Routing through bridges/culverts 1

[Crystalct1]

When establishing a relationship between outflow and elevation, it is easy to see which equations to use for culverts or bridges flowing full or overtopping the roadway. What is the best way to calculate outflow from a bridge in low flow conditions? Would that be considered weir flow? or open channel flow?

Weir flow seems appropriate since it accounts for the constriction of flow at the opening, but H (height above "weir") wouldn't make sense - unless you use an average height above stream bottom? Would that make sense?

Open channel flow would be complicated because you'd have to take wetted perimeter and area at each elevation. Although, HEC RAS could do that - I already have a running model.

What has everyone else done?

 

[RWF7437]

Use Hec_RAS. If possible, assume normal depth at a cross section downstream from the last bridge or culvert, unless you know the water surface elevation at some point downstream.

 

[LHA]

You must analyze as a culvert to account for headwater and tailwater, unless it is blatently obvious that neither exists. During very low flow, that could be valid; but you should demonstrate why and under which flows it is valid, likewise, at which flow it becomes invalid.

If HW and TW have no effect, weir analysis might be logical only if you have a very wide, flat bottom, and very low head...like a box culvert during base flow conditions. Otherwise, it is certainly a channel, not a weir.

You state "I already have a running model". If I were reviewing, and your modelling software doesn't have culvert and step analyses built in, I would require you to use HEC-RAS, or another one that does.

Engineering is the practice of the art of science - Steve

 

[Crystalct1]

What are you guys talking about???? You can't use HEC RAS to obtain outflow from a structure. That's the job of programs like HMS and HEC 1. I'm talking about outflow after attenuation - you know - taking into account storage upstream of the structure. You have to use the topography of the basin to obtain storage at different stage values. Each stage has a corresponding outflow rate. Outflow - not inflow. HEC RAS accounts for energy changes and backwater effects-changes in WSE with flow held at the value which the user inputs. What I need from HMS is the flow value after the water has routed through the structure. Remember? Change in storage = Inflow minus Outflow. I have inflow and storage/elevation relationship. I need elevation/discharge relationship for the "outlet" (bridge exit) so I can obtain Outflow.

By the way, I think the procedure is to apply the outflow obtained from HMS at all the cross sections downstream of the bridge in HEC RAS to properly model the stream. Flow changes at a bridge/culvert/dam, etc.

For bridge flowing full, orifice equation is used; For overtopping roadway, orifice and weir equations are used. What equation is used for low flow through structure - no pressure flow? Has anyone done this?

 

[blueoak]

You can use HEC-Ras to model outflow from a structure. The only item is that you do have to have some flow throughout your stream. Such as when modeling a breach you normally have some minor base flow to model a "sunny day" failure. Typically you don't just because setting up the storage versus stage items is a pain in RAS compared to HMS. However, HMS is limited as a hydraulic model. I saw a beautiful RAS model of unsteady filling several reservoirs that backed up to a storage pond that then released after the storm was over. You had flow going two directions.

As far as modeling a bridge, it depends on what your bridge is. Is it a culvert, piers, or is the bridge not affecting flow and you really just have open channel? You have to balance energies. RAS for bridge design includes a lot of trial and error. You have to set your ineffective flow areas upstream and downstream, both widths and heights, if you want to model the whole spectrum of expected flows. This is why sometimes you make two models so you don't have to fine tune a single model as much.

If you want to model the bridge outside of RAS, a lateral constriction is similar to a flume. Instead of contracting the bottom you force critical flow by a side contraction.

As to your last question, I would again recommend you review the manual. You have to trial and error to determine which method is critical.

 

[Crystalct1]

Ok. I appreciate your answer. Perhaps I misspoke about RAS. It is capable of unsteady modeling, flow in two directions and all that stuff. We've always just used peak flows and ran steady conditions. But I am using HMS or HEC 1 for flood routing. Think of the section upstream of bridge as a giant reservoir with the bridge as the reservoir outlet. This is how the project client expects it to be done. It's called Modified Puls method for reservoir routing. What you (blueoak) have been talking about involves HEC RAS modeling and I am sure that what I need to do is to be done with HEC 1 or HEC-HMS (per company vice president with 20 years experience in storm/flood engineering). Perhaps this is a question for someone experienced in HMS or HEC1 or a hydrologic engineer in particular.

 

[cvg]

either method could be used to generate your stage-discharge data for the routing to be done in HEC1.

weir flow with avg flow depth might be appropriate if you have a drop / grade control structure at the bridge such that flow is going through critical depth / supercritical.

However, for most bridges you probably have subcritical conditions. Open channel flow analysis using hec-ras, HEC2 or other program is probably more accurate.

 

[blueoak]

As cvg says you can output any stage discharge data you want from HEC-RAS. If you have the model as you say, just output the information and use it for your stage disharge curve in HMS.

As for determining your bridge model you may wish to review pages 5-9 and forward of the Hydraulic Reference Manual. This gives a pretty good overview of what goes on.

I am familiar with Modified Puls and HEC-HMS. Most attenuation is in the floodplain and using a detailed HEC-RAS model will more accurately model this. You normally apply modified puls in HMS to help account for this, it depends on where you want your accuracy. I know a lot of people shy away from running unsteady in RAS because you need a stable model and have to check your stage discharge curves to make sure. But if you are trying to look at floodplain attenuation it is the better tool.

Saying all of that, I did have a reservoir recently that I ended up going to a 2D model to make sure that the reservoir was adequately modeled along with attenuation.

 

[Crystalct1]

It was my understanding that the stage discharge curve in RAS would not be useful for a bridge routing since the flow you have imposed going into the bridge is not the same flow coming out. The point is to find out what that outflow value is so that you can accurately model the storm in RAS. The new flow is to be applied at the cross-section just before (some say just after) the bridge. HEC RAS (unless you use unsteady - which I have been told is not the way to go in this case) can only use the flow that the user puts in. HMS or HEC1 are used to obtain the value of flow after attenuation. People, this is not a river reach routing. I understand HEC RAS can do that pretty well. This is a bridge routing (check the thread title) When it comes to attenuation due to changes in storage from backwater effects - such as a river flowing through a very constricted opening - HMS is needed. The flow in is not equal to the flow out.
Original question: After determining storage versus elevation, how can you establish the relationship between elevation and outlow (not inflow - which we already know) for the low flow stages?

 

[RWF7437]

You have three choices:

Either treat the bridge as a outlet control structure and do an unsteady flow analysis in HEC-RAS using an inflow hydrograph at the upstream end of the system. This is theoretically "correct" but unsteady flow examples in HEC-RAS are notoriously difficult.

Or, treat the inflows as a series of steady flows and review each resulting profile to see if it seems reasonable. This is time consuming and tedious but easier understand and check for errors.

Or, Model the system in say, HydroCad, treating the bridge as a weir with a "lid" on it.

This may be the easiest to setup and do a couple of trial runs to see if results seem reasonable. Post your question in the HydroCad forum too. Peter Smart can direct you to and axample or two which can help a lot in setting up the model.


Good luck

 

[blueoak]

You are wrong. RAS will perform what you need. If you choose not to use it, that is your choice.

The other option is to do it by hand, which is the most fun, but you have to know what a weir is. It is actually worse than your original post lists, because you have to run an iterative step method on multiple cross sections at multiple flow rates.

http://www.fhwa.dot.gov/engineering/hydraulics/library_listing.cfm

The weir equation is only an estimate of a very special case.

 

[Crystalct1]

You know what? Everyone has good ideas on how to do this, but I have to do it the way I was advised - HEC1 or HEC HMS. Everyone seems to agree that HMS will do it. That is what I am using. Both programs (hec1 or hms) need the same input that I am asking about - relationship between elevation and outflow from bridge. I understand that Hydrocad or HEC RAS unsteady will do this, or that it can be done by hand. There are a lot of ways to skin a cat. Unless there is some technical reason that I shouldn't use HMS to do this, I don't understand why the original question was so difficult. We know the orifice equation is used to determine outflow during a pressurized condition such as bridge flowing full; We know the weir equation determines the flow for the part overtopping a roadway. All I asked is: What equation is used to determine outflow form a bridge in low flow conditions?

 

[cvg]

you have been given the answer in previous posts - no need for attitude, just re-read the posts