Does anybody have experience with s

[Magalhaes]

Does anybody have experience with sampling LPG for mercury analysis ? We have had repetibility problems with marcury analysis in LGP and wonder how sampling care would help on that.

 

[EmmanuelTop]

ASTM D1265 defines sampling procedure for LPG. See also attached article on comparison of LPG sampling techniques for Mercury analysis.

Can you describe the problem in more details?



Dejan IVANOVIC
Process Engineer, MSChE

 

[Magalhaes]

Thanks Ivanovic. The issue we are trying to overcome is Hg analysis inconsistency in LGP: the same sample varies its result if done in duplicate or three times. We also have trouble trying to sample a gás-liquid phase stream (C1 to C6).

 

[EmmanuelTop]

I've got that part. It would help a lot if you could post a sketch of how your sampling system looks like, as well as the difference in Hg analysis results and which method you use.


Dejan IVANOVIC
Process Engineer, MSChE

 

[Magalhaes]

We use UOP 938 method. Sampling is performed with stainless steel PTFE bomb (stainless steel bombs were formerly used in the past). Analysis is also performed according JLPGA-S-07. We also tried direct sampling with tedlar bags. Total Hg values vary from 0,2 to 800 micrograms/m3 but we believe the highest values are by far the most reliable ones.

 

[EmmanuelTop]

UOP 938 explicitly prohibits sampling in metal containers because Mercury will adsorb on the metal. Perhaps you should verify (or replace) the SS/PTFE sample bombs. Tedlar bags are normally used for VOC's and I am not sure if they can be used for gas/Mercury sampling.

You also need a proper sampling system arrangement (ASTM D1265) - how does your sampling system looks like?

Variations of the results you get are in the range of 3 orders of magnitude so there must be something done incorrectly, either during the sampling or in the sample processing.

Dejan IVANOVIC
Process Engineer, MSChE

 

[Magalhaes]

I will check sample system arrangement with ASTM D1265 as suggested. I personally guess that most part of our issue is the problematic stream to be analysed. It is a two phase stream (C1-C6)which makes a representative sample vey difficult to succeed. Since mercury tend to concentrate in tha gas phase, sample makes all the difference. Anyway, thanks for the tips.

 

[EmmanuelTop]

If you are sampling at a location where you have a 2-phase stream, that definitely is the heart of the problem. In such case, sampling at another location needs to be arranged - you need a full-composition stream. Other remarks as mentioned above are still valid.

Dejan IVANOVIC
Process Engineer, MSChE

 

[Magalhaes]

I agree with you. I will try to figure out another sample point as suggested. Thanks.

 

[ylep02]

Any update ?
For LPG, direct sampling into Tedlar bags is preferable. Avoid sampling bombs, even teflon lined.
What type of mercury analyzer do you use?
And what volume of gas do you inject in your analyzer ?