Water pressure sensors?

[JohnRob]

@mfalkvidd,

I'm guessing this is not a hobby request (based on the photo of the manhole cover). I've looked at pressure sensors for level detection many times for automotive applications and can offer the following:

1. o-rings don't seal. For your application you need either a hermetic seal or fully potted assembly

2. open tubes will fill / drain based on condensation temperature changes etc.

3. Freezing is often an issue but I'll guess at you depth you don't have to worry about it.

Have you considered a pressure switch? It wouldn't necessarily solve the seal problem but might work for you. Still thinking.... the pressure switch might be hard to self test. At least on the pressure you can see small changes, suggesting the transducer is functioning.

you might find this link useful: submersible transducer

If you are looking for a lower cost solution, look for "absolute" pressure sensors as opposed to "gauge" pressure sensors.

If you want to go the potted route, you might look at this: [link text](absolute xducer). If the cabling was completely potted in a suitable material, this could work for you.

[NeverDie]

Maybe just measure the height of the effluent? It should be proportional to the pressure. It seems like what you care about most is overflows anyway.

[RWoerz]

I use a pressure sensor on the output of the pump that feeds my solar hot water panel. I think I got mine on Amazon. They use a lot of them in automotive applications. They come in lots of different pressure ranges. On mine I think I first had to figure out how to convert the analog input port reading from a number to a raw voltage. Most important you will first need to take a reading of your sensor in open air. FYI most of these sensors will run with any gas or liquid i.e. air, oil, water, etc.

The math for the sensors looks something like this:
PSI=(Sensor Voltage - open air voltage)/7

This is why we need the open air reading. If your current voltage reading and your open air voltage are the same that means you current pressure would be 0 zero, right?

The next part the 7 is a little more complicated but remember it's just a number and I think I got mine right off the spec sheet for the sensor. If you plotted a graph for every reading from 0 psi to the sensors max pressure the slope of that line is where the number 7 comes from. Technically it's the slope of the linear regression. But we don't care how we got it because someone even nerdier than than me did the math and it's on the spec sheet. That's why we buy a new one instead of taking one off an old motor at the junkyard.

With that info I first had to determine the raw sensor voltage. The Arduino returns a value of 0 - 1023 for a voltage of 0 - 5 volts (I mostly use 5V Arduino's) so to find the voltage I use the following line in my code.

First get the raw sensor value.
float rawSensorValue = analogRead(tankPumpPressurePin); // Read Pin A15 Value range 0-1023

Then convert that value into a voltage.
float voltage = rawSensorValue * (5.0 / 1023.0);

Then to get the actual pressure I had to do a little more math with the following line. On my sensor the open air reading was 109.

tankPumpPressure = ((rawSensorValue - 109) / 7);

The whole thing looks like this.

void readTankPumpPressure()
{
// Read Pin A15 Value range 0-1023
float rawSensorValue = analogRead(tankPumpPressurePin);
float voltage = rawSensorValue * (5.0 / 1023.0);
tankPumpPressure = ((rawSensorValue - 109) / 7); // Should be in PSI.
currentTankPumpPressure = tankPumpPressure; send(msg_tank_pump_pressure.setDestination(GW_ID).setSensor(Tank_Pump_Pressure_ID).set(currentTankPumpPressure, 1));
}

I hope this helped I also take flow reading.

[RWoerz]

@rwoerz
Sorry to say you can't use a tube submerged in the water. That will only work for a very short time. I was going to use that method to take a pressure reading to see how much head (water depth) was in my water well. The problem as I was told is the air in the tube is eventually absorbed by the water ending with a pressure of zero. If you had a way of blowing the tube clear of water just before you took each reading that would work.

[zboblamont]

@rwoerz Partly true.... There are multiple effects on a compressed air pipe in water (condensation, gas absorption, etc) which reduce effective air volume and pressure over time, but these are in reality very small.
Blowing the pipe clear before every measurement is an ideal datum but is generally impractical other than in industrial locations. In reality, a periodic purge with a footpump to an inserted tyre valve on the pipe will restore pressure accuracy for many months at a time for a static well scenario.
The circumstance posed by @mfalkvidd originally is quite different in that a CSO or SSO is a surcharge arising from periodic rainstorms rather than a constant submergence. If the sealed pipe is installed at a level with free discharge when flow abates, it will function very accurately as it compresses from a natural state. All that is required is to add the height of the pipe to Invert and the actual level can be derived. The bigger problem would be installing a 10m rigid airtight pipe..

[bjacobse]

You can also measure the pressure, how much pressure to creating air bubbles

Below is written by Benoit Drooghaag (just to ensure I provide credit to the author)

http://playground.arduino.cc/Main/Waterlevel

[gohan]

I use a similar setup for my well but I used a compressor from a broken refrigerator and an old pressure gauge :D

[RWoerz]

@gohan
gohan great solution wish I had thought of that! Not sure where I got the info that it wouldn't work. They didn't say how long it would take to get bad readings.
Because the pressure transducer I used is sealed you wouldn't have the same problem. As long as the connection were made watertight the sensor could just be installed at any level and it should work without the extra pump. Assuming you don't install it with the open end pointing down trapping a tiny bit of air in the opening. Not sure that would affect the readings but why take the chance.

[gohan]

I have the plastic pipe for the compressed air pointing down so it doesn't get clogged and it runs down 60 meters with the pump. Of course the pressure reading is not very precise on an old "analog" gouge but at least I get the idea if I have 40 meters of water instead of 30, I don't really care if the actual level is 41 or 39

[mfalkvidd]

Thanks everyone for your ideas.

As most of you figured out, this is not for a home DIY project. It is not my project so I can't share a lot of details.

The "industrial" solutions are expensive (for example, the "submersible transducer" linked by @JohnRob costs $400+) and require big sensors. Often, DIY users have found clever solutions - and several such solutions have been suggested here which has given me a good perspective of the possibilities which I really appreciate. Thanks!

[gohan]

I don't know if it may fit in the project, but some of these could be used to get a rough idea of the water level

[JohnRob]

@mfalkvidd

If I were to look for a lower cost solution (vs a full industrial offering) I would talk to Epcos EU.

I have worked with them in the past and they have a small absolute pressure sensor that may work for you.

John

[dbemowsk]

I just received an email about these Honeywell PX3 series pressure transducers today:
https://www.sager.com/external-freeze-thaw-resistance/default.aspx

https://sensing.honeywell.com/sensors/heavy-duty-pressure-sensors-and-transducers/PX3-series
I didn't read into them much, but you may find a solution with them.

[ybirch]

Does anyone have an idea on cost efficient water pressure/level measurement in 1000 ft deep groundwater wells? There is a new law in California that made these measurements mandatory and my good friends have to drive some crazy mileage to do manual cord measurements of water level.

[Tommy Petersen]

If you want to measure level in a sewer and get a measurement that can be used for something "that have a high degree of accuracy" it needs to be atmospheric compensated, otherwise the measurement will be off by a lot, because of pressure in the pipe system, normally a sewer system is not open to the atmospheric pressure because of gases.

If your application is anything like the picture, that would not be a good solution, since a atmospheric sensor would have a small air tube and getting water into those would make it unstable.

if you can put your vent tube so no water gets in then a pressure level measurement would be a good solution.
I normally use alibaba for that.
https://www.alibaba.com/trade/search?IndexArea=product_en&CatId=&fsb=y&SearchText=pressure+sensor+for+sewer

But if you only want to know when the water level reach a certain height there is the option of using a float switch.

I often find that there is a need for a combine solution, both pressure/ultrasonic and a float, to keep up a high accuracy, sewer systems are a whole other world when it come to sensor technology, but again it depends on what you need.

[boozz]

@ybirch

https://www.alibaba.com/product-detail/CS-Pressure-measuring-instrument-diffused-silicon_60734702750.html

above link is an example of a sensor which could be used. Not sure what the level change will be, but the "new law" as you refer to will tell you what accuracy is needed. This sounds to me as a perfect example of a LoRa sensor.
The sensor goes in the bore hole (same as the manual cord measuremens) and stays there. The electronics (4-20 mA preferrably as 1000 ft is quite a distance) can remain at ground level.

If you have any idea what the change of level is in the well, you can easily find the range for the sensor.

BR,

Boozz

[bjacobse]

@ybirch Hi did you see the post I posted?
https://forum.mysensors.org/topic/9313/water-pressure-sensors/30
cost efficient, use a salvaged refrigerator compressor and purchase a long aqaurium airpump hose and a pressure transducer - might be fairly cheap