Remote sensing techology for leak detection

Hello, I was wondering if anyone had experience in remote sensing of leakage in water distribution systems using drones.

Preferably a technology that doesn't require the infrastructure to be in an arid region. TIA!

Taxonomy

• Leakage Detection
• Leakage
• Leakage Reduction
• Drones

15 Answers

1. We are talking about leakage in water distribution systems and a major issue is small, nonsurfacing leaks that do not produce a significant pressure differential.  Because these leaks can run for months, water loss is significant (and some eventually cause damage when the leak evolves into a major break or undermines the soil)  For nonplastic distribution pipe, correlating acoustic monitoring is a well practiced technique available from several vendors (including Echologics, Guterman, Aquarius Spectrum).      

2. Pipeline leak detection by mass / standard volume balance is the only truly rigorous method, but is currently   virtually impossible due to very lax measurement practices.  In high value petrochemical applications, it has proven to be of utility.

   In nominally incompressible flows, the computationally lowest detectable leak will be approximately 2.5*(calibration uncertainty + meter repeatability).  Once leakage loss is detected, the location is approximated via downstream-upstream pressure drop convergence techniques. 

3. Mr Wu Hao is correct about drone collection of noise logger information.  However, it is far more common today to use AMI networks that typically transmit meter data through radio frequency or cellular systems to collect this data.  

   1 Comment

   1. Spot on David - 24/7/365 not just when a drone is in the air. Frank is right too that good staff, well trained and motivated, with some simple kit can (and regularly do) achieve fantastic results

4. The leakages need to be pinpointed，are invisible and underground. Drones might be useful to survey water consumption points along the main.

   Or drone can reveive radio data sent by Noise loggers inside valve chambers. Those radio data can be used for leakage detection.

5. The leakages need to be pinpointed，are invisible and underground. Drones might be useful to survey water consumption points along the main pipe.

   Or drone can receive radio data sent by Noise loggers inside valve chambers. Those radio data can be used for leakage detection.

6. The use of drones is promising however it is important to be aware of restrictions imposed by the FAA especially over populated areas by nonrecreational users.  There are major restrictions.  That said, for transmission mains running through largely open spaces, the potential is there.

    https://www.faa.gov/uas/commercial_operators/

7. Utilitis use satellite tech, as highlighted here by others, and were at recent Aquatech talking on this. Also VertexAir have worked with UK water companies on leakage to:

   undertake subsurface water leak surveys using UAV mounted thermal imaging camera. On cold mornings (0C-2C) any escaped water (10C-12C) is clear on the thermal image. The benefits of using a UAV mounted thermal imaging camera is that it allows survey of large areas in a short space of time and with access typically difficult or inaccessible areas such as rural land.

   Might be helpful 

8. Thermal Imaging and Remote Sensing using Satellite imagery technique (such as Utilis) have been tried.  However, drone technology could be compared with the thermal imaging using helicopters.

   THe success much depends on the accuracy of pin-pointing leakpoints and just finding areas of likely leakages is not so important to arrest losses.  Therefore the accuracy should be +/- 50 cm, i.e. half of the pit width.  Also how many invisible leaks are in the system?  It should be found what is the efficacy of locating invisible leaks.  If the pressure in the system is low to the tune of about 1 kg/cm2, it will be worth continuing with the conventional electronic equipment.  In any case, the conventional equipment can't be set aside.

   Now a days, some people are also claiming that leak detection is possible through hydraulic modelling. One should weigh such claims vis-a-vis the usual methods in vogue ans ask the modellers to come and detect leaks for you at site.

   Kindly Note : THERE ARE NO SHORTCUTS TO LEAK DETECTION AND BUILDING UP PERSEVERANCE IN THE TEAM IS IMPORTANT.

9. Hi, lot of interesting answers. Yes, you can use thermal imaging, I was recently and INTERGEO 2019 and saw really impressive drones with long flying time which is necessarily to cover large areas.

   However, my suggestion is to start simple and slowly build complexity - and that is to use available satellite images that are free.

   Landsat 8 has thermal band with resolution of 90 meters and its revisit time is 16 days. Sentinel 2 A and B have 10 meters resolution, no thermal band, but more other spectral bands and 5 days revisit time. Images are free to use.

   If you want to learn how to get those images and learn how to use them here is the online course

   https://groundwater-modeling.teachable.com/p/remote-sensing-basics-with-application-to-water-resources-geology

10. Really the answer comes down to this, it’s all about data collection and transmission. First determine which method of leak detection your current system monitoring equipment is capable of interfacing with.

    Once that has been determined, review your options and choose the most appropriate for your system needs. Then you can determine if drone use will be an effective means of collecting & relaying system data. 
    Don’t over think or over engineer a solution that may cost more than You’re saving.

     

11. I have heard of it, and saw a case done by helicopter (expensive) using thermal imaging. With thermal imaging, the water will have to be a different temperature than the surroundings. It would also only detect leaks coming to the surface. Leaks going down into sewers or drainage systems would likely be missed. I'd think depth of the pipe would be a factor as well as deep leaks are even harder to detect acoustically than shallow ones from ground level. I'd think it would be a good shotgun approach survey tool, but you would need others such as loggers and acoustic devices to fully survey.

    1 Comment

    1. Not quite Mark, leaks that do not surface do change the temperature profile in space and/or time.

12. Hi Jonathan, yes it can be done; I've seen some very impressive results - and some not so. It would/could/should work well within the right parameters... Happy to help just contact me. Regards Nick

13. There are several companies offering this kind of approach.
    One of them that does leak detection via analysis of satellite pictures is: https://utiliscorp.com/
    From what I heard, assuming the water network maps are up to date, their solution can provide an indication about existence of a leak in a certain area but you will need local leak detection equipment to do accurate pinpointing.

    1 Comment

    1. I was also about to suggest Utilis.  I have never used them personally but they seem to have achieved some great results on the scale of entire cities, finding leak hotspots (?damp spots, I guess?).  As Aviv says you then need other technology to pinpoint the leaks - acoustic, pressure anomaly etc etc.

14. I'm no expert, but I'm told that thermal infrared on a drone can detect leaks because the surface remains cooler than normal in the vicinity of the leak. But it's less effective / difficult if the leaks are deep

15. Hi. I was not part of the team, but I am aware JOAT carried out a pilot project using satellite technology on behalf of the City of Cape Town during the recent drought

    1 Comment

    1. If you are possibly interested in a non drone device Ketos offers a remote sensing device called the Wave that provides real time monitoring of water metrics including leak detection with automatic shutoff, flow, pressure, usage and volume. Bill@Ketos.CO