Measuring river depth, water level and flow speed (farm flooded)

[MrOmnos]

Hello guys, how have you been.

So, There is a small stream that goes through our property. Our country has rapid elevation changes in quite small distance, we live in a small valley. Sometimes it rains in distance, and stream gets flooded rapidly. Last night we had some rain uphill and suddenly the stream was flooding. We live quite far from the stream so, our house is not at that much risk but our animals have sheds near the stream, since it is easier for us to get water from the stream. Stream is not big enough to wipe away the sheds but I would not like to take any risks. I only come home once in a while and I would like to make sure my father and my animals are safe when I am not around. I would like to have an early warning system about 2-4km upstream from the sheds which would alert my father on his phone when the water level upstream rises unexpectedly. I need something that can measure stream depth and water level. I tried ultrasonics sensors from Arduino kits but they are not good enough for my application. Has anyone here worked on something like this before. I have decided to spend $300-$500 (USD) on this and I am looking for a sensor that can measure depth of the stream and if it can measure flow speed then it would be best.

[nctnico]

How about using a floating switch like they use on pumps and lots of wire? If the level rises the switch closes and you can get an alarm.

[CatalinaWOW]

If your situation is like mine you need to think about how to make the sensor (and its communications) survive the conditions you are trying to warn about.  Floats are simple and robust but can easily be washed/scoured away by high speed water and its accompanying debris.  As long as a fault condition (open circuit detected) is reported that might be ok.  A buried pressure sensor might be an option.  If you have the option of a bridge or use of an existing bridge the ultrasound approach might work, but the surface of flood waters is chaotic and noisy.  Reflections are bound to be very different under those conditions.  Another option might be a simple vane which is heavy enough to ignore wind and similar false alarm sources with a switch to detect movement.

[jmelson]

Get some large PVC pipe, like 3" drain line.  Glue ends on.  Then, put it in a basket or maybe a larger size pipe, with some holes drilled in it to let the water in.  Drive the pipe into the river bed, or clamp to convenient structure.  When the inner pipe floats upward, it trips a microswitch to send the message.

Jon

[nctnico]

Get some large PVC pipe, like 3" drain line.  Glue ends on.  Then, put it in a basket or maybe a larger size pipe, with some holes drilled in it to let the water in.  Drive the pipe into the river bed, or clamp to convenient structure.  When the inner pipe floats upward, it trips a microswitch to send the message.

I second this suggestion. And it doesn't need to be in the middle of the raging water. Just along the side where it is somewhat sheltered. A steel pipe may be a better solution because PVC probably gets brittle over time due to the UV from the sun.

[mleyden]

Have a look at these guys for ultrasonic level sensing: https://www.maxbotix.com/

Nice project described here, using a pressure transducer and mobile: http://riverspy.net/buildhistory.html

[Beamin]

Is there a basic telemetry device that can do this like a radio Arduino? I actually bought one but it doesn't have a lot of support for noob's and seems only to be short range "in your house" type radio. Since you are in Nepal I'm guessing you can transmit how ever you like since you don't have a FCC.

As far as the phone goes there is a project where some one takes an Arduino attaches a door open sensor and it connects to wifi and them a server that sends text, that's half of it but that's not going to go 1 or 2 km away! But it would be a start. You might have to make a box that lights up or sounds an alarm if the text and internet are not available.

[Mr. Scram]

Get some large PVC pipe, like 3" drain line.  Glue ends on.  Then, put it in a basket or maybe a larger size pipe, with some holes drilled in it to let the water in.  Drive the pipe into the river bed, or clamp to convenient structure.  When the inner pipe floats upward, it trips a microswitch to send the message.

Jon

Exactly my idea. Some kind of pipe with a floaty bit inside.

[MrOmnos]

Hi, we have cellphone coverage here, that is why I was thinking GSM module with a simcard. if that doesn't work, may be invest in a LoRa network that could also be used by future generations in the region.  But for now, communication is not my issue but sensing part it. Thank you

[magic]

Power should be your issue too. Maybe a waterwheel

[beanflying]

Float switches in all forms will fail or be torn away so don't. Been there and done that 30 years ago for work and a PITA.

There are now some very good 'Radar Level Sensors' used by water boards would suit your application. A Lora based solution of some sort for the data and mount the lot on top of a davit away from the fastest flow. Anything evilbay/aliexpress is not really where you want to look for the Sensor you need to look at people doing environmental logging or fluid control systems.

[hendorog]

I like the idea of a covered pressure sensor mentioned before, or a capacitance sensor.

Neither of these have large moving parts which can be restricted by crap, or significant holes which can be blocked by crap.

[Brumby]

I like the idea of a covered pressure sensor mentioned before

I first liked that idea, but then all you need is for a big rock to get rolled onto the spot to cause problems.

[beanflying]

There is a problem with anything you put into potentially fast flowing streams or rivers. Rocks, branches and silt as well as just the simple pressure of the water will break, silt up and otherwise stuff up at the worst possible time. The business I used to work for did a roaring trade after even mild flooding in replacing pump suction lines.

Low pressure activated switches universally suck too on top of the other issues above and are on or off only. Capacitive multilevel sensors are great for all sorts of industrial uses from clean to sewage and limited chemicals but it is still a protrusion into the flow and subject to the same damage.

[MrOmnos]

Have a look at these guys for ultrasonic level sensing: https://www.maxbotix.com/

Wow, thanks a lot. I found my sensor.

https://www.maxbotix.com/Ultrasonic_Sensors/MB7363.htm

I am going to try this one. It fits my budget and has range of 10m which means I can fit it on pole like the above picture. For power a small solar panel with 12v lead acid would do. If I have any further questions I will let you guys know.

If you guys ever get time, please visit Nepal. Nice place to trek. Do visit Mt. Annapurna. I love Everest but Annapurna, it just a different experience, especially when the sun is setting.

[soldar]

First thing I would do is build a box or well with bricks and cement so you are measuring water level in quiet water. 

A float sensor to detect a certain level has been reached is trivial to build. A proportional signal requires a bit more complication.

Transmission point to point over 2-4 km can probably be done directly via radio without need for GSM although if GSM is cheap and reliable it probably makes sense.

You will need electrical power at the sensor/transmission point.

This seems an ideal setting for installing a long header pipe and a turbine electrical generator.

[electromotive]

I would like to have an early warning system about 2-4km upstream from the sheds which would alert my father on his phone when the water level upstream rises unexpectedly.

That would only be useful in the event of a Tsunami. It wouldn't help in the event of a flash flood. By the time you got the message, it would already be flooded near the sheds. My degree is in geology of all things. Ultrasonic is quite literally the industry standard outside of visual reporting, although Arduino sensors aren't going to cut it. You need something more like this Senix ToughSonic 30 (disclosure: not my listing or anyone i know): https://www.ebay.com/i/272490641948?chn=ps

[MrOmnos]

I would like to have an early warning system about 2-4km upstream from the sheds which would alert my father on his phone when the water level upstream rises unexpectedly.

That would only be useful in the event of a Tsunami.

Hah, last time we had a Tsunami was in 2012 and it reached the Everest. Anyways would you be able suggest me any reading materials that I should read before going into this project. I would like to be well informed on the geological sides of things before I do this. I wouldn't want to waste my time and money on something that would not be effective.
Thanks

[beanflying]

In your watercourse does the water mainly come from Monsoon storms or from Snow/Glacier melt?

If it is Monsoon events adding a rain sensor to the system might be a benefit as an even earlier warning. Also if possible take in the most local feed from your nearest government Meteorological station if that is possible.

Interesting Nepal has some community programs in regard to Flood alerts. Might be some funding or even opportunity to come up with a system for more than just you? http://www.hydrology.gov.np/#/?_k=uhqckz In particular http://hydrology.gov.np/cm/files/Establishment%20of%20a%20regional%20flood%20information%20system%20in%20the%20HKH-region%20(HKH-HYCOS)_1509011430589.pdf

[Zero999]

If your situation is like mine you need to think about how to make the sensor (and its communications) survive the conditions you are trying to warn about.  Floats are simple and robust but can easily be washed/scoured away by high speed water and its accompanying debris.  As long as a fault condition (open circuit detected) is reported that might be ok.

If a one shot solution is acceptable, then how about a float fixed to a connector which is unplugged, when the water level rises?

Use a relay with normally closed contacts and a coil of low enough resistance, that the small current flowing through the water isn't enough to keep it energised. A lower voltage unit would be better say, 5V. At higher voltages, the risk of it remaining on, due to the current flowing through the water, will be higher.