Helium ppm to mbar*l/s conversion for Fug. Em.'s 2

[NGiLuzzu]

In order to perform Fugitive Emissions tests, we use a VARIAN sniffer probe helium leak detector, which gives readings in mbar*l/s. Our usual European reference specification (Shell MESC SPE 77/312) also gives limits in the same measurement unit.
First of all, is anybody able to explain the physical meaning of "mbar*l/s"? Is that a flow rate or a concentration?
Then, how is it possible to relate this with "ppm"? As many leak detectors manufacturers have instruments with both scales, we think there must be a simple relation between the two units.
Many thanks in advance, 'NGL

 

[cranekiran]

Hi,

Just read this site,maybe it will help you.

http://www.epa.gov/ttn/emc/promgate/m-21.pdf

Bye!

 

[NGiLuzzu]

Dear "Crenekiran",
thanks for your help.
According with the ideal gases law p*v = n*R*T, if temperature is constant, the "mbar*l/s" reading is a measure of the number of moles (or molecules) of Helium detected in the unit of time (1 s). Then, this is proportional to the helium concentration (ppm) in the intaken flow of the instrument, if the intaken flow rate itself is constant.
As per the mentioned website, this constance is surely true for VOC analyzers in EPA method 21 (point 6.4: "The instrument shall be equippen with an electrically driven pump to ensure that a sample is provided to the detector at a constant flow rate..." is it true for helium mass spectrometers too?

 

[cranekiran]

Hi,

I am not a professional in this unique field of N.D.T but I have come across helium testing since we get it done for few of our valves only if it is a customer requirement.

I cannot answer your specific question but after a hectic web-search I have found two sites close to your query:-

http://www.genesismission.org/educate/scimodule/SSWPrOptPDFs/Appendices/AppendixB-PO.pdf

http://www.industrialtechnology.co.uk/1998/oct/leakdetection.html

 

[NGiLuzzu]

I found in a Company Specification, and precisely in Shell MESC SPE 77/312 (Revision: 26-11-2002), note 7, par. 6.1, page 16:
"... For leakage rates conversion factors see EN 1779 annex B
(1ppm He = 2.5 * 10-8 cm³ /s = 2.5 * 10-7 Pa*m3/s) ..."

This equivalence has been added in this latest revision of the document: if it's true, it means that
1 ppm = 25 mbar*l/s, or 1 mbar*l/s = 0.04 ppm of He.

Unluckily, EN 1779 (1999)standard does not explane how this could be derived. Is anybody able to do that?

Thanks in advance, 'NGL

 

[NGiLuzzu]

See also Thread124-48813.
Thanks and regards, 'NGL

 

[NGiLuzzu]

OK: in other words, for example, let's say that I can count, on the road near our factory, 25 red cars a day passing by. If this is equivalent to say that red cars represent 25% of total daily amount of vehicles passing by, it is only because I KNOW, somehow, that this total amount is 100 vehicles a day.

Now, let's put helium moles (or molecules) instead of red cars, seconds instead of days, ppm instead of % and total intaken volumetric flow rate instead of total daily amount of vehicles: I think the reasoning is still valid.

Our manufacturer (

http://www.varianinc.com/)

confirms that the leak detector (VARIAN 979-70) we use in "sniffer mode" CAN NOT measure the total intaken flow rate and therefore CAN NOT give readings in ppm.

Regarding the above mentioned Shell statement and the relevant "=" signs, I think they are NOT CORRECT (by the physic and dimensional point of view!)! As a matter of fact, it CAN'T be a real equivalence, conversion or equation: it may be an "equipollence", true under particular conditions only. The problem is that such conditions are not specified.

What do you think about it?

 

[25362]

To anegri, correct me if I'm wrong. The sniffer assembly operates by injecting He into the testing piece or by having a He pre-filled chamber, then the gas is sucked by the sniffer pump at the surface leak point into the tester. The "diluted" He is measured by a mass spectrometer and so its concentration can be expressed in ppm, or otherwise translated to cc/s or mol/s of the leaking gas. Am I right ?

 

[NGiLuzzu]

Dear 25362,
we usually perform helium tests to detect external leakage of valves, pressurizing with 99% pure He the valve inside and slightly moving sniffing probe close to stem packing and body flanges (as per the above mentioned SPE 77/312); sometimes, instead of this "sniffing method", we also use the "vacuum hood method" (that is vacuum in the valve inside and Helium in a hood around the possible leakage points; see Shell MESC SPE 77/307, for example).
In any case, the VARIAN 979-90 leak detector, to which the probe is connected, gives output readings in mbar*l/s only.
The problem is that some Customers express leakage limits in ppm: the best we can do, at the moment, is to establish an "instrumental correlation" between the readings of our detector and two or more ambients at a certified helium concetration (e.g.: 10 and 100 ppm). I wondered if there could be a theoretical relationship too...
Bye, 'NGL

 

[Burdy]

Leakage flow rate conversion charts can be found on

http://www.leaktesting.co.uk/leak_spec_guidelines.htm

and your mbar/l.sec is on there.

Try talking to the manufacturer if you need more www.varianinc.com.

Br
B

 

[NGiLuzzu]

Dear All,
about this issue there's an interesting paper on the latest "Valve World" Magazine (web:

http://www.valve-world.net/index.asp):

see Volume 8, Issue 3, June 2003 at page 27: FUGITIVE EMISSIONS REDUCTION BEYOND "DEFAULT ZERO" by ADRIAN JEFFERIES (James Walker & Co. Ltd., UK).

To Burdy:
thanks for the link. As you can see from my post of June 11, I've talked to the Manufacturer already.

Bye, 'NGL

 

[scalleke]

The Varian 979 cannot measure the flow rate of the gas coming in and thus cannot count the helium passing through and simply state the concentration reading or "PPM"

However a smaller test device also From varian, the Helitest, can. The Helitest allways has the same fow rate of gas being aspired so measuring the quantity of He atoms makes that the instrument can detect concentrations "PPM".

mbar l /sec or I prefer to have the reading in atm cc /sec means that at one atmosphere pressure there are X cc of helium atoms passing per second through the machine.

THERE IS NO CONVERSION POSSIBLE BETWEEN mbar l / sec, atm cc / sec and PPM. One is a flow reading and the other is a concentration reading. There are tricks to establish a link but they are empiric and have no scientific basis.

The Varian 979 has a callibrated leak source internal.

At start the unit wil self callibrate. in fact what it does is it lets the callibrated leak pass through the unit. ionises the helium atoms and separates them off with a magnetic field. the helium atoms fall on a capturing electrode and create a small current. The machine then establishes that this current is equal to the flow rate of the callibrated leak.

When in sniffing mode the machine will draw in helium atoms and see how many atoms pass through the unit by measuring the detection electrode amperage.

Allow me to continue later Lunch break is over.

scalleke

 

[scalleke]

When the unit self calibrates all of the leakage from the calibrated leak source passes through the instrument.

When sniffing helium only a part of the leakage that is coming from the leakage source passes through the instrument.

When one calibrates the instrument with a calibrated leak source of say 1.7E-8 atm.cc./sec. we are sure that all the helium leaked passed through the machine.

Now it depends on the method of measurement whether or not we will be measuring all the leakage.

Method TA Luft; good old German Grundlig und Punktlich.
The leak source is completely encapsulated and all leaking gasses are passed through the leak detector. This is great, all leak gass is detected and we can determine how much helium has escaped into the atmosphere.

Method 21: EPA Uncle Sams way; a sniffer with a determined opening cross section is placed at a location 5 mm from the leak source and parts of any whiffs of Helium that are leaking out pass through the leak detector. We can only guess at which percentage of the leakage is captured.

Method TA luft may be good and correct but as many times is the cases with our German friends it is impractical. Who would like to go out and make 5000 adaptors to be able to sniff all the leakage in a chemical plant.

Method 21 is inaccurate but practical. the idea I guess is that if you standardise on the way and the instruments with which to measure you will be able to pick up trends. And also it is easy to equip a helium detector with a correct sniffer and a distance spacer and to go out measure 5000 valves in a plant.

One more thing: Helium intake by a varian leak detector is influenced by pressure and temperature. the detector compensates for temperature changes since the last self callibration. I still need to get from Varian whether the detector compensates for pressure changes also.

Scalleke

 

[NGiLuzzu]

I wish to thank Scalleke for the clear and precise answers, that confirm my suspects.
I think there is a lot of confusion about this matter, aven among those who should be "experts" and write specifications...
As a matter of fact (and as it can be easily sensed), sniffing method is good to reveal and locate leaks, not to measure them; but, because of its easiness, there are many attempts to extend it beyond a simple "classification" or "ranking" of leaks, even without scientific basis.
Regards, 'NGL

 

[EVSP9000]

A bit late on this Thread

BP have a program to measure and calculate the relationship between helium and Methane, as all new gate and globe valves are tested with Helium for safety reasons, the max emission is 50 PPM methane that equates to 100 PPM helium this is a bit contravercial however it makes delivery large volumes of valves emission tested a reality, the test equipment is made by Varian

Hope this helps

Mike

 

[NGiLuzzu]

I found other interesting references about this issue.

* First of all, see ASME Boiler & Pressure Vessel Code (2001), Sect. V, Art. 10, Appendix IV, Par. IV-1010 SCOPE: it is clearly stated that sniffing method "... is a semiquantitative technique used to detect and locate leaks, and shall not be considered quantitative."

In any case, the whole Appendix IV deals with concentrations, distances and speeds of sniffing, and so on.


* A quite old reference, but very useful in order to clearly understand things, could be the paper
Valve steam leak-tightness: evaluation of measurement methods for qualification and QA tests
by Simon E. Leefe from BHR Group Ltd., pubblished in Valve World magazine, August 1999.

Bye, 'NGL

 

[FGreen]

Hi there, not sure if this will have any relevance to the original question but the general "standard" leak rate for vacuum valves is 1x 10 -9 mbarl/s. This is for for use with vacuum valves in a vac system, although it sounds like you're looking for leak rate "limits" for valves being used at pressures above atmosphere so this may not have any relevance!!

Regards

F

 

[NGiLuzzu]

FGreen,
you're right, but, although we usually deal pressures above atmosphere, your tip is interesting anyway: for information, and also because sometimes we're requested for valves in vacuum service (even if in low grade vacuum...).

May I ask you what instruments and type of tests do you use to detect leak rate levels you mention?

Many thanks, 'NGL

 

[FGreen]

anegri,

we use an EDWARDS Spectron 5000S leak detector (nitrogen filled) this allows us to test with relative ease to leak rates of 1 x 10 -9 mbarl/s. Products each tested individually by holding at vacuum and then applying Helium to external area of product. That pretty much covers the main testing we do.

Regards

F