Vortex breaker for pump suction in floating skimmer

[MFJewell]

I am designing a floating skimmer that will draw off the top 1" of fluid in a pit. The skimmer has a sump that will hold ~4" of water (total volume of ~5 gallons). The suction line is connected to the bottom of the sump. The pump suction line is 2 inch schedule 40. My operating point on the pump curve with the pump I plan to use (using my estimated system curve) will be about 100 gpm. Using submergence rule, I should have ~6 feet of water to prevent air entrainment due to vortexing. I put a small vortex breaker at the suction line inlet, but have no experience with actual design of a vortex breaker, so I have concerns that I may still get air in the suction. I attached some images to show my current layout and design. CFD isn't an option to confirm the design. I've read some older threads on here and some articles online and think I have some basic direction I want to go. That said, anyone have any tips on avoiding air entrainment or vortex breaker design?

Edit: also want to add that the pit is open atmospheric and the fluid is water typically 70-95 degreeF.

 

[bimr]

For the system that you have decribed, you would have to use a pd pump, either a perstalic or ADD.

This company has another type of skimmer that may work for you.

Oil Skimmer

 

[MFJewell]

Already talked to Skimoil. They don't have anything that fits my application that I would not need to modify to make it work.

 

[bimr]

Then use a pd pump. The pd pump will handle the air that will enter with just the 4-Inch water depth. Don't think any vortext breaker will work with 4-Inches of water.

 

[MFJewell]

Not sure what a PD pump will get me? I plan to use a self priming centrifugal, but whatever pump type I use, to get the flow and head I need, I end up with high suction line velocity. I could go with much larger line size but there are cost and weight penalties to that. The velocity is what causes the potential for vortexing. I believe I can avoid the vortex with a breaker, but have never used one in any application.

 

[LittleInch]

Are you sure on some of this data?

100gpm to me is 6.3 l/sec. In a 2" pipe that equates to 0.35m/sec?

So where on earth did 6 feet of submergence come from?? 6 inches maybe, but 6 feet sounds way too big for that velocity.

Having said that you are flowing too fast with only 4" of submergence of the incoming fluid.

You need to start with something like an 8" inlet reducing down to a 2" or multiple (say 4) 2" outlets gradually coupling up to a single pipe or header which is below the floating skimmer.

You're into open channel flow here and it's more like a drain pipe design. At 4" submergence I think you're looking at entry velocity into your pipe of 0.1m/sec or less to avoid air entrainment, which is your issue rather than vortexing.

You might find some useful information in this thread

http://www.eng-tips.com/viewthread.cfm?qid=409750

or other search terms which might prove useful.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MFJewell]

Yes, I am sure of the data.
100 gpm => 0.222 ft^3/sec
Area = Pi*(2.067/2)^2 = 3.3556 in^2 = .023303 ft^2
V = Q/A = .222/.023303 = 9.536 ft/sec.

For general submergence recommended depth of suction inlet is ~ 6 foot (see chart). The same chart and equations (which give the same velocity as I calculated) are what seem to pop up over and over again when researching this topic.

I also have an alternate design where I use a 6x2 reducer on the bottom of the skimmer. It adds 5.5 inches of height to the skimmer, so it limits my low level ability. I'm also not sure it would stop the vortexing, it may just move it a little further down the pipe.

 

[LittleInch]

My error - 3.2m/sec. 6 feet sounds about right at that velocity

That is way too fast for what you're trying to do - you simply need a much bigger pipe. You're really looking at a pipe inlet area equivalent to a 102 or 12" pipe, either one pipe or a series of pipes leading into a manifold which is at least 6" diam.

When you do that chart or calculation in reverse for a height of 4" or whatever is the min, what diameter do you get?

You will simply not be getting that amount of water down a 2" pipe with only 4" of freeboard.

This is not alternative facts, but the physical reality of life on this planet.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MFJewell]

Using the chart, velocity needs to be right around 1 ft/sec for 4.5" of water depth. I can obtain this by using a hat style vortex breaker (see below). If I size it to be 4D and locate it D/2 above the suction inlet, the annulus area will be 24.632 in2 (.171 ft2). 0.2222/0.171 => 1.3 ft/sec. I can adjust the dimensions of the hat to get the velocity down more. I believe this has the same effect as putting multiple connections and manifolding them together.

 

[LittleInch]

That's starting to look better, but my feeling remains that the 2" line at that point remains too small.

I don't know how to work it out exactly, but the acceleration of the liquid from 1.3 ft/sec to 10 ft/sec in a distance of 3 inches may require such a low pressure at the inlet that you'll either get a vacuum developing or create such a difference that the air will simply be sucked in direct even without a vortex.

From what I've found from a search on this site and elsewhere is a recommendation that the annular inlet speed into the hat is 0.5 ft/sec and no more than 1 ft/sec. Note that in general inlet lines to pumps are recommended to be around 3-5 ft/sec, not 10, therefore a lot of the guides need to be treated with caution for what you're trying to do. Why can't you make it a 3" pipe??

At 0.5 ft/sec, your diameter comes to 20" or 10D? , 1 ft/sec is 10" so maybe 16" is a bit more like it, especially if the top of the inlet to water level is actually only 3.5".

The issue as I see it given that there is not much about for this sort of thing is how much of a factor of safety do you want / need. If this thing starts gulping and entraining air, what is the impact.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MFJewell]

I could go larger, but it adds weight and cost. Even at 3 inch, I am still prone to vortexing in this application. I may look at some additional configuration with in the sump to slow velocity and/or make the transition better. Possibly replace the first elbow suction elbow with a mini sump (maybe 4x4 or 4x6 cross section) to give additional space to accelerate the fluid.

 

[LittleInch]

Sounds like a better design.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bimr]

Regarding: Not sure what a PD pump will get me?

A pd pump will pump air while a self-priming centrifugal pump will not. If you are trying to pump from a skimmer with only 4-Inches of water level, the centrifugal will quickly lose prime.

 

[MFJewell]

I briefly looked at AODD pumps when I was beginning this exercise, but then it adds air supply considerations (and cost) to the project. I also didn't find any AODD pump curves that seemed to fit my application very well. Maybe if I go larger, but that adds more cost. Possibly another type of PD pump (gear, recip??) might work, but I didn't explore any of those options.

 

[bimr]

A gear pump will also work well if the application is for solids free liquids.

 

[Artisi]

what are you talking about 'submergence rule", as the inlet is below the water level by 4" then the submergence will be 4" so your problem is air entrainment - absolutely nothing to do with 6ft of submergence and vortex.

Turn the skimmer over with the inlet underneath - you then have a chance of not entraining air.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[ione]

HI 9.8 and specifically para 9.8.7.2 reports the following relationship:

S/D = 1.0 +2.3Fd

where:

S = submergence
D= Inlet O.D.
Fd = V/(gD)^0.5 Froude number
g = Acceleration of gravity
V = velocity at inlet bell face

With your numbers S is slighly more than 8"

 

[MFJewell]

Regarding "what are you talking about 'submergence rule'? There are some empirical rules out there that use flow rate and/or velocity to recommend a critical submergence depth. See the chart I posted above. There are also several formulas that have been published over the years:

Hydraulic Institute: D·(1 + 2.3·Fr)
Pumping System Manual: (V2/2g) + 0.5
Prosser: 1.5 * D
Paterson & Noble: D·(1.5 + 2.5·Fr)
Hecker: D·(1 + 2.3Fr)
Knauss: D·(0.5 + 2·Fr)
Flyght: 1.7·Fr
Werth & Frizzell: D·(2.1 + 1.33·Fr0.67)

 

[MFJewell]

Regarding: "With your numbers S is slighly more than 8"

I get:

Fr = 9.56(ft/sec)/(32.2(ft/sec2)*.17225(ft))^0.5 = 4.06

S = D*(1.0+2.3Fr) = 2.067*(1+2.3(4.06)) = 21.37

Using the formulas above, I get a range of values from 3.1005 inch to 24.08 inch. Seems like there is no correct answer.

 

[LittleInch]

"Seems like there is no correct answer." - Apart from 2 inch [edit] isn't it...

As I said before, you need to balance the data from rules and formulas with some level of safety factor which only you can determine.

For the pump you seem to have (self priming centrifugal by the look of it) it may be fairly robust to air entrainment / gulping of air, but know nothing about your downstream system.

what about a secondary sump within your initial sump and a horizontal plate at the end then draw off the liquid at both ends of the second sump horizontally in a 3" pipe and join both up before going up.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.