Backyard Aquaponics
http://byap.backyardmagazines.com/forum/

Detecting pump failure
http://byap.backyardmagazines.com/forum/viewtopic.php?f=8&t=12455		 Page 1 of 3
Author:  tcshad [ Apr 23rd, '12, 09:47 ]
Post subject:  Detecting pump failure
I'm looking at picking up a water and air pump to run in my system but i'm very much interested in redundancy and was wondering how people detect their pump failure. ie not power failure but just a mechanical failure of the pump.

I'm trying to design my redundancy system so if any component carks it, a redundant one will kick in and carry on.
Author:  Simo [ Apr 23rd, '12, 10:08 ]
Post subject:  Re: Detecting pump failure
Two pumps with float switches, one set to come on at a higher water level, lower pump does all the work and if it dies the water level rises and second pump starts working until you notice and replace pump one.
Author:  Charlie [ Apr 23rd, '12, 10:45 ]
Post subject:  Re: Detecting pump failure
Yea as simo said.....its really a visual thing, pumps generally die over time and reduce in flow or in some cases just straight out failure.

You could monitor current draw but it wouldnt be an accurate tell tale sign anyway because in some cases the current draw would increase as bearing etc wear but then if the impeller was wearing away then you would probably see less current draw. The point is if you have the time to test these things...you might as well just check your flow and inspect the pump itself, which you should be doing regularly anyway.

Im a maintenance planner and a big part of my job is to monitor component life and in the mining industry this can be difficult as there is so many variables.

In the situation of AP with a submursable water pump under the same continuous load you would still need a history to make even a ball park avaluation.

I think with regular maintenance and inspections you should get a good couple of years from a mid range submerable pump.
Author:  Marc d W [ Apr 23rd, '12, 10:46 ]
Post subject:  Re: Detecting pump failure
+1 Simo
Creative pumps in SA (http://www.creativepumps.com.au/work/float_switches/luise/luise_float_switches.htm8 have a special on at the moment for Float switches ($59 - reduced from $79) which is a bargin. No excuse for anyone not to have a float switch if only to prevent a pump out due to a pump failure in the sump etc.
Cheers
Marc
Author:  DunderOZ [ Apr 23rd, '12, 10:48 ]
Post subject:  Re: Detecting pump failure
[quote="tcshad"] ie not power failure but just a mechanical failure of the pump.quote]
I guess you need to detect water pressure and/or water flow, for the water pump side of things.
You can get free switches/sensors from discarded dishwashers.
Without putting too much thought into it, I would go for a mechanical sensor that Incorporated a magnet and reed switch.
This reminds me of a sensor found in vehicle brake fluid bottles.
When the container is full, the magnet, sealed and on a float, is at the top of the container. Empty, magnet at bottom. Please yourself whether you have the reed switch sensor top or bottom.
So, your pump fills the FT or GB. The delivery pipe end first empties into a small container. This container contains the magnet and has a drain, which is smaller then input flow, so container is continuously full and overflowing and also draining. If flow stops, container empties and magnet drops to bottom.

You need to interface this so that, when the power is on, and the magnet is at bottom, the fault is detected.

The air pump is the easiest. All you need to do is to place a T adaptor on your hose and plug in an ex washing machine pressure sensor.

You can also use these sensors to monitor your water pump pressure. But I don't have time to explain this right now, so hope someone else will or will do myself later. One needs to be cautious here as, if the seal leaks, you could end up with water in the electrics.

This is the type of switch you are looking for. Dishwashers can have 2 or 3. You can even adjust the sensor's pressure settings. Once you have pulled everything off you want, take it to the recyclers :thumbright:

Image

Cheers
Author:  JORGY [ Apr 23rd, '12, 12:29 ]
Post subject:  Re: Detecting pump failure
Some good ideas there DOZ :thumbleft:
Author:  DunderOZ [ Apr 23rd, '12, 16:45 ]
Post subject:  Re: Detecting pump failure
tcshad wrote:
I'm trying to design my redundancy system so if any component carks it, a redundant one will kick in and carry on.

Carrying on with previous idea - Redundancy air pump.

You need a short extension lead, shortest and cheapest - Chip as Chips.
A pressure sensor - Ex dishwasher.
A relay - dishwasher circuit board http://www.jaycar.com.au/images_uploaded/relaydrv.pdf
A plug pack - ex disused printer http://www.jaycar.com.au/images_uploaded/plugpack.pdf
A diode.
Plastic junction box.

Cut the extension lead so you have length of cord left over, that you need. Or, use the full length and simply cut into the outer shell and expose the active wire (brown).
Cut the brown wire and expose the wire at each cut/ends. Solder wires onto the relay's 'open' contact's terminals. Apply 5 min epoxy to joints to insulate. Build up epoxy with a few layers. Place relay in plastic junction box.

You have just made your 240V switch.
Now you need to be able to switch it on.

The switching voltage is a low voltage, for safety.
You now need to make a series circuit that joins - the 12V power pack outlet, the sensor and the relay's coil.

When there is no pressure, due to pump failure, the sensor contact will bridge (make contact). This switching will send the current to the relay's coil, and relay will switch 'on', switching the redundancy pump on.

Always a good idea to have a diode across relay coil to negate 'back EMF'.

The Jaycar PDF file, I have sent link for, shows a transistor in the place where you will have the 12V power pack connection. :thumbright:

You could use the same electrics for the water pump.
To sense the pressure difference in the water pump system, you need to place a T connector close to the pump outlet. Off the T, you place a length of pipe with a cap. This length of pipe needs to be vertical (very important). You have created an enclosed air space. Drill a small hole in the cap and fit a connector to join the pressure sensor hose to.
This hose needs to be long enough to allow the sensor, with electrical switch, to be above all water levels. This will help to mitigate water entry into the switch, should the diaphragm leak. Having said this, its not a danger problem as you are only using 12V at the sensor switch. :cheers:

Now that I have done all the working out, I might even make one myself, as I have all the bits in my shed :headbang:

The last dishwasher I found, on the footpath (hard rubbish), was a european brand with stainless steel skin. I got $52 for it at the recyclers. :shock:
Author:  Amph [ Apr 23rd, '12, 19:09 ]
Post subject:  Re: Detecting pump failure
Nice.. I recently got a busted F&P washing machine for a wind generator, will have to check tomorrow if there was a sensor like that on it..
Author:  kthignight24 [ Apr 24th, '12, 06:14 ]
Post subject:  Re: Detecting pump failure
Another simpler idea is to just buy 2 pumps that each supply 1/2 the power your need and run them at the same time. When one dies, you'll still have 1/2 power. Then just switch that one out.
Author:  DunderOZ [ Apr 24th, '12, 08:29 ]
Post subject:  Re: Detecting pump failure
kthignight24 wrote:
Another simpler idea is to just buy 2 pumps that each supply 1/2 the power your need and run them at the same time. When one dies, you'll still have 1/2 power. Then just switch that one out.

I like that idea. :headbang:

There are some issues to consider though -

If the water pump is on a timer, one pump may not fill the growbed during the allocated time.

With regards to the air supply. One pump may be OK in the short term, but if fault is not detected quickly (one could be away for a few days), and the fault happens during hot weather, the fish could be in trouble.
Author:  DunderOZ [ Apr 24th, '12, 12:55 ]
Post subject:  Re: Detecting pump failure
I should also add - In the dishwasher, you will also get two water pumps.

One pump will be approx 35A and the other approx 120A.

Initially, I used the smaller one to run my system, on the outside of the sump.

Can be used as the redundancy pump, if you don't mind making appropriate joints/connections.

If you find a dishwasher with an SS outer skin, you'll get enough money at the recyclers to buy the redundancy air pump.

In the IBC of aquaponics manual, one of the systems is run by the larger pump. Having seen this, I started collecting dishwashers and found they contained lots of goodies. :D
Author:  tcshad [ Apr 24th, '12, 21:07 ]
Post subject:  Re: Detecting pump failure
You certainly are one hardcore cookie dunderoz. Thanks for all the info!
Author:  vwgti87 [ Feb 8th, '14, 16:34 ]
Post subject:  Re: Detecting pump failure
WOW, great thread.

Wish I had come across it earlier when it was first started.

I had the exact same concerns as it doesn't make much sense after so much hard work and sweat to leave it all up to en electrical device to "do the right thing".

After much thought and several drawings when I was first planning out my system I realized that the easiest answer was most correct for my situation.

The plan was to run my system completely off the grid by using 2 120w 12v solar panels going into a solar charge control to charge 2 6v Deep cycle batteries. all up for that portion of the build I was out of pocket $340. I have a never ending supply of deep cycle batteries so I didn't have to pay for those. Then the plan was to run 2 12v bilge pumps off my batteries one of them runs day and night off the batteries. I hooked up an inline water pressure sensor like the ones you find in a dishwasher that runs to a normal automotive relay. When the water pressure fails the sensor kicks in and switches on the relay in turn switching on the second pump. I also have what's called a live bait aerator that runs off the batteries, but if that fails there is a "T" piece with a reducer that shoots the water with high pressure back into the tank to aerate it.. The bilge pumps draw 2A each, the aerator draws 1.2A and the sensor draws less then 1A. The 2 batteries run in series to make 12v have a combined 510Amp hours so basically it can run one pump for about 250 hours but more likely about 200 hours with out being recharged but the sun takes care of that for me every day pumping 15amps into my batteries and keeping them happy.

all in all I'm completely off the grid with my aquaponics setup and not having to pay for the power to run it makes the strawberries taste that much sweeter.


the 2 pumps plus the aerator plus the sensor and relay were about $150 for the lot.

hope this helps
Author:  DunderOZ [ Feb 11th, '14, 08:50 ]
Post subject:  Re: Detecting pump failure
Nice work vwgti87 :wave:
Author:  tom77 [ Feb 11th, '14, 09:05 ]
Post subject:  Re: Detecting pump failure
I would just run 2 water pumps on timers 30on/30off (which would supply 100% of the flow you need) if one breaks you still have 50% of the flow which should keep your system kicking over until you notice the dead pump. The only problem is if your using auto siphons or something which are flow sensitive, but not an issue with constant flood.
Page 1 of 3 All times are UTC + 8 hours
Powered by phpBB® Forum Software © phpBB Group
https://www.phpbb.com/