Dry Run Protection for DC solar pump, sourcing from a borehole

[Dream_Renewables]

We are installing solar power irrigation systems for smallholding farmers in Rural Ghana,

We have 12V DC 0.18kW (max 20A) pumps, there only power source will be a voltage varying from 15-20V coming from solar panels.

The boreholes are narrow, therefore is not much room for a float switch,

The max flow is 20 Litres per minute but more typical flows would be 5-10 liters a minute

Can anyone advise on flow or pressure switch solutions (I have done a google but I don't have experience in this area and am not confident myself)

Ideally, there would be no second power source or battery needed, and the solution should be as low-cost as possible. something under £20 per unit for 20 units would be ideal.

(It's a charity project replicating 20 units - so cost is important)

let me know if i missed out any key info

[Gyro]

In terms of a pressure (loss) switch, the sort used for water level sensing in washing machines might be cheap and reliable answer. They have reasonably high current snap action switch contacts (not sure about 20A DC though) and low pressiure operation. Note that they are not intended to be filled with water on the pressure side (vapour must be ok), so would probably need to be fitted on the end of a thin rubber hose (also washing machine spare) above the output pipe.

Just an idea.


P.S. Welcome to the forum. 

[coppercone2]

I see possibilities if its very clean water, but any mechanism that allows a float switch to be like.. linear.. is going to be susceptible to corrosion and jam up.

Like I can imagine something along the lines of a teflon shaft connected to a float, kind of like a "corn dog" food, but if you fix it, it can get jammed up.. the hanging nature of a traditional float makes it a very robust mechanism.  I thought it might be useful to investigate if its possible to make a traditional float switch that works in a narrow bore, but how do you keep the shaft from binding. Longer shaft and a free bearing (like a gyroscope style, or floating style, made out of ptfe) that allows a float switch to work at a arbitrary angle with minimum resistance, made with teflon. And then connect it to a mechanism that linearizes the upward travel via ball joints to activate a switch. Or just connect the switch to the shaft with a chain that stops pressing the switch as the float shaft falls, but it might be hard to get that right. The idea is if you make it a cylinder instead of a sphere, it can be rather long, then you can get alot of boyancey in a narrow well tube. Kind of like how a locomotive piston works 

https://www.123bearing.com/bearings-126TN9?gclid=EAIaIQobChMIrIa_qPWN_QIV5P3jBx3ONglDEAQYASABEgLlPPD_BwE

and there would be another bearing inside of that bearing that is the roller kind that allows the shaft to move freely through the thing too. Just like it was fixed in the middle of a gyroscope

[jpanhalt]

If the pump runs dry, does its current draw decrease, at least until damage occurs, or rpm increase due to the reduced back pressure?  If so, what type of controller is on the pump?  A low-current shutdown or high rpm shutdown might be easily made.  A current-based control might be quite simple.

How much air/bubbles is generally in the pumped flow?

[coppercone2]

Also, won't the RPM increase with some motors as the resistance drops? Maybe you can measure a spike in RPM to coincide with low water. I know they get loud, but I am not sure if thats rubbing or if its just faster. Then you can use a window detector on a rotation signal.

But this won't help much if it is self priming pump with limited prime life, because it would turn itself off as it primes. And then if you do have a prime failure, how does the pump know to turn itself back on.. you can only guess with a timer. That depends on how consistent the well is. That is like power grid management, you guess how many times is reasonable and then disable the pump. But the nice thing with a float is that you know for sure when the water is back, and it requires no management. Like this type of control system needs alot of good guesses to work...vs a float. If its possible the float would be better.

[james_s]

What is controlling the pump? Do you have a microcontroller involved already? If so you could use a flow sensor, the sort with a paddle wheel and a magnet that passes over a reed switch. If the pump is on and there have been no pulses for a period of time then that means water is not flowing for one reason or another and the pump should be shut off.

[coppercone2]

how do any of these solutions allow a pump to turn back on other then random guessing? Like the way I understand it water builds up in a tube at atmospheric pressure and gets syphon up or has a pump on the bottom. If the pump is on the bottom near a pressure switch that makes sense that there is pressure acting on the pump. But if he said its a narrow tube I assume that means it has to syphon and pump wont fit down there, so you would need a remote pressure switch that runs a wire to the pump, or a float. Is this right?

[Gyro]

A heavy duty reset button bridging the pressure switch contacts. If the pump (pressure switch) has tripped out through lack of water in the borehole then it will presumably take a couple of minutes for somebody to notice and press the button. By then there ought to be enough water depth accumulated in the borehole to temporarily protect the pump. If it starts but doesn't restore pressure within a few seconds then the well is dry - presumably not too likely a situation if it has been bored deep enough.

Ok, it's not perfect protection, but it is simple and probably good enough. Maybe protect the button with a key lock so that only somebody responsible can activate it.

[coppercone2]

well if its used by low tech people, you know that someone is going to suspect the sensor is bad and try to override it. Frustrated people might do so even more, thats why I really like the idea of a float for this, because its hard to argue that there are other factors at play when a float is not floating. Low tech obvious mechanical design is king... if its feasible.

[Gyro]

Oh, I agree, there will be nothing more effective than actually monitoring the water level in the borehole. Unfortunately it is the OP who said that this was a problem due to limited space - frankly I'm surprised that pumps of this type don't have level switches and associated protection already integrated into their housings, that's the only truly tamperproof solution (although probably a real pain if you have to pull the pump up and probably replace it if goes wrong).

Yes, reading between the lines, I think the OP is looking for an easily maintained system with minimum fault finding difficulty, maybe just a $2 meter. If the level switch is really not possible then the pressure switch could be made reasonably tamper proof - put the whole thing in a locked box with only a responsible person (who would also be held accountable for tampering). Basically Solar panel cable in, pump cable out, and the reset button inside the box too. Just a thought that even if the level sensor can be made to fit in the borehole, a pressure switch might also a good idea to prevent loss of all the water due to a broken pipe!

We can only suggest solutions at this point as the OP is the only one with actual on-the-ground knowledge of conditions.


P.S. The OP hasn't been active since 10 minutes after he posted so I doubt this is actually being read.

[beanflying]

Rather than speculating on solutions how about you talk to people locally who do this for a living.

This was done typically in Rural Oz when I had a real job was using either a Pressure LOP (Loss of Prime) switch or a flow switch with a timer to detect flow above ground. Sensors in small diameter boreholes over time depending on groundwater quality run into false triggering issues with deposits on probes. Float switches are a complete no go!

If what you are installing by the sounds of it are likely diaphram pumps then they can also handle a fair bit of time running dry regardless but the motor heating can be an issue if the level gets below that (not likely).

Slightly different application but some of the topic might be helpful for background
https://www.eevblog.com/forum/projects/over-temperature-circuit-for-water-pump-controller/msg2969412/#msg2969412

And for Tautech we did a quick and dirty selection for his pump system. $13-14 of bits flow switch and a timer to allow flow to start before tripping the contactor if it doesn't https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg2461185/#msg2461185 The background to this solution started here and continues on the next page https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg2282234/#msg2282234

The Pressure LOP switch is basically a manual lockout pressure switch so if you lose prime once it trips and needs a manual reset. We tended to use Flowswitches over these but both work. https://www.southernpumping.com.au/pressure/pressure-switches/sqd-lop