Modeling a large obstruction

One of the branches in the storm sewer I am modeling is a 4'H x 6'W box
culvert, with a 30" pipe running straight across the width of the box.
The bottom of the pipe is located about 1' above the floor of the box.
I hope the sketch below will not confuse anyone. Direction of flow in
the box I am modeling would be "out of the screen".

Any suggestions on the best approach to modeling this obstruction? I am
concerned that simply raising the n-value will not give an accurate
picture of low flows (before WSEL reaches bottom of pipe), high flows
with orifice flow under the obstruction, or high flows with orifice flow
above and below obstruction.

Thanks,
Chris

************************************************************* top of box

modeled flow is "out of the screen"
------------------------------------------------------------------------
---- top of 30" pipe
-----> direction of pipe obstruction
------------------------------------------------------------------------
---- bottom of 30" pipe
************************************************************* floor of
box

Chris Perry, P.E.
christopher.a.perry@usace.army.mil

Chris - The following report looks at this type of situation with the
cross-pipe in a junction box, presenting experimental data and design
parameters. Not exactly the same as your situation, but you could
probably use the results with some judgement, safety factors, etc.

"Hydraulic Performance of Conflict Junctoin Boxes", S.C. Krane,
F.Romano, K. Cecil, D. Olin, A. Perez, S. Ethier, and E. Droz, Report
No. 710, Dept. of Civil and environmental Engineering, University of
South Florida, Tampa, Fl 33620. Prepared for Florida DOT & US FHA,
available from NTIS. 29 pp.

Ron Kilmartin
ronkilmartin@attbi.com

If this is a proposed design, my first recommendation would be a
redesign to reroute one or both of the structures to avoid obstruction
and maintenance issues. Perhaps the box could be transitioned through a
squash box, or the pipe could be rerouted above or below the box.
Failing that, there should be access from the ground surface upstream of
the cross pipe to remove debris when the system fails.

If this is an existing storm drain or combined sanitary/storm drain that
receives surface stormwater flows with debris, then there are safety
issues in addition to modeling issues. I am familiar with two similar
examples, both of which trapped debris that obstructed flow.

Modeling assuming a sealed pipe upstream and bridge pier with debris
loading might be an approach, but in this case there is more pier than
conveyance section.

Neil Jordan
NJordan@EXPONENT.COM

If this is a proposed design, my first recommendation would be a
redesign to reroute one or both of the structures to avoid obstruction
and maintenance issues. Perhaps the box could be transitioned through a
squash box, or the pipe could be rerouted above or below the box.
Failing that, there should be access from the ground surface upstream of
the cross pipe to remove debris when the system fails.

If this is an existing storm drain or combined sanitary/storm drain that
receives surface stormwater flows with debris, then there are safety
issues in addition to modeling issues. I am familiar with two similar
examples, both of which trapped debris that obstructed flow.

Modeling assuming a sealed pipe upstream and bridge pier with debris
loading might be an approach, but in this case there is more pier than
conveyance section.

Neil Jordan
NJordan@EXPONENT.COM

Hello Chris;

Having no access to the doc Ron referred to, and to get a general idea
of the importance and the region of influence of the obstruction effect
on the system behaviour I would have done the followings:

1- define the culvert's us node as a storage junction.

2- set the headloss k values for the joining links on the C cards just
as there's no obstruction there.

3- run the model, examine and spare the result.

4- decide on a range of headloss k values judged to be appropriate for
the losses at the storage junction defined (Entrance,Exit & Other
losses).

5- run the model for the assumed max and min increase in the k range
decided on in step 4 above.

6- observe the sensitivity of the model results obtained in steps 3 and
5 to the assumptions made on k values.

7- Now see if you're content with the outcome so that you can peacefully
come to a plausible final decision on the k values, especially
considering the step 6 above OR, you should seek a higher level of info
(e.g., what Ron's proposed).

8- let other list members benefit from your interesting case,if you may
post a summary of what you've found.

Best of Luck!
Mostafa Bozorg-Zadeh