Hydrologic culvert design in general terms may be considered an art more than a science in my opinion. There certainly are hard facts and science involved but more often than not, when it comes to practical applications, the amount of data is typically very limited; this is especially true during the early stages of a projects design cycle. Therefore, it becomes crucial for the culvert designer to take on a more creative approach but maintain balance with conservative assumptions.
The quality or accuracy of the culvert design is highly dependent upon the quality of the input data. The phrase, “garbage in, garbage out” rings true on this topic. More often than not, when asked to perform a proper culvert sizing analysis, the data is usually limited to a peak flow rate.
The phrase, "garbage in, garbage out"
rings true on this topic.
There are many variables to consider with a culvert sizing analysis, the maximum allowable rise and span, slope, natural bottom (arch) versus full invert pipe, end treatment, shape geometry, roughness coefficient, bed load considerations, just to name a few. Also, there are two main design approaches of inlet control versus outlet control that must be considered. There are many powerful software tools available that will give a highly accurate culvert design. However, sufficient input data must be available to run the programs and as discussed earlier, those input fields are typically lacking.
Let us consider a case where the data is limited to a peak flow rate only, the pipe slope and road elevation relative to the streambed may be adjusted to prevent over topping and it has a free flow discharge (assuming no tail water). Consider inlet control first since the data input requirement is much less. Nomographs, as found on page 192 of the NCSPA book, are very helpful in this endeavor. Many different culvert sizes may be proposed at a given flow rate. Drawing a line on the chart intersecting those two data points will result in the expected Head Water (HW) in terms of pipe rise for the given end treatment. The minimum pipe size with adequate height of cover that exceeds the HW as given in the chart will be the proper minimum culvert size.
An outlet control analysis will be impossible with this limited amount of data. One way to make an educated guess of the flow capacity of the culvert is to consider the pipes conveyance (K) as given on page 172 of the NCSPA book. This property is constant for each pipe size of the same roughness coefficient (n). To arrive at the flow capacity, simply take the square root of the slope in decimal form and multiply it by the conveyance (K). This will determine the minimum slope required to convey the designed peak flow rate. If the slope exceeds what is practical from the site conditions, increase the pipe size until a solution is found. Whichever method, inlet or outlet control that results in the largest pipe is the one to be considered.
This method does not give “the answer” for a properly sized culvert, but it does allow for an idea of the pipe size that may be expected given the limited data set available. It is important that the assumptions be properly disclosed, so the Engineer of Record knows that the final design must be calculated when a more complete data set is available. However, it may not be economically viable for some projects to ever obtain all the required data. In this case, a second review for sensitivity of the projects risk assessment for the pipe function may be all that is needed.
So what do you think? Hydrology .... more science or art in the design?