Ultra high purity gas pipeline

[leel]

Hi,

Can any one give me some info for below question:

1) Guideline to choose correct surface roughness for different gas purity.

2) The maximum gas purity can be maintained for Threaded and Swagelok Compression connection.

3) The porosity of teflon tube can catch moisture up to how high level?

Thanks
Lee

 

[Onno]

I am working for some years in high vacuum technology now and I can try to answer some of your questions.

Surface roughness: normally machined vacuum parts have average roughnesses: typically 3.2 micron (3.2e-6 m) and do really fine in vacuum sytems down to 1e-10 mbar if you consider other material characteristics: cleanliness, bulk material inclusions, structural material defects etc.

For pratical applications down to 1e-10 mbar use clean and good quality materials: e.g. steel LN316 electroflux refined (EFR type)with a known track record on composition.
Surface roughness seems to play a less dominant role but opinions differ considerably.

If the gas is high purity it can be contaminated by adsorbed gas on the walls, leaks and gas impurities itself.

To clean the materials and components from adsorbed gasses and contamination before entering the high purity gas the system is baked out at temperatures between 125 and 450 C under vacuum (<1e-5 mbar). This procedure takes time( 1day to sevearl days) but is standard procedure to clean before admitting your vaulable high purity gas.

Threaded Swagelok connections are definitely not to be used. Compression Swagelok is good but much better (orders of magnitude!!!) is their VCR face seal connection system. Choice depends on your design criteria.

Teflon can indeed adsorb a lot of water. High purity and teflon are not the best of friends all the time.

Succes.

 

[leel]

Hi Succes,

Thanks for your prompt reply.

Anyway, most of the time I am dealing with the surface roughness from 40 Ra till 10 Ra (micron inch). And also gases purity from 99% till 99.9999%. So, I am facing a problem, at what kind of gas purity I need to use what kind of surface roughness and also fitting. And it is for sure orbital welding and VCR can give a very good performance but it is too costly.

Besides, why copper pipe cannot be used for high purity gas? is it because of impurity caused by the brazing job?

Thanks

 

[Onno]

I have difficulty with the surface roughness you use.

1 micron inch = 1e-6*24.5 mm =2.45e-2 micro meter.
This is just incredibly low. I can only say that the ultra high vacuum equipment we use to measure gas quantities down to 1e-14 standard cm^3 is machined to about 3 micrometer (3e-6 m) is doing really well. This is an easy attainable value for any mechanical workshop cutting. So your roughness value is of another magnitude lower.

We use the same gasses 99 to 99.9999 %. Orbital (TIG) welding is a must. Of course you can use copper pipe also. But, be sure to make brazing connections use flux free vacuum brazing. Any use of flux is just too bad to put into words since the flux is a heavvy contaminator. NEVER DO IT.

Yes about the cost. At our university it is a non stopping discussion: too high, too high. However an investment in VCR, orbital welding and high cleaning and machining standards are the only possible solutions considering the high purity of the used gasses.

Remember this (out of my personal experience): people will forget all about the initial financial investment if you show them you have produced a non contaminating system with minimum downtime. Consider the consequences of downtime in your design because this time needs to be paid for also.

 

[Montemayor]

Onno:

Please inform us what analytical instrument or set up you (or others) have used to accurately identify gas purities up to "six nines" (99.9999 %). I assume you mean Volume %, although you fail to mention the exact unit.

Have you applied this analytical method to Hydrogen?

 

[Onno]

I work in a mass spectrometry department where we use mass spectrometers to measure gasquantities and composition for earth science reasearch. We use quadrapole and magneticfield separation spectrometers.

 

[leel]

Hi Onno,

It is really, I am using such incredibly surface roughness fpr my gas piping job. It is because we need to do the certified our line up to <10ppb O2 & H2O, <0.1 micron m3 particle.

Anyway, can you please give me the work procedure or info about flux free vacuum brazing,

Hi Montemayor,

Actually, I am working at a gas company, and most of the time we are mentioning in %volume. And the purity of gases is set and certified by manufacturer.

 

[Onno]

To give you an idea on vacuumbrazing.

Pressure <1e-5 mbar

Use gold, silver, NiCr solders depending on temperature, materials, assembling procedure etc.

Electrical heating to the melting points of the pure metals or alloys. Usually a small clearance, gap of about 0.02 mm is machined between components to be connected. This gap will be wetted with solder thoufg its capillary action. You will end up with 2 metals being brazed together using a very thin soldering film.

Several things are not discussed until now.
1. Cleaning the gas extra before it enters your line.
2. Filter for dust particles.
Other surfaces of your materials are outgassed in the high vaccum oven. Usually clean N2 is used to bring the oven to atmospheric pressure.

Use a serch engine with the words: high vacuum brazing and you will find enough additional info.

 

[leel]

HI Onno,

Actually, the piping job I am doing is most for N2, Ar, O2 , H2 and other special gases. So, our procedure is purging the line among the installation. And I can find the procedure for doing piping installation by using orbital welding, but I still cannot get the actual guide line to choose proper material and type connection.

Anyway, I have just checked in internet about vacuum brazing, I don't think it is pratice in my country. We are still using flux (but low flux) brazing for most piping job.

 

[Onno]

Now I would like to make some final statement.

If you consider flux you better use the Swagelok system instead. This company can also supply a machine to perform orbital welding on their pipes and components. You can "hire" it to do just the welding you need.

I am not up to date with your situation. What are your plans exactly, what do you do with your gasses, do you have manpower, do you have a helium leaktester, what is your diagnostic equipment to check on gasqualities, what is the length of your piping, pressures and flows.

You might benifit from some information on the use of gases in the semi-conductor industry.

 

[leel]

Exactly, most of the time I am doing piping for semi-con industry.

So, depending on the requirement from my customer, we may do orbital welding, helium leak test, particle count, O2 and H2O analysis before charging the gas.

Actually, what I am considering here is how to choose the material (surface roughness, type of connection and etc) for diffenrent type of gas purity?

 

[Onno]

With regard to surface roughness you already know that for our ultra high vacuum application the average 3 micrometer roughness really works well.

Rath Manufacturing Company sells tubing and pipe for ultra high purity applications and offer roughnesses of 10-40 Ra micron inch.

Swagelok ultra clean tubing is in the same range of roughnesses (see their catalogue on tubing). Also from their catalogue: "Chemically Cleaned and
Passivated Tubing This tubing complies with ASTM G-93, Level A requirement for non-volatile residue levels and also meets requirements of CGA G4.1." I do not have any idea what is in these standards. Do you?

 

[leel]

HI Onno,

Actually, Swagelok also purchases tubing from Rath.

And by using chemically cleaned and passivated, tubing can be polished up to 10 to 40 micro inches. And ASTM G-93 is standard for fabricating tube according to thier thickness, material and so on. And CGA is Compressed Gas Association, which will be the guide for us to do piping job.

But, there is one problem on my side, ASTM and CGA standard book is very costly, so, I can't get the full spec of them.

 

[Onno]

You say you are working in the semi condcutor industry. Might I suggest your employer does an investment in the copies of the standards?
This seems very reasonable to me. Your employer cannot expect that you will solve this problem without any info at all.