Upsizing storm pipe to increase velocity

[jgailla]

I am reviewing a drainage assessment report in which the following statement is made.
"Larger pipes are able to self-clean due to their ability to achieve a minimum cleansing velocity of 2.5 feet per second based on the engineering principals (sic) behind storm drain design."
It seems to me that this statement is incorrect. If the pipes are assumed to flow full, it would be correct.
But let's say the hydrology indicates a flow of 2 cfs to an inlet. If my slope is limited to 0.1%, going from an 15" pipe to a 36" pipe will not raise the velocity in the pipe. I am assuming the hydraulic grade line here equals the pipe slope.
Any thoughts?

 

[msquared48]

The velocity is a function of the pipe material, wetted perimeter, hydraulic radius, and area of the pipe.

V = (1.486/n)(R^.67)S^.5

Where AR^.67 = Qn/1.49S^.5

Do the math here and answer him.

Mike McCann
MMC Engineering

 

[cvg]

all things being equal, flow in a larger pipe will be slower and assuming a pipe flows full doesn't make it so...

"HGL equals pipe slope" is also a dangerous assumption

 

[Ron]

The HGL for low flow will approximate the pipe slope.

The velocity is inversely proportional to d^2, so as the diameter increases, the velocity decreases for constant flow.

Further, if your slope is limited by the pipe geometry, then using a smaller diameter will allow you to locally increase the slope, thus increase the velocity to a greater extent than other variables.

jgailla....your report writer is wrong...but you already knew that. Hope things are going well.

 

[jgailla]

Thanks all. I appreciate the confirmation.

cvg, good point about the HGL equals pipe slope assumption. I always do calcs off the HGL as is correct for Manning's, but for this discussion it opens up a can of worms that I didn't want to get into.

Ron, things are finally going well. I had been doing carpentry for over a year now while looking for engineering work. I'm working for Darrell, so I'm pretty happy about the way things are going right now.

 

[Ron]

jgailla...glad to hear it. Haven't spoken with Darrell in a while...I'll have to call both of you. Your Dad told me about the other business venture.

 

[beej67]

But even the "full flow" assumption is wrong. If you've got a 36 inch pipe that's flowing full, and you neck it down to an 18, then it moves to pressure flow and flows even faster. For a constant Q, the V always goes up when A goes down. The author needs to be versed in basic continuity relationships.



Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[zabrab]

My initial reaction was to generally agree with the prior responses, but after thinking about it for a moment, I am not so sure. There are conditions under which upsizing a pipe would increase the velocity.

I’ll have to pull a book and bang out some numbers to verify. I’m assuming constant Q and s with varying D. I’ll try to do it at lunch.

 

[tumbleleaves]

My recollection of sizing sewage pipes (for small subdivisions) is that the reviewers would typically redline for a bigger pipe than the one I correctly "engineered". Never understood it, but they all seemed to have functioned without problems.

 

[zabrab]

I ran some calculations by hand at lunch and then realized I could use a program we have, Bentley Flowmaster, to run the numbers. Right from the start the results did not match my calculations. Long story short – don’t use the hydraulic elements graph in the Wisconsin DOT Facilities Development Manual. Even though it has a nice grid, it doesn’t match the hydraulic elements graphs in other sources.

When I posted my original response, I was thinking that if a pipe was flowing full, increasing its diameter slightly would create a condition where the flow in the partially-full larger pipe would be in the area where V/Vfull is greater than 1 and that this would offset the larger diameter, but that the effect would quickly fade and that would be that. What I didn’t consider is that Vfull for a larger-diameter pipe at a given slope is larger than Vfull in a smaller-diameter pipe. This second point seems to extend over a large range of diameters.

Starting with numbers close to those in the OP, a 15” pipe @ 0.001 has a Qfull of 2.21 ft3/s with a corresponding Vfull of 1.80 ft/s (n=0.012).

A partially-full 18” pipe @ 0.001 would convey 2.21 ft3/s at a depth of 0.85 ft with a corresponding V of 2.14 ft/s (19% increase).

24” Pipe: d= 0.73 ft, v=2.13 ft3/s (18% increase)
36” Pipe: d= 0.63 ft, v=2.05 ft3/s (14% increase)
48” Pipe: d= 0.58 ft, v=1.98 ft3/s (10% increase)
72” Pipe: d= 0.52 ft, v=1.87 ft3/s (4% increase)
96” Pipe: d= 0.48 ft, v=1.80 ft3/s

The Flowmaster results are at this link:

http://files.engineering.com/getfil...ac-4d3e-869b-e2c63c828730&file=10-25-2011.pdf

 

[Drew08]

Zabrab, what you’re not considering is that a pipe under gravity flow is not at its maximum flow rate at full flow, but closer to about 0.9D. Therefore, two possible flow depths exist for a flow rate of 2.21 ft3/s. The lower of which (1.07 feet) has a velocity of 2.96 ft/s, and all larger pipe sizes that you have shown have a decreased velocity, not increase from the 15-inch pipe.

 

[zabrab]

Drew08, I was looking at the conditions at a given flow rate, not the maximum capacity of the pipe. Point taken, I was comparing the 15" that I arbitrarily set as flowing full as that was implied in the OP, but I did not impose any type of constraint on the larger diameters. However, depth for maximum flow rate is not applicable in this instance as there is a given flow rate.

Using a flow of 2.21 ft3/s as it is close to the 2 ft3/s mentioned in the OP and conveniently is equal to the flowing full capacity of a 15" RCP @ 0.001.

There are two possible depths the 15" pipe conveys 2.21 ft3/s, d/D=1.0 and d/D=0.81 (1.01 ft). The corresponding velocities at these depths of flow are 1.80 ft/s and 2.08 ft/s.

A partially-full 18" pipe @ 0.001 would convey 2.21 ft3/s at a depth of 0.85 ft with a corresponding V of 2.14 ft/s (3% increase)


21" Pipe: d= 0.78 ft, v=2.14 ft/s (3% increase)
24" Pipe: d= 0.73 ft, v=2.13 ft/s (2% increase)
27" Pipe: d= 0.70 ft, v=2.11 ft/s (1% increase)
30" Pipe: d= 0.67 ft, v=2.09 ft/s
33" Pipe: d= 0.65 ft, v=2.07 ft/s

 

[pilesmakesmiles]

jgailla - Some obvious ones which you've probably already checked...
- Is it a circular (not ovoid/egg shaped) pipe?
- Is the gradient staying the same? They are not proposing to relay a long length of pipeline at a steeper gradient?
- Are the materials similar?

 

[jgailla]

pilesmakesmiles,
It was somewhat of a blanket statement in the draft report, so there was no reference to squash pipe or using different materials.
I not sure what they intended to say, but the statement has been removed from the report.