Another water pump automation project (suggestions ?)

[redwire]

Re: Another water pump automation project (suggestions ?)

From OP's pic, it's a submersible pump in the well, with solenoids to divert flow to either of the two tanks. Usually there is a backflow valve to prevent siphoning/water running back down into the well.

One hassle is the tank filling can make a lot of turbulence, sloshing waves etc that make the float sensor malfunction.
I have used a sprinkler kind of head, or a plastic fill pipe runs in and almost to the tank bottom and has holes drilled all around it. This stops the water falling down hard (on the float sensor) and it causes the water level to stay steady.

[stj]

Re: Another water pump automation project (suggestions ?)

here's a question.
is the water on the pump inlet under *any* type of presure at all?
can it free-flow slowly through the pump when it's off????

[keeney123]

Re: Another water pump automation project (suggestions ?)

So, there are different types of float switches. One can use a limit switch as a float switch. If you think about the float in a toilet bowl. The float goes down until it gets low enough for the value to turn on and then continues to operate until the float turns off the valve. Instead of a valve one can use a limit switch. So, the float would go down to a point that the limit switch opens. The limit switch would be on an arm to the float. The upward pressure of the float would have to overcome the arms force to close the limit switch. This then has a predictable on and off measurement and would not be subject to unstable condition of the water.

[redwire]

Re: Another water pump automation project (suggestions ?)

There's no money with water, it comes out of the well for free. If this was diesel, gasoline, food being pumped you could spend more than $10 on Aliexpress lol. If the budget is only for one relay, and the customer is providing other parts well good luck with that. You'll get boxed in and unable to make anything decent.

That float has a ball inside to give some hysteresis I believe.

[Dannyx]

Re: Another water pump automation project (suggestions ?)

Here you need a small relay to act as latch because the float switches will chatter due to water waves and slosh. You don't want a large pump freaking out and switching on and off quickly.

I thought of this method, but this would require TWO switches per tank: one to latch the relay, the other to UN-latch it when it's full, since one SPDT is not enough, at least not without some headaches. Since this guy only bought two, I can't do this, so I'll just have to hope the float is stable enough that it doesn't do what you describe. Here's what they look like, by the way.

You can use a contactor+fuses or a "motor starter" or overload - these are electronic with a trimpot to dial in trip current. Something is needed to protect the pump motor against overload. It's likely a centrifugal pump which have little drama unless they get jammed.

It's a small 230v submersible pump, but it's a good idea to over-prepare for future projects which may use larger 400v pumps...

[stj]

Re: Another water pump automation project (suggestions ?)

if it aint broke, dont fix it

[redwire]

Re: Another water pump automation project (suggestions ?)

Contactors commonly have a mains-voltage AC coil. This is because they need a lot of power to pull in hard and fast, there is high inrush current until the armature pulls in and the magnetic circuit is closed. How many HP is the motor? You can use low voltage but they are hard to get and watch the coil current requirement. Get the aux side switch with the contactor for feedback if it pulled in or not.

Limit switches and logic is usually done at the 24VDC level with lower current. Here you need a small relay to act as latch because the float switches will chatter due to water waves and slosh. You don't want a large pump freaking out and switching on and off quickly.
So an intermediate 24VDC relay is used to switch the contactor mains voltage, and the coil can connect to pushbutton switches, logic etc.

You can use a contactor+fuses or a "motor starter" or overload - these are electronic with a trimpot to dial in trip current. Something is needed to protect the pump motor against overload. It's likely a centrifugal pump which have little drama unless they get jammed.

[keeney123]

Re: Another water pump automation project (suggestions ?)

I began sourcing some parts for this project, at least to get the basic form off the ground and someone was right: I'm indeed unable to find contactors with a coil voltage of 24VDC - only 24VAC, so what is the most common way of switching large motors in automation projects in this situation ?
Should I go with a 230v contactor coil which you then switch via a smaller relay (probably the most cost-effective), or add a dedicated 24VAC transformer, in addition to the 24VDC supply I'd need for the logic and PLCs and other stuff IF they ever get added (or we assume they're there, for the sake of argument)....

TE Conductivity sells worldwide and they have 24 volt DC relay with contacts you should be able to find those.

Here is TE relay shop site. They are very good people and will help you any way they can. You can email them or chat with them and they will actually get back to yo.

https://www.te.com/usa-en/plp/relays...=false&instock

[Dannyx]

Re: Another water pump automation project (suggestions ?)

I began sourcing some parts for this project, at least to get the basic form off the ground and someone was right: I'm indeed unable to find contactors with a coil voltage of 24VDC - only 24VAC, so what is the most common way of switching large motors in automation projects in this situation ?
Should I go with a 230v contactor coil which you then switch via a smaller relay (probably the most cost-effective), or add a dedicated 24VAC transformer, in addition to the 24VDC supply I'd need for the logic and PLCs and other stuff IF they ever get added (or we assume they're there, for the sake of argument)....

[keeney123]

Re: Another water pump automation project (suggestions ?)

He is the pump control all you would need is a relay switch to connect the AC voltage. For some reason when I uploaded schematic it rotated.

Explanation on operation of my pump circuit below. When float Switch in last tank falls below a given point the float switch closes. With the normally closed switch of relay coil 2 the coil 1 energizes and closes coil 1 switch. That then turns on the pump. The pump forcing water through the pipe and closes the flow switch. This will happen before the time delay can open relay coil 2 switch. As long as water is flowing, and the flow switch and the float switch are closed relay coil 1 switch will remain closed because it is energized by coil 1 relay. The energized coil 2 will keep its switch open which makes the pump dependent on the flow switch to close and the float switch to close. If the flow switch opens and the float switch remains closed the coil 2 switch will remain open and the pump will stop. If the float switch were to open the pump would stop because Relay coil 1 would deenergize and open relay coil switch 1. It would deenergized Coil 2 relay and the coil 2 switch would close. As soon as the water drops to a point in the last tank and closes the float switch the process with start all over.

So, this is a basic circuit. Circuits can be added for instance one looks and says what happen if Coil relay 2 burns out or does not open Switch 2 how can one protect against this. Can one add another connection to Coil relay 2 switch that when one open another closes? Can I add another circuit for overflow of tank 2?

[Dannyx]

Re: Another water pump automation project (suggestions ?)

see how silly and costly this is going to get!

True - the tried and tested way would be a proper PLC-based system and those CAN get pricey, at least for this guy's demands, which are fairly small, since it's his own homestead so to say - it's not an industrial application. This is why I was planning to go fully "analog" first and MAYBE try to add some brains later on if they really want to...

[stj]

Re: Another water pump automation project (suggestions ?)

client is a dickhead - get a deposit before doing anything
if they really want a web interface then get an ESP8266 module and a router.
(and a telephone line, and an internet contract)
see how silly and costly this is going to get!

[Dannyx]

Re: Another water pump automation project (suggestions ?)

You can also split the project in two. First get a PLC and panel/HMI built, then later add the webserver/gateway - as long as the PLC is capable of external comms.

Even less than that: do it purely using relays and buttons at first and then add the "brains". I'll of course plan ahead and prepare the necessary connection points for where a PLC or whatever would go.

Omron CP2E looks interesting

Doesn't seem to do web-server, so I don't know how I'd get it online the way the client wants...

Since our company happens to excel in alarm systems, I was even thinking of using a Paradox EVO192 alarm panel as my controller. It supports an app out of the box, through the addition of an ethernet module, although we haven't deployed too many of them where the client expressly requested this. I'd use the zones as my inputs and the PGM (programmable outputs) as my....outputs It's silly and pretty primitive and it's only got the 4 PGMs to work with, but it works - I tried it yesterday on the bench. It's not 24v tolerant, obviously, so it's not quite industry-standard, which made me steer away from it.

[keeney123]

Re: Another water pump automation project (suggestions ?)

One does not want a timer on a well pump unless the pump is being used for irrigation. If the demand on the water is not predictable the pump needs to pump whenever demand requires. Irrigation systems are different because the demand is predictable and timed. As long as the pump has water going through it, it will continue to run no problem. The starting and stopping of a pump will put more wear on the pump. The only problem a pump will come across is when debris from a well enters it that is if the piping in the well was not done correctly. Or the installation was old, and the materials use are capable of deteriorating. In that case one would need a filter. This can be done using another tank before the pump if the pump is above ground. Most motors now a days have a resettable thermal shut-off built into them

Also, what happens if the timer fails?

When a person designs a circuit with switches and relays one tries to make it fail safe by using the Normal open switches as a fault indication and stoppage. In a normally open position one needs to close the switch to make it work and it relies on the connection to continue to run. It is more likely this switch will fail by not closing than it will fail by fusing together. Although this does happen. That is why one has an additional backup system. Like when car brakes fail, we have a manual emergency brake.

[sam_sam_sam]

Re: Another water pump automation project (suggestions ?)

Omron CP2E looks interesting

I have seen there PLC before but I have never used one how easy it to use and what is the cost for the simple version of the PLC and what is the cost of the software this website does not want to give you this information

[redwire]

Re: Another water pump automation project (suggestions ?)

You can also split the project in two. First get a PLC and panel/HMI built, then later add the webserver/gateway - as long as the PLC is capable of external comms.
I looked at PLC Internet Gateway Modules but they start at $700-$1,200 and then software for it is another few grand but supports a ton of stuff, like they threw everything in- when you don't need it. Some people even use a small PC for that.

The pump timer is just to keep the motor from running forever/valve always on, if there is a sensor malfunction. Timers are free in a PLC. But both tanks/pumps would be dependant on the one PLC which makes shutdowns a bit more complicated. You can manually run it with mechanical switches (not touch screen).

Omron CP2E looks interesting

[keeney123]

Re: Another water pump automation project (suggestions ?)

The run timer is a good idea indeed, assuming the tanks fill up at the same rate each time and they're not being drained at the same time by whatever they're connected to, which would also affect how long it takes for them to fill to the brim, if ever, if the demand is high enough.

The valves sit above the tanks and water pours down into the tanks from what I understand from the client, so there's no risk of backflow whatsoever.

An indicator light for each tank and the pump would indeed look nice. I'd also add a master stop switch and maybe a manual override toggle switch to force the tanks to fill up if you want to go past the max level for whatever reason.

I decided to take a step back do it all using relays at first (see the highly professional MS Paint schematic), which was easy enough to do and then worry about how I'd get this thing online...

I would agree with stj in his filling one tank with the other tank. This would reduce redundancy. On the last tank filled on would have an overflow line. There could be some type of indicator to monitor if overflow ran out the pipe and stop the pump.

Even-though one fills from the top if the pipe is within the tank it can back up. In my schematic a flow switch which is Normal open would stop the pump once flow was stopped. I would think that one would want to have the tanks enclosed completely to prevent anything outside from entering them. One would not need valves to open and close if one monitors the output. If the last tank to be filled started to lower and the float switch opens, then the pump will start to pump. One will want the pump to continue to pump until the float switch closes meaning the tank if full. The flow switch will open if the well one is pumping from starts to go below the points that are put in the ground that allow water in. This would mean the pump is pumping water faster than it can run into the points. That then would indicate a need for a smaller pump.

A check valve is also used in a well system to prevent loss of priming of the pump. So, when the pump stops the prime of the pump will not flow back into the well.

[Dannyx]

Re: Another water pump automation project (suggestions ?)

The run timer is a good idea indeed, assuming the tanks fill up at the same rate each time and they're not being drained at the same time by whatever they're connected to, which would also affect how long it takes for them to fill to the brim, if ever, if the demand is high enough.

The valves sit above the tanks and water pours down into the tanks from what I understand from the client, so there's no risk of backflow whatsoever.

An indicator light for each tank and the pump would indeed look nice. I'd also add a master stop switch and maybe a manual override toggle switch to force the tanks to fill up if you want to go past the max level for whatever reason.

I decided to take a step back do it all using relays at first (see the highly professional MS Paint schematic), which was easy enough to do and then worry about how I'd get this thing online...

[redwire]

Re: Another water pump automation project (suggestions ?)

This is a fire pump panel I've always found interesting. Uses WEG CLIC-02 PLC and a custom back board with a familiar MCU

[redwire]

Re: Another water pump automation project (suggestions ?)

I could see the control logic needs some thought, customer going from manual control to automatic.
Typical pump control panel has START/STOP, HAND/OFF/AUTO (and likely needs LOCAL/REMOTE) as well as indicator lights for the level switches, motor overloads etc. https://www.sjerhombus.com/products/...ontrol-panels/ see catalog page/library

There's usually a run timer for the pump motor so it does not run forever if there is a fault. It would limit motor on-time to however long it takes to fill up as the smallest timer value.
The valves are a bit weird, usually it's a check-valve. If a motor was commanded on (water level is low) but did not run say due to a fault/trip/fuse etc. the valve would be open though, would the other tanks' pump backfeed into it? i.e. one pump/tank just fills the other by accident.
Or the valve is stuck closed is another fault that would likely cook the pump.
All this means is extra I/O...