detecting the presence/absence of water in a pipe

Discussion in 'The Projects Forum' started by latunga54632, Jun 18, 2014.

  1. latunga54632

    Thread Starter New Member

    Jun 14, 2013
    7
    0
    I am working on a project where I have to detect when there is no water in a pipe. This is water in pipes from boreholes that are located far away from town. So I'll be able to detect when a borehole pump is off if there is no water in that pipe. I wanted to use ultrasonic clamp water flow sensors but according to this link http://www.omega.com/prodinfo/ultrasonicflowmeters.html , they don't work well for drinking(pure or distilled) water. Is there a simple circuit or something I can use to detect if water is not present?
     
  2. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,548
    2,373
    Is the water in boreholes going to be that pure?
    Max.
     
  3. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,800
    1,104
    Can you be sure that the pipe empties when the pump stops? If not, the problem becomes one of distinguishing between flow and non-flow.
     
  4. Externet

    AAC Fanatic!

    Nov 29, 2005
    760
    57
    More information/disclosure is needed.
    If the situation is need to know the presence of water in the municipal mains in order to start a water pump to fill a raised household reservoir or similar use*, and to know when to turn the pump off when if is no water to be pumped, would be good to know if the point of detection is in plastic or metal pipe, if it is accessible or underground with no access. Is the pump started/stopped manually ?

    Many ways to implement the indicator depending on the missing details.

    * If it is for the situation described above, with no access to underground municipal pipe to install a sensor, the best I can suggest is a pump current sensor, or an ampermeter.
    If the pump is manually turned on/off (and that is why you want to know if there is water)
    - Turning the pump on will show a low current if there is no water.
    - Turning the pump on will show a higher current if there is water.
    Watching the ampermeter will tell you to turn it off, or not.

    A Hall effect/current sensor in the pump power lead and a comparator circuit can decide to turn off the pump when supply pipe empties, responding to lower motor current.
     
  5. latunga54632

    Thread Starter New Member

    Jun 14, 2013
    7
    0
    @Externet this are underground boreholes located quite a far distance out of town but they are the one supplying the town with drinking water. They all pump the water to a main reservoir where the water is dozed (treated). Now the problem is that this pumps go off (unexpected breakdown or power failure) at random times , so people have to drive there 3 times in a day to go check if the pumps are off. So I need to detect when a certain borehole pump is off then send this signal through a GSM modem to their office, than they don't have to be driving there unnecesarrilly. These pumps run continously unless disturbed by power outage or pump breakdown, they dont get switched off. So I thought if i there is no water in the pipe from a pump than that pump is not workin. that is why i thought of detecting the water in the pipe.

    @Alec_t the water stops flowing completely in pipe a few minutes (about 20 minutes) after pump is off.
     
  6. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    how about a pressure switch?
     
  7. inwo

    Well-Known Member

    Nov 7, 2013
    2,433
    315
    Pipes feeding a reservoir will always have water in them. Unless they are top fed and have no check valves. Not a normal situation.

    The pumping stations I work on use the tried and true vane type flow switch.

    Does it have to be external sensing? No tapped port or process fitting?
     
  8. crutschow

    Expert

    Mar 14, 2008
    13,016
    3,235
    Why can't you just detect the pump motor voltage?
     
  9. Externet

    AAC Fanatic!

    Nov 29, 2005
    760
    57
    OK. If the pump is pushing water to the reservoir first, there has to me some pressure that can be measured at the pipe feeding the reservoir, and/or the reservoir level. But as there has to be a check valve, measuring pressure there may not tell there is flow.

    Checking the pressure in town with a simple sensor, will tell the reservoir has already emptied, not a good strategy, no prevention measures.
    The flow has to be sensed at the reservoir filling pipe.
    It is not as your post title "presence/absence of water in a pipe" It is about water flowing or not.

    A plain vane at the reservoir inlet can move a switch to tell there is water coming in, and that switch can turn on whatever signals the dispatch office. Can be even a simple radio transmitter, carrier only, or tone. When the signal is received in town, water is not entering the reservoir.
    Then the crew should go to fix the trouble.
    If there is a power outage at the pump, another radio signal can tell the crew to not travel as they cannot fix it, but wait for power to return.
    Is there visual from the reservoir to town ?

    Simple warning flag, beacon, whatever can be implemented on the reservoir site if it is in visual range.
     
  10. wayneh

    Expert

    Sep 9, 2010
    12,126
    3,048
    +1

    I'd go back to the source of the problem and monitor that. It might even be nice to have a current-draw measurement to keep an eye on the pump performance. Put a Kill-a-Watt on it.
     
  11. crutschow

    Expert

    Mar 14, 2008
    13,016
    3,235
    Yes, using a motor current sensor (current transformer or Hall-Effect) would give a more accurate indication of the pump operation than voltage. Too low or too high a current reading would indicate something is failing or has failed.

    Don't know how you'd remotely monitor a Kill-a-Watt device. :confused:
     
    latunga54632 likes this.
  12. #12

    Expert

    Nov 30, 2010
    16,298
    6,810
    Better to reverse that logic. Tone must be received to say that the water is flowing and the transmitter is operating correctly. Missing tone indicates a failure of the flow OR the transmitter. It's a basic fail-safe design.
     
    latunga54632 likes this.
  13. crutschow

    Expert

    Mar 14, 2008
    13,016
    3,235
    I second that. ;)
     
    latunga54632 likes this.
  14. latunga54632

    Thread Starter New Member

    Jun 14, 2013
    7
    0
     
  15. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,800
    1,104
    Monitoring pump motor current gets my vote. An alternative would be a vibration/hum sensor on the pump outlet.
     
  16. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,392
    1,605
    Just to point something out... the ultrasonic flow sensor latunga originally looked at is close to the best sensor, but there are others that would work here. I used to be in the ultrasonic world and we made such sensors to simply detect presence or absence of liquid. Most of the work was on tiny tubing for medical uses (or nuclear submarines) where we could actually detect bubbles above a certain size.

    These sensors go on opposite sides of the tube. It’s non-invasive, meaning you clamp to the outside of the pipe. I have some memory of steel pipes having a side issue where the signal could go around the tube in addition to thru the tube, that has to be dealt with or it always looks wet.

    A complete path of liquid from transmitter to receiver is necessary to get then to detect liquid as the signal will bounce off air and not get from one side to the other. Assuming a horizontal pipe then mounting this on top to bottom may give false negatives if the pipe is not completely full or pumps an air bubble down so a side to side.

    Just putting this out there, measuring the motor current may be the simpler approach in this application.
     
    latunga54632 likes this.
  17. Externet

    AAC Fanatic!

    Nov 29, 2005
    760
    57
    The pump current sensing is a convenient way, as expressed in post #4; transmitted on telemetry.

    But there will be ocasions where pump works fine and there is no water in the reservoir : pipeline cracked.
    Still believe sensing the reservoir inlet is more important. amount of flow must be known; or several simoultaneous monitoring approaches for power presence, current, pressure, flow, level.

    One of many ----> http://www.rshydro.co.uk/data-acquisition.shtml
     
    inwo likes this.
  18. wayneh

    Expert

    Sep 9, 2010
    12,126
    3,048
    I completely agree with the wider point. My only quibble would be with the word "more". Which process parameter is more important depends on a bunch of statistical issues of the particular installation; things like which failure mode is more common, more costly, and so on.

    The OP said his problem is that the pumps turn off randomly. You could argue that fixing that might be wiser than monitoring that.
     
  19. crutschow

    Expert

    Mar 14, 2008
    13,016
    3,235
    The motor current will significantly drop when it's not pumping water so if you detect low current (not just no current) you should cover both cases of no water in the pump or the motor is off.

    Measuring too high a current will indicate if their's excessive load on the motor such as a bearing going bad.
     
  20. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Bore pumps are usually reciprocating, so instead of measuring the motor current I would measure the reciprocating movement.

    A reed switch or other pulse sensor would work fine, and if microcontrollers are not an option a 555 monostable can be used as a "pulse is good" detector.

    Personally I would use a microcontroller as the pulse frequency could be a very good indicator of pump good/bad status. ie; with an empty or clogged pump piston the frequency would change.
     
Loading...