42 replies to this topic

[Zauberberg]

If you are interested, there are two sub-sections of the Eng-Tips Forum dedicated to ASME Codes, and AUS/NZ pressure vessel codes:

http://www.eng-tips....der.cfm?pid=744
http://www.eng-tips....der.cfm?pid=292

You can register at Eng-Tips and post your queries - it is very likely that you will get some quality answers.

Cheers,

[Alistairm]

Hi,

I've looked through the documents and am happy with them. They've helped me set up my own spreadsheet that is in accordance with my own countries standard.

A few quick questions:

When selecting this steel, A 516-70 (or equivalent AS 1548 7-490 in Australia) I need to justify why I chose it. For example, why I chose this over a stainless steel, which would give me a smaller thickness and safeguard me from corrosion if an abnormal case with H2S and CO2 were to occur. Can anybody help me with the justification of this low-alloy steel?

When designing my unit, which design cases should I design for and what would they involve?

In terms of co-adsorption, which species would adsorb preferentially after water on a 4A molecular sieve? My composition contains N2, CH4, C2H6 and heavier hydrocarbons.

I have calculated the bed height to be 6.63 metres. How much do I need to add for the total height of the vessel?

How does the desiccant get removed after it has been spent? Do I need to design a 'lid' at the top of the vessel?

I have chosen to go for external insulation - should I go for a wool style insulation?

Regards

tonyflow1

Edited by tonyflow1, 07 September 2010 - 10:06 AM.

[Alistairm]

ah, and thanks Zauberberg - I just saw your post. I didn't see it until now because it was on the 2nd page.

Cheers

[Zauberberg]

1. 516-70 is much cheaper than stainless steel while it serves the purpose completely. Normally, stainless steel adsorber is never employed for drying sweet gases. Even at H2S/CO2 breakthrough from the Amine Unit, there should be no corrosive environment since there is no aqueous phase inside the vessel. Acid gases are only capture some of the adsorbent capacity.

2. Make sure to consider all cases when you have turndown conditions, so as to avoid channeling inside the bed. There is a minimum gas flow rate at which you have to switch to 1-bed operation, which requires modifying dryer sequences (adsorption/regeneration) in the DCS as well. Also, from mechanical point of view, ensure bed integrity (particular attention to be paid to bottom mesh screen, and ceramic rope - to prevent adsorbent downfall and plugging the outlet basket), and the vessel inlet piping to minimize flow maldistribution.

3. 4A Mol Sieve adsorbs Hydrocarbon C2- components, which includes Methane and Ethane. Account for co-adsorption of Hydrocarbons at approx. 50% quantity of adsorbed water. This information is readily available on the internet.

5. There should be sufficient space above the top layer of ceramic balls in order to ensure proper gas distribution across the bed. Absolute dimensions cannot be given because the vessel diameter varies from case to case. Here's a loading diagram from my plant (6 beds total, 4 in adsorption, wet gas rate 1,500 MMSCFD) where you can see relative dimensions.

6. There should be a catalyst unloading nozzle on the vessel close to the bottom of the bed (usually 2-3 feet above the bottom support grid). Rest of the adsorbent is usually removed by vacuum - for dry unloading, which I strongly recommend.

7. For evaluating different types if insulation (including cost), there is an excellent software available for free download from: http://www.pipeinsul...4/download.html

Best regards,

Attached Files

•  Loading Diagram.JPG   53.21KB   92 downloads

[Alistairm]

Thanks for the swift reply.

What do you mean by "Rest of the adsorbent is usually removed by vacuum - for dry unloading". And how do I determine the nozzle size for desiccant unloading?

Regards

tonyflow1

[Zauberberg]

There are two ways to remove spent adsorbent during plant turnaround: a "dry", and a "wet method".

The dry method comprises of purging (or even better, pressure-popping) the adsorber with Nitrogen till the LEL is brought down to a sufficient level for opening the vessel and dumping of the desiccant. Simply, the side nozzle is open and adsorbent allowed to flow out by gravity, while being collected in a sort of hopper. Size of the nozzle is irrelevant - just don't make it unreasonably small and make the desiccant unloading last forever, or too big to affect mechanical design of the vessel. As far as my memory serves me, it is very similar size to the feed gas nozzle size. Make it large enough to accommodate vessel entry by the qualified personnel - not less than 24".

"Vacuum" refers to the procedure similar to vacuum-cleaning in your house. I'm sure you know how this thing works, although we prefer women doing this particular job. A simple hose that is connected to the vacuum device, used to suck the remaining desiccant out of the vessel, instead of painful shoveling.

The "wet" method comprises of flooding the adsorbent vessel with water, which is supposed to create an inert atmosphere. But - according to my practice - it just creates a mess: it makes unloading of adsorbent extremely difficult, and sometimes you can even create a kind of "cement/concrete" inside the vessel and around the outlet basket. Maybe someone has good experience with wet unloading. My experience is equivalent to disaster.

A good thing is also to provide small connections along the vertical length of the vessel, for adsorbent sampling purposes. Top, middle, and bottom of the bed should be more than sufficient.

[Alistairm]

I think I will design for dry purging as it seems the logical choice. What does 'LEL' stand for?

I've had an extensive look online for ethane isotherms to determine co-adsorption but couldn't find any, or any information on how to account for this co-adsorption. I even rang up the local UOP adsorbent supplier but couldn't help with any isotherms and said that co-adsorption is of low concern to the commercial world. Is there any further information that you can supply?

Regards,

tonyflow1

[Zauberberg]

LEL = Lower Explosive Limit, is by the definition "the smallest amount of the gas that supports a self-propagating flame when mixed with air (or oxygen) and ignited. In gas-detection systems, the amount of gas present is specified as a percentage (%) LEL. Zero percent (0%) LEL denotes a combustible gas-free atmosphere. One hundred percent (100%) lower explosive limit denotes an atmosphere in which gas is at its lower flammable limit."

Normally, you want to ensure that the atmosphere around adsorbent is such that the LEL can never be reached. As said, pressure popping witn Nitrogen, or even hot Nitrogen stripping can be applied to drive the combustible gases out off the desiccant.

Hydrocarbon adsorption: getting the isotherms would be good, but I don't know if they are available. If you find them, please publish that information on ChE Resources. On the other hand - looking from a practical standpoint - design the system with a conservative margin: assuming co-adsorbed HC's equivalent to 50% of adsorbed water will result in bigger quantity of desiccant, and in higher regeneration heat requirements (don't forget that heat is required to drive off the Hydrocarbons from adsorbent as well). If you assume this figure too low, you may end up with an undersized adsorption system.

[Zauberberg]

And here's a snapshot of Adsorber general arrangement drawing, showing 24" manhole which is used for catalyst unloading.

Attached Files

•  Adsorber GA.JPG   89.72KB   97 downloads

[Zauberberg]

Tony,

In addition, attached are some technical materials that may be of interest - particularly an example of Adsorption process datasheet published by Porocel, and the "compound bed" technology which is more and more present nowadays - employing Alumina or Silica Gel on top of the Mol Sieve layer, to act in two ways:

- Protection against liquid carryover from upstream unit operations (we have excellent experience with Sorbead WS, a grade of Silica Gel). Mol Sieve itself is very vulnerable in presence of liquid water or even worse, the amine;

- Additional adsorption capacity provided by these layers which is even higher than for the Mol Sieves, since Alumina/Silica Gel have higher adsorption capacity at high relative humidity regions of feed gas (see their adsorption isotherms).

Ola,

Attached Files

•  NG Dehydration Datasheet - POROCEL.pdf   93.51KB   87 downloads
•  Optimizing Solid Bed.pdf   802.9KB   102 downloads
•  Effects of Contaminants on Molecular Sieves.pdf   122.02KB   79 downloads
•  Axens Adsorbents.pdf   799.75KB   81 downloads

[Alistairm]

Wow, those attachments are great. Good to see an adsorber specific spec sheet, now I know what direction I need to take for my own.

As for the ethane/methane isotherms, there don't seem to be any readily available. Here is my correspondence with a local supplier for UOP mol sieves:

Co-adsorption of Ethane and Methane onto the Molecular Sieves in a NG dehydrator does not seem to be a major issue for the LNG industry. This is not to say that it does not occur in any form, rather that it is not a big issue for operational plants. The reasons for this is likely to be the following

1) The regeneration gas (and any co-adsorbed Ethane/Methane) is usually returned to the feed stream and recovered.

2) The Molsiv’s affinity for Ethane/Methane is very low and will only be co-adsorbed in the MTZ or WUB (fresh) section of the bed and this will also be influenced by what other more preferentially adsorbed contaminants are present in the gas stream.

Regards

tonyflow1

[Zauberberg]

Thanks for posting the correspondence details. I have had similar discussions with both UOP/Grace, and they had very much similar answers. I hope that Don Pfennig - who has recently joined ChE Forums - will see this excellent and very educational thread, and post his valuable comments. Don is a veteran in adsorption business (previously worked with Grace, then starting his own business), and his suggestions are a true gold mine. He used to guide us frequently through e-mails, and also performing two Mol Sieve performance test runs, in the Equatorial Guinea LNG plant, West Africa, operated by Marathon from US.

Nevertheless - considering your post under (1), bear in mind that we are not speaking that much about recovery of Hydrocarbons, but rather about up to which extent they rob adsorbent useful capacity. By recycling the regeneration gas, HC's are recovered - that is true, but how much of them is actually being co-adsorbed by the Mol Sieve during active cycle? Because that is an important design information to consider, as it affects the required quantity of desiccant for complete dehydration of feed gas. As I wrote in my previous post, assuming a figure that is too low will lead to undersized adsorption/regeneration system. Being a little bit more conservative can save you from lot of trouble, downtime, lost revenue, and eventually a personal feeling of shame that you have screwed up the unit design (which is the most difficult element, I suppose).

In any case, that critical part is left to the catalyst/adsorbent supplier, and you just have to hope that everything will be as good as it usually is, everywhere in a similar units worldwide.

[Zauberberg]

I found this file in my library, showing how the catalyst unloading nozzle looks in some other applications. It is only 2" size and 45 degree oriented downwards.

We used to dump spent Mol Sieve through the 24" manhole without any problems - of course after purging the bed and lowering the LEL down to acceptable limit.

Attached Files

•  Catalyst Unloading Nozzle.pdf   61.49KB   68 downloads

[Art Montemayor]

Zauber:

Your dump nozzle appears to be 8 inches in nominal size, not 2. I doubt if a 2-inch nozzle could be effective in removing or "dumping" adsorbent - in beds >24" ID. An 8" nozzle is about the average that I have specified and used in the past, although I've gone as high as 12". I don't like to specify a larger size because I prefer to dump molecular sieve under controlled conditions:

1) I insist I be present during the dump - or someone with engineering management authority; I never leave it up to just maintenance to do the job thoroughly, safely, and under direct supervision. Mol Sieve is just too expensive - in my opinion - and when the time comes to dump the sieve it is an important event in that it has either come to the end of its effective life or it is being dumped because of process troubles. In either case, a careful analysis of what is present in the bed, where it shows bad signs, and the taking of samples for learning purposes is part of any dumping.

2) I always dump under a batch system, storing the product in clean, plastic bag lined, 55 gal drums that are immediately sealed as they are filled, tagged and stored for later analysis and/or disposal.

3) The degree of attrition has always been a concern with me, as well as the color and size distribution. The process of adsorbent "aging" is still an unknown and arcane art and it helps to record any evidence that may lead to better future adsorbent performance.

4) Bed poisoning or "pluggage" is another concern that haunts the operator when performance starts to diminish. Careful sampling and analysis for future study is an investment that can pay back in better performance in the future.

As we who have had operating experience with adsorbers know, adsorption is a very direct and simple process, but there is still much that can go wrong and much left to learn on successfully controlling its performance in the field. In my opinion, you are doing a first class job of pointing out the importance of the details of what one would think is a mundane item - the removal of adsorbent. This is a key operation that, if designed and operated correctly, need not be done very often. But when it is required, it is a key and critical job in a gas processing unit.

[Zauberberg]

Art,

It is an 8" nozzle, you are absolutely correct. Apologize for misreading the drawing.

Best regards,

[vne]

***

[Art Montemayor]

VNE:

Please start your own, new thread on your specific topic. Do not sky jack another member's thread for your own purposes. This is a Forum, and not a Blog.

I will erase this post after a day.

Thank you.

[vne]

Sorry for posting to the wrong place.

New thread created.

Thanks