8 replies to this topic

[process101]

Hi every one,
I was asked to find out that if we can use one stage desiccant dryer (physical absorbent) for the following gas conditions.
Gas flow rate: 1 MMscfd
Pressure: 1000 psig
Temperature: 120 F
Gas is under saturated. Water contents in gas is about 30lb/MMscf in feed gas
Calculation says that we need three desiccant dryer to get less than 7 lb/ MM.
I was surprised that for 100% saturated we also need three stages. Why can’t we use one stage to get dew point?
But funny thing is that even by increasing the size of the vessel it doesn’t change the number of stages.
I asked vendor about this but they could explain me except that they told me “it is because of Temperature and pressure”
Can someone explain me in detail why it required three stages?
Thank you in advance

[ankur2061]

Process101,

GPSA databook has a complete section devoted to Solid Dessicant Dehydration (Section 20: Dehydration) for water removal from natural gas. I have developed a spreadseet based on the GPSA method using USC units (it uses a modified "Ergun" equation) and if you can provide me the following data I can provide you the bed diameter (ft) and bed length (ft).

Data Required:

Gas Flow Rate, MMSCFD

Mol Wt or gas composition

Inlet pressure to dessicant dryer, psia

Inlet Temp. to dryer, deg F

Required dewpoint of outlet dry gas, deg F

Particle type: 1/8" bead or 1/8" extrudate or 1/16" bead or 1/16" extrudate

Allowable pressure drop per unit length of the dessicant bed, psi / ft
(Note: Typical value for design as suggested by GPSA is 0.33 psi / ft)

Send me the data and I can work up some numbers for you.

Regards,
Ankur.

[process101]

Ankur:
Thank you for your reply. Actually we are not working on Mol Sieve Desiccant; we are working on desiccant RG, SG and hydrolith. Their properties are different then mole sieve.

I appreciate if you can help me to understand that. if we have gas with following conditions
Flow: 1 MMscfd
Pressure: 1000 psig
Temperature: 120 F
Specific Gravity: 0.65
Water content in feed gas is : 30 lb /MM
Gas is not 100 % saturated

When it goes to vendor he tells me that we need 3-stages to dehydrate this unsaturated gas to 7 lb/MM
Although gas has only 30 lb/ MM of water in feed gas.
When vendor size the desiccant for 100% saturated gas. He still tells me that we need 3- stages.

I want to understand why do we need 3-stages when water contents in the gas is 30 lb/MM

Thank you

[ankur2061]

Process101,

I don't understand your terminology of unsaturated gas. The gas is saturated with water vapor at the inlet conditions of temperature and pressure to the dessicant system,is my undertsanding.

For a proprietary dessicant media, you will have to depend on the vendor to provide you the basic design for the dessicant system. I cannot provide you the answer for the system design using a proprietary dessicant media.

The GPSA method for sizing a dessicant bed is very generic in nature and the final sizing and design of the dessicant system will have to be confirmed by the vendor. Why don't you ask the vendor about his design methodology, although I have doubts that the vendor will provide you this information.

My idea of replying to your post was that the GPSA engineering databook does provide a design / sizing method for mole sieve type of dessicant system and this method maybe used to estimate the footprint (layout) for the dessicant system rather than arrive at a very exact design which only a specialized dessicant system manufacturer would be able to provide.

Your best bet is to go back to manufacturer / vendor and ask for more calrifications.

Regards,
Ankur.

[process101]

Anker,
What i mean of unsaturated is “That Gas stream is 20 % saturated" I am sorry for the miscommunication I should have mention that in the first place.
The question is “why do we need 3- stages of descant to dry gas when gas is 20 % saturated. For 100% saturated gas it makes sense to use 3- stages of desiccant but for 20% saturated gas 3 stages requirement. Is it because of equilibrium?

Thank you

[ankur2061]

process101,

As I mentioned earlier proprietary dessicant systems will necessarily be sized by vendor. The dessicant properties play a big role in the adsorption characteristics of the dessicant. Most vendors will not provide details of their proprietary dessicant systems. Dessicant systems are still not considered an open art in engineering and you will have to go with what the vendor proposes.

Regards,
Ankur.

[process101]

Thank ankur for your help

[kkala]

Following interpretation is not complete, but probably of some usefulness. "Digging" by a knowledgeable person can give a much better answer. Advice / comments would help understanding and probably improve the interpretation.
1. One stage reported is understood as one column. A fixed bed column can represent numerous equilibrium stages.
2. Regenerating columns beside the operating ones are necessary, but only operating columns are considered here.
3. Reference used for adsorption is Fig 391b of "Unit Operations" by G. G. Brown (Wiley, 1950), where vapor pressure of water adsorbed by SG (=silica gel) is plotted against (actually) temperature. This vapor pressure varies according to (h) how much water has been adsorbed per unit mass of dry SG. The example is based on the specific SG in reference.
4. Plot of equilibrium vapor pressure of H2O adsorbed by SG, versus quantity of this H2O adsorbed per unit mass of dry SG (h), can be seen in attached "dess.xls" for 120 oF temperature. As water content of SG increases, said vapor pressure also increases reducing SG potential as drying agent.
5. As indicated in "dess.xls", outlet requirements of 7 lb/MMscf grossly mean H2O vapor pressure of 0.00030 in Hg (abs). In order to have adsorption, equilibrium pressure of absorbed water has to be lower than 0.00030 in Hg, meaning max h=4.5E-5 lb H2O per lb of SG. Indicated SG has a range of h=0.01 to 0.40. So SG is not proper as a desiccant for the last column, while it could be satisfactory for the first column. This is also concluded from http://www.sorbentsy...ts_charts.html. Another desiccant, e.g. CaO, could fit the last column.
6. So vendor intends to place different desiccants in separate columns in order to efficiently realize the required extent of drying. Desiccants may be three, depending on cost optimization considered by Vendor (we do not know them). This does not change if inlet stream contains 30 or 80 lb H2O / MMscf, same desiccant for their first column is used.
7. Same desiccant of different grade / properties is considered as another desiccant. There may be other reasons for placing 3 columns in series, e.g. migration of absorbed humidity by diffusion into the desiccant, due to local difference in absorbed water. No idea about significance of this.
Lack of experience on the matter is pointed out. As mentioned, advice would be highly appreciated for better education and interpretation.

Attached Files

•  dess.xls   29KB   57 downloads

Edited by kkala, 29 October 2011 - 10:04 AM.

[kkala]

In relation to previous post, attached you can find mentioned fig 391b of G. G. Brown's book, in case it has some interest for the subject.

Attached Files

•  equipment.jpg   1.41MB   26 downloads