16 replies to this topic

[apex]

dear friends,

                     we have a gas processing facility which sweetens, dehydrates & dew point depresses the natural gas received from offshore. I am attaching snapshots of one of the dehydrated units & want to know the  moisture dew point achievable by maintaining the parameters shown. I have simulated the same in hysys & it shows moisture dew point to be higher than the chilling temp of dew point depression unit (-4.5 C). but there is no hydrate formation problem in our dew point depression units.

 

Inlet Gas Composition:   C1       C2       C3      iC4      nC4      iC5      nC5      C6+      CO2       G                                                                     |
|                                     81.73     7.27     3.92     0.73     0.89     0.18     0.15     0.05     5.08

 

stripping gas is used & the flow is shown in the snapshot.  20140910_095642.jpg   144.09KB   6 downloads  20140910_095651.jpg   152.72KB   5 downloads  20140910_095702.jpg   128.54KB   2 downloads

 

[srfish]

If you are interested, I could make a dew point run on our in-house software. But I would need the operating pressure first.

[Art Montemayor]

Apex:

 

This is a funky way to submit your basic data.  You also fail to state your simple application:

 

I believe all you are doing is running a TEG dehydration unit on a natural gas source and you are using high pressure steam to heat your TEG reboiler.

 

You expect us all to read your copied instrument panel, but you don't furnish the UNITS that you are using in your instrument readings.  Is pressure in atmospheres, bars (g or a), or kg/cm2?   Is temperature in celcius?  what is the flow rates in?

 

All we really need is the composition (which you furnish), the temperature and pressure at the top of the TEG contactor, and the purity of the lean TEG to predict the expected dew point (assuming the appropriate lean TEG flow rate is circulated.

 

I suspect that you are well above the normal 7 lb H₂O/MMScf of product gas in your product gas because you are running your reboiler well below the usual (404 oF) and you are using wet stripping gas instead of dry product gas for stripping.  Tell us what you are obtaining as your dew point or lb H₂O/MMScf of product gas.

[apex]

Dear Art,

               I am sorry that I presumed that the things you are asking for would be available through these snapshots. we are receiving sour natural gas along with condensate from offshore & after separating condensate the gas is sweetened with MDEA absorption & TEG dehydration unit is downstream of MDEA sweetening unit. After dehydration the gas is chilled to around -5^oC through propane refrigeration. We are using High Pressure steam in the TEG reboiler. The TEG unit absorber is having 9 bubble cap trays of which the top tray is dry. After TEG reboiler there is a packed column where stripping gas from outlet of Dehydration unit is used after heating in the reboiler to strip of the moisture. The temp & pressure are in ^oC & Kg/Cm2 respectively. The flow rate is in KNM3/Hr. The problem is that chemistry section in my plant is not able to provide a realistic dew point (according to them the dew point is -58^oC). The lean TEG purity is 99.0% which is also doubtful & considering simulation results based on the operating parameters this doubt becomes certain. The thing which is disturbing me the most is that there is no hydrate formation in dew point depression unit while simulation shows my outlet gas to be in that zone.

[apex]

The flow rate of TEG is in SM3/Hr.

[Bobby Strain]

I expect that you can achieve a dew point of about -30 C with proper regeneration and circulation. I wouldn't use HYSYS to simulate the glycol unit. If you have nothing better, use NGPA databook.

 

Bobby

[processengbd]

 

"I wouldn't use HYSYS to simulate the glycol unit"

 

Bobby can you explain?

[Art Montemayor]

Bobby is right on with his comment regarding the use of HySys in simulating the TEG dehydration process and expecting design-quality type of results from the output.

 

Hysys is renown, in my opinion, for spitting out erroneous results in TEG dehydration simulation attempts.  There is much more reliable and accurate software in the marketplace today.

 

My further remarks and comments follow in the next post.

 

 Water Removal Predictions using Glycol Dehydration Simulators.doc   163.5KB   126 downloads

[Art Montemayor]

Apex:

 

Please carefully review the attached flow diagram in the workbook and tell us if the depicted flow diagram is correct and all the corresponding values related to the flow streams are accurate.  Your submitted data has some questionable data and it is difficult to follow a printout of a control room monitor.  We need to have accurate description(s) of what you are writing about in order to furnish accurate comments of value.

 

Why do you mention you are chilling the process gas down to -5 ^oC?  Your data states the feed gas is entering the contactor at 36.6 ^oC.  Obviously you are warming up the gas ahead of the contactor, so why even mention it?  As PingPong states, all we really need is accurate, related information:  What are your process pressure, temperature, dewpoint, and flow rate conditions and what are your product specifications or expectations.

 

Also, who designed and built your unit?  The reason I ask is that it is next to difficult to produce pipeline quality gas the way your diagram shows.  I’ll detail my comments when you confirm the correctness of the diagram and data.  To comment now is useless.  You would have saved us a lot of trouble and time if you had just created your own, detailed flow diagram at the outset.

 

 

 

Attached Files

•  TEG Regenerator with Steam Reboiler.xlsx   815.6KB   111 downloads

[apex]

Dear Art,

                Your depiction of flow diagram & data is accurate & I am thankful to you to for that. Minor correction is that there are 4 nos. bubble cap trays in place of packed bed in the TEG stripper.

                 I mentioned that the dehydrated gas is going to a chilling unit downstream of dehydration unit. In this gas chilling unit the dehydrated gas is chilled to -5^o C & the condensate generated due to chilling is used to recover LPG. As I mentioned earlier that simulation in hysys of this dehydration unit shows moisture dew point to be much higher than -5^oC. But we are not facing any hydrate formation problem in chilling unit. This really is surprising me & raises question that either dehydration simulation results are not ok or the hydrate prediction utility is incorrect.

[Art Montemayor]

Apex:

 

Who designed and built your TEG unit?  Why don’t you correct some errors in your control room display?  TI 1209 = 69 ^oC?  FIC 1208 = 0.3 m³/h?  PI 1219 = 8.4 kg/cm²g?  TI 1220 = 58 ^oC?  FI 1202 = 10 Nm³/h?  Are you certain that you personally checked the accuracy of this data?  I don’t believe that it is accurate.  Allow me to explain ....

 

You now explain that you are reducing the water dew point (or water content) of the feed gas going into the TEG contactor by pre-chilling it to -5 ^oC.  You fail to explain how or where you heat up the feed gas to 36.6 ^oC prior to its entrance into the TEG contactor, but I have to assume that you do.  This feature allows you to reduce the amount of water content in the feed gas before subjecting it to TEG absorption.  It also allows you to cool down the lean TEG entering the contactor to 15 ^oC and obtain better and more efficient dehydration.  Why this is not done I don’t understand.  That is why I ask who designed this system.  Or could it be that the pre-chilling is an after-thought and a process modification AFTER the unit was installed?

 

The temperature of the inlet gas into the contactor will dictate the total amount of water fed to the unit – if the gas is saturated with water (which is the normal case).  However, in your case, this is not happening.  You are feeding un-saturated gas.  That is why I specifically noted the term “saturated” in my flow diagram.  Look again at my diagram.  Yet you say that my flow diagram & data is accurate.  That is not so.

 

You are operating your reboiler at 191.5 ^oC (376 ^oF) and using stripping gas from your flash drum.  At this reboiler TEG temperature, the very best lean TEG purity you can obtain (without stripping gas) is 98.5%.  And that is at atmospheric pressure.  You haven’t given us the reboiler pressure and I have to assume it is at least 5 psig – or perhaps more.  Any reboiler pressure increase over atmospheric reduces this TEG purity value substantially.  I would estimate your reboiler TEG purity to be 98% or less.  Using steam to heat TEG reboilers is not normal – and this application shows why this is so.  Normal operations call for heating the reboiler at 400 ^oF and atmospheric pressure to obtain about 98.4 to 98.7 wt% TEG.

 

Therefore, you need to apply stripping gas to the 98% TEG in order to have a decent 99.5 or better purity at the TEG contactor.  But your process shows that you are using flash gas to strip – and not dry product gas.  In order to reduce the TEG water content it is important to use DRY stripping gas.  Your flash gas is a result of flashing wet TEG – which means the resultant flash gas is not dry.  This makes your stripping step very ineffective.  That is why I ask what I ask.

 

You normally would use 3 standard ft³ per gal of TEG circulated to bring the TEG purity up to 99.5.  If you are circulating 5.9 m³/h (26 gpm) of TEG, then I calculate you should be using approximately 125 Nm³/h of dry stripping gas.  You report a stripping gas rate of only 10 Nm³/h.  This indicates you are not regenerating your TEG sufficiently to meet pipeline specifications – which you haven’t identified.

 

I can’t address why your simulation doesn’t agree with your results.  We haven’t been given a copy of your simulation inputs and outputs and we don’t know how you ran it.  I also can’t explain why you don’t have hydrate problems.  We don’t know the temperature and pressures you are operating at.  You also don’t identify the basis of your dewpoint.  Is it atmospheric dewpoint, or is it the process pressure dewpoint?

 

All in all, this analysis can’t be specific because we lack all the required basic data on this application.  Your pre-chilling may be doing more benefit in removing water than the TEG unit.  See the attached dewpoint curves to see where Bobby Strain estimates the expected dew point.  Although I am unable to answer your direct concerns, I hope that some of the details I have outlined help you understand what is happening in your described process.

 

 Equilibrium Water Dewpoint in TEG Contactor.pdf   1.03MB   120 downloads

[apex]

Dear Art,

                I am sorry that I could not make you understand that we are not chilling the feed gas to -5^oC & then reheating it to 36.6^oC. The feed gas is fully saturated with water at 36.6^oC. The TEG inlet temp = 69^oC is correct as I have checked it in the field. The other thing of importance i.e. stripping gas flow of 10 NM³/Hr seems to be correct as we have 8 dehydrating units & all are having stripping gas flow in the range of 5-20 NM³/Hr. The stripping gas is taken from dehydrated gas(unit outlet ) through a pressure control valve. The moisture dew point (-58^oC) mentioned is at atmospheric pressure.

                          We recently commissioned a dehydration unit which is almost similar to the previous one in capacity & design.Its  almost all instruments are working. I am attaching snapshots of that unit which might through some more light. In these snapshots  a working moisture analyser (048AI1101)is showing reading of -27.05^oC at system pressure. I would be thankful to you if you can examine these snapshots & help me to understand the mystery. 

 

 

 20140919_101121.jpg   145.82KB   5 downloads  20140919_101132.jpg   149.52KB   4 downloads  20140919_101142.jpg   132.17KB   4 downloads

 

[srfish]

If I use an operating pressure of 54.2 kg/cm2 ,our software shows a dew point of 13 C. That is based on a 50/25/25 split of the C6+ and a convergence pressure of 2470. Should I have used a lower operating pressure?

[Art Montemayor]

Apex:

 

You originally stated “….(I) want to know the moisture dew point achievable by maintaining the parameters shown.”  I have shown you:

• your data is not correct:
• Your process is not designed correctly to yield P/L quality product;
• The correct temperatures and process conditions to yield P/L gas.

You have not given us the design product gas dew point or water content expected from your design.  I suspect that your process was not designed using U.S. TEG technology because your diagram shows things that are not part of the original, U.S. TEG process as designed and incorporated in USA.  That doesn’t mean your unit won’t dehydrate gas; it means it will work inefficiently.  Refer to Rev1 of my workbook and read my comments on the flow diagram that you checked as correct.

 

To give you the dew point using your process information as shown means I or our members would have to make all the calculations.  We are not going to do this – at least not I.  I charge for engineering services rendered.  If you have purchased these natural gas dehydrating units, you should have complete and detailed operating and maintenance instruction manuals indicating what the temperatures, pressures and resulting product should be.  As indicated by some members, you should not have to run any simulation program to find out what the process conditions and results should be – for that you should have the operating manual.  And if the manual is deficient, you have the designer and fabricator (unidentified) who you can consult and demand answers.

 

I seriously doubt that you can produce pipeline quality natural gas with your units operating in the manner shown in your diagrams and with your information.  I hope my attached workbook shows this clearly.

 TEG Regenerator with Steam ReboilerRev1.xlsx   2.2MB   83 downloads

 

[RockDock]

I've also found that Hysys will not give very accurate results for this. The inaccuracy is also compounded when CO2 is present. These days, most TEG designs are done with ProMax. Hysys is good for dry, sweet processes only.

 

I'm with Art on this. This dry gas cannot go into any standard US pipeline. The process appears to be highly inefficient.

 

Going back to your original question about the water dew point, my answer is that Hysys will not give you accurate results. If you want to know the water dew point, take a measurement of the gas, produce a ProMax model or visit the many resources Art has provided. My recommendation is to do all three.

[apex]

What will be the moisture dew point temperature of natural gas containing 85kg water/MMSCMD at 74kg/cm2 abs pressure & 110 kg water/MMSCMD at 51 kg/cm2 abs pressure ?

[P.K.Rao]

By your saying Dew Point, what I understand is you want to know the moisture (water vapor) content in the gas. You can simply determine the moisture content by Moisture detectors or dew point meters. Please refer to ASTM D 4888 and ASTM D 1142. You can also directly determine moisture in gases using KFR.