2 replies to this topic

[extradoctor]

Dear All

An internal floating roof (IFR) tank is used to store nearly pure ethanol. According to API 1626, an air dryer should be installed on the shell to avoid ingress of wet air during inbreathing (partail vacuum). Also all vents on the fixed shell should be kept close.

API STD-2000 guides for calculation of inbreathing rate based on the summation of maximum empting rate+thermal effect and it seems that it is for general type tanks and there is no reference for venting calculation of IFR tanks.

Question: How can i calculate the capacity of air dryer? Should i consider the effect of both movement and thermal effect?

Regards

[proinwv]

Dear All

An internal floating roof (IFR) tank is used to store nearly pure ethanol. According to API 1626, an air dryer should be installed on the shell to avoid ingress of wet air during inbreathing (partail vacuum). Also all vents on the fixed shell should be kept close.

API STD-2000 guides for calculation of inbreathing rate based on the summation of maximum empting rate+thermal effect and it seems that it is for general type tanks and there is no reference for venting calculation of IFR tanks.

Question: How can i calculate the capacity of air dryer? Should i consider the effect of both movement and thermal effect?

Regards

[proinwv]

Since you need to prevent the ingress of moisture then it seems to me that you must consider that there can be product withdrawl at the same time as there is a temperature drop. Of course the ullage space situation in an IFR tank is different than that of a fixed roof tank. The API equation (7) in para.4.3.2.3.3 takes into account the total volume of the tank (to the 0.7 power) whereas in a functioning IFR tank the ullage volume is constant and much less than the tank volume.

This seems to mean to me that the API equation will give you much larger volumetric flow rates than would be experienced. My inclination would be to use the maximum ullage volume (or a multiplier for F.S.), but I am not sure that it would be correct.

I would hope that others might be able to add to this query.