Hydrocarbon Dew Point - Measure instead of estimate

[Mr2jz]

thread135-253698
The discussion of the proper use of EOS based on GC data is often one that overlooks the obvious - why calculate it using theoretical methods when it can be measured. There are many papers that describe the deficiencies of using EOS for determining this parameter and you can find a good one at:

http://www.ceesi.com/docs_library/events/ishm2008/Docs/5110.pdf

But in any case please review ISO23874 and a GE publication GER-3942. Accuracy in this measurement is critical to profitability and getting shut-in because the calculation is based on insufficient data is very expensive.

 

[dcasto]

The issue is: will a pipeline company accept an instrument just because a vendors says it'll work. The gas industry has Southwest research studying all the ways of finding HC dew point.

 

[Mr2jz]

You are absolutely right so the report that SWRI should bear close review as well. It was funded by PRCI (the gas industry) and is under Contract 18-45425-A, between PRCI and Southwest Research Institute released for fional publication January 30, 2008. It compared a C9+ with EOS to other methods the industry uses.

In an article "The Need For Accurate Hydrocarbon Dew Point Determination”, D. George, Ph.D. et al., Southwest Research Institute, published in Pipeline & Gas Journal, September 2005, he states "Based on comparisons to date, however, the C9+ characterization most often appears to predict measured dew points to within ±25°F."

I just am concerned that the information I have heard at some of the gas schools indicating the ability to get within 2° is highly doubtful at best data has been presentad that suggest the calculations may be even more inaccurate than the 25°. That could get suppliers shut in quickly.

 

[dcasto]

I hope you mean C6+. If you have measureable C9's, you won't make any pipeline specification.

I had a pipeline try to shut me in when their $49 EOS package was predicting HC dew point. I asked the pipeline dispatcher to read the flowing temperature and pressure to me. It was 48F and 650 psia. I said your transmitters are off. He said they were exact and they are the gospel. I said, good, now get your engineer in and explain how a gas at 48F can have a measured (calculated from their cheapy BWR EOS) dew point of 55F and the spec is 50F.

Needless to say, the gas stayed online.

 

[PaoloPemi]

Chilled mirrors are useful for several applications but I know many plants where cricondentherm (HCDP) and several other parameters are calculated from GC composition, results are quite accurate (for cricoT usually within +/- 2 C), costs are generally competitive and the system quite reliable (compared with the chilled mirror).

 

[belbon]

we have two EOS based on GC composition systems and they match (1 C) our laboratory mirrors, in-line chilled mirrors have been found quite critical so we decided to calculate dew points in DCS, our GCs go up to C6 and we characterize C7-C8 with a simple splitting, results match those of our laboratory instruments and we have no in-line instruments to maintain,
Daniel

 

[Mr2jz]

Daniel,

Thank you for your contribution.

You are very fortunate in that you have a supply of gas that has been processed very well. A common characterization practice is often based on a laboratory analysis on a properly extracted sample (per API 14.1) to a sensitivity of 0.1 ppm or at least 1.0 ppm to C14. Once a history of the field is established with multiple analysis reports over time, you can get to a reasonable set of actual characterization values for that field/supplier. Why analyze to C14? I have seen conditions where even a change of 1 ppmv of C10 in a natural gas system produces a 10°F change in dew point. One thing that I have seen is that if I take real analysis data and plug it into the EOS and then play with the values just outside the actual analyzed values, the impact of missing them is revealing.

When I assume that there is total efficiency in the processors facility to remove all of these heavies beyond C6 can get me into trouble. Also, characterizing the C7, C8, … is fine until the source changes as in an upset condition at the processing facility.

The other wrinkle is having a well trained operator using the chilled mirror properly per ASTM D1142. Not just anyone can properly interpret the image on the mirror. I know with my older eyes, you would not want me to take this measurement and make decisions on my readings. The ASTM procedure recommends a procedure that typically takes a full hour to get right even for the experienced technician.

 

[PaoloPemi]

according my experience it doesn't make much sense to include a 1 ppm C14 as input, chilled mirrors have a relatively low sensitivity and they tend "to average" values, differently any decent VLE package works with 16 digits precision and tolerances of 1.0E-8 or better, this means that a 1 ppm C14 doens't produce any change in chilled mirror but the software calculates a higher dew point of 20 C or more... if you include 1 ppm of C22 a decent package as that mentioned in

http://eng-tips.com/viewthread.cfm?qid=253698

can calculate 50 C (or more) difference...
It is important to filter values to get results comparable with those of chilled mirrors.

 

[dcasto]

I can not disagree more with Mr2jz. A 1 ppm change in C8 on a typical composition has less than a .05 deg F. Injecting C14 at 1 ppm shows a 10 deg F change, but that ain't happening in real life. That C14 molecule is not going to be in the gas phase because there is no such thing as steady state.

 

[belbon]

Paolo55 is correct, VLE package detects very small amounts of heavy components as C9+, we split C6+ into C7 and C8 fractions to get results comparable with those of our laboratory instrument, results are reasonably accurate, a laboratory chilled mirror is required to verify regularly calculated values (we store values in DCS),
Daniel

 

[dcasto]

In the 1990's, Jack Grynberg sued gas companies claiming that the filters on the fron of chromatographs would remove larger molecules and hence lower the heating value. This conspriacy theory is typical in the industry and leads to all kinds of new solutions. Well, not so new. see page 16

http://www.bwenergylaw.com/documents/TenthCircuitAffirmanceonGasClaims.PDF

 

[PaoloPemi]

all instruments (VLE packages, chilled mirrors etc.) must be utilized properly, differently you will get wrong result, all the same....
there are many papers comparing results of chilled mirrors and VLE packages, both solutions have been extensively tested....