Epsilon Likelihood Gumbel
Epsilon Likelihood Gumbel
Epsilon Likelihood Gumbel
Try it today
A Step-by-Step Approach
If you’re a newbie to hydraulics & hydrology then you’re in a good place right now. This blog is
about how to design a detention pond outlet device without pulling your hair out. I can tell you
first-hand, there are a lot of moving parts to a detention pond. Not just the pond itself but the
devices that allow water to flow out of the pond. If you’ve ever had the task of designing one
then you’re at least familiar with standard design criteria that usually goes something like this:
The storage facility must be designed to detain the post-developed flows to that of the pre-
developed flows for the 2-, 10- and 100-year return periods.
Simple enough. And, in all likely hood, the resulting outlet structure will something like this:
This is known as a “Multi-Stage” structure. Multi-Stage refers to the multiple stages or depths in
the pond that this structure must address. In our case depths corresponding to the 2, 10, and 100
year runoff events.
You typically see as many devices in a multi-stage outlet structure as there events. In the figure
above there are three. An Orifice, a Rectangular Weir and a Culvert. As for the Riser, well, it’s
just there to hold the first two guys… Orifice and Rectangular Weir while conveying their flows
to the culvert.
Since this article is more about the outlet structure design “procedure”, I won’t go into a lot of
detail about the storage part. But it’s important that you get this first step right. Ninety-nine
percent of pond design failures are due to not having enough storage.
So let’s assume you already have done the storage estimating by developing the pre- and post-
developed hydrographs. And then estimating the required storage by computing the area between
the two as shown in the blue cross-hatched section here. Be conservative, you can always scale
the pond size back. I just want to make sure you at least get a solution on your first try.
You have several outflow devices, which one do you size first? The riser? Intuitively one wants
to start with the orifice, then the secondary weir, then the riser and finally the culvert. In other
words, you natuarally want to design in the order of return periods, i.e., 2-yr, 10-yr and then 100-
yr. Don’t do this! You could be at it all afternoon. The size of one affects the ability of the other,
which affects another, and so on.
A Better Procedure
Here’s the secret to a no-fail outlet structure design: Design your devices
starting downstream and work your way upstream. Not the other way around. Starting at the
most downstream end, add a device at a best-guess size, location, slope, etc. Then perform a trial
routing to determine if that device is sized to meet its criteria. If not, make adjustments until it
does. The next step is to add another device, just upstream of the previous one, with its invert
elevation equal to the maximum stage reached during the previous trial routing. Perform a new
trial routing to determine if that device is sized to meet its criteria. If not, repeat until it does.
Continue this process of adding new devices until all event criteria have been met.
This is a simple and worthy procedure… unless you start with the wrong device. In which case
you could end up chasing your tail.
The next upstream structure is the orifice. In this case, the 2-year event is up to bat. (Orifices
seem to work the best in these cases as they have a somewhat linear stage-discharge curve and
unlike weirs, are more predictable under large heads.)
At this point you can add a Riser structure but just for use as a container for the secondary
structures. Set its crest elevation high for now.
Set the orifice at, or just above, the bottom of the pond. (Note: If you’re using Hydrology Studio,
be sure to check the multi-stage (“Flows through Culvert”) option.
Pick a size and perform a trial routing and observe. Adjust the size of the orifice until you’re
satisfied with the maximum discharge, i.e., 2-year Target Q. When that’s done, make a note of
the corresponding maximum water surface elevation and move upstream to the next device.
Here you have a choice of adding a weir or even another orifice. I prefer to use rectangular weirs
but you should feel free to experiment with orifices as well. V-notch weirs can be difficult to get
dialed in. Leave those for the researchers. In this example we’re using a rectangular weir. Set its
crest elevation to the maximum elevation reached in the orifice design. Next, set a best-guess
crest length and do a trial routing. Per the general procedure, adjust the crest length, not the
invert, and perform trial routings until you achieve the 10-year Target Q.
In most cases, this device will also satisfy the remaining, and intermediate, target discharge
requirement(s). In our case the 100-year. If not, repeat this same step with an additional weir.
(Basically turning it into a compound rectangular weir.) Simplicity should rule. One device is the
better solution.
Don’t Be Tempted
You may be inclined at this point to use a Riser to satisfy that last criteria but I don’t recommend
it. First of all, the riser crest will need to be large relative to the what you really need. After all,
it’s holding the orifice and secondary weir(s) while attaching itself to the upstream end of the
culvert. Its crest length could very easily surpass 12 feet when you need only one foot. The crest
elevation then becomes very sensitive to changes to head. Remember, weirs are very powerful
and can allow a lot of water to pass with a very shallow head. Think down the road when this
pond becomes silted up, causing the maximum water surface elevation to rise slightly higher
than designed.
However, there will be cases when you don’t have an economical alternative other than to use
the Riser’s crest. Remember to subtract any other secondary weir crest lengths from the riser’s
crest length, so you’re not double-dipping.
Now you can add the riser. I like to add the riser last but honestly, you could have added it first
or whenever, but you don’t know where to set its crest elevation until the other structures have
been set. You’ll end up moving it up and down a lot during the process, wasting time. If you
insist on adding it earlier, make its crest length large, set it high, but not above the top of the
pond, and adjust those parameters last.
Size the riser for practical reasons and not so much for hydraulics. Set its crest above the final
maximum water surface elevation reached during the last structure design. Set the crest length
just large enough to hold the multi-stage devices as well as connect adequately to the upstream
end of the culvert. Perform a final trial routing to insure it is not hindering the performance of the
other structures. If it is, enlarge the crest length.
Done
Hopefully this post has given you some insight and guidance on how to approach your next
detention pond design. The procedure should help save you some time and frustration regardless
of what hydrology software you’re using. For more information on detention pond design
software check out Hydrology Studio. Hydrology Studio’s Help contains a tutorial on how to
use its Trial Routing feature which follows this same procedure.
Further Reading