Jacketed Tank Puzzle

[gne77]

I came across an interesting problem with a jacketed tank recently hoping for some perspective on this - the company who built them vanished long ago.

The tank is 2000 gallons and presumably was a dairy bulk tank. The jacket has (3) 2" pipe inlets along the side of the tank (all one above the other), and one on the tank bottom, opposite side of the rest. There is (1) 2" outlet at the bottom of the tank (jacket) and (1) 1" outlet (again in the jacket). A common manifold feeds all (4) of the jacket supply locations it's inlet is slightly above the top of the tank. Another manifold drains the (2) outlets.

The owner likes to connect 100 PSI steam to the jacket (I questioned the safety of this and was given a shrug) from up top at the inlet manifold. Now they want to connect cooling water to the jacket instead of steam, I suggested connecting from the 'drain' manifold to ensure the entire jacket is filled with cooling water. We made a test run and recorded the flow rate. Just for kicks we also connected from the top of the 'supply' manifold and recorded the flow rate. The flow rate when feeding from the bottom of the tank was twice what we saw when feeding from the top. The same pump was used obviously.

So it seems that the water is finding a shortcut when feeding from the bottom, and despite the fact that we have an additional 10 feet of static head to exit out the top the head pressure is less than half.

Based on this description does anyone have a guess of what is going on inside the tank jacket? I have to assume that this tank has at most 2 zones, the owner thinks only 1 zone despite all of the connections. The only ideas I can come up with all lead to the tank having a lot of stagnant flow areas in the jacket and are probably wrong. The tank is also insulated so feeling for cold areas from the outside is a no-go. Nobody has been lowered into the tank for lack of having proper safety in place.

 

[JohnBreen]

Sounds like a "casual" owner who does not understand the difference between a tank and a pressure vessel. If the outer jacket and inner wall are not "U" stamped you may have an atmospheric tank. If a vessel is pressurized to a pressure of 15 psig or greater (even if it is cold water) and it is not a "Code" pressure vessel it is likely that the owner is violating the local law - and that doesn't even address the liability of injuring someone.

In what State or Province will this thing be used? Go to the National Board Internet site and see what the "local" law is. Check with the chief inspector of the locality (the NB site has all the phone numbers). Also check out the design of double containment tanks (Google is sooooooo handy). You will have to put pressure (and vacuum) relief devices on that tank to protect it against changes in ambient temperature and other vagaries. Has anybody done any corrosion checking to find out how much of each wall is left after all these years?


John

 

[rmw]

When you pump up from the bottom the liquid level fills the entire jacket until it exits from the top. When you pump in from the top, there is air in the jacket and the liquid has to fight its way around the air pockets to get to the drain. The difference in resistance to flow due to the presence of the air pockets. is what is giving you differing flow rates.

rmw

 

[quark]

RMW hit the nail. It is air that is creating the difference. Check whether it is full jacket or limpet coil.
I pressume the following nozzle configuration. If one nozzle in at midway then there is a split limpet and you should have one more outlet in the midway.

2" bottom nozzle - chilled water in
2" top nozzle - chilled water out
2" top nozzle - safety valve+vacuum breaker+pressure gauge
1" bottom nozzle - condensate out

 

[gne77]

Thanks all.

John-
I understand where you're coming from but the issue of pressurizing the tank is not why I posted. I have subsequently learned that they only run 20 psi steam in the jacket by throttling it down with a hand valve. The jacket would have logically been designed from some amount of pressure (at least one would imagine) I just don't know what it is.

RMW-
That makes sense I just didn't think it would amount to that much resistance.

Quark-
I'm pretty sure it's a conventional jacket. If it is a limpet they skinned it with a shell. Also the nozzles don't make sense for a limpet design. As I mentioned before all 3 "inlets" are vertically in line with each other and all are tied to a common supply manifold that appears to be part of the original tank design. It could be a split coil though with the upper two side ports servicing one zone and the lower side port and bottom port servicing another.