Ditches and Culverts

[Findlen]

Hi. It is my understanding that for a typical plant layout with roads, roadside ditches, and culverts crossing roadway intersections and accessways, one should model the culverts as ponds, and should not model the ditches as reaches, but rather connect each culvert (pond) to the next downstream culvert (pond). However, it seems that doing this will not allow one to see the velocity in the ditches? I looked thru the posts and didn't see much discussion on this. Any insight is appreciated. I've attached a 2 page file, page 1 showing the preliminary site plan, page 2 showing my initial thought on how to model.

https://files.engineering.com/getfi...-5168-489b-9889-7343b3a4d239&file=concept.pdf

I'm evaluating the software to see if we will purchase it. Please let me know if you think this software is applicable. I was thinking it would help size ditches, size culverts, size ponds. Thanks!

As mentioned, page 2 shows my first thought (catchments and ponds), however, I'm thinking I need a more complicated model with reaches to see what's happening in the ditches. Am I wrong? Do I model the pond storage areas based on the ditch cross sections, and can I then see the water levels in the ditches by looking at the pond diagrams? Is there a way to know the velocities and water elevations in the ditches without reaches? Sorry for so many questions all at once, but any thoughts may be helpful.

 

[psmart]

As you stated, we generically recommend that you route pond-to-pond, partly to ensure that the model can handle tailwater effects for each culvert. For details see

http://www.hydrocad.net/pipes.htm

If there were NO tailwater effects, and each ditch were free-flowing according to Manning's equation, then you could use a separate reach to model the ditch and determine the flow depth, velocity, etc. as discussed at

http://www.hydrocad.net/reach.htm

But in most situations the culverts tend to be the control points, causing water to "back up" at the inlet and inundate some or all of the ditch. In this case the channel/ditch acts more like pond storage (with a culvert outlet) rather than normal Manning's flow. Hence the general recommendation to model pond-to-pond, without using an intervening reach.

If the reach volume is significant, it can be included as part of the pond storage. Otherwise you can use the zero-storage option. As for the question of how you define the ditch storage, you could use the custom storage option and enter contour elevations in the ditch. Or perhaps the pipe storage option (yes, for the ditch) which allows you to set the slope directly. But this will only work for a square ditch, being calculated as a box pipe.

To determine flow velocity, you can use the reported velocity for each culvert. Determining the actual velocity in the ditch is much complex when you no longer have normal Manning's flow. In fact, the velocity will vary along the length of the ditch, probably being highest at the culvert outlet, and lowest where there is ponding at the entrance to the next pipe.




Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[Findlen]

Peter, thank you for answering so quickly.
I did think one option was to use Reaches and slope everything enough to avoid tail water, but we want to keep this as flat as possible (hence tail water is likely).
Is my sheet 2 essentially correct (modeling culvert 1, culvert 2 & area 3 pump directly to the Southern Pond)? (In other words, does the sheet 2 model sketch look like it would make an accurate model of the sheet 1 physical sketch?) (Note North is up on Sheet 1 sketch, and what I call "Catchment East of Road" should be called "Catchment West of Road")
How would you size the ditch feeding the Southern pond without a Reach (flow from culvert 1, culvert 2, area 3 pump & catchment West of road all feed this ditch, but in the model they would each go directly to the pond)? There will be a lot of flow in this ditch, and I imagine it would need to be quite wide.

 

[psmart]

Yes, your sketch looks reasonable.

Any ditch overflow will probably occur at the pipe inlet(s), so I would evaluate the ditch based on the peak WSE calculated by that pond/culvert.

You may want to include the upper ditches as flow segments within the adjacent subcatchment. As a side-benefit this will cause the ditch capacity to be reported, which you can compare to the peak runoff at that point.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[Findlen]

Thank you Peter. The model seems to work correctly, and customer service is outstanding! I will be recommending we purchase this software.

 

[Findlen]

Peter, I was hoping you could give me a little further guidance. The preliminary sketch in the link below shows basically how I'm thinking to model a new 75 Acre Plant Site. I'm thinking to have a "pond" for each catch basin and culvert (with ditch volumes modeled as pond storage), and a "subcatchment" for runoff areas to each catch basin and culvert. This results in about 82 nodes. I've read tips on your site, and understand that not everything necessarily needs to be included in the model, and that it is best to keep the model as simple as possible; however, I'm also thinking that each element needs to be included in the model in order account for tailwater effects. I've looked online and have seen models of similar sized sites (75 acreas more or less) with just a few nodes used (but I don't know if there were any culverts, as they only have the model without a physical sketch). I'm hoping you will comment if using about 82 nodes to model this site makes sense to you, or if the model should be much simpler. Any advice is appreciated. Thanks in advance.

[URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1687438753/tips/Prelim_Drainage_Sketch_lfvx17.pdf[/url]

 

[psmart]

Your approach sounds reasonable. I would setup a portion of the model, examine the results, and modify as required. Once you refine your modeling approach you can setup more of the project.

Do have a look at our recommendations for modeling storm sewers at

http://www.hydrocad.net/sewers.htm

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[Findlen]

It appears the ditches should be included as flow segments in the subcatchments, and also as "pond" storage at the culverts. Is that correct (I don't think this is "double dipping")?

 

[psmart]

A channel can appear as a Tc segment, as a separate reach, or as storage for a "pond" (such as water detained at a culvert inlet.) Since the pond storage will introduce a routing time lag, I would avoid inclusion in the Tc, which will tend to double-count the effects of the channel.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[Findlen]

Peter, just to verify: If I have several culverts, modeled as Ponds, all feeding an actual pond, it is not possible to model reverse flow from the actual pond to all of the culverts feeding the pond (rather, it is only possible to model reverse flow to 2 of them, 1 as secondary, and 1 as tertiary) - is that correct?

 

[psmart]

Since this is a completely different question, please start a new thread with the appropriate subject line.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[Findlen]

Hi Peter. I'm curious - do you plan to release a version anytime soon that would have Reaches that are Tailwater Aware?