6 replies to this topic

[Paty]

Hi, I am working in a molecular sieves design, and I would like to know, some criterias to select bed numbers for a adsorption system with (200 MMscfd)of natural gas. Some literature suggest to use 3 beds (one in adsorption mode, one in regeneration mode and one in cooling mode).

Thank you....

[garyc148]

I would agree with the literature that you mention stating 3 beds. When you have breakthrough on your operating bed, you need to be able to switch to a regenerated bed. Delays result in off spec products.

Paty,
May be this will help :
1- First af all you need to specify the water spec in the outlet gas and you determine accordingly the maximum quantity of water need to be removed.
2- based on the above you will determine the quantity of the molecular sieve required, may in this stage you will need vendor support.
3-Based on the above you can know if two beds on adsorption and the third in regeneration will be adequate.

You need also to design the regeneration system by identifying first the source of the regeneration gas which generally taken from the mol sieve unit outlet. The flow required will be advised by the Molecular sieve vendor.

I hope this clarifies

[shan]

3 beds design is a typical molecular sieves arrangement. It does nothing with your flow rate 200 MMscfd. The bed diameter is sized per allowable gas velocity through the bed, which is related to 200 MMscfd.

[fallah]

Regardless of the literature,actually in many situations in real world there are two beds in hydrocarbon drying systems,operating in parallel.

At any time,one of the beds is in adsorbtion mode and the other is in regeneration (heating/cooling)or stand-by mode.

[Art Montemayor]

Paty:

You have received four responses to your query. I have designed, built, and operated adsorption units in the past and I disagree with this post on various points. You are posting in what is supposed to be an industrial, professional forum, and not a student-level forum. Yet you state that you are “working in a molecular sieves design” and don’t know how to decide how many adsorption towers to use. This is not what occurs at an industrial-level of process design. What normally occurs is that an experienced, knowledgeable process engineer is assigned to such a project. What is even more confusing is that you don’t tell us WHAT you are adsorbing. Is it water vapor (i.e., are you drying natural gas?) or are you selectively removing certain molecules – such as CO₂, H₂S – or perhaps purifying the natural gas stream itself for removal of certain components or recovery? This type of information should not be left to anybody’s guess.

Fallah´s response is one that makes the most sense because it reveals real-life adsorption applications - but even he is forced to guess at what you are trying to do. He, like the other responses, also guesses that you are drying the natural gas. But you should be SPECIFIC and detailed in your request. If you are a student, then I can understand your basic mistakes in communicating; but then, again, you should post your query in the student Forum.

If you are a professional engineer and you have been assigned to design an adsorption unit without any previous experience or knowledge, then you are potentially going towards failure in your assignment – but not because it is your fault. I would fault your supervisor for assigning such a specialized design to someone who is lacking in hands-on experience in this design assignment.

The basis of any adsorption unit is primarily dependent on WHAT you are trying to remove and under what operating conditions. Even a common and ordinary Unit Operation such as water removal (“drying”) is subject to a variety of key decisions in planning the basic design and operation. For example, if you are indeed simply drying natural gas, it makes a world of difference whether the drying is meant to remove water down to a level of ½ PPM(V) in order to subsequently liquefy the product. This would constitute a typical LNG process plant and one that is so dependent on cash flow and capital intensive that it behooves the designer to always consider safeguards against any potential off-spec product that could conceivably shut down the LNG liquefaction facilities due to a freeze up downstream. This simply is not acceptable and, in that case, the process design incorporates a number of towers with the capability of ensuring that a moisture break-through doesn’t happen.

On the other hand, if you are simply conditioning natural gas for pipeline service (7 lbs water/MM Scf), you have a totally different scope of work and a different design scenario. I would use a 2-tower design as has always been the case for this simple (and relatively benign) application.

As Fallah has inferred, you must be prepared to economically justify any adsorption tower over and above the basic two vessels required. If your scope of work dictates that you require the flexibility and added security of operation, then you should follow those concepts and design accordingly. Otherwise, I would stick with the basic, 2-tower design. I have designed and employed the 2-tower design for drying 3,000 psig air down to less than 1.0 PPM(V) of water as feed for air separation plants operating 24 hours a day for over a year in continuous service – and they performed just as I designed them to work. No one designs an adsorption unit for break-through of the adsorbed compound. This is something that is simply not allowable for an operation where it is vital to maintain a constant product purity.

[Zauberberg]

Number of beds is usually governed by the velocity-pressure criteria, i.e. the crush strength or attrition resistance of mol sieve (a good rule of thumb can be found in GPSA databook). Three vessels is very likely the most economical approach for higher feed gas rates, sometimes you may end up with four - if two vessels in operation result in unreasonably high vessel diameters.