2 replies to this topic

[gmilligan11]

Hi There,

 

I currently work in the Scotch Whisky Industry and recently the Competent Authority (CA) has asked all COMAH sites to undertake an assessment of all vent sizes for our ethanol storage tanks. This is to ensure there compliance for a fire engulfment scenario.

 

All tanks which are to be assessed are low pressure storage tanks and therefore I have been able to calculate the venting required for every tank we currently have on each site using API 2000 and NFPA 30 (this gives additional reductions for tank farms containing only ethanol). At this point I have been going a little round in circles to actually size the vents based on this required rate and to compare this to what we have physically installed. I will detail and example of a tank below:

 

Tank Design: Vertical (supported 1.2 m above grade)

Tank diameter: 3.8 m (inner)

Tank height: 4.3 m

Wetted area: 82.64 m² (889.52 ft²)

Design Pressure: 10 mBar (we have received guidance on this if no tank details/records exist using hydrostatic testing)

 

Venting Requirements:

API 2000: 289532 SCFH (21717 lb/hr)

NFPA 30: 131340 SCFH (9852 lb/hr)

 

After this point I have used API 520 Part I to size which gave a result of 5 inch, which to me, seems pretty small for a vessel of this size. I have not long left university and have consulted my previous textbooks and references. I have also looked at Crane TP 410 for guidance but do not know exactly what equations to actually use or modify to suit my needs (steam sizing for example). (I know the required relief rate, the design pressure (upstream) and downstream pressure (atmospheric))

If someone could point me in the right direction it would be apprieciated.

 

Thanks,

Gary.

[ankur2061]

Gary,

 

I am presuming that the vent configuration is an open gooseneck pipe type. Correct me if I am wrong. Normally tanks holding flammable liquids have dual vents. One is a normal vent sized based on normal inbreathing / outbreathing rates (refer API STD 2000) and the other is an emergency vent (such as a relief valve / rupture disk / blow-off hatch) for external fire scenario. This is standard industry practice in chemical process plants.

 

Am I presuming correctly that you want to have a single vent which will cater to the normal venting and emergency venting?If yes, please confirm.

 

I am detailing a simple method to size a gooseneck pipe type vent. Here goes:

 

Minor loss in head or pressure takes the form of the equation:

h = K*v² / 2g

where:

h = head loss in meters

K = head loss coefficient (consider 1.5 for your case of gooseneck)

v = velocity in m/s from the goose neck (= Flow rate in m³/h (Q) / C/S area of the goose neck pipe in m² (A)

g =acceleration due to gravity, 9.81 m/s²

Once you have calculated the head loss in meters you can calculate it in terms of pressure drop using the equation"

P = rho*g*h

where:

P = pressure drop in Pa (N/m²) (1 Pa = 0.01 mbar)

rho = density of ethanol vapor / air mixture @operating temperature, kg/m³

g, h are described above

Once you have calculated the pressure drop from your vent, you need to ensure that this value is below the design pressure of the tank. Suppose if your design pressure of the tank is say "X" mbarg then you need to ensure that the pressure drop is well below this value, say "0.5X" mbarg or less. If your pressure drop through gooseneck is close to the tank design pressure or exceeds it then you have to select a bigger gooseneck size and re-calculate the pressure drop. This exercise may have to be repeated many times till you are satisfied that the pressure drop from the gooseneck is below the design pressure and there is enough margin between the pressure drop from the vent and the tank design pressure so as to ensure a safe tank operation.

 

I have used Metric units in the above formulas. You would require to use appropriate conversions for calculation based on the above.

 

Hope this helps.

 

Regards,

Ankur.

Edited by ankur2061, 03 November 2015 - 01:01 PM.

[gmilligan11]

Ankur,

 

Thank you for your response. We currently have a few different venting arrangements across our sites. The majority are of the open gooseneck type as you described above, however, we also have some tanks which are internally stored and therefore these are vented to atmosphere with end of line flame arrestors fitted (these can be directly through the roof or can be directed through a wall using 90° bends depending on location and engineering constraints).

 

I assume the calculations for this is fairly similar however the length of the vent and also the amount of 90° bends / fittings etc will effect the pressure drop/head loss. Again, assume a vent size and make sure the calculated pressure drop is considerably lower than the design pressure?

 

Currently all tanks have only one vent fitted and not two as is common throughout chemical process plants. I am therefore completing the calculations based on this to see if at the time of install these were designed correctly for both. If this turns out not to be the case and the vents are too small, we will make recommendations to either install new vents to insure we have sufficient venting, resize the existing vent or utilize mandoors where possible.

 

Thank you again, its appreciated! (I'm a young chemical engineer and currently the only one within my company so I don't have anyone senior to ask these kind of questions too) 

 

Regards,

Gary.

Edited by gmilligan11, 04 November 2015 - 04:17 AM.