4 replies to this topic

[Fbion]

Hello,

 

I was assigned to evaluate the Nitrogen gas flow rate requirement given by our contractor. This N2 gas is used for the regeneration of our molecular sieve dryers which during pre-commissioning phase will only contain moisture (no hydrocarbons). This regeneration will be carried out during air dry-out of our plant in which time, only air will be circulated in the system to ensure the removal of moisture after the hydrotesting phase. 

 

During normal operation, methane is used as the regen gas. Nitrogen is only used for pre-commissioning and start-up. The calculation of N2 gas requirement by our contractor assumed that the volumetric flow rate for methane and nitrogen should be the same. Which i think is based on the superficial velocity of the gas through the beds. But they have also assumed that heating time is the same for both (they have calculated the N2 regen gas flow rate based on the heating time using methane, which is 9 hrs). Knowing that the heat capacities of both methane & nitrogen is very different, 0.59 Btu/lb.F and 0.25 Btu/lb.F (respectively) , my engineering instincts tell me that a longer heating time should be required when using Nitrogen as regen gas. However, I do not have any past experience on using N2 as regen gas. 

 

Please give me your valuable opinion and comments on this analysis.... What is the typical heat-up cooling curve when N2 gas is used for regeneration?

 

Thank you very much,

 

[Art Montemayor]

Fbion:

 

I believe you are correct in your observation of the basic differences between nitrogen and natural gas when applied as the heatup AND COOLING media in an adsorption bed.

 

In the past I have done many similar calculations with various gases – especially using nitrogen as the heating and cooling media in regenerating adsorbers for air separation plants (nitrogen is a natural by-product of an air separation unit).

 

You are not correct in deducing that the volumetric flow rate is “based on the superficial velocity of the gas through the beds”.  The volumetric flow rate of any regen gas (in both heating and cooling phases) is based on the time requirement for such an operation, the heating and cooling temperatures applied, and the heat capacity of the carrier gas.  The superficial velocity is a criteria applied for other reasons.

 

In my opinion, you should start the required calculations for the heating and cooling cycle.  You have failed to consider the cooling phase of the total bed regeneration and this is an important phase you must calculate as well as the heating phase.  You should have all the necessary basic data already compiled for you by your contractor regarding the adsorption unit – data like the dimensions of the vessels, the beds, the total weight of the vessels and the adsorbent (Mol Sieves, in this case), and all the attached piping and valving.  You certainly should know the heating temperature generated by the heater and the design bed final heating temperature as well as the cooling temperatures.  You should assume that the adsorption bed will be saturated to some extent before beginning the heatup.

 

The calculations are pretty straight-forward and with all the basic data in place should not take you more than a day.

[Fbion]

Thank you very much Sir Art. The calculation given by our contractor is definitely wrong. In your past experience with using Nitrogen as regen gas, for how long does heating time usually take? I have already calculated the volumetric flow rate of Nitrogen as 4560 cubic ft/min based on a heating time of 9 hours. After calculating the pressure drop across the beds, I found that the pressure drop for the guard bed exceeds the design. In this case, I would have to increase the heating time to lower the volumetric flow rate...

 

Thank you Sir.

[Art Montemayor]

 

Fbion:

 

The regeneration (the same as the adsorption) calculations are done once the basic data is established for the drying unit.  Part of the basic data is the established drying and regenerating time that you wish to impose on the unit.  This used to be called a NEMA cycle (I don’t remember why; I think it is in respect to a complete sine curve defining a cycle) and it details all the steps in the complete cycle of an adsorber bed: switching, purging, heatup, cooling, pressurization, and switching.  The number and type of steps (and the time allotted to each) may vary, depending on the application.

 

Normally, I always designed my adsorbers to operate in accordance to my operator’s working schedule.  This is done a lot because the adsorption cycle is semi-batch in that beds have to be switched on a scheduled basis.  It is very wise to have your operators present to administer or supervise the switching of the beds – whether manually or automatically.  If that is the case, then popular NEMA cycles are 16 hours (4 hrs heating, 4 hrs cooling; 8 hours drying), and 32 hours (8 hours heating, 8 hrs cooling; 16 hours drying).  As you can probably surmise, this basic data has already been set by your engineering contractor who did the original calculations and specified the fabrication of the unit.  You need to have the original adsorber calculations and specification (data) sheets.  Those will tell you what you have inherited plus what contingency there exists (if any) in the basic design.  You cannot vary any part of a set NEMA cycle without affecting another part of the cycle.  If you extend the heating time, you have to reduce the adsorption time.  The beds are already sized for a given NEMA cycle and you can’t play around with the establish cycle beyond the contingency allowance given in the basic bed design.  If you are forced to extend the heating cycle, you are in trouble with your operating requirements because your drying capacity will be reduced per bed and this will force more regenerations and result in less adsorbent life, and other negative results.  This all means that what you are doing is indeed very important prior to accepting the turn-over of this dryer from the contractor.

 

What you can do is increase the heating temperature exiting the heater.  Raising the regen temperature means you need less regen gas flow rate.  However, you are limited by the allowable design stress in your units materials of construction.  This information should also be in your data sheets.  Normally, I would expect an engineering contractor to design in such a manner so as to maximize the profit – by using conventional steels (such as A516 Grade 70).  This limits the allowable temperature to approximately 500-600 ^oF.  When applying molecular sieves, I have always used as high a temperature as my vessels, piping, and valves would allow and this usually turned out to be 600 – 700 ^oF.  Molecular Sieves regenerate better at higher temperatures.

 

I presume you have used the Ergun equation to calculate the pressure drop across the bed – both for adsorption and regeneration.  Make sure that you evaluate the up-lifting effect on the bed as it relates to the calculated pressure drop.  Any bed movement caused by an excessive pressure drop means adsorbent attrition and dusting – both of which mean a lot of trouble for the operations and maintenance.

 

You haven’t furnished any basic data so that I can’t specifically comment on any one topic within your drying unit’s design and operation.

[Bobby Strain]

If it were my operation, I would ask my contractor to supply the necessary information from the adsorber system vendor. That is the best way to "check" your contractor's calculation.

 

Bobby