I am looking for practical ideas such as how to mark and document leaks. Have any of you seen "tags" that are designed for leak testing? A 2-part tag would be great so it would be self-documenting. I am thinking of using different colored surveyors tape for visible flags. Have any of you used a digital camera to document leak locations? I'm just looking for practical ideas.... I think I am up on leak detection techniques, but if you can't communicate what you find, it's not much good doing a survey. Thanks in advance for your help.

Rusty,
I have used custom preprinted Tyvek tags for marking compressed air leaks. Here are two links to ordinary printed Leak tags:

http://www.emedco.com/search2.asp?dept%5Fid=

http://www.seton.com/seton/catalog/refineResults.do/~0/....html?sblid=0.401439

Both companies have a wide range of tag types, materials, colors, and sizes. There is a lot more to a leak survey than finding a leak and tagging it.

Walt

quote:

There is a lot more to a leak survey than finding a leak and tagging it.

I'm aware of that... my approach will be to develop my own "standards" for measurement that will be more accurate than the "Guess-timator" that UE Systems provides. I plan to buy a small air-flow meter and then simulate various types of typical leaks that are found in the plant. I will be able to correlate the actual loss to that predicted by UE (based on the dB level, or meter sensitivity setting). I plan to develop a realistic priority system so that the large leaks are fixed first, then medium, and then small maybe not at all. I think the 80/20 rule will apply.... 80% of the leakage will be from 20% of the leaks.

"Believe nothing you hear, half that you see, but all that you do."

Rusty,

I know of folks who tried to get a correlation of sound level to leak flow rate and found it to be a daunting task. Even flow through a simple round orifice is not easy to quantify. Give me a call if you want to discuss further. Otherwise, good luck with your experimentation.

Walt

SDT did some extremely good work using various pressures, flow rates, orifice shapes, areas and of course angle to and distance from the leak.

Best Regards,
Tom Murphy

Tom,

Have the test results been published? Are they available? Was the meter(s) calibrated to a sound pressure standard (Pascals) and not to a voltage? Do you know of any ultrasound meter on the market been calibrated to a sound pressure standard?

Walt

Hi Walt,
was the device calibrated - yes of course.
To a reapeatable standard - yes of course.
Is there an equivalent in ultrasound to a pistonphone - no, not that I am aware of.

I believe the results were at one stage published on the SDT website. I will try to find a link to post.

Tom Murphy

No ultrasonic device is calibrated to an accepted standard. The SDT is calibrated to an internal standard and the DB is with respect to voltage, which has nothing to do with the sound pressure level. You simply have a "repeatable" voltage across the meter. UE is less repeatable. CTRL UL101 is another unit to consider. They sell carbon copy defect tags. In short, don't waste your time relying upon the sensor as none of the sensors on the market measure sound pressure level. There are tools available to measure electrical consumption at the compressor.

Hi Rusty,
This probably goes beyond what you want to do, but here is how I do it.
I do the inspection and leave half of a maintenance Tec-Tag at the leak. These tags are two part cardboard with a string attached. I put the date and leak information on both parts. When I get back to my office, I create a workorder. I put the workorder number on the half of the tec-tag I brought back. I then open a autocad drawing of the facility footprint. I have the facility broken down into the areas I check each trip out. I mark the location of the leak with the tec-tag number and the workorder number with a red dot. I write the report using MS Word. I cut and paste the autocad drawing into the Word document. If a mechanic comes by trying to locate a leak, I give him a copy of the drawing for that leak. When I am done with the entire survey for the facility, I print out the Word document and post it on a public area of the server. An e-mail is then sent with the URL to all department heads and the planners.
I hope this isn't too much detail.
On the workorder, I indicate the tec-tag number and the severity of the leak, A-D with A being the least severe and D being a high volume leak. I don't quantify it beyond those letters.
Jon

Benji,
I agree with your comments 100%. Ultrasound is a fine method for detecting/locating leaks. The meters are not calibrated to sound pressure level standards, and sound level does not correlate well with leak flow rate in real-world plant conditions. There are guesstimating methods for leak flow rate using ultrasound, but the best bet is to measure flow rate at the compressor or in isolated sections of the distribution system.

Walt

This is in response to Stan's question from over a year ago (when the thread was started). Dr. Penrose quoted an efficiency of ~14% at best for compressed air systems. For the record, this is very realistic.

Stan, HP per SCFM for particular pressure might be a good metric for a compressor itself, but the key here is the word "system". This would take into account actual compressor efficiency, leaks, viscous pipe losses, and the big kicker, the efficiency of the tool/process being driven.

A 300ftlb @ 350 rpm (equalling 3.2 HP of actual power) impact wrench might require around ~6.5 SCFM. If it takes 3.5 HP compressor power per SCFM of air, the real world efficiency is 3.2HP (tool) per 22.75 HP (compressor), or 14% total system efficiency. And that's probably not counting line losses and leaks.

Expensive.

Compressed air has many strong, redeeming points, however energy efficiency is probably its weakest. All the more reason to be miserly with it.

Enter ultrasound leak detection...
Personally I can't wait to get this into our plants, although I'm a bit afraid that what we hear initially will be overwhelming.

~Sam

Hi Sam,
it is frequently quite a shock for people when they start out on their leak detection programme - a feeling of where to turn next and how to prioritise the hit list.
One tip from a user last month: start off doing your survey for the first time with a low setting of amplification on your instrument so that you can only hear the bigger leaks - test this out in your workshop to decide how bad a leak you want to fix the first time around.
Once they are all out of the way, you can then increase the sensitivity and find the lower level leaks.
Tom Murphy

inspection tags

My Inspection tags offer a variety of popular calibration and inspection tags on guaranteed lowest prices. These tags help to keep a systematic record of your Inventory with in your organization. Order NOW for top quality inspection tags!

http://www.myinspectiontags.com http://www.myinspectiontags.com

Maybe a little expensive for some but: We recently installed meters for different production zones off mian supply headers. When all the tweaking and calibration was completed, one zone measured 700CFM during a "lights out" condition. Using ultrasonic leak detection we have reduced that amount by 75%. We also adjust the settings from low to high as we progressed. Low settings find those big leakers and so on ...

I "borrowed" an orifice size to cfm leak calculator from here and it proved to be quite useful when detirmining the average leak size when compairing total repairs made against the reduction in "lights out" meter readings. The next inspection will be done during a "lights out" condition to eliminate false positive exhust cycles inherant to production.

Really like the quick payback and ease of use of ultrasonic leak detectors ...