Hello Every body, i need to use a low cost sensor to measure in real time the level of a 150m deep well, anybody knows about it??? its possible to modify an ultrasonic sensor to get that range????
What accuracy do you need? In what range does the level change? What kind of access do you have to it? And define "low cost"!
Hello pylon, i need a precision of 1m, tha range is between 20m to 130m, i have a 2" diameter hole and low cost is no more than $100US
Thin rope, with a weight and a float works well.
fsalazar:
Hello pylon, i need a precision of 1m, tha range is between 20m to 130m, i have a 2" diameter hole and low cost is no more than $100US
Wow! if the casing is 2" inside diameter, what size is the pipe going down to the pump? The wells I have been around are 6 inch casing and 1 1/4 inch pipe to the pump. The actual volume of water available in your well is really, really small.
Paul
I'd explore the option of putting a waterproof pressure sensor at the bottom of the well.
To clarify, are you saying the level of the water is between 20m to 130m below ground reference point ?
That's quite some depth and a lot to ask from any sort of cheap sensor.
Ultrasonics will not work in such a small diameter hole for the depth you have, don't even try.
For this sort of application, normally you would use a suitable pressure sensor.
But, you need to overcome some challenges such as how will you connect to a pressure sensor, that needs to be at around 150m below ground reference point with around 130m of head of water above it ?
At that depth, water ingress will be a concern with any cable connection. Any cable connection will need to be professional. A basic pressure sensor that may work down at that depth and provide the sort of signal you want will alone cost a few hundred dollars.
The cable will be long, and so you either use a current loop, so something like standard 4-20mA signal, or you use a smart sensor that has EIA-485 communications.
Very quickly you will find that this is going to cost you more than $100 USD.
Here is what I have and use, it is a Level Troll 400, it has things like 4-20mA as well as EIA-485, in fact, it is a complete data logger is built into these devices. Level Troll 400
These are used in well yield test pumping applications.
Senors like this don't come cheap, by the time you have cable and connections, you are looking at around $2000AUD. Have a look on that Troll 400 page and you might get some insight into what you might need to think about.
I really do not know of any feasible way to do this for $100USD that is going to use electrical signals, hence the thin wire rope idea is about the best for that sort of money.
Paul - VK7KPA
If you can restrict your want to "tell me if it falls below xx, so I can turn pump motor off", different options would open... long string, float, weight, switch.
long string, float, weight, switch.
If you add "rotary encoder" to this bill you'll have almost exactly what the OP wants. It's cheap, the accuracy of 1m is easily accomplished and the range should also be no problem. Putting all that into a two inch casing might be a bit more of a problem.
Why are pumps being talked about when the OP hasn't mentioned it ?
When you think about it some more, a 2" (50mm) hole is not going to accommodate a pump for such a well. Even if there was a pump, then there would be no way to have a level measuring sensor in the same 50mm hole because the pipe that connects to the pump would interfere with any sensor. And even if there were a pump to do this, it would likely cost many thousands of dollars, so why limit the level sensor cost to a measly $100USD ?
I believe this is a hole where the OP wants or needs to measure the depth of some liquid, maybe water, as no mention was made if it was water of some kind.
Pylon, yes, you could use a rotary encoder which would have very good accuracy, but, how would you envisage the mechanical side to work ? You simply can not have a string with a float on it going to a rotary encoder, and having a weight on the other side to counter the float wouldn't work either.
The only way to use a rotary encoder with not much else it to use solid rod that will rise with the well level and cause the rotary encoder to turn. But, this in itself is not possible for this situation. The rod would need to be 120m long and the weight of that alone would preclude this sort of idea.
Using a rotary encode suddenly becomes a lot more difficult.
The other way to use a rotary encoder is to make use of a stepper motor as well, which would wind the cord to the float.
It would need to be a large stepper motor to be able to, again, lift the weight of the level cord with float, so again high cost and complexity. Added to that you need to determine when the float was on the liquid surface and not above it, which again is not an easy thing to detect.
All of this means high cost and frustrations with an idea that really was not going to work form the start.
I see the only way for the OP to measure the level in the well is using a pressure transducer.
Either pay good money for a proper purpose sensor like I pointed to or try and mess about with a standard industrial pressure sensor giving 4-20mA output. This sensor will be encased in stainless steel and for the depth, will fairly expensive.
I don't believe the OP will attain what is asked for for a mere $100USD, I will be very surprised.
Paul _ VK7KPA
As is almost always the case, OP has told us too little about the DETAILS of the circumstances of the problem, and the PRECISE nature of his wants... but this is understandable. But it does lead to people going with their best guesses, in order to make some kind of sensible constructive comment.
The shaft encoder could easily be made to work with string. You just pass the string over a pulley, and attach weight to the free end. Of course, the narrow bore of pipe is a problem, and the range of water levels which could be reported is limited... but if well were 100m deep, and water level was always at least 50m below ground, such a system would work. (Yes, I know OP gave his dimensions. Not needed, to make my point, are they?)
What is "real time"? Can you afford to dip the well once a day, or once a millisecond?
tkbyd:
The shaft encoder could easily be made to work with string. You just pass the string over a pulley, and attach weight to the free end.
Yes, except the free end needs to have at least 110m of travel to move up and down. You would need a tower at least this tall to stand over the well. Think about the weight of rope dangling down the well. 100m of rope will have a significant weight - greater than the buoyancy of any reasonable float. So the 'counterweight' needs to change depending on how much rope has gone down the well.
And putting the counterweight inside the well (using that 150m of free space) won't work because the specified dimensions mean the weight must pass the float at some point. Then part of your counterweight rope is in the unknown depth of water and its buoyancy must be taken into account.
What about a free float and a very carefully aligned laser sensor?
So, instead of using a pressure sensor on the bottom as suggested, another approach is to use a long vertical PVC (or flexible) pipe. To build the sensor close one end and place the pressure sensor inside it, on the closed end.
Now you have a sensor that measures the pressure inside the pipe correct? If you "blow" the air pressure inside the pipe will raise and you can measure it.
The next step is to submerge the open end inside the water. Make sure you submerge it vertically so no air will escape from the pipe. That's it... if the water level raises the pressure will raise as well, if the water level lowers, the pressure will reduce.
Another option is to submerse with both sides on the pipe open and close the "top" after half or all pipe is submersed, so you'll create some low-pressure/vacuum when the water level drops... 150M is quite a lot, so make sure you build a strong setup and plan if you will submerse as a single piece or assembly the piece as you submerse it.
You can find a variety of pressure sensors, just need to calculate the min and max pressure you can have inside the pipe to find an appropriated one. But you should be ok to experiment with the popular BMP180....
ps.: Could you draw a model how the well and the 2" access looks like, a "side" and "top" view would help.
A continuously-blown sensor would be a good idea. Any mud at the bottom of the hole would always be blown out. Just set up a small air compressor at the top and measure the output pressure. 130m of water is 13 bar pressure, or 190psi. Some tire compressors can do this. You don't need a lot of air, just occasional bubbles.
Schedule-40 pipe can be used to hold this pressure. Choose a small diameter. 1/8" nominal diameter is going to be about 50kg of steel pipe for 130m. 1/4" is going to be twice the weight. You could also use thinner pipe or PVC for the bottom sections that don't have as much pressure (or weight hanging off them.) Just assembling the pipe and passing it down the hole will be difficult, but the pipe itself isn't expensive. You should probably have a plan in place for retrieving a dropped section of pipe, up to the maximum calculated weight.
I haven't done the calculation for pipe under pressure and tension at the same time. It might burst. This would be a good idea to check for your chosen pipe first. Even metal pipe will float if it's full of air like this one, so also consider that it may be trying to float up out of the well when the compressor is operating.
It may also be possible to do this with small diameter pipe that comes on coils. Just unroll it and feed it down the well. Floating will be the main problem - you won't know exactly how deep the end of the pipe is.
All interesting ideas, but from my experience, I would not choose to do it this way, to use an air pressure sensor.
The BMP180 will be no use at all, how would you couple a pipe to this small sensor ?
Plus I think such a sensor would not cope with the pressure range.
What about the effect of atmospheric air pressure changes, which can have significant effects on any sort of pressure sensor if not accounted for by using a compensating pressure sensor at ground level.
Are we to assume the well has a bottom at 150m, it could be that it extends far deeper, as all we are told it is a 150m deep well. Like was mentioned, we can give lots of advice based on assumptions and the little the OP has provided in terms of details.
I've just finished commissioning a project for an automated rig for well pumping down to 150m depth with real time level and pump control all mounted on a 8 wheel drive truck for remote sites in Australia. Not including the vehicle, the costs were in the order of half a million dollars. I know this is different, but once you understand how far 150m is down below ground level you get some insight into the difficulties of the situation, regardless of the sophistication of the components used.
Thanks MorganS for bringing some sanity 
Anyhow, I wonder if the OP is still interested ?
Paul - VK7KPA
Hey @rockwallaby, very likely BMP180 wouldn't be able to cope the pressure as it's designed to measure atmospheric pressure. As I said, that would be just for proof of concept. Such sensors can be found for many range of pressures (http://au.element14.com/pressure-transducers and http://au.element14.com/pressure-sensors).
Any atmospheric change would not interfere as the sensor would in within it's own "atmosphere", here how it would look like:
┌────────┐
│ Air │
╔══════════╡ Box │
╔══════ ║ ══════╗ │ Θ───┼────[Micro-Controller connected to a sensor inside the AirBox]
║ ║ ║ └────────┘
───║ ║ ║──────────────── Ground Level ────────
║ ║ ║
║ ║ ║
║ ║ ║
║ ║ ║
║ ║ ║
║~~~~~~~║~~~~~~~║
║ ║ ║
║ ║ ║
║ ║ ║
║ ║ ║
╙───────────────╜
I've added an external chamber so the pipe pressure gets "diluted" on a bigger volume and you don't need to support higher pressures. Probably the only concert would be temperature which could affect the closed system, but this can be compensated.
Now, as a closed system, you just need to measure the pressure inside the "Air Box". It sure needs to be calibrated and properly sealed. I guess the biggest issue is the self-support for the down pipe.
Regarding floating, the pipe doesn't need to be full or air, you could let both ends open while installing so you you have the pipe filled with water before sealing it. If the well water level drops, the system pressure is reduced. Just need to make sure the bottom end of the pipe never gets off the water.
Apart from the scale of the project (pipe sizing) and the initial calibration, I can't see major issues. For sure it's not an industrial grade solution but I guess for $100 is the best you can get.
The pressure sensor should be a differential type. That way it's always referenced to the local atmospheric pressure, so there's no compensation required. A box should not be required - the volume of the pipe is big enough.
The pipe must be totally filled with air when operating. A compressor fills the pipe until air starts to bubble out the bottom end of the pipe. At that point, the pressure will be equivalent to the depth of water from the water surface to the end of the pipe. You can even tolerate some leaks in the pipe, so long as the compressor is pumping enough air that it has some bubbles coming from the open end of the pipe. A "sealed" pipe, where you measure the pressure from pushing it into the water will not work for the long term. Leaks will always occur and the measurement will drift.
For a short term 'dip' measurement, the sealed pipe would probably work well enough but you may as well drop a weighted rope and look for the wet patch if you're doing that.
@musskopf, I really don't know what you would achieve with a BMP180 sensor for this project, except playing around with code and using it on the work bench only. As I say, they is no real way of connecting the BM180 to anything except to open air.
And then the possibility of condensation within the sensing environment is high, which does not bode well for the poor BMP.
Like MorganS writes, you really need to compensate for atmospheric pressure.
The recent project where I was doing commissioning tests had a well, that was around 700m deep, fully cased to around that depth. Now, what was interesting was noticing the water level on different days as the atmospheric conditions changed.
Most days, when we unbolted the well lid, water poured forth in great abundance, gushing over the rim of the well case, about 600mm above ground level. This indicated a low pressure atmospheric condition. Then, a few days later, the water was a good 3m to 4m lower. It made commissioning very difficult, not to say creating a lake at the test site most days.
Depending on the site conditions of where the OP is wanting to measure, there may well be similar conditions, where the atmosphere may well have an effect on the aquifer table quite some distance away from the actual well site.
Any ocean nearby may have tidal effects as well on the aquifer, as do lakes.
What this means is that the water below ground level is very unlikely to be sealed from any atmospheric conditions above ground, and this is why it is important to use differential pressure sensor if using this concept.
Now, rather than playing with a BMP type sensor, it would be more worthwhile for the OP to use this Honeywell sensor.
They come as single or dual port and come in various pressure ranges. They have temperature sensor for temperature compensation and use a 14 bit internal ADC for the I2C variety, PDF link: Honeywell ABP series pressure senors.
They can be purchased from RS and maybe Element14 for around the $20AUD I recall.
I have some on the workbench for a new PCB being designed based on a 1284p MCU, and they appear quite rugged, and are used in various industries.
Using such a sensor allows the electronics to be keep out of the environment of water and condensation.
The Honeywell sensors can also be purchase with a gel membrane for situations where moisture is present.
All up, I am not convinced using any form of air pressure sensor will be suitable, though the idea may be worth a small scale test, where I think it will be found that too many difficulties will present themselves, in terms of calibration and accuracy and repeatability of data, which will only be magnified when applied to the scale the OP wishes.
Speaking which, where is the OP ?
@musskopf, I looked at your site the other day, still have it open a few tabs across to the right, very interesting indeed
Paul - VK7KPA