2
$\begingroup$

Reprhasing the question is a more open ended manner:

Method 1: Use a pipe and measure it's resonance.

Given a closed pipe from 10 to 300 feet in length, and roughly 5/8" in diameter, made of 1/16" polyethylene, can the length be easily determined from it's resonate frequency?

(This is really an acoustic engineering question, but there isn't such a stack exchange group yet)

Issues:

  1. Edge damping: The very high length to diameter ratio (up to 600:1) may cause the signal to fade into mush before the reflection off the end and propagation back.

  2. Q factor. Resonance may not not be sharp enough to provide useful information.

  3. Detection. The nominal resonant frequency is around 1 Hz. This is difficult to measure. In theory it would be possible to use a harmonic instead, but it may be difficult to determine which harmonic is being used.

Method 2:

Perch a suitable tank on the well casing and make a helmholtz resonator. This uses an unrealistic amount of power, I suspect to move the required amount of air. Not clear how to build it to resonant at the fundamental, and not a harmonic.

Method 3. Run a length of wire down the well, and measure the capacitance between the wire and the steel well casing. Water has a high dielectric constant so the capacitance should vary substantially as the water level runs up and down. Make the capacitance part of an LC circuit, and count oscillations.

3a. You already have two wires going into the well to supply power to the pump. Should be possible to use one of those, if you can tune the LC to a frequency that is high enough that the motor ignores it. (Might put a 1 KHz signal on every device for a long way around. Isolation is a big issue here. I think I prefer a separate wire.


Where this comes from:

I want to measure the drawdown of a well continuously. (Once a minute) Ideally this device logs to a memory stick, and has a LCD display of the current level to make it easy for the well owner to track changes in his well behaviour over time.

Pressure sensors are expensive for this application, are at best only good to about 2% of range, are subject to hysteresis, require an A to D converter...

When testing wells, the common way is with an insulated tape with bare end and a continuity tester. You have to keep making and breaking the contact.

My first kick at the cat was to use an ultrasonic pulse and bounce it off the water surface. However there is a lot of competing junk in the top end of a well. Once you are below that, you still are working in the annular space between the riser pipe and the well casing. You have possibilities of spurious reflections off of casing joints, the torque arrestors, the well pump, and the bottom of the well. And ultrasound attenuates quickly.

My second thought was to use a wavelength that was a small multiple of the casing diameter. The transducer is placed below the wellhead junk, and emits a short chirp -- about 20 cycles. That won't work. The reflection off the well head will overlap the chirp. The chirp has to be under Vsound/20 or about 1/50 of a second. This would mean a 1 foot wavelength. Given well casing diameters of typically 4 or 6 inches, this is close enough that I can see all sorts of weird modes, and an unbelieveable amount of junk in the return signal.

This is my third thought. It's easy to lower a weighted 1/2" (which is really 5/8) plastic water line down to the bottom of the normal drawdown zone. The water level in the pipe is the same (within a second or two) of the level in the well. And there is no junk in this smaller pipe.

Ok, then, you say, do the 'bounce the sound wave' trick in the small pipe. My intuition is at a stretch here: I think to get efficient transmission, the conduit has to be large, or at least not small compared to the wavelength.

Resonance has the advantage that you are getting a very specific amplification. But it takes a long time (at least several cycles) to pump it up.

For this application the ability to get one measurement a minute would be sufficient.


"But what about the temperature? Vsound is temperature dependent."

The temperature once you are more than a few feet underground is constant. If the well is in frequent use, there is a 1.25" diameter column of fresh water in the middle of the casing. That water is at average ground temperature. In mid winter, the steel casing, which stick out of the ground a couple feet, will create an annulus of cold air, but I don't think that this can chill any signficant length of the pipe.

$\endgroup$
  • $\begingroup$ In the numbered list the first two items are the same thing. $\endgroup$ – DanielSank Apr 3 '15 at 6:27
  • $\begingroup$ Excellent, challenging question. Indeed length determines the N set of resonant standing wave frequencies for N = 1,2,... So how would you differentiate N from the changes in L? Your proposed device would have to either scan a range of frequencies or generate white noise, and from that would have to measure the resonant (peak) modes like a spectrometer. Tricky problem perhaps though if the tube gets really long. Non-linearity may turn out to be the demon in this situation. I just don't know how to approach it without thinking further ... $\endgroup$ – docscience Apr 4 '15 at 2:34
  • $\begingroup$ Daniel, yes 1 and 2 are very similar. Although a low Q may have multiple causes. Doc, the attenuation between the white noise generator and the return would make the signal hard to pick out. $\endgroup$ – Sherwood Botsford Apr 4 '15 at 14:33
  • $\begingroup$ What about making a classic helmholtz resonator and supplying it with air? This would not be a low power solution. Ideally I want a method that works with deep wells, as I think there would be a lot of interest in such a gadget. My well at present has a static water surface at about 30 feet and a drawdown of about 12 feet when pumping at capacity. $\endgroup$ – Sherwood Botsford Apr 4 '15 at 14:37
1
$\begingroup$

There are distance measurement devices commercially available, but if none will do, then I recommend making your own using a pulsed laser, a detector, and accompanying electronics.
Another method would be a small buoy with a radio transmitter and a sound generator inside. The radio transmitter sends a signal once per minute and the sound generator emits a sound at the same time. The time difference between the radio signal detected and the sound signal detected, can be used to give the distance from the buoy to the detectors (water depth).

$\endgroup$
  • $\begingroup$ Possible gotchas: a 6" steel casing will act as a wave guide. Wouldn't any of the common bands, (900 MHz) with it's foot long wave length have severe propagation issues in 6" pipe? Fc= 1.8 c / pi *d = 570 Mhz, and that's ignoring the riser pipe. Higher frequecies are hard to work with. $\endgroup$ – Sherwood Botsford Apr 9 '15 at 18:33
  • $\begingroup$ Gotcha2: You have to fish the bouy out on a regular basis for battery replacement. $\endgroup$ – Sherwood Botsford Apr 9 '15 at 18:34
  • $\begingroup$ Gotcha 3: You cannot guarantee a clear line of sight to the water level. Oh, it's there somewhere, but you have the pitless adapter, the riser pipe, torque arresters. $\endgroup$ – Sherwood Botsford Apr 9 '15 at 18:36
  • $\begingroup$ That said, you have the germ of an idea. A floating buoy that was powered from the surface might work. Turn the power on, it goes Chirp! The issue then is keeping the line from tangling $\endgroup$ – Sherwood Botsford Apr 9 '15 at 18:39
  • $\begingroup$ A final way would be to put the chirper at the bottom of the well. This way it doesn't move. Speed of sound is much faster in water than in air, so as the water gets deeper, the time goes down. The gotcha here is that water will couple well to the casing, and you may get sound propagating along the casing with very little signal making it through the water air interface. Speed of sound is even faster in steel, (another factor of 3) so putting a time gate on the microphone may be sufficient. $\endgroup$ – Sherwood Botsford Apr 9 '15 at 18:42
0
$\begingroup$

So far the best commercial unit I've found for this is this one:

http://www.enoscientific.com/well-watch-600.htm

They use a 60 hz audio signal, with some signal processing to reject false returns and do edge detection. It records to a flash card, has a USB port for downloading, and optional radio access.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.