A Rain Garden Is a Drainage Calculation, Not a Flower Bed

The bed that is really a basin

A rain garden looks like a flower bed and behaves like a piece of drainage. Set a shallow, planted depression at the low point of a yard, route a downspout or a driveway's runoff into it, and it captures the first flush of a storm, holds it, and lets it soak into the ground instead of rushing to the street and the storm drain. That single function, catch and pond and infiltrate, is what separates it from ordinary landscaping, and it is measured in inches per hour and hours to empty, not in bloom time. Get the hydrology right and almost any suitable planting will thrive. Get it wrong and the finest native palette drowns, or the water never arrives at all.

Start with the contributing area

The first number is not a plant; it is the area draining into the garden. A roof section, a driveway, a compacted lawn each shed a different fraction of the rain that lands on them. Designers weight these with a runoff coefficient: a roof or paving sheds roughly 0.90 to 0.95 of its rainfall, while an established lawn sheds closer to 0.30, and compacted or clay soil behaves far more like pavement than like open ground. Multiply each surface by its coefficient, sum them, and you have the effective contributing area the basin must handle.

From there the garden is sized as a fraction of that area. On free-draining sandy soil, or on soil you have excavated and replaced with an engineered mix, a garden of roughly 20 to 30 percent of the contributing area is typical. On heavy native clay left in place, the same job can demand a basin closer to 60 percent, which is usually the moment a designer decides to amend the soil instead. The ratio is not aesthetic; it falls straight out of how much water is coming and how fast the ground will take it.

The soak test decides everything

Before any plant is chosen, the ground itself has to pass a test. The standard soak test is humble: dig a hole about 12 inches deep, fill it with water to saturate the surrounding soil, let it drain, then refill it and measure the drop with a ruler each hour for at least four hours. The resulting rate, inches per hour, is the single most important figure in the whole project.

Most guidance wants to see at least half an inch per hour. Below roughly that, the basin will hold water too long; below about a tenth of an inch per hour, essentially pure clay, infiltration is effectively off the table and the design has to change. This is why two identical-looking gardens, planted from the same nursery flat, can succeed on one side of a street and fail on the other. The plants did not decide it. The soil did.

The 24-to-48-hour rule

The governing performance target is drawdown: how long ponded water takes to disappear. Residential guidance generally wants the surface pool gone within 24 hours, and most engineered stormwater manuals set a hard ceiling of 48 hours from the storm peak. The reason is not tidiness. Water standing longer than about two days breeds mosquitoes, drowns the roots of all but true wetland species, and signals that the system is no longer infiltrating as designed. Ponding depth is set to serve that clock, commonly 4 to 8 inches, occasionally up to a foot where the soil is fast enough to clear it in time. Depth and infiltration rate together, not the flowers, decide whether the garden empties on schedule.

When the ground is too slow

If the soak test comes back sluggish, there are two honest moves. The first is to excavate the basin and backfill it with an engineered bioretention media, broadly 50 to 60 percent coarse sand blended with topsoil and compost, a recipe tuned to infiltrate quickly while still feeding plants and hosting the microbes that break down pollutants. The second, where even amended soil cannot clear the pond in time, is an underdrain: a bed of gravel and a perforated pipe beneath the media that carries excess water away to a safe outlet, converting the garden from a pure infiltration basin into a filtering one. Neither fix is a compromise on planting; both are hydraulic decisions made long before the shovel meets a root ball.

Why clogging, not drought, kills it

The most common way a rain garden dies is not thirst or the wrong species, it is clogging. Over seasons, fine sediment carried in with the runoff settles across the basin floor and seals the surface, and the infiltration rate quietly collapses. The tell is the drawdown clock: a cell that once emptied overnight but now takes more than 72 hours to drain is telling you its surface has silted over. The maintenance that keeps a rain garden alive is therefore mostly upstream and hydraulic, a vegetated buffer or a forebay to drop sediment out before it reaches the bed, and periodic removal of the accumulated fines, rather than the deadheading and dividing a border demands.

The planting comes last, and still matters

None of this makes plants ornamental afterthoughts. Deep-rooted native grasses, sedges and rushes keep the media open, lift water through transpiration, and hold the surface against erosion, and a garden that reads as intentional is one a neighborhood will actually keep. But the planting plan is the last decision, not the first. Size the basin to its contributing area, prove the soil with a soak test, set a ponding depth that clears in a day or two, and fix a slow site with media or an underdrain. Do that, and the plant list becomes a genuinely free choice rather than a doomed gamble laid over drainage that was never going to work.