27 replies to this topic

[chandan20]

hi i m assistant process engineer in a fertilizer industry. We use molecular sieve drier for the removal of co2, nh3 and water vapour content from synthesis gas. We have 2 drier one of which rwmains in line for 24 hrs and the gets regenerated at the same time. We have an online analyzer to measure dew point of the outlet gas from the drier. It is being observed that after regeneration of the drier wen it is taken in line dew point remains close to -65 deg C but gradually it goes on decreasing upto -88 deg C at the end of the cycle of 24 hrs. Pls justify the same as i think it should behave in vice versa manner.
Thanks in advance.
Chandan

[Bobby Strain]

Actually the equipment is a dryer, not drier, which is an adjective that indicates a condition of dryness. What you are likely witnessing is that the probe is drying out.

 

Bobby

[chandan20]

thanks bobby and pardon me for using the wrong terminology.
i agree with u that i may be witnessing drying of the probe. how does the probe gets wet every single time as regenerated dryer is taken in line provided that dryer is regenerated with completely dry gas?
my observations are
at the start of the cycle dew point - -65 °C
at the end of the cycle dew point - -88 °C
time of each cycle is 24 hrs
dessicant used is zeolite 13-X.
pls suggest.
chandan

[Bobby Strain]

And, to make you an expert, molecular sieves are desiccants. I wouldn't worry about this behavior. But if you want to pursue it, contact the instrument supplier.

 

Bobby

[chandan20]

thanx bobby. i ll take this up with insrrument vendor.

[shan]

The dryer performance depends on dryer bed temperature. Is your dry bed temperature higher at start of cycle than that of end of cycle? It seems to me that you have potential to increase your dryer on line time because it still performs very well to dry the stream to -88 C at the end of current cycle.

[chandan20]

yes bed temperature is 9 °C at the start of the cycle and 7 °C at the end of the cycle but does 2 °C create so much of a difference?

[shan]

Actually, it is not that much.  Please check your stream water content at -65 C dew point and water content at -88 C dew point.  I bit the difference of water content is not a significant value.  At that range of water content, a little bit of water content change will lead to a noticeable dew point variation.

[chandan20]

dear shan,

 

 You may be correct from this viewpoint, but i have no means to find the absolute humidity as my medium is not air, but syn gas ( hydrogen and nitrogen in the ratio of 2:1 respectively as well my pressure is 29 kg/cm2 g).  Most of the literature provides humidity charts for air at armospheric pressure.

Still, ur viewpoint has given me a better prespective.

thanks

Edited by Art Montemayor, 20 March 2013 - 11:49 AM.
terrible spelling errors; composition

[shan]

 You may input the gas composition and process conditions into Hysys to obtain a decent dew point vs. water content table.

[thorium90]

What are you using for the regeneration? Just to confirm, the regen gas goes in from the top? The online bottle also has the gas going in from the top? I dont think the reason is due to 2C difference in the bed inlet temperature. I think the answer lies in the regeneration. Could you share the time length of each major step in the regeneration cycle? It would be even more helpful if you have a trend to show. y-axis should be temperature and x-axis should be time, preferably at least 24hrs. The trend should show the outlet temperature of the regeneration gas in that time frame. This trend should instantly confirm or disprove my hypothesis of why you observe the dew point decreasing while it is online. Hope its not too much to ask.

Edited by thorium90, 20 March 2013 - 09:24 AM.

[Art Montemayor]

Shan is right.  The dryer performance depends on dryer bed temperature.  However, an adsorption dryer’s performance is designed to yield a final product dew point that is relatively constant.  Based on field experience operating adsorption dryers with Mol Sieve adsorbent, I have never found a product gas dew point to vary beyond what I know to be the dew point meter’s accuracy or reliability.  There are dew point meters – and then there are “drop” point meters.  There are only one or two makes of meters that I would trust for reliable, consistent, accurate results.  Certainly an on-line dew point meter is not one of them.  That is my factual experience.

 

Chandan20 doesn’t explain what he calls the “cycle”.  Is it a NEMA cycle?  Or is it the adsorption cycle or regeneration cycle?  Even the regeneration cycle is split into two parts: the heat up and the cool down.  Additionally, we don’t know what type (or setup) of regeneration is being used on this unit.  Is dry, product gas being used on the total regen?  Or is another gas being used?  Also, what is mode of regeneration – is both the heat up and cool down done counter-current to the adsorption flow?  All this could be very simply explained in a detailed sketch, showing the conditions and locations of the various valves and instruments being mentioned.  That would be a lot more productive and quicker than the speculation being done.

[thorium90]

Yes Art, I agree with you, many things are not specified. The setup of the regeneration, the regen gas etc etc. It is therefore hard to prove that the difference is due to the 2C in bed inlet temperature. As the final dew point should be relatively constant even with slight fluctuations in inlet temperature, the difference could either be due to the dew point meter itself or how the regeneration is done.

Evidently using hysys will not help, as it is unlikely he has some online GC or other analyzer to measure the full composition of the outlet stream. He most likely only has a dew point meter am I right? Furthermore, at that kind of values, it is not an operational problem but rather, just a subject of curiosity. Even getting the dewpoint from hysys will still not explain the phenomenon that OP is interested in.

Edited by thorium90, 21 March 2013 - 12:33 AM.

[shan]

Per Hysys,
In the condition: H2: 0.66667 mole fraction, N2: 0.33333 mole fraction at 29 kg/cm2 g (427.2 psia)
The water content is 0.0754 lb/mmscf for -65 C (-85F) dew point and 0.0464 lb/mmscf for -88C (-126.4F).
The water content difference is 0.0754 - 0.0464 = 0.0290 lb/mmscf, which is less than a mouthful water (a common bottle of water = 500 g = 1.1 lb, 0.029 lb is 2.6% of a bottle of water) in 100 ft x 100 x ft x 100 ft cube.
Please verify if your dryer is able to adsorb additional 0.0290 lb/mmscf water if the dryer operating temperature is 2C lower.

Edited by shan, 22 March 2013 - 06:51 AM.

[chandan20]

dear all professionals, sry to provide u insufficient data,
there r 2 dryers A & B. one remains in line for 24 hrs and other gets regenerated in 24 hrs.
the cycle i m referring to is regeneration cycle.
individual time for each step in regeneratikn cycle is
Parallel Mode( where both driers r in line) 10 min
isolation step(wen one of the drier gets
isolated for regeneration purpose) 5 min
depressurisation of the 1 which gets regen 120 min
heating step 660 min
cooling step 540 min
pressurisation step 90 min
standby mode 15 min

as shown regeneration involves both heating and cooling down of the dryer.
Now the flow pattern
for inline dryer flow is in top to bottom direction.( for adsorption)
for heating step the flow is in bottom to top direction.
for cooling step the flow is in again top to bottom direction.

the gas used for regeneration is completely dry as it comes from cryogenic cold box at the temperature of 2°C. major component of this gas is nitrogen, methane and argon.
for heating step this gas is heated to 250°C in an exchanger where heating medium is steam. at the end of heating step the temp of bed reaches 210°C.

and yes we r nt facing ny difficulty in our operation. this is just for better understanding of mine for better process control.
the dryer works gud even at inlet temperature of 12°C as in summers the bed inlet temp reaches this temperature.

and at the end guys well this is quite embarassing but i cud not find any tab to attach files that is why i had to write all this. can anybody tell me how do we attach files.
sry for inconvenience

Edited by chandan20, 21 March 2013 - 02:13 PM.

[chandan20]

the volume of the bed is 44 cubic metre and the regeneration flow is 16000 Normal cubic metre per hour giving us the space velocity to be nearly 400 hour inverse.

[Art Montemayor]

Chandan20:

 

It is important to know how to upload a file on the Forum.  Follow the instructions below:

1. First thing: go to the “more reply options” button at the bottom of the Editor window;
2. After writing your message and leaving your cursor resting at a space below your message, go to the bottom of the Editor page and click on the “Choose Files …” button directly under the title “Attach Files”.
3. A window will appear where you can select the file to upload from your filing system.  Select the file you want and click “Open”.  This will load the file into the Editor and you will see its name directly above the “Attach Files” title.
4. To the right hand side of the file name you will see two commands – “Add to Post | Delete”.  Click on “Add to Post” and the file will be loaded onto the message in the Editor where you last left the cursor.  It is best if you leave your cursor at the end of your message, with one line of spacing.
5. The message – with the attached file – will be posted when you click the “Add Reply” button.

 

[Art Montemayor]

Chandon20:

 

I had to edit your prior post because I couldn't understand it.  Please write your posts in English, not in texting or street vulgarization of the English language.  Compare my product with yours.

 

Dear all professionals,

 

Sorry to provide you with insufficient data.

 

There are 2 dryers, A & B.  One remains on line for 24 hrs while the other gets regenerated in 24 hrs.

 

The cycle I’m referring to is the regeneration cycle and the individual time for each step while in the regeneration cycle is:

1. Parallel Mode (where both dryers are on line) = 10 min
2. Isolation step (when one of the dryer gets isolated for regeneration purpose) = 5 min
3. Depressurization of the regenerated bed = 120 min
4. Heating step = 660 min
5. Cooling step = 540 min
6. Pressurization step = 90 min
7. Standby mode = 15 min

Total Regeneration time = (10+5+120+660+540+90+15) = 1,440 mins = 24 hrs

 

As shown above, the regeneration involves both heating and cooling down of the bed.

 

Now the flow pattern through the beds:

• For the online dryer bed, the flow is in top-to-bottom direction.(for adsorption)
• For the heating step the flow is in bottom-to-top direction.
• For the cooling step the flow is in again top-to-bottom direction.

 

The gas used for regeneration is completely dry as it comes from a cryogenic cold box at a temperature of 2 °C.  This gas contains nitrogen, methane, and argon.  For the heating step, this gas is heated to 250 °C in an exchanger using steam.  At the end of the heating step the temperature of the bed reaches 210 °C.

 

And yes, we are not facing any difficulty in our operation.  This is just for my better understanding and better process control.  The dryer works good even at inlet temperature of 12 °C as in summers the bed inlet temp reaches this temperature.

 

And at the end, guys, well this is quite embarrassing but I could not find any tab to attach files.  That is why I had to write all this.  Can anybody tell me how do we attach files?

 

Sorry for the inconvenience.

 

[chandan20]

Dear all,

please find below the heating and cooling cycle flow pattern in the dryer.

Attached Files

•  MOLDRYER.pdf   10.3KB   120 downloads
•  MOLDRYER2.pdf   95.16KB   104 downloads

[thorium90]

So you have 11 hrs of heating time and 9 hrs of cooling down.

Is the temperature already stable at 210C for some time (a few hrs?) before cooling down starts?

What temperature does the cooling down ends at? About similar to the process gas inlet temperature of 7-9C?

[chandan20]

Yes temperature remains stable at 210 °C for nearly 3-4 hrs and then cooling down starts.
Cooling down steps end at similar process gas inlet temperature of 7-9°C.
Regeneration ( heating up and cooling down) is done at 1.5 kg/cm2 g pressure.

[thorium90]

Well, your cycle and steps look pretty normal. Im gonna have to agree with Bobby on this one. The probe takes a long time to dry out. Or it could just be that the sampling tubing going to the meter is really long and the flow is low enough that it takes time for the entire tubing and equipment to reach a new "dryness" equilibrium with a different dryer bottle online. I assume there is only one dew point meter and it is located downstream of V-5 or MOV-608 (curiously both are labelled differently?)

Perhaps a trend showing the dewpoint wrt the dryer steps over a 48hr period would help visualize the trend and demonstrate the long time constant of the dew point meter thus confirming Bobby's hypothesis and Art's advice on certain makes of meters.

[chandan20]

Thorium90

 

please find below the attached file showing behaviour of dew point across 48 hr timeline.

 

hour 1 is the time when fully regenerated dryer was taken on line and accordingly after 24 hrs other dryer was taken online.

Attached Files

•  DEW PONT VARIATION.xlsx   11.51KB   67 downloads

[Art Montemayor]

I have organized and formatted the data submitted in the same workbook and submit it here as Rev1.   I have included my notes and comments to fully explain in one page what is being done.

 

As will be read, I don't have any problems reversing the cooling flow pattern as compared to the heating direction.  It obviously requires more instrumentation and controls and accomplishes the same regeneration level and dew point result as simply leaving the regen gas to continue on without any heating.  Everyone has his/her way of doing the same thing.

 

The same applies to the repressurization step.  I normally have used a restriction orifice and flowed a portion of dry product gas through it and into the cooled, regenerated bed.  This accomplishes the same repressurization as that shown, but with one important caveat: by repressurizing with dry product gas, there is no adsorption (drying) capacity loss as there is in repressurizing with wet process gas.  That is the way I interpret the flow diagram that shows a by-pass valve that allows wet process gas to enter the regenerated bed.

 

The variations in the on-line dew point meter could easily be the result of a faulty or un-calibrated meter or the fact that most on-line meters can't be trusted to yield continuous, accurate results.  This has always been a problem in natural gas glycol dehydration units producing gas for custody meters.   The dew point could also be varying because of leaking valves within the unit.  This also is a continuous problem, depending on the type and quality of valve used.  These valves should be very special shut-off valves that can withstand the continuous heat and cool-down stresses that they are exposed to on each and every cycle.

 

I hope this experience helps.

 

 DEW PONT VARIATION Rev1.xlsx   148.79KB   93 downloads

[chandan20]

thank you all professoonals for your valuable input in this topic.

i think i understand its operational behaviour better now. Once again thank you all.

Regards
Chandan20

Edited by chandan20, 26 March 2013 - 09:50 AM.