7 replies to this topic

[nathans]

Howdy everyone / anyone!

I'm looking for some help from someone more knowledgeable than me. This past weekend my
plant finished installing a scrubber system to scrub the exhaust from a rotary salt dryer.
The scrubber tank would not drain properly, building up past the cone of the tank into the
air duct. When we would turn off the supply, the water would eventually start to flow. In
the interest of time, they eventually plumbed the scrubber into the old scrubber gravity
drain with success. This same size line empties to a different (not preferred) place and
does not have a below ground portion as the new one does (you'll see in the sketch - it
basically forms a trap). I studied Art's gravity flow excel file a few times and it is
helpful, but I am not confident in how to apply it to my problem.

It was suggested that the line was airlocked and the behavior seems to support this, so a
vent was installed, but it looks to me like the wrong place (it was a convenient place).
Based on my friction loss tables, there is more than adequate vertical head to move the
required flow rate of water. I think that what is happening is that water drains into the
open top tank, leaving water standing in the vertical line on that end and water flowing
until water in the upstream vertical leg reaches that same equilibrium elevation. This
leaves air in the line from that point all the way to the scrubber. I suspect that when
water is added it flows until is passes the first vent (allowing that amount of air to
escape), but then due to a small (6") step up with two elbows, the air from that point
forward cannot escape, airlocking the system until some air manages to burp out of the
system somewhere (they were manually cracking a joint below to relieve it).

I believe that the elbow where the pipe turns vertical should be replaced with a tee and a
vent stack and that should allow the system to flow. Can someone familiar with these sort
of systems give me some advice? Alternately it has been suggested we add a pump. Some
other advice that has been given doesn't make sense to me - it was suggested this could be
solved by adding a dip pipe at the discharge end, ending at least 16" below the surface of
the water. I am not certain that the engineer that gave that advice had a full description
of the system. I will send them the same sketch as it attached. Any help would be
appreciated before we start modifying.

Thanks!

Nathan

Attached Files

•  Scrubber Gravity Drain.xlsx   14.58KB   43 downloads

[sgkim]

Just remove the open vent line and attach a normally-closed vent valve. Then the liquid drain head will become 12'(=22'-10').  If 12' liquid head is not still sufficient compared with the suction pressure of the 'fan',  then make the drain pipe dip below the surface of the open tank and maintain the liquid level below the point required.    

 

The open vent, presently, makes to provide only 2' of liquid head.  Even if you extend the drain tip below the liquid level of open tank, the open bent will always drain 35' horizontal PVC line 10' below the horizontal line.  So the present open vent shall be removed.

 

Stefano

 

 

P.S.  Sorry... I had replied without fully understanding of the system-I did not noticed, and neglected the " 6" step-up elbows" then. If you cannot remove two 6" step-up elbows, then Mr. katmar's comment would be the best choice to solve the 'air-lock' problem on the downstream side of the elbows.       ~Stefano Oct 31, 2016  

Edited by sgkim, 31 October 2016 - 05:09 AM.

[nathans]

Thank you for the response, Stephano.  The suction from the fan is minimal (inches of water gauge).  The scrubber was initially tested with the fan off and the door open (vented) without the vent stack and it wouldn't (immediately) drain.  They added the vent stack and again it wouldn't immediately drain.  Does this change your opinion at all?

 

Nathan

[Pilesar]

If the scrubber liquid has absorbed salt, is it possible that there is a denser liquid in the low point of the loop than in the scrubber downpipe? I could imagine this happening especially on startup of the scrubber. Hot liquid from the scrubber and cold liquid in the line would be a density difference but there may also be composition differences. The higher head of the less dense liquid entering the transfer line may not have enough force to push the dense liquid out. Once the flow is started, the density of the liquid would be more uniform and the system should work more smoothly.

[nathans]

Good question.  I think the impact of this is minimal.  The gravity drain was not working before the scrubber was put in full operation - fan not on, no salt absorbed, no temperature difference - just trying to run the water through it...  We have hooked to the old drain line and it is draining well while we are in operation - fan on, some salt/solids dissolved, slight temperature increase in water.  However the old drain line exits a little lower (more vertical head) and does not go underground, thus it does not have this "trap."

[Bobby Strain]

You most likely have an obstruction in the drain line.

 

Bobby

[katmar]

I agree with your conclusion that there is an air lock in the 20' vertical downflow section, but there are aggravating factors that have not been mentioned yet. If the liquid level in the conical section is not controlled and you want the liquid to drain away as fast as possible then air will continuously be entrained with the draining liquid. Your piping is on the large side  of normal (i.e. lower velocities) and very little air will be flushed out of the underground section until there is a very high level in the scrubber.

I also agree that a tee installed at the top of the 20' vertical section with an added vent will solve the problem, but there is a pipe sizing problem as well. At 30 gpm in a 3" pipe your Froude Number (Fr) is about 0.45. This is too low to allow easy flushing of the air, and too high to allow self venting in vertical downflow. The 20' vertical section should be increased in size to at least 4" diameter to allow the air to be separated from the liquid and discharged through the proposed vent. You do not have to increase the diameter of the whole 20' section but you should increase it from the proposed new tee (including the tee itself) down for about 13' so that the lowest 3' of larger diameter piping would always be flooded (because of the 10' vertical section further downstream).

I would leave the vent you have already installed in place, because if air is continuously being entrained from the bottom of the scrubber then you want maximum capacity for venting.

Once the vent is installed the pipe has to be analyzed in two sections because the new vent will fix the pressure to atmospheric where it joins the pipe. The existing vent will not have any impact because the 6" step will force the connection between the vent and the pipe to be continuously flooded.  The section from the scrubber to the proposed vent will be 72' of pipe and at 30 gpm will cause a backup into the scrubber outlet of only a few inches.

The section from the propsed tee to the end will need only a foot or two of head.  This means that the level in the 20' vertical downflow section will find its equilibrium about 8 or 9 ft down from the tee.  This makes self venting flow in this section a necessity.

The dip pipe into the open top tank would serve no useful purpose and would probably be detrimental because any entrained air would have to be forced down this vertical downflow section as well.

 

BTW, congratulations on a very well formulated problem statement and a clear drawing.  Asking the question in the right way is always a large step towards getting the correct answer.

[nathans]

Thank you for a great answer, Katmar!  Very thorough and well explained.  I thought the pipe was oversized, but was uncertain what the net result of that was and I knew that as the water fell vertically and accelerated, there would be a space issue, but I didn't know if it would really be critical.  I think you're sending me down the right path and I might have another question or two for you after I study this a little further and see about implementing it.  Thank you, Sir!

 

-Nathan