2 replies to this topic

Hi. My name is Ingemar Quintero. I am working in a refinery revamp, specifically in the product storage unit for fuel oil. We are trying to define the closed drainage system for the storage tanks.

We designed 4 atmospheric storage tanks (fixed cone roof) of 433,000 Bbl (nominal capacity). Typical arrangements for a closed drain system for a storage tank consist of a drain ring connected to a main header that goes to a closed drainage drum, in this case, for heavy hydrocarbons, but, reviewing the literature, especially API 650, I haven’t find a criteria regarding the number of drain  nozzles to be specified for fixed roof tanks. In fact, I found that there isn’t a clear criteria for that… http://www.cheresour...ospheric-tanks/.

Howerver, according to others typical arrangements for tanks of similar capacity we specified 4 nozzles of 6" each (4 sumps). So, the problem is that I would like to know, what are the typical operation times for tank draining procedures in refinery plants? Of course, I know this depends on the capacity of the tank, the service factor of the plant (maintenance) and the properties of the fluid. According to the proposed settlement (6 pipes of 6 "and a common header of 12"), our calculations show a drain time of almost three days for a unpumpable volume of 33,320 BBl of a fuel oil of 570 cP @ 130°F.

I also wanted to know what are the advantages or disadvantages of using water drawoff sumps or bottom drain sumps. Take into account the tank diameter (220 ft) and the capacity. With water drawoff sumps the remaining volume is large enough to drain with vacuum trucks.

Thanks.

Regards.

Sorry. Are 4 pipes of 6" and a main header of 12" to the closed drain drum. 

Thanks.

[Art Montemayor]

Ingemar:

 

Without any scope of work, accurate layout, and piping sketches as well as complete basic data it is very difficult to estimate whether your nozzle sizes are appropriate for the tank(s) you describe.  The primary objective in any project for an engineer such as yourself is to follow the precise and detailed instructions of your client.  And if your client lacks the ability to instruct you, then it is your moral and professional obligation to bring out those criteria that can negatively affect your client during and after the project’s completion.  By common sense we should all be aware that any tank installation of the size that you describe is a formidable one and therefore of prime importance in ensuring that it will operate to the client’s needs and requirements as well as meet all safety standards and requirements – whether noted by the client or not.

 

One basic criteria related to storage tanks is the fundamental importance of establishing a means of positive fluid containment within the vessel at all times – especially when it comes to hydrocarbon fluids.  Based on this criteria then, we can surmise that the less the number of exposed, below liquid level nozzles that the tank has, the lower the risk in having a future spill, leakage, or nozzle rupture within the diked area.  In fact, the “ideal” storage tank (in my opinion) has NO external below-liquid level nozzles.  The problem with this basis of design though is that it requires a submerged, internal pump.  However, this is the philosophy used, for example, in LNG storage tanks.

 

The point I would make here, then, is that the less the number of pump-out or drain nozzles you have, the less the risk surrounding the tank’s operation.  The capacity of the drain-out is a subject that should be contained in the client’s scope of work.  If it isn’t, then the matter should be brought out immediately to the client because it is directly in the client’s interest to fix the capacity of drain flow rate required.  Only the client can know what is the indicated flow rate needed for day-to-day operations or for the eventual turn-around.  The engineering contractor is in no position to conjecture or guess what a “normal” drain out should be.  This could wind up being a very important item in the final design of the tank and it is best to bring it out in the open with the client in constructive meetings and discussions.

 

I would point out to you that the 130 ^oF temperature that you mention MAY NOT be the correct design parameter to use in estimating the design viscosity of the fuel oil during draining.  Always check with the client and make sure that they can maintain that temperature AS A MINIMUM.  During a turnaround or a shutdown, it is not unusual to simply not supply heating steam to the tank and allow the fuel oil to reach ambient temperature – and in Venezuela this could be 100 ^oF or perhaps 90 ^oF after a rainshower.  This could raise the viscosity and make drain-out even more difficult.