Measuring water level in Well

Thread Starter

MikeML

Joined Oct 2, 2009
5,444
I live where my water comes from my own well. Since the entire Western US is in a drought, I would like make a project to keep track of the static water level in my well as an indicator of the water table in our aquifer. When my well was drilled in 2005, the static water level was about 100ft.

A crude drawing of my well is shown below. The well casing is 300ft long. The casing is 3"id Black ABS pipe. It is perforated with small holes all the way to the bottom. After the pump stops running, the water from the surrounding aquifer refills the casing so that the static level is even with the top of the water table.

The 1.5 h.p. pump is suspended 280ft down the bore, effectively hanging on 280ft of Sched 80 White PVC water (pressure) pipe, putting the pump about 20ft from the bottom of the bore. The pump is powered through a four-conductor rubber-covered cable with four #10AWG conductors.

Three of the conductors go to the motor windings (run winding & start winding, with one wire common to both windings) and the fourth wire is a (green) safety ground which is earthed at the top, and connects to the stainless steel body of the pump at the bottom.

I tried lowering a #28AWG insulated wire (with a small stainless steel plumb-bob attached) thinking I could tell when the plumb-bob makes electrical contact with the standing water. Knowing the length of the wire would tell me where the static water level is. However, the pump cable meanders down the hole sort of loosely wrapped around the PVC water pipe, and what happened is that the wire and plumb-bob got jammed between the pump cable and the inside side-wall of the casing.

I am planning to get a coil of 3/8 PEX water pipe, and I will insert that down into the well. The PEX is stiff enough that I think it will push past the random coils of the pump cable, and strong enough that I could withdraw it from the well if I ever need to... I show the PEX pipe as the Standpipe on the drawing.

Here is what I am thinking. The PEX pipe comes premarked with length ticks. I can keep track of how much I have inserted down the casing. Imagine that I insert 90ft, and then apply very low air pressure at the top end. If the bottom end hasn't reached the static water level yet, there will be very little resistance (pressure rise) because the air can escape at the bottom. Now I keep pushing down more PEX, keeping track of the total length so far.

As the bottom of the PEX reaches the static water level, two things should happen. First, I should be able to hear bubbling as the air comes out around the now submerged end. Second, the air pressure required to keep air flowing will rise compared to prior to the end reaching the static level.

This establishes the static water level right now. Suppose I shove in another 33ft of PEX, and then secure it. That will put the open end of the PEX standpipe 33ft below the static water reference level.

Since 1 atmosphere of pressure (14.7 psi at sea level) will support a column height of 33ft of water, it should take about 14psi to push the water level inside the standpipe down to where the air starts bubbling. More over, if I slowly increase the air pressure inside the standpipe by bleeding it though an orfice, the air pressure will increase as the water column inside the standpipe is pushed down and then the pressure should stop increasing as the air begins bubbling around the open end of the standpipe.

So, as the static water level (aquifer level) fluctuates above and below the reference static level established above, it should be possible to measure the current static level by measuring the air pressure required to cause bubbling. Suppose in the future, the water table drops ten feet (I hope not). Now, the pressure required to bubble would be 14*(33-10)/33 psi, and it should be possible to calculate the new level just from the pressure measurement.

I would like to use an Arduino interfaced to a LCD, a gauge pressure sensor, and an electric pump. This is similar to measuring blood pressure. I have one of the inexpensive automated cuffs which I might be able to cannibalize for parts.

Help me here. What allowance do I need to make for my elevation (4316 ft MSL)? What allowance do I need to make for temperature? (The temperature 100ft underground is going to be constant, but do I need to know what it is)? The temperature of the water pumped from the well should be a good indicator.

Any other suggestions on alternative approaches welcome...

I have thought about echo location, measuring capacitance, etc, but none seems as clean as measuring air pressure....
 

Attachments

Last edited:

jpanhalt

Joined Jan 18, 2008
11,087
First, I agree with the approach.

Assuming you will be reading gauge pressure, atmospheric pressure at your location should fall out of the calculation. There is a very small effect caused by depth of the measurement. One hundred feet is approx. 0.1" Hg change at sea level ≈ 0.33% (0.05 psi) change. That can probably be ignored compared to the changes in pressure due to the water depth.

Edit: What I mean is the change in air pressure at the measurement depth as the water level changes over just several feet is insignificant.

As for temperature, a change from 10°C to 20°C causes only a 3.5% change in pressure in a closed vessel. I need to sleep on whether the temperature difference between the water and ambient gauge will make much if any difference given that the tube is open.

John
 
Last edited:

ericgibbs

Joined Jan 29, 2010
18,767
hi Mike,
Have you considered this method of water height above transducer measurement.
They come 1 or 2 Bar versions, giving a 10m or 20m range.

Its a differential sensor with a cable which has an inbuilt air tube to atmospheric pressure.
I have used this type of pressure transducer in my Tidal height Hydro-graphic products

http://www.globalw.com/products/levelsensor.html

Eric

Note there are cheaper versions available,
 
Last edited:

jpanhalt

Joined Jan 18, 2008
11,087
My thought on the temperature is that since you are measuring the pressure, not the volume of air needed to displace different volumes of water, temperature will not make a difference. That is, when static, the pressure in all parts of the tube should be the same; although, the density in different parts may vary with temperature.

Of course, as the air moves, there will be a pressure gradient, but that gradient should be small as I understand it.

John
 

Thread Starter

MikeML

Joined Oct 2, 2009
5,444
...

Note there are cheaper versions available,
First, this validates my bright idea. The difference between what I proposed and this commercial version is that I put the pressure sensor at the top, and they put it at the bottom. That would get the slight pressure drop due to flow down the standpipe that John mentioned out of the equation. Their system is vented to atmospheric pressure continuously, while in mine, I would have to open my standpipe to atmospheric pressure just before commencing a new measurement.

Second, would their cable be stiff enough to be able to force it past the coils of pump cable?

Third, the price puts me off. I would like to do this for less than $100.
 

ericgibbs

Joined Jan 29, 2010
18,767

Thread Starter

MikeML

Joined Oct 2, 2009
5,444
Hey, something just occurred to me...

When the well pump runs, the pumping rate is about 45 gal/min, causing the water level in the bore to drop about 100ft. This will drop the water level way below the bottom of the standpipe.

Suppose the top of the standpipe is connected to a pressure gauge but otherwise sealed from atmosphere. When the pump runs the water in the standpipe will all drain out, and the air therein will be at atmospheric pressure. After the pump shuts off, the water in the bore will rise, eventually covering the bottom of the standpipe, and then rise higher.

Wouldn't the final air pressure in the standpipe reflect how far the water rose above the bottom of the standpipe? Wouldn't the measurement reset to current atmospheric pressure every time the pump runs (which it does about twenty times per day)
 

inwo

Joined Nov 7, 2013
2,419
Been there done that!:D String/wire stuck between cables and drop pipe.

Also used a bicycle pump and gauge to measure water depth at end of a small tube.

Pex is good choice for the right stiffness. Takes a while to get the curls out. Works for jetting drain lines too.

Short term, measuring pressure at the top of stand pipe should work. IMO
Long term, air goes into the water.

What about sensor at the bottom?
Is price the only issue? Or getting stuck? No small thing!

I've had so much junk stuck over the years. I once had a plastic ring stuck that took a week of slide hammering to get out. :(

edit,

I found a 0-5psi sensor I was going to post, but I see you just need a gauge.
 

Attachments

Last edited:

Thread Starter

MikeML

Joined Oct 2, 2009
5,444
Ok, here is the next question: (see attached)

Hey, what can I say. I slept through Fluid Mechanics (>50 years ago).
 

Attachments

wayneh

Joined Sep 9, 2010
17,496
Just a couple random thoughts:
1) I'd be nervous about pumping air into my well unless that air is filtered. I wouldn't want to inoculate my well with something from the surface. A small issue but one I would take steps to fix.
2) How about measuring pump performance? I mean, the pressure developed by your well pump, measured at your well head for instance, should relate to the head it is pumping against, which in turn depends on the height of the water column above it and the (known) pressure of your system's reserve tank. It may be tricky to calibrate to any absolute values, but I think it could give you relative changes. You should be able to see the ∆P of pumphead-minus-reservoir drop during a pumping cycle.
3) Water temperature matters only because the density changes slightly. 30 feet of water indicates slightly different pressures if the water is at different temperatures. I think this would be negligible for well water since the range is so narrow.
 
Last edited:

wayneh

Joined Sep 9, 2010
17,496
X and Y are the same, since the pressure indicated by the raised water columns is the same. You could be on to an interesting solution.

Oh wait, I misunderstood the drawing. Need to think about it. The deflection from atmospheric (Y/2) is the same at the top as in the well, but that is only part of what you need to get ∆X.

You should be able to make a PV=nRT correction to the observed deflection, Y/2, to correct it to what that deflection X would be at atmospheric. The RHS can be assumed constant, and V depends on the PEX tubing diameter, which should also be nearly constant.
 
Last edited:

jpanhalt

Joined Jan 18, 2008
11,087
I think you need a piece of data, "Z":

View attachment 72050

Doesn't X-Z = Y?

John

Does this work:
Capture.PNG

I always check my attachments after posting and got no sign that it didn't work. Resized a little smaller. Lines are crooked as I am a lefty, but use mouse in right hand. Basically, how I see it is the manometer measures pressure. It simply measures the pressure to "blow" x-z of water out.
 
Last edited:

THE_RB

Joined Feb 11, 2008
5,438
We used a similar method when I worked with the instrument fitters in industry for sump liquid depth measurement.

We used a small air bleed valve to allow a tiny bit of air flow from a compressor into the pipe.

That results in a tiny bubble per second or so, from the pipe. Continuously. That stops the pipe being blocked by sludge.

The depth of the liquid is given directly from the pressure, and the pressure sensor can be anywhere dry on the pipe, like just after the bleed valve.

The pressure represents the vertical distance from the pipe opening to the top of the liquid.
 

BR-549

Joined Sep 22, 2013
4,928
This post is very interesting, I wonder if there might be a commercial demand for this in the future. I have measured level and flow for many years. No one has ever asked me this question before. I love this question.
Your first approach is know as a bubbler system. It is very accurate. It will tell you the level within 1/8 inch very easily. 24/7. Be sure to get a dry, or reed, or oil-less bubbler. Wipe the weight and the bubbler tube with Clorox rag when lowering. Lower it to the bottom. Install pollen filter on bubbler intake. I would think you should be able to get in-expensive bubbler. If you can’t....look for a surplus ISCO bubbler sampler. All you need is the bubbler...not the chart recorder. If you can get the recorder with it for fair price....grab it. When I was in the world I used to call mfg. and ask for samples of products for projects. Find any 10 transducers you want and start calling....... you will be surprised. There should be a chart that comes with any bubbler.....it will show you the curve of pressure to level. Have your Arduino record pressure every five mins. and calculate level. During pump cycle...record every 5 secs. After pump cycle, record every min for 15-20 mins or so to see well refresh time. With this setup.....you can easily see the water level and refresh times....also seasonal changes.
I never had much luck with ultrasonic level control, but it’s been awhile. If you could fine a pinger that can get thru all the wire, I would try that.
Now that that is done and we know the true level........there’s another way. For those who don’t like the idea of putting something down the well. Monitor, sample and record the current and the pressure during pump cycles after surge current. Plot these two curves. These curves can be symmetrically related to the pump curves. These two plots can be adjusted with transducer sensitivity and sample rate times to get a recognizable curve. Comparison of these curves will give you the head. It will not be as accurate and it will show level only when pumping.
 

wayneh

Joined Sep 9, 2010
17,496
...there’s another way. For those who don’t like the idea of putting something down the well. Monitor, sample and record the current and the pressure during pump cycles after surge current. Plot these two curves. These curves can be symmetrically related to the pump curves. These two plots can be adjusted with transducer sensitivity and sample rate times to get a recognizable curve. Comparison of these curves will give you the head.
That's the idea I suggested above, and...
It will not be as accurate and it will show level only when pumping.
...that's the problem with it. May or may not be a problem for the application.
 

BR-549

Joined Sep 22, 2013
4,928
You are absolutely right Mr. Weyneh, please forgive me. For ease and simplicity, the ideal would be a pinger. I was watching this video and I heard the words “sounding tube.”
http://www.youtube.com/watch?v=Fgki7esvOO0

Why can’t we use the pump discharge line as a sounding tube? Mount pinger at elbow at well head. All we need is the transducer and the Arduino.
 

wayneh

Joined Sep 9, 2010
17,496
You are absolutely right Mr. Weyneh, please forgive me.
For what? No infraction to forgive. I just wanted to point out that two folks had a parallel idea.

I like the sounder. Minimal invasive. Now I'm wondering what my well level is! ;)
 

BR-549

Joined Sep 22, 2013
4,928
OOPS....where's that check value? If it's at the bottom, where it probably is, this wont work. If it's at the top, we could vent slowly(so as not to disturb well) and then measure. The sounding tube needs to be empty for measurement.
 
Top