An inverted siphon is designed just like any other stormwater pipe between structures (manning equation for pipe flowing full with adjustments for minor losses such as entrances/exits and bends) with the following differences due to operational/maintenance issues:
1) Use a inlet structure with a minimum 24" deep sump (but no less than the 75% of the pipe diameter) as a sediment trap. Clean this trap regularly - twice annually depending on the sediment load You do not want this trap to fill with more than 50% silt/debris. Flush/clean the invert siphon regularly- twice annually depending on the sediment load.
2) Use an outlet structure with a a minimum 24" deep sump (but no less than the 75% of the pipe diameter) as a sediment trap. Slope the pipe to this sump at the minimum slope for a 2.5 fps flowing full velocity. Clean this trap regularly - twice annually depending on the sediment load You do not want this trap to fill with more than 50% silt/debris
3) If the pipe is less than 24" diameter, consider dual parallel pipes if it is critical that the pipe function at 50% capacity or greater. (Critical being flooding of habitable structures, excessive roadway flooding, etc.)
4) Use a structure with a minimum chamber flow area of 200% of the pipe flow area or include the chamber lengths in the head loss calculation and consider deepening the sumps. In any case, you should include the chambers in the head loss if their aggregate length is 10% of the pipe length. Also use rounded entrances/exits for the pipes to minimize extrance/ exit losses. Do not forget to include the 90 degree bends (structure throat to pipe) in all cases as a minor head loss.
5) You may consider to use the high range of Manning N factors or 20% of excess of normal design N to account for poor maintenance/pipe siltation.
6) As usual, remember to check the grating capacities with a 3"-6" head, depending on your jurisdictional requirements or DOT reccomendations. Also remember, that the total head loss between the inlet grating and outlet grating plus the 3"-6" head at the outlet grate should be reflected in the difference in the elevations of the inlet and outlet gratings. Remember to specify the proper grating type for the appplication. Also consider a grating with 2 non traversable slots as an inlet (and outlet if backflow is possible.) In other words, an inlet with 4" slots along the short walls, just below the grating. This would help prevent a person being stuck to the grating.
7) Check with your local DOT drainage engineer and your local city/county engineer for their reccomendations. Even if this applicationis not under their jurisdiuction, persue the "what if" scenario.
Inverted siphons are a dirty practice to be avoided at all reasonable costs (but they are preferable in my opinion to stormwater pumps). Backflow through an inverted siphon is an even dirtier practice. These items suffer badly from poor maintaince because they can not be self-cleaning. Remember to verify that there are no reasonable alternatives, and factor in the additional cost of the maintenance and failure. Pipes can remain in the ground for 50 years plus.
Best of Luck... Clifford H Laubstein
FL Registered PE 58662
I just recieved my "Design & Construction of Urban Stormwater Management Systems" (ASCE Manual of engineering practice #77, WEF manual of practice Fd-20)from
1) Use sumps on the legs, vertical legs are more difficult than sloping legs to maintain. some agencies limit the legs to 15% max slope.
2) Where redundancy is required for maintaince purposes, one additional equal capacity barrel is sufficient.
3) flow velocity of 3 fps for the 5 year return interval design flow. If abraisive suspended materials, keep flow to under 10 fps.
4) Upstream surcharging to be avoided - ie minimize head loss through siphon. Est head losses via Darcy-Weisbach and Manning equations -
Hf = (29.1 *n^2*L*V^2)/(r^{4/3}*2*g)
Hf = head loss - ft
n = Manning friction factor
L = length of conduit- ft
r = Hyd radius - ft
V = velocity - fps
g = acc of gravity - ft/s^2.
Minor losses such as bends, contraction, expasions, entrance and exit can be estimated using tables and charts from books like Brater & King (see also ch 6)
I hope that this helps....
Clifford H Laubstein
FL Registered PE 58662
