at the shaft, with a piece of reflective tape on the shaft. This would provide a once/rev pulse and could also provide RPM, which could be useful to know if the pump is operating abnormally.

There may also be some indicator lamps and gauges that could be monitored.

You could also put an optical sensor on the main shaft. Or a hall-effect current clamp on its power cables. The failure of the main pump is the real problem.

 

[Nigel Goodwin]

The diesel engine may be air-start....

I hadn't even considered that?, but I don't think they are - next time we hear from him, I'll try and get some information.

 

[augustinetez]

I would be inclined to add a "mains power fail" and "no diesel start" indication for that 1 in 100 occasion where the backup fails.

A remote fuel tank level monitor would be a nice addition too seeing as you are going to add telemetry to the system.

 

[For The Popcorn]

Measuring water pressure may be a good idea too.

 

[schmitt trigger]

A lot depends how mission-critical a proper water flow is. The best but not cheapest option would be a flow meter, to indicate that water is actually flowing. The pump may be running but a valve may be closed, something might be clogged …minimal or no water flow is the result.

My understanding is that in nuclear power stations a coolant flow meter is mandatory, as other devices only show an indirect indication (like the diesel engine’s generator output).

 

[ZipZapOuch]

My understanding is that in nuclear power stations a coolant flow meter is mandatory, as other devices only show an indirect indication (like the diesel engine’s generator output).

Remember, you can't add too much water to a nuclear reactor.

 

[AnalogKid]

Because any electrical things that are on the engine (alternator, oil sensor, cooling water temp, whatever) might vary from one installation to the next, my first thought was a vibration sensor of some type because it is 100% electrically isolated from, and independent of, the engine itself.

ak

 

[ZipZapOuch]

Because any electrical things are on the engine (alternator, oil sensor, cooling water temp, whatever) might vary from one installation to the next, my first thought was a vibration sensor of some type because it is 100% electrically isolated from, and independent of, the engine itself.

ak

Yes, or electret microphone or temp sensor or...

 

[AnalogKid]

A temperature sensor might be fast enough depending on where it is attached. I thought abut a microphone, but we have no way of knowing the acoustic ambiences and how they vary. Overall, I think vibration is a more reliable indicator.

ak

 

[For The Popcorn]

my first thought was a vibration sensor of some type

An optical sensor and reflective tape on the shaft will be a more positive indication, at a tenth the cost of an industrial accelerometer.

I would however recommend vibration monitoring as predicitve maintenance on the main electric pump to detect and correct mechanical deficiencies before they lead to failure (vibration analysis on reciprocating machines is a much more complex topic).

 

[ZipZapOuch]

A temperature sensor might be fast enough depending on where it is attached. I thought abut a microphone, but we have no way of knowing the acoustic ambiences and how they vary. Overall, I think vibration is a more reliable indicator.

ak

Microphone actually works quite well if you can do an FFT on the audio clip. There are Android and iPhone apps for this very purpose that you can test drive before implementing. I think there's an FFT library for Arduino if you want to go that route- or Pi if that suits you.

Some cars have a resonance in the muffler that causes trouble at some rpms, I doubt a diesel will have any issues at any rpm used for pumping water.

 

[AnalogKid]

An optical sensor and reflective tape on the shaft will be a more positive indication, at a tenth the cost of an industrial accelerometer.

My experience with optical sensors in an uncontrolled, outdoor environment leads me away from them. It becomes another item on the "regularly-scheduled" maintenance list that is more like "rarely-scheduled", and often overlooked after the "real" work is done.

ak

 

[For The Popcorn]

Is the pump outdoors? That remains to be seen.

IF the pump is outdoors, a microphone will be subject to environmental noise and weather conditions.

A waterproof industrial accelerometer could be used - an ICP-type with integral signal conditioning (ICP is a registered trademark of PCB Piezotronics). Some types have a DC output proportional to broadband (overall) vibration level. An ICP-type accelerometer requires a constant current power supply of 2–5mA with a compliance voltage of 18 – 27 volts. Plan on about $100 for this type of accelerometer.

As an alternative to an optical 1/rev detector, a magnetic pickup with a small magnet glued to the pump's shaft.... or possibly 2 magnets 180° apart to maintain balance. Like the optical method, this is also a low cost solution that's simple to monitor.