|
The Rational Method is a very simplistic method that was developed by
British civil engineers in the middle of the 19th century to estimate peak
runoff rates from small watersheds in order to provide a "rational method"
of figuring out how big storm sewers had to be to carry the load. There has
been a great deal of empirical studies over the years to develop runoff
coefficients from various land types and the Rational Method does a
reasonable job of predicting peak flow rates under the conditions for which
it was developed (Reference: Water Environment Foundation/American Society
of Civil Engineers Manual #77- Design and Construction of Urban Stormwater
Management Systems, 1992).
That said, the Rational Method has many limitations, is not recommended for
drainage areas over 200 acres and has no capability to calculate detention
or reservoir routing flows. That requires a method that prepares a
discharge hydrograph based on variable rainfall inputs such as the Soil
Conservation Society TR-55 method. One of the problems with a methodology
that requires "no more flow than before" is that it tends to lead to every
development building similar detention basins that detain the flow for
something like 12-18 hours and then release it. This results in all the
detention basins releasing the flow at the same time. Whereas prior to
development, the flow was released gradually everywhere and built up slowly
in the river with slow rise and fall, now you have a situation where the
same amount of flow is being released all at once, resulting in much higher
peak flows. That coupled with increasing amount of impervious area
(pavement) within the watershed, speeds the flow of water into the river
significantly compared to slow percolation through vegetated areas and leads
to flash floods.
· After extensive evaluations and verifications, most, if not all by now, Puget Sound communities have done away with the Rational and Y&W methods for pond sizing. In a nut shell, the ponds are severely undersized, the rainfall distributions are fantasy, the rainfall amounts are unrealistic, and the methods have to be severely modified to make them even give the type of results you need. The Rational Method is still used for small catchments for pipe sizing as it is conservative (gives big pipe sizes) and most engineers do like simple and quick calculations.
The rational method fails up in your area for five main reasons:
-
The predevelopment C Factor is conservatively high for peak flow calculations (often 0.2 to 0.4) but is non-conservative when sizing a target, pre-development flow value - then the pipe is sized too big.
-
The runoff hydrograph used for the rational method ignores the tail of the hydrograph thus tending to undersize the basin volume.
-
Sizing for a single value peak flow (e.g. 10-year flow) does little to attenuate flows less or greater than the design flow
-
Delaying flow in the lower parts of a watershed often serve to move flow back in time INTO to peak from upstream causing the peak flow to be worse.
-
Detention does not deal with the VOLUME of flow, only the peak. When we are sending two to three times the volume into the stream system flooding will occur.
In general the rational method will undersize detention ponds without
modifying the base method for determining storm duration. The rational
procedure provides a peak flow rate and in order to size a detention basin,
a hydrograph must be developed. Therefore, some type of hydrograph timing
must be applied to the rational method.
If the time of concentration is used as the time to peak for the hydrograph
and the site is small, the time of concentration will be small, and the
volume for the design hydrograph will be small, and therefore the detention
pond will be small, and usually not sufficient. The critical storm duration
for peak discharge is assumed to be the time of concentration, which is true
to calculate the peak, but not true to calculate the critical hydrograph for
detention.
A modification to the rational method has been developed. See link:
http://www.spokanecounty.org/utilities/stormwtr/guidelin/swglap-p.pdf
One of the problems with a methodology that requires "no more flow than before
is that it tends to lead to every development building similar detention basins that detain the flow for something like 12-18 hours and then release it. This results in all the One of the problems with a methodology that requires "no more flow than before" is that it tends to lead to every development building similar detention basins that detain the flow for
something like 12-18 hours and then release it. This results in all the
detention basins releasing the flow at the same time. Whereas prior to
development, the flow was released gradually everywhere and built up slowly
in the river with slow rise and fall, now you have a situation where the
same amount of flow is being released all at once, resulting in much higher
peak flows. That coupled with increasing amount of impervious area
(pavement) within the watershed, speeds the flow of water into the river
significantly compared to slow percolation through vegetated areas and leads
to flash floods. In general the rational method will undersize detention ponds without
modifying the base method for determining storm duration. The rational
procedure provides a peak flow rate and in order to size a detention basin,
a hydrograph must be developed. Therefore, some type of hydrograph timing
must be applied to the rational method.
If the time of concentration is used as the time to peak for the hydrograph
and the site is small, the time of concentration will be small, and the
volume for the design hydrograph will be small, and therefore the detention
pond will be small, and usually not sufficient. The critical storm duration
for peak discharge is assumed to be the time of concentration, which is true
to calculate the peak, but not true to calculate the critical hydrograph for
detention.
A modification to the rational method has been developed. See link:
http://www.spokanecounty.org/utilities/stormwtr/guidelin/swglap-p.pdf
The Rational method often over predicts the flows for small
urban watersheds and under predicts the flows for large undeveloped
watersheds. It is a good tool to get you in the ballpark for design but the
actual design should be based on statistical approaches. There are a number
of models developed by NRCS and the Corps of Engineers that have factored in
The realistic factors of future storms.
First I will explain why I think the criteria of "rate control" is inadequate even if it was truly applied then I will talk more about the rational method and TR-55.
Basins and other detention facilities, even if properly designed (which they rarely are), will never be enough to prevent flooding in developed watersheds. The reason is that basins do essentially nothing to mitigate the huge increase in runoff volume from development. When we convert natural, pervious landscapes to impervious and semi-pervious ones the volume of runoff increases dramatically. Over an average year it may typically increase 400-500% and during many individual storms it may be even more drastic.
This increased volume causes bankfull flow conditions much more often than receiving streams are able to handle (which changes the entire morphology of the streambed). When a multitude of basins are all releasing to the same watershed, the cumulative effect is a huge increase in overall amount of stormwater conveyed by streams, the peak rate of the flood, and the duration of the flood peak.
What we advocate, and many government groups are beginning to realize, is that rate control alone is entirely INADEQUATE. Both VOLUME and rate control is really the only way to protect both quantity and quality of our receiving waters.
Now on to the actual methods:
The two most common methods for sizing basins (to "ensure" runoff rate is less) that I have encountered are USDA's TR-55 and the Rational Method. Use of the rational method is generally limited to fairly small sites. In fact, many municipalities require that if the rational method is used to size a basin its volume must be at least as great as the volume calculated using TR-55.
A major area of uncertainty for both methods is the time of concentration (TOC) under both pre- and post-development conditions. In many runoff analyses that I have reviewed the "calculated" TOC changed very little when a natural landscape was made largely impervious. It is fairly obvious that the TOC of a woods (for example) should be vastly longer than a typical development site with considerable impervious area.
Determining an overall runoff coefficient (C) for the RM or a curve number (CN) for TR-55 introduces more error. With TR-55, if a site was 50% pavement (CN=98) and 50% a semi-pervious surface (CN=78) the method tells one to use a weighted average CN for the whole site (CN=88). This obviously makes the calculations simplier, but it also represents a totally DIFFERENT site! In the case of proposed development, what weighting a CN does is essentially to evenly distribute the impervious areas throughout the entire site (physically the builder would have to break the pavement up into a few thousand pieces and spread them evenly over the site to produce the runoff predicted by weighting the CN). The only way to accurately apply TR-55 is to handle each impervious area with confined drainage (i.e. a curbed parking lot or a building whose leaders enter a storm sewer, etc.) as a separate entity.
Attempts to use unit hydrograph methods resulted in similar results - too small ponds and frequent overtoppings. It gets old telling your citizens that the reason they flooded, again, was because we had another "100" year storm. Since most methods only estimate what the "100" year storm is supposed to be based on previous (and typically short) rainfall records, odds are you are not really designing for the 1% event anyway. Compound that with a method which is inherently inaccurate and you have a recipe for a (natural) disaster.
The Rational Method is popular because it is very simple to use but the
application is actually very complex - a differentiation that is misunderstood by many designers and laypersons. I have dealt with many drainage designers who use the Rational Method in exactly the same way for every job they run into - because they don't understand hydrology they don't see how a simple equation with only 3 terms in it can be used differently for different sites.
A good example is your description that development should not cause
runoff to exceed natural rates. This sounds simple but is not, in fact.
Engineers will restrict outflow from their projects to the natural rates by
using storage; in ponds, on parking lots, etc. However, the increased
imperviousness of developments results in greater volumes of runoff. While
the rate of flow can be held down the extra volume means that the flow
lasts longer. This longer duration means that the receiving stream peaks at
a higher level. Before the use of storage, the downstream flows in a
channel would flow into it and out before the flows from the upstream areas
arrived. Holding back the runoff for a longer time means that the
downstream catchments are still at their peak when the upstream flows
arrive and thus they add up to a higher flow than the old natural
condition. Storage did reduce the final impact on the stream but did not
prevent an impact on the stream as a "no increase in discharge" policy
would lead people to expect because the increase in volume was not
addressed.
Many jurisdictions in the USA now require that runoff volumes as well as
runoff rates be no higher than the pre-development values.
Virgil's comment about a full system model being required indirectly
addresses the above - the absence of a full system approach is what I mean
by inadequate or faulty hydrology. A full system model, however, is quite a
different problem than site design. The full system model addresses every
catchment draining to a receiving stream and the impact of development on
the stream as a whole. A full system model is well beyond a reasonable
requirement for site designers and, indeed is neither warranted nor useful
on a site by site basis. Since a full system model addresses all sites it
is correctly the responsibility of the governing jurisdiction, not site
developers or their engineers. Thus, lack of a full system model is an
inadequacy of the jurisdiction.
Another design mis-application is that some designers, especially in the
past, did a quick and dirty design by only applying the Rational (or any
method) to a single storm condition, say a 5 year 2 hour event. This is way
too simplistic. The correct procedure would be to determine the governing
rainfall depth and duration for the catchment for pre-development and then
the governing rainfall and duration for post-development and ensuring that
the flow rate and volume do not exceed those allowable. This means
analyzing the 2, 5, 10, 25, 50 and 100 year rainfall depths for each of 1,
2, 6, 12 and 24 hour durations. For each return period, a different
duration will result in the highest flows and greatest storage requirement.
|
