Deep well, measure water level

Hi all,
I want to measure the water level of a deep well.
I want to be able to do it permanetly but also easily for any well.

I have done some research but noone seem to have a good solution.
Most are suggesting measuring the pressure level at the bottom. This will be impossible to do without renoving the pump, as the pipe, wires, arrestors, etc would make it impossible to get a sensor down to the bottom.

Also, some suggest using a floater. For the same reason, this wount not work. Others suggest ultrasonic close to the water level. This would not work either as the waterlevel will change too much.

The reason I say this is that I don't want any of these ideas. I need to have something that easily installs at the cap of the well.

So, some info:

  • Well is 350ft deep
  • Pump is at 270ft
  • If no water has been used for a long time, the water level is about 50ft from ground level
  • If water is pumped continously for a long time the water will run dry, i.e. below pump.
  • This means the water level to measure is 50-270ft from where the sensor will be.
  • The well is drilled 8" in diameter, it has a cap on top.

If I tap the cap with a rock or hammer, I can manually listen for the echo. This means pipe, wires, etc does not obstruct the sound.

So I am thinking of using echo. A sound pulse is sent by a speaker, expect echo to return 0.05-0.5 seconds later.

I could build the HW myself (small speaker or to reduce size, just a magnet and coil to make "tap" sound, mic with notch filter for return), but I would expect there are stuff out there that could be used to make it easier.

The cap has a 3/4" PVC pipe for air inlet. I would prefer to just remove that and replace with my sensor. That would be ideal, but of course I have the 8" if needed, but would then have to open the cap, making it harder to use this sensor for temporary measurements.

So what do you all think?

On the SW side, in case you want to know, I am planning to log water level, maybe every hour or so and store in GSheet and Blynk.
All of this is mostly for the nerd in me, but it is also useful to find water level and recovery rate for the well, especially during dry periods.

Just curious. What does the drillers log show as the water level on the newly drilled well?

Paul

Does the pressure or flowrate from the pump vary with water level?

Water well drillers do use ultrasonic rangers. Not the cheap hobby type, of course. Here is one commercial option, among many.

The borehole conducts sound in 1-D quite effectively, so making your own is certainly an option. Doing so would require considerable expertise in electronics.

Paul_KD7HB:
Just curious. What does the drillers log show as the water level on the newly drilled well?

Paul

I don't know, this well is old. I replaced the pump and measured the depth with a string and a weight.

david_2018:
Does the pressure or flowrate from the pump vary with water level?

The old pump was 3/4 hp and would not be able to pump the well dry. It would only pump at 30 psi at about 150-200ft (measured by listening to echo). I turned pressure switch down to 20/40 but would still turn off.
I replaced it with a 2hp some weeks ago and has not run dry yet. I have it set to 40/60. Flow is limited by pressure limit and pipe size. I do messure power consumption but it does not seem to be accurate enough for depth mesurement as it varies depending on water usage.

I'd use loss of prime (pressure) at the top.

Electret microphones are tiny and very sensitive.

Use something that snaps or pops to make the sound to echo have a nice hard wavefront.

This 8" dia hole has a pipe and wires going down to the pump at the bottom?

When the pump runs don't you get more flow the deeper the water is?

jremington:
Water well drillers do use ultrasonic rangers. Not the cheap hobby type, of course. Here is one commercial option, among many.

The borehole conducts sound in 1-D quite effectively, so making your own is certainly an option. Doing so would require considerable expertise in electronics.

Yes, I have seen professional ones, but I don't want to spend that much, and also I want to be able to log.

I have more than "the required expertice in electronics", thats not a problem at all. Was just hoping for a quicker solution. But when I think of it, making the board and all for the arduino, it would not add alot do just do it myself.
Will post it here if I do, but ir is not top prio and will rake a while... Too many project, too little time...

If you can figure out the acoustic measuring/timing, the actual setup might be somewhat simple. Unscrew the vent pipe and stick in a microphone, then tap the cap with something and measure the time between the audio peaks. Do the math to find a depth value. You would need to work up a display if desired, which might have a couple of ways of doing.

bluejets:
I'd use loss of prime (pressure) at the top.

I dont understand. Then I would have to also meassure pressure at the pump, right?

GoForSmoke:
Electret microphones are tiny and very sensitive.

Use something that snaps or pops to make the sound to echo have a nice hard wavefront.

This 8" dia hole has a pipe and wires going down to the pump at the bottom?

When the pump runs don't you get more flow the deeper the water is?

Yes, thats what I was thinking. Some click/pop and a mic.

Yes, it has pipe and wire and also maybe 20 or so pipe guides. But the echo is clear, not obstructed.

The deeper the waterlevel, the more pressure is needed to pump the water, i.e. less water flow, not more.

zoomkat:
If you can figure out the acoustic measuring/timing, the actual setup might be somewhat simple. Unscrew the vent pipe and stick in a microphone, then tap the cap with something and measure the time between the audio peaks. Do the math to find a depth value. You would need to work up a display if desired, which might have a couple of ways of doing.

Yes, but I want to be able to log.

If I sit there to tap it anyway all I need is to count the number of exhos in a second and then calculate.

I do know how to make the electronics, its quite easy. But would be even easier if it was available for a few $..

The deeper the water is in the well? Full well == deep, empty well == shallow.

A snail shell (shape) cam goes from close to the axle to far in 360 degrees and then drops straight back down.
Turn it with a stepper to raise a strong spring end and let it snap back down.

Arduino Nano has pins you can jumper. Hang sensor(s) and active parts on jumpers and glue into place.

This ain't rockets or rocket science.

GoForSmoke:
The deeper the water is in the well? Full well == deep, empty well == shallow.

A snail shell (shape) cam goes from close to the axle to far in 360 degrees and then drops straight back down.
Turn it with a stepper to raise a strong spring end and let it snap back down.

Arduino Nano has pins you can jumper. Hang sensor(s) and active parts on jumpers and glue into place.

This ain't rockets or rocket science.

Ah, see what you mean with deep/shallow now. Was thinking you meant deep as in low water level.

Yes, snail shell cam is a good idea, but moving parts. Think a clicker, buzzer or piezo would be more reliable.

Just to test the concept you might use your cell phone. I've never really used it, but my smart phone has an audio recording function that records audio files in the m4a format. You might remove the well head vent pipe, start the phone audio recording with the phone mike at the vent pipe hole, tap the cap, and record the echos. Then open the captured file in an audio editor (there seem to be a number of free ones) and measure the time between the tap/echo peaks. This should somewhat test the concept of measuring echo time to calculate the water level.

Hi,
Can you monitor the pump current?
If the pump runs dry can you measure a drop in output pressure?

I can see two other ways, change in motor current or change in outflow pressure.

Tom.... :slight_smile:

richardax:
I dont understand. Then I would have to also meassure pressure at the pump, right?

Yes, thats what I was thinking. Some click/pop and a mic.

Yes, it has pipe and wire and also maybe 20 or so pipe guides. But the echo is clear, not obstructed.

The deeper the waterlevel, the more pressure is needed to pump the water, i.e. less water flow, not more.

If you run out of water, the pressure drops below say a minimum system range level ( perhaps set it to 10psi as 20-40 is normal range)

richardax:
Ah, see what you mean with deep/shallow now. Was thinking you meant deep as in low water level.

Yes, snail shell cam is a good idea, but moving parts. Think a clicker, buzzer or piezo would be more reliable.

The flow rate at the top of the pipe might be all the sensor you need every time the pump runs.

A buzzer will give you a whole train of small clicks. A bare piezo, I get weak sound only at 5V.

A clicker? So how do you make that work w/o moving parts? The cam lets the work load to snap anything be spread out for a whole turn, that lets it work with a smaller motor.

Try a speaker first and see how well that works.

At least down in the covered well the air temperature changes little, right? Otherwise to use echo time needs a thermometer since speed of sound varies a little per 10C. If you're not trying to measure close, don't worry about that.

There is another trick using sound, see how low a note the space above the water can make. That would be more expensive and harder to do for sure.

Totally different way; may or may not be feasible of course. If that well is a tube of constant diameter, pump in air and see how fast the pressure rises.

The lower the water level, the more air above it, the slower the rise. Ratio between air inflow and pressure change is a measure of the air volume, which in turn tells you how far away the water is.

zoomkat:
Just to test the concept you might use your cell phone. I've never really used it, but my smart phone has an audio recording function that records audio files in the m4a format. You might remove the well head vent pipe, start the phone audio recording with the phone mike at the vent pipe hole, tap the cap, and record the echos. Then open the captured file in an audio editor (there seem to be a number of free ones) and measure the time between the tap/echo peaks. This should somewhat test the concept of measuring echo time to calculate the water level.

Yes, I have done that before on other projects. That is what I was planning in order to filter out unwanted noise.

TomGeorge:
Hi,
Can you monitor the pump current?
If the pump runs dry can you measure a drop in output pressure?

I can see two other ways, change in motor current or change in outflow pressure.

Tom.... :slight_smile:

Yes, I have built a energy monitor for the whole property. 16 channels Arduino, logged to GSheet and Blynk, works really well. The pump is one of them. The power consumed will depend on the pressure the pump is pumping. However, I have several taps from the pump. If no tap is open, the pump turns on at 40psi and off at 60psi. If a tap is open it will pump at higher pressure without reaching the stop psi. Pipes are connected to different places and different altitudes, so pressure and flow depends on many factors. I have been thinking of looking more into analyzing power consumption but I really don't think it will be accurate enough.

bluejets:
If you run out of water, the pressure drops below say a minimum system range level ( perhaps set it to 10psi as 20-40 is normal range)

Correct, if I run out of water, the pressure will drop and pressure switch turns it off. The pressure switch works at 40/60 and turns completely off at about 20psi.

But I would like to know the water level. In that way I would know how much water I have and how long I can expect it to run and if I need backup solution, like switching to city water or taking water from pond/creek for the horses, etc.

wvmarle:
Totally different way; may or may not be feasible of course. If that well is a tube of constant diameter, pump in air and see how fast the pressure rises.

The lower the water level, the more air above it, the slower the rise. Ratio between air inflow and pressure change is a measure of the air volume, which in turn tells you how far away the water is.

Thanks for that suggestion but I don't think it sounds feasible.