The Rational Method
Peak runoff is how hard it rains, times the share that runs off, times the area.
A hard thunderstorm settles over a 10-acre parking lot, and every drop on it heads for one storm drain in the low corner. Somebody has to size that drain, and it all comes down to one question: at the worst moment, how much water arrives each second? The surprise is how little you need to know to answer it.
The peak is a flow balance
Start the rain and watch the drain. In the first minute, only water that landed right beside it has arrived, so the flow is small. A few minutes later, water from farther out is arriving too, and the flow is bigger. It keeps growing for as long as new parts of the lot are still reporting in.
Then the water from the farthest corner gets there. That travel time has a name, the time of concentration, . At the whole lot is contributing at once, and if the rain holds steady, nothing is left to change. Every second, the same amount of rain lands and the same amount of water leaves through the drain. That's steady state, and that's the peak.
So the peak is a plain flow balance: what comes in each second goes out each second. What comes in is the rain rate times the area it falls on, since a depth per time multiplied by an area is a volume per time. Not all of it reaches the drain, though. Some soaks in, some sits in puddles, and some just wets the leaves. Lump all of that into one fraction, the runoff coefficient , the share of the rain that makes it out. The result is what the Handbook's Hydrology/Water Resources section calls the Rational Formula:
Here is the peak discharge in cfs, is the dimensionless runoff coefficient, is the rainfall intensity in in./hr, and is the watershed area in acres.
Those units should bother you. Inches per hour times acres doesn't look like cubic feet per second, yet the formula has no conversion factor. Work it out for one inch per hour on one acre:
That's a pure accident of how inches, hours, acres, and feet happen to be defined, and engineers gratefully round it to 1. The 0.8 percent they drop is far smaller than the uncertainty in .
The accident only works in those units, so get every input into them first. The Handbook's conversion table gives 43,560 ft² per acre. It has no row for square miles, so use acres per square mile. And is a rate, not a storm's total depth: a storm that drops 2 in. in 30 minutes is falling at 4 in./hr.
What C stands in for
is a fraction of the rain, between 0 and 1. At , nine-tenths of the rain reaches the outlet. When a problem says 70 percent of the rain runs off, is 0.70, not 70. Everything that keeps rain from reaching the outlet goes into that one number, whatever the reason. That's crude, and it's what makes the method so quick.
That lower row is the parking lot you started with. At 2 in./hr, its drain has to pass 18 cfs. Three numbers, multiplied.
The Handbook prints the formula but no values of , so a problem has to hand you the coefficient. What's worth carrying in your head is the ranking: sealed surfaces sit near 1, and soft, planted ground sits down near 0.
Source: typical published ranges from general drainage practice, not the FE Reference Handbook
Three knobs on one product
is a pure product, with nothing added and nothing squared. So each factor is a knob that works the same way: multiply any one of them by some number, and the peak is multiplied by that same number. Many questions can be answered without ever finding the peak itself, and you should take that route when it's open, because there's less to get wrong.
Paving is the classic case. It leaves the rain and the size of the site alone and raises . Take the meadow and the lot from the flow-balance figure: goes from 0.20 to 0.90, so the peak rises by times, and that would hold at any and any .
Real paving also speeds the water on its way to the outlet, which can raise the design intensity too. Time of Concentration and Design Storms takes that up. With held fixed, the ratio of the values is the ratio of the peaks.
When the land is mixed
Few sites are all one thing. Rain falls at the same on the roof, the driveway, and the lawn, but each passes its own share. So treat each patch as its own small Rational calculation, and add the flows:
Now ask what single , applied to the whole site, would give that same peak. The answer is the composite runoff coefficient:
It's an average weighted by area, which is what you'd want: a big lawn should count for more than a small roof. The Handbook prints no composite- formula, but this is its weighted arithmetic mean, from Dispersion, Mean, Median, and Mode Values, with area as the weight. The result must land between the site's lowest and highest , pulled toward whatever covers the most ground, which makes a quick check.
Running it backward
Forward, sizes something new: the peak becomes a storm sewer's design flow in Flow in Sewers. Backward, it tells you how much land something that already exists can serve.
Think of it per acre. At the peak, each acre sends cfs toward the outlet, and a ditch of known capacity has room for only so many of those acres:
What the picture assumes
Everything so far comes from one picture: steady rain, the whole area contributing, and a single fraction for everything lost. The Handbook prints only the formula and its definitions, with nothing on these assumptions and nothing on the time of concentration, so you have to bring them with you.
- The storm lasts the time of concentration. Stop the rain any sooner and the near water has drained away before the far water arrives, so the whole area never contributes at once. Time of Concentration and Design Storms covers why a longer storm doesn't make a bigger peak either.
- Rain is uniform, in time and over the area. One number, , has to describe the rain everywhere and from start to finish.
- is the only loss term. There's no separate infiltration calculation and no soil that gets wetter as the storm goes on. There's no storage either: water held back on its way to the outlet has nowhere to appear in the formula.
- The drainage area is small. This is what makes the others believable. One storm can cover a small watershed evenly and outlast its travel time, and a large basin breaks both.
How small is small? There's no sharp line, and local rules vary, but a common cutoff is around 200 acres, about a third of a square mile.
The method also gives you only the peak flow. The total volume of runoff is a job for Curve Number Runoff, and the way the flow changes over time is a job for Hydrographs and Detention.