How do you delay dam failure?

I have a small dam I want to fail at a given time after the start of simulation, to that I put on the dam window initial level equal to breach level, and on the boundary window i put time to failure 60 hours which the peak of my flood, the dam fails almost at t=0 hours. What is wrong?

Sincerely
luis carballada

Dear Carballada, Luis :

You need to make certain that the inflow hydrograph into the upstream reservoir is steady- state, meaning that it has a constant flow value. This is the only way you can directly fail the dam at a specific time using DAMBRK.

However, if you need to fail the dam at a specified time with a dynamic infflow hydrograph into the upstream reservoir, there is a trick that you can perform. Set the "Water Surface Elevation at Time of Breach" data entry to a very high value (i.e., 20,000 ft or meters). Then, run your simulation and see at what elevation the reservoir water surface is at, at the time you wanted to fail the dam. Then, set the "Water Surface Elevation at Time of Breach" data entry equal to this value. Then, re-run the DAMBRK analysis. The dam will then fail at "approximately" the time you wanted.

If you have any additional questions or need any further information, please contact me again.

Sincerely,

Chris E. Maeder
Senior Technical Engineer
chris.maeder@bossintl.com

Dear Maeder Chris;

Thanks for your answer it works really well. I have another problem on the model cliped on when I use option 1 i get quite a lower peak discharge that when I use option 13

do you know the reason?, or it is just an stability problem and the results of option one are the good ones

luis carballada

Dear Carballada, Luis :

Note that Problem Specification Number 1 is using a different method in computing the analysis. It uses a sequential computational method, and as such it first computes the dam failure and then routes the computed breach hydrograph downstream. Therefore, it does not account for tail water submergence effects. Usually this gives you "higher" flow values and not "lower" flow values. Strange!

Problem Specification Number 13 is a more accurate simulation method, since it performs a simultaneous solution of the dam failure and the routing. Usually, you get very close results with both methods. However, when the two methods diverge, use Method 13. This has all of the "new" numerical analysis methods and corrections incorporated within it. In fact, the NWS does not recommend Number 1 anymore--although we do. There are times when Method 1 is preferred over 13, especially when you require a reservoir depletion table to be generated. None of the other methods provide that.

However, in looking over your model, I noticed that you are using a 0.2 hour time step in the Boundary Conditions. This is MUCH too large. Try a value equal to 0.02. I think that then the 2 methods will give very similar results--although not identical since they use different computational code.

If you have any additional questions or need any further information, please contact me again.

Sincerely,

Chris E. Maeder
Senior Technical Engineer
chris.maeder@bossintl.com

Dear Chris Maeder:

The file who is cliped on does not breach, I am not able to break the dam can you suggest what is the problem.

Dear Luis :

I would strongly suggest reviewing the DAMBRK Troubleshooting Guide that is available on our WWW site in the DAMBRK Download Area.

In reviewing your model, I found the following problems.

1. You set the WSEL (Water Surface Elevation) at Time of Breach equal to 164.4m in the Reservoir and Breach Description screen. However, in reviewing the output in the reported Reservoir Depletion Table and with a Reservoir Initial WSEL of 161.0m, the routed WSEL in the reservoir never reaches the trigger elevation of 164.4m to initiate the dam failure.

2. After routing the flow through the dam structure, DAMBRK fails to route the flow downstream. It stops nearly right away. This can be caused by the following:

2.1 The time step is set at 30 minutes. This is much too large.

2.2 In many locations in your model the river is extremely steep. This is reported as a warning in the output, and DAMBRK states that it is highly likely that part of the model will be supercritical flow. Therefore, it will be necessary to define in the Channel Valley Boundary Conditions screen, for the Flow Type Parameter is to specify Mixed Flow. Any of the available options (i.e., #1, #2, or #3) are viable selections.

2.3 In some reaches in your model, the river is in a steep adverse slope (i.e., negative slope). Again, DAMBRK reports this as a warning in its analysis. This can potentially cause modeling problems.

As is explained in the DAMBRK Troubleshooting Guide, I recommend that you break your model down and simplify it so that you get the basic model running. As explained in this guide, first try to get the model to run steady state, then see if it can route a hydrograph, then see if it can also fail the dam. In addition, you may want to break it into smaller reaches and get each reach running. Much too often, first time users attempt to get the entire DAMBRK model done the first time and do not try to build the model in a piece-wise fashion, making certain that each piece works correctly.

If you have any additional questions or need any further information, please contact me again.

Sincerely,

Chris E. Maeder, M.S.
Senior Technical Engineer
chris.maeder@bossintl.com

Dear Chris;

Thank you for your answer, but what you show me is the limitations of the model which are many, especially the limitation on taking real data, that is what you have when you deal with real rivers.

When I just finish a job and the screen asks press any key, when I do the job returns to execution again very often and is a mess because when you cut the job one looses the 'primitive' from the graphics files Is there any way to do to avoid this.

On the file that I sent to you the submergence effect are important, when I put option 1 I get as much as 1500cms, but myself I know the river and the downstream water will imped such huge discharge, because when I go to option 13 the 3000 steps allowed on my model make it to finish before the peak of the flood PMF reaches the reservoir,the breach flow estimate given by option 1 is completely out of thouch or is an estimate neverthless.

Best regards.

I am considering buying RIVERCAD, but inspite of reading all the brochures I am not still convinced of his advantage, lately I made the study of three rivers concerning floods with HEC-RAS and DAMBK, can you show me what I could add of net value to my results and reports if I had RIVERCAD--

Dear Chris:

The DAMBRK version that you sent to me it was an special one with 3000 time steps allowed .It is not enough I need 6000 steps, it works after all very well on very dificult rivers, but when the stability limits the time step and significant Quebec hydrgraphs are very slow even if already I use only the peak of the hydrograph the 3000 steps are not enough otherwise your model works very well.

Could you make the changes to have 6000 steps on my version and send it by e-mail, since you have the togle number I will still have only one version of your product.

By the way the question tha I asked you about option 01 and option 13 we can conclude that option 01 is very dangerous in our case with option 13 as it happens on reality the breach effect is almost null

Best regards
luis carballada.

Dear Luis :

I do not understand your question:

> When I just finish a job and the screen asks press any key, when I do the job returns to execution again very often and is a mess because when you cut the job one looses the 'primitive' from the graphics files Is there any way to do to avoid this.

Can you describe this a bit better so I can properly respond?

You're correct, Problem Specification Option #1 does not account for tailwater submergence. If that is a concern you have, and it sounds like it is, then use Problem Specification Option #13.

We used to have a version of the analysis engine that has 9000 time-steps. I will ask if we still have that.

What RiverCAD will provide you is a graphical interface that will allow you to cut DAMBRK cross-sections from any type of digital topography, and will automatically generate a flood inundation map from the DAMBRK results. This is most useful from a time-saving stand-point.

If you have any additional questions or need any further information, please contact me again.

Sincerely,

Chris E. Maeder
Senior Technical Engineer
chris.maeder@bossintl.com