Water Content

[alchemist]

Hi there...

I am looking for a method for the water content in LPG instead of the current method we are using, which is the Karl Fischer Method.

Anyone has idea?

Regards,
Alchemist

[Willy]

Hello there,

It depend on the matrix....I would like to know the reason(s) why Karl Fischer isnt working? Are you using the right solvent to dissolve your matrix. Are the parameters of your apparatus not optimised (like mixing time)? You probably try few things before posting your question but ill need to know first what kind of matrix is LPG.....

Other than that GC with TCD detector is a good way to analyse water....

Willy the GC

[chromatographer1]

LPG is liquified propane gas.

Water analysis is troublesome because of sampling issues. That is why your Karl Fischer titration is a problem.

There are water content meters commercially available which are often used in pipeline gases. There are also indicating tubes which will give you a good estimation of water content available.

If you want to try GC-TCD the R polymer packed columns in glass or fused silica SS with special heating of sample lines and valve sample loops can give you good reliable answers at even low levels.

The real issue is how can you sample reproducibly and accurately. Once you solve that problem the other issues are easy to solve.

best wishes,

Rod

[AICMM]

Alchemist,

Rod's comment about a PoraPak are appropriate. I have also done a very similar analysis to this using a very thick film CP-Sil 8 column and an HID detector. Peak shape is quite reasonable for this combination. I am pretty sure the water came clear of the standard alkane series but I can dig up a chromatogram if you need me to. The biggest problem was getting a good flash of the water in the inlet along with the hydrocarbon which results in a broader water peak.

I would be interested to hear how this is going for you.

Best regards.

[alchemist]

Thanks all,

I just wanna highlight what the reason behind searching for another method instead of the current one utilized by us in the determination of watre content on our product which is LPG.

With the reference to ASTM method E-700, we are having some obstacles ragarding the repeatability of our results.

We are using the following matrix: Idoine, Sulfur dioxide and Pyridine with the water by reduction.

Our issues:
- We are using cylinder bombs of 300cc for our samples, when we inject the sample using the aqua test titrator for 10 minutes, we have to put the sample amount in order to calculate water content.
our practice is by weighting the whole sample in the bomb in put it as the sample amount to be calculated. on the other hand the injected amount is much less than the whole sample in the bomb.
- We are not having the same problem with the liquids. ??? and that we enter only the amount of liquid weighted by syring to be calculated for water content.

looking forward any comments or clues.

Regrads,
Alchemist

[AICMM]

Alchemist,

I assume, from what I read, that you are bubbling LPG into KF from sample bomb? The problem then becomes how much LPG? Instead, think about doing repeated samplings from a sample loop of known volume, similar to the way that Thermo does total sulfur analysis on liquid hydrocarbons. You set up a loop and inject with a gas stream (to push sample into KF) some number of times according to the required detection limit.

On the other hand, you application is very, very similar to the one I describe earlier except that the product of interest was toxic as well and they did not want to bubble through KF. Thus, a thick film column, blocked liquid sample valve, and appropriate detector should work quite well.

Best regards.

[chromatographer1]

Alchemist

That method was withdrawn from the ASTM for good reason, reasons of which you are struggling now.

Your sampling cylinder, the 'bomb', must be anhydrous and must quantitately transfer water to any measuring device you use.

This is difficult to achieve.

It might be more appropriate that you take a dry sample of mole sieve or silica gel and pass a fixed amount of LPG through it. Then take the sample, desorb the water and titrate to determine water content.

Using the proper detector you should be able to do very low levels of water content in LPG by GC with sampling and multidimensional valves.

The former would be cheaper and easier to implement.

best wishes,

Rod

[Bruce Hamilton]

If the analysis is for a tradeable product, then look for a standard method, usually ASTM.

The most common used to be the dew point meter ( eg Shaw, Stanhope-Seta etc., ), which probably comply with ASTM D1142 or similar.

Alternatives are the inline moisture meters ( careful to ensure the sensors aren't poisoned by any components ). A Google search on ASTM D1835 or D5454 might be a good place to start.

You may be best informed by asking the ASTM to point you to a member of their LPG subcomittee, to discuss the options.

If it's just for research, or internal, then...

You could keep using your current system. The critical aspects are removing tramp moisture ( caused by condensation ), and using a moving piston sampling cylinder to obtain representative samples.

When I was measuring water content of vapour and liquid, for in-house research, I had a stream of nitrogen going through the titration vessel and used a needle valve to to replace it with LPG vapour.

It was a very time consuming process, as the water content was very low, and we found a Shaw Moisture Meter ( dew point ) was much faster, even allowing for the initial equilibrium time when used in batch mode.

There has been a lot of work on using indicating tubes, as a simple field test for impurities ( including water content ) in LPG, and I think there is a draft/final ASTM Dxxxx method for their use on gaseous fuels.

If the ASTM method is not obvious, there was a comprehensive SWRI report in the last few years ( possibly for the Propane Council ) which should be available on WWW, as I recall reading part of it online for free.

Please keep having fun,

Bruce Hamilton

[alchemist]

Hi again,

If i would say I'll go for the GC-TCD method. Could anyone provide me with any reference to that concerned method? From GC conditioning till the result interpretation.

But at the same time, i would like to share experince with you about the Karl Fischer and water in LPG.

regards,
alchemist

[chromatographer1]

A thick film capillary Wax column is suggested for the water analysis, ie 15-30m 0.53mm 1µm Inject the sample via sample loop valve at a temperature of 120°C injector, and have the TCD as cool as possible, or use pHID detector (AIC, GOW-MAC or Valco). You could use a low oven temperature (40°C) to focus the water peak and then program it for elution.

This should give you a starting point to develop your GC method.

Bruce's suggestion about using indicating tubes is well founded, however.

best wishes,

Rod

[Bruce Hamilton]

You really need to clarify whether these results will be used on a tradeable product CoA because, if so, you need to follow a standard method agreed between supplier and purchaser.

Given how low the water concentration should be in some forms of LPG, I think GC with a TCD might be a struggle without serious valve switching configurations.

It may be more possible on a standard GC confiruration if some form of sample concentration was used, as suggested above by Rod for KF.

My guess would be that you may have to put a backflush in place for water in LPG by GC column, depending on local LPG quality, as some grade can have significant amounts of heavier fractions and sulphur compounds. I'd also be rather wary of TCD filament coating ( leading to poisoning ) by sulphur components that might elute from a simple column system.

The article on indicating tubes that I referred two above can be found at the address below...

http://www.propanecouncil.org/files/112 ... t_rev2.pdf

Please keep having fun,

Bruce Hamilton