Best way to detect if a water pump is dead heading

[brendan-petrol]

Hey Guys!

I was wondering if anyone had some good ideas on how to detect if a water pump is dead heading? I was considering using a pressure based switch, however the problem with that is I cannot find a decently priced differential pressure sensor designed for water use. I've looked at the MPX2010 line of sensors (well, the similar ones which handle higher pressure than the MPX2010) and using them with an air gap like those normal tank sensors but they specifically specify that they are designed for "dry air"... I was considering using this as a backup switch? The fact it mentions dry air makes me nervous - for just normal old water level sensing it's no dramas if it fails, but if this fails it will probably take our bore pump with it!!!

My other option was a flow based switch (which is preferable, in my mind), however I don't know the LPM rating of our bore pump which is a bit of a problem. I've heard the Doppler water flow sensors are quite good in that respect, is this correct?

A temperature based sensor would probably work, perhaps based on the fantastic DS18B20, however this will be going into drinking water. Perhaps one of those stainless steel waterproof versions? Basically the theory is that as the impellor is spinning it will heat up the water... But that's not foolproof!

Basically, over this summer our fantastic old Davey pump went belly up (On the weekend, mind you)! So we had to do some emergency plumbing and connect the house up to the bore pump for water. However, that pump has no pressure switch, so we had to keep turning it on and off in order to not dead head the pump.

Now I want to make a sensor for it which will cut the power if it detects dead heading.

Thanks, and I look forward to hearing from the lovely EEVBlog community!

Brendan

[IanB]

I'm not quite familiar with such a setup, so a bit of extra clarification would help. Am I right in supposing that your pump feeds a header tank, and the header tank has an inlet shutoff valve (like a ballcock) that shuts off the feed when the tank is full? And this causes the bore pump to run against the closed valve?

I see distance/lack of proximity might be a bit of an issue here, but could you not attach some kind of microswitch to the ballcock or float valve that turns the power to the bore pump on and off as required?

[brendan-petrol]

Basically the bore pump is normally fed into a network of underground pipes to provide garden water, and it is normally switched on when required. What I would like to do is determine when water is flowing through the pipes in order to turn the pump on and off.

The other reason I want to do this is for if our household pump dies again, we can connect the house to the bore pump and just let the controller turn it off and on.

Does this make sense?

[AlfBaz]

 Not sure if this is what you mean but I've seen similar settups for tank water on farms using a pressure swith. When some one opens a tap the line pressure drops and the pump starts. You close the tap the pump keeps pumping to a preditermined pressure and turns off. Due to leakage the line pressure may drop by itself occasionally so the pump turns on to up the pressure.

[brendan-petrol]

That's the sort of thing I am after. Do you know if they used commercial devices? Or do you know what kind of pressure switch I could look at buying to interface with an AVR?

[AlfBaz]

I have seen discrete ones built but you can just buy pumps already set up for this precise purpose.

As for the pressure switch any mechanically adjustable one will do. They have a voltage free contact and usually rated high enough for mains switching (AC3?) with almost always a normally open/normally closed (NO/NC) option. Any halfway decent electrical wholesaler (the ones that cater for leco's) should have one on the shelf for a reasonable price and most of them will come with a thread to screw straight in to a water pipe fitting

[brendan-petrol]

Yeah, unfortunately our bore water pump has no built in device to regulate pressure. !

[IanB]

Yeah, unfortunately our bore water pump has no built in device to regulate pressure. !

AlfBaz said pressure switch, not pressure regulator. The switch is wired to turn the pump on when the line pressure drops and off when the line pressure is high.

I'm not sure, but I think you have a bit of a misunderstanding about what you are trying to achieve. You need to turn the pump on when there is a water demand, and turn it off when the demand is not there. Your statement, "What I would like to do is determine when water is flowing through the pipes in order to turn the pump on and off" does not hold together. Water will not start flowing through the pipes until the pump is turned on, so you have a chicken and egg situation.

The best answer is the pressure switch as noted. When the pressure is high the pump is turned off. When the pressure is low the pump is turned on. This means that as soon as the irrigation valves are opened the pressure is let out of the system and the pump turns on. It keeps running while the irrigation valves are letting water out the other end of the system. When the irrigation valves close the pressure builds up and the pump is turned off.

From time to time the pressure will leak out of the system and the pump will turn on for a few seconds, but it will soon turn off again.

Unless you directly switch the pump on and off under the control of the irrigation timer or controller this is the best you can do.

[AlfBaz]

Neither do the comercial ones, admittedly the pumps are weak enough not to blow your plumbing. You adjust the final pressure in the line by setting the pressure at which the pump turns off. Seeing how you are going to connect it to an AVR, you can first implement contact debounce and then some simple hysteresis to prevent the pump from hunting.

For an all out electrical solution you would connect the press. switch to the gate input of a timer which in turn activates the pump

[fcb]

Is there a measurable/significant increase in pump current consumption during the condition you want to detect?

[brendan-petrol]

IanB,

Yeah, now I read back through that it doesn't exactly make sense, though from what I read of what you're saying we are talking of the same thing.

Basically I think I have the pressure side sorted using an MPX2* series sensor or a switch like the one that AlfBaz describes, but I would like a backup switch wherein if the pump is running and no water is flowing it would switch off the pump.

Does that make more sense?

AlfBaz,

I'll have a look for a pressure switch like that, thanks mate!

fcb,

I actually hadn't thought of it that way! I'll have a to have a look!

[brendan-petrol]

I forgot to mention, the reason I want this to be linked to an AVR is I want to link this device and my tank water level sensor project together if possible!

I was going to put in an RS-485 bus and link them both to a controller in my shed.

Brendan

[IanB]

Basically I think I have the pressure side sorted using an MPX2* series sensor or a switch like the one that AlfBaz describes, but I would like a backup switch wherein if the pump is running and no water is flowing it would switch off the pump.

OK, so I think you want a safety backup system. If the pump continues to run for a period of time with no water flowing and the pressure switch fails to activate, then your backup system will switch off the pump.

However, I think you have a little bit of a logic problem to overcome to make that work. Consider:

IF the pump is on AND no water flow is detected THEN cut power to the pump
ELSE

The ELSE part is the problem. If the pump is off OR no water flow is detected OR both, what do you do?

If you restore power to the pump it will start running again and you are back where you started.

If you don't restore power to the pump, then when the pressure switch tries to turn the pump on it won't start.

Also, the pump can only be running because the pressure switch was trying to turn it on. So as soon as you release the override the pump will come right back on again.

What happens in industrial situations like this is that the override safety system is a cut-out switch requiring manual reset. Often such a switch is a thermal switch inside the motor that is activated if the motor windings overheat. Presumably if this safety switch activates then something went wrong and needs investigation before returning the pump to service.

[brendan-petrol]

Exactly!

That's the only part I'm struggling with! My problem now is trying to pick a safety system. Should I go for a simple turbine flow meter? What happens if the pump tries to push too much water through? The flow limits on turbine sensors are fairly low. I still need to find the specifications for the bore pump to check the max flow...

[David_AVD]

My bore pump has a simple pressure switch on it that turns the motor on and off.

It looks like the one here: http://inspectapedia.com/water/WaterPumpAdjust.htm

[IanB]

Exactly!

That's the only part I'm struggling with! My problem now is trying to pick a safety system. Should I go for a simple turbine flow meter? What happens if the pump tries to push too much water through? The flow limits on turbine sensors are fairly low. I still need to find the specifications for the bore pump to check the max flow...

You could over design this.

How about you just mount an indicator light somewhere near or in the house where you can easily see it that tells when the pump is running? You can just look at it from time to time: hmm, the pump is running but the sprinklers aren't operating...must be a problem with the pressure switch...

If the indicator light not feasible, how about a spring operated pressure relief valve? If the pump runs dead headed the relief valve opens up and you get a water fountain in a prominent place where you can see it.

[AlfBaz]

The current sensing option woulkd be the go.
Several years ago I serviced ships that came into our local port. They were burning out expensive seals and pumps on there ballast pumps. What was happening is that they were, say pumping from one tank to another and leaving the pumps running. Once the tank was empty the pump would run dry. The seals the were using were spring loaded ceramic types.

Since the pumps already had CT's moinitoring current I installed settable current sensing relays the would stop the pumps if the current dropped below a certain value.

With current monitoring under software controll you could do lots of things, especially if you have level sensing from the tank. You could stop the pump if the laod current was what youd expect if the feild valves were closed, you could stopp the pump if the load current dropped to far indicating that it was running dry, you could even calculate flow if you accumulated enough data from your level sensor and current drawn.

As for trying to get a spec from the pump for its maximum flow, the flow is limited by the load you hang off it. A lot like the current drawn from a supply is dictated by the load. so the most likely spec would be max pressure

... how about a spring operated pressure relief valve? If the pump runs dead headed the relief valve opens up and you get a water fountain in a prominent place where you can see it.

Thats a good idea the releaf valve can just pump back to the tank

I was typing this up when you posted about over-designing... LOL