We finally received our UE Ultraprobe 10,000. So far it's been taken out of the case twice since Thanksgiving for orientation purposes only. During the walk throughs we have discovered lots of leaks in a 2 year old air system. This leads to several questions:

1. What dB range is acceptable?
2. What steps do you take to pinpoint a leak?
3. Someone somewhere mentioned a liquid solution, what is it?

I'm sure I'll think of more, but that'll do for now.

Thanks,

Eric

Eric,
1. There is NO magic number or db range. However, with your UE unit you should have a "GUESSTIMATOR" CHART. This chart is a wild and crazy attempt to tell you how many cfm's you are losing from a leak. However, remain cautious when using this chart. You will get in trouble with it. Why? Because you cannot take airborne ultrasound and measure a leak for cfm's, period. But, you can use this chart as a reference. You may want to say that for your plant, so many db's (or cfm's) is a small, medium or large leak. If you use the CFM's as one fellow did, and who later approached me at a trade show and told me that I got him in trouble. He found over 1300 cfm's of leaks, but he only had a 1100 cfm compressor! So you need to establish what is a large leak for you. As the saying goes, "IF you can feel it, FIX IT".
TIP: Take a wrench with you when you are looking for leaks, YOU WOULD BE SURPRISED HOW MANY LEAKS CAN BE FIXED INSTANTLY WITH A SLIGHT TURN OF THE WRENCH OR TIGHTENING A CLAMP.
2. As far as pinpointing a leak. Using the UP10,000 scan the area WITHOUT the rubber cone on the unit. The rubber cone is named "the focusing cone". Once you hear a leak and proceed towards the leak, lower the sensitivity or volume (but NOT TOO LOW), as you get closer to the leak (within 18-24 inches) you then place the cone on the scanning module (module with wire screen on front), sweep side to side focusing in on the loudest signal. If you have more than one leak in the area place a glove on your other hand and block competing sounds or use a clip board or piece of cardboard.
3. There is a liquid called "Liquid Leak Amplifier" this is a wetting solution that is meant for low-level leak detection. When sprayed or squirted upon a leak it will produce small bubbles the size of "spider eggs" (tiny bubbles) and they when the trigger is keyed on the Ultraprobe (or any ultrasound receiver instrument in the scanning mode) you will hear the bubbles popping or collasping. I often tell people I'am training, "that it sounds like a bowl of RICE Krispies, but a thousand times louder". This liquid senses a leak of 1x10-3 std/cc/sec down to 1x10-6 std/cc/sec. The ultraprobe and MOST ultrasound receivers are rated at 1x10-3 std/cc/sec. Typcially when this liquid is squirted on YOU DO NOT have to have line of sight to know there is a leak, you can hear the bubbles. For instance if you have a 2 inch pipe vertical or horizontal leaking and you squirt it, if the leak is on the backside of the pipe and out of sight, you may still hear the leak.
Hope this helps.......
Jim Hall
Ultra-Sound Technologies
www.ultra-soundtech.com

Eric,

I have not seen a universal dB-value criteria for leaks. UE may have a guide to possibly correlate dB-value to leak size, but be careful about relying on this.
Pinpoint a leak by verifying that it "sounds like a leak", move closer to get higher sound level, alter direction/angle to verify sound is not being relected off surface.
Leak verification can be by feel/touch, direct audible sound, or visual from liquid bubbles or dust/liquid traces.
Leak sound level can be enhanced by spraying water or light oil (like WD-40) over suspect location. The popping bubbles increase the ultrasound emissions. Leaks can be visually enhanced by liquid soap solution (such as Sherlock by Winton Products or your kid's bubble soap).

Happy leak hunting in the New Year!
Walt

Jim,
I really like how you put that.

My two cents: even the smallest leaks when multiplied by hundreds will pay off. Big hitters first! If the compressed air leak is strong enough to feel and not be restricted by a palm or finger grip, I label as a priority 1 and tag accordingly. Leaks that can be felt and stopped by a palm/finger grip are priority 2. Those that can be felt best by the back of the hand are three and audible only leaks are a priority 4.

Maybe when all the heads get together, a perfected method of quantifying cfm based on db's along with certified calibration will aid us all in calculating cost to repair verses cost of loss.
Good In = Good Out

Eric,

These guys have given you some really good information here. Quantifying how much air or steam you are loosing is rather difficult. Size shape and pressure are also contirbuting factors, this is why people call it a GUESSTIMATOR chart.

Jim has explained what they call the "GROSS TO FINE METHOD". This technique is used to help locate the source of the leak. Using Barriers and shielding techniques is also a vital step in testing as often compteting ultrasounds can give you strange results.

Be aware if you are trying to quantify by using your hand remember SAFETY FIRST !

COMPRESSED AIR CAN KILL YOU. EVEN AT LOW PRESSURES AN AIR BUBBLE CAN GET INTO YOUR BLOOD SYSTEM.
You do not get second chances with this kind of problem.

Try using a piece of paper or a leather glove as your guide rather than your hand.

Hope this helps,

Hooch

IIRC, this is not the first discussion of quantifying leaks located with ultrasound. However, it got me thinking --- and I believe that the flows could be measured.

I would like the group's comments on the value of being able to measure the leaks. This would involve some extra equipment and some cost, but would potentially give as good as +/- 10-15% accuracy. The way I envision it would require access to the leaking area.

There would be some cost involved.

Any value to you? Would additional cost and an additional piece of equipment be acceptable? What level of cost would work for you?

Thanks for any feedback.

Jack

We would do well to remember that when it comes to Ultrasound all of these factors affect the sound level:

• Pressure
  
• Flow
  
• Orifice configuration (the shape of the leak) = a smooth surface will generate much less turbulence that a jagged edge leak even at the same flow rate
  
• Barometric pressure
  
• Type of gas (helium is a small molecule compared with compressed air)
  
• Distance from leak
  
• Sensitivity of the sensor
  
• Angle of approach
  
• More...
  

I have seen a number of people create lab leaks that did not correlate well to the real world.

If it leaks - fix it - oil ain't going down anytime soon and if it does - it ain't staying down. Power costs for compressed air is significant. The more it runs - the more maintenance it will require as well - compounding the issue.

Make sure you confirm any leaks you "fix" as new leaks can be created in the process.


Terry O

An Idea
Since this thread started I have been contemplating a method to quickly collect leaking inert gasses, primarily compressed air, for the purpose of calculating loss. Naturally what follows is limited in applications but should work in theory for nearly all compressed air delivery systems.

Purchase or construct a given volume container to collect leaking gas within. Vent container with pop off valve set at specific pressure. (even the smallest leaks will eventually reach system pressure if contained) Encapsulate leak at the source (proprietary method) and direct flow to the container via hose with a valve adjacent to container. Mark the time the valve was opened and mark time required to reach pop off pressure. Boyles law can then be applied and divided by the collection time for cfm loss per seconds / minuets / hours / days.

All of this can be done fairly inexpensive and quick. For a little extra $$, time stamping can be done as an automatic function when desired pop off pressure is reached. To borrow a phrase “Set it and forget it.” I believe after several collections are made and calculated in this manner the user or Ultrasonic Technician could “Guesstimate” subsequent leaks fairly accurately.

This is a good forum and I have gained much knowledge from the “free flowing” information shared here.
Good In = Good Out

There is an ultrasound detector on the market (SDT) which accepts a Mass Flow Sensor so the detector actually reads SCCM or SCFM (depending on the setting for metric or imperial).
First find the leak with the internal ultrasound sensor, then quantify it with the mass flow sensor.
Caution 1: The sensor is meant for very low flow rates (1 litre per minute, or 1000 sccm). Higher flow rates can damage the sensor.
Caution 2: The sensor is only accurate if a tight seal is made between the sensor and the leaking orifice. This makes quantifying a leak on a pipe thread quite difficult. One way to capture that leak is to put a plastic bag around the pipe thread. Fit one part of the bag with a small outlet pipe that connects to the flow sensor. Once the bag fills from the leaking air the flow sensor can quantify.