All tools

Driveway Culvert Sizing — A Worked Example

A start-to-finish design for a typical rural driveway culvert under the 25-year storm. The math is simple; the trap is in the assumptions. We walk through delineation, time of concentration, rainfall intensity, peak flow by the Rational Method, HDS-5 inlet vs. outlet control, and the outlet velocity check that catches most plan-review comments. Each step links to the live PE-Calc calculator so you can swap in your own numbers and run it.

The site

A new single-family driveway crosses an unnamed creek in rural western North Carolina. The county requires the culvert pass the 25-year storm with 0.5 ft of freeboard below the road centerline. USGS quad and a topo survey give us:

Drainage area (A)	12.0 ac
Land cover	~70% pasture (well-grazed), 20% mixed deciduous woods, 10% gravel road + bare
Soil group	HSG B (sandy loam over residuum)
Longest flow path (L)	1,420 ft — sheet 90 ft, shallow-conc 530 ft, channel 800 ft
Average watershed slope	6.5%
Channel slope at crossing	2.2%
Allowable headwater (HW_max)	3.0 ft above culvert invert
Tailwater (TW)	0.8 ft (downstream control: riffle 60 ft below)

Step 1 · Time of concentration

TR-55 segmental (sheet + shallow + channel)

The watershed has all three flow regimes, so TR-55 segmental is the right call — see why here. Kirpich would underestimate the overland portion.

Sheet flow — 90 ft over pasture (n = 0.24, P₂ = 3.4 in NOAA Atlas 14, S = 0.04):
  T_t,sheet = 0.007 · (0.24 · 90)^0.8 / (3.4^0.5 · 0.04^0.4) = 0.161 hr = 9.6 min

Shallow concentrated — 530 ft, unpaved, S = 0.06 → V = 16.13 · √0.06 = 3.95 ft/s:
  T_t,shallow = 530 / (3600 · 3.95) = 0.04 hr = 2.2 min

Channel flow — 800 ft, n = 0.045, R = 0.85 ft, S = 0.022 → V = 4.4 ft/s (Manning's):
  T_t,channel = 800 / (3600 · 4.4) = 0.05 hr = 3.0 min

Total T_c = 9.6 + 2.2 + 3.0 = 14.8 min ≈ 15 min

Run this yourself in the time-of-concentration tool. Manning's n for the channel came from the n reference card — "winding, some pools and shoals" with light vegetation.

Step 2 · Rainfall intensity

NOAA Atlas 14, 25-yr / 15-min

The Rational Method needs intensity at the storm duration equal to T_c. NOAA Atlas 14 (precip.nws.noaa.gov) gives, for this site at 25-year recurrence:

i_{25, 15-min} = 5.84 in/hr
i_{25, 30-min} = 4.32 in/hr
i_{25, 15-min} = 5.84 in/hr (T_c lands on a tabulated duration — no interpolation needed here)

PE-Calc's linear interpolator handles this kind of duration interpolation in either lin-lin or log-log space. Atlas 14 IDF curves are slightly non-linear in log-log; for typical drainage design the difference is < 5%.

Step 3 · Composite runoff coefficient C

Weighted by sub-area

From the runoff coefficient reference:

pasture (heavy clay-loam, 6.5% slope) ≈ 0.30 · 8.4 ac
woods (light brush, summer) ≈ 0.15 · 2.4 ac
gravel + bare ≈ 0.55 · 1.2 ac

C_composite = (0.30 · 8.4 + 0.15 · 2.4 + 0.55 · 1.2) / 12.0
C_composite = (2.52 + 0.36 + 0.66) / 12.0 = 0.30

25-yr frequency factor: C_f = 1.10 → C_design = 0.30 · 1.10 = 0.33

Step 4 · Peak flow Q

Rational Method — Q = C·i·A

Q₂₅ = 0.33 · 5.84 in/hr · 12.0 ac = 23.1 cfs

Open the Rational Method calculator with these inputs to confirm. For a 12-ac watershed, Rational is appropriate; if this were 200+ ac, we would switch to NRCS curve number for runoff volume + a unit hydrograph for routing.

Sanity-check against the NRCS method. Rational gives peak only, no volume. For a quick volume sanity check, the 25-yr 24-hr depth at the site is 5.2 in, CN = 71 (B-soil pasture-woods mix), giving Q-depth = 2.05 in × 12 ac = 2.05 ac-ft. This is the runoff volume the downstream channel will carry over the storm — not the peak. If your design must hold or attenuate this volume, you've moved past culvert hydraulics into pond design.

Step 5 · Trial culvert size — HDS-5

Inlet vs. outlet control

FHWA HDS-5 requires checking both inlet control (entrance is the bottleneck) and outlet control (the barrel + outlet conditions are the bottleneck). The governing design is whichever produces the higher headwater. We try a 24-in CMP first.

Trial	Inlet HW (ft)	Outlet HW (ft)	Governing	Pass?
24-in CMP, projecting	4.05	4.07	Outlet	No (HW > 3.0)
24-in RCP, square edge	3.47	2.80	Inlet	No (HW > 3.0)
30-in CMP, projecting	2.55	2.76	Outlet	Yes (0.24 ft margin)

Run any of these in the culvert hydraulics tool. Both 24-in trials fail: the projecting CMP entrance loses head to entrance contraction and barrel roughness, and even the smoother square-edge RCP exceeds HW_max on inlet control at this flow. A 30-in CMP works comfortably and is usually the same install cost as a 24-in once you account for headwall and bedding. Recommend: 30-in CMP, projecting entrance, 35 ft long at 2.2% slope.

Don't pick the smallest pipe that "passes." Plan reviewers want freeboard in the design, not at the design. Note how little it took to move this design: correcting one sheet-flow term shortened T_c from 21 to 15 minutes, raised intensity from 5.20 to 5.84 in/hr, and pushed Q₂₅ from 20.6 to 23.1 cfs — enough to fail a 24-in RCP that would otherwise have looked acceptable. The recommended 30-in CMP clears HW_max by only 0.24 ft, so if the watershed or the T_c assumptions are soft, price the 36-in.

Step 6 · Outlet velocity & scour

Manning's check on the outlet apron

At the design Q = 23.1 cfs, the 30-in CMP flows partly full — normal depth y = 1.54 ft (y/D = 0.62). The outlet velocity at the design event is V_out ≈ 7.3 ft/s. That's above the typical no-scour threshold of 6 ft/s for unprotected earth, so we need an apron.

Apron rock size (HEC-14 simplified):
D₅₀ ≈ 0.044 · V² = 0.044 · 7.3² = 2.3 in
Apron length ≈ 4 · D_culvert = 4 · 30 in = 10 ft

Use 4-inch nominal Class A riprap, 10 ft of apron, geotextile underlay. Verify the downstream channel can convey 23.1 cfs without overtopping at the design event using Manning's equation; for this site, R = 0.85 ft, n = 0.045 (winding pasture stream), S = 0.022 → capacity ≈ 22 cfs at bank-full, marginal but adequate.

Step 7 · What we did NOT do, and when you would

The bigger picture

No hydrograph routing. The Rational Method gives peak-only. If the road is a critical lifeline, if the downstream property has structures within the floodplain, or if the regulator requires a hydraulic grade line, you need a routed hydrograph (NRCS unit hydrograph or kinematic wave) and a stage-discharge analysis through the culvert.
No 100-yr check. Most counties also require checking the 100-yr event with road overtopping allowed. The 100-yr peak here is roughly 1.7× the 25-yr (Atlas 14 ratio at 15 min); the 30-in CMP would overtop the road by ~0.4 ft, which is acceptable in most rural-driveway standards.
No FEMA floodplain check. If the creek is a FEMA mapped Special Flood Hazard Area, this is a Conditional Letter of Map Revision (CLOMR) job, not a driveway permit. Check the FEMA Map Service Center first.
No water quality / SCM credit. Rural NC driveways under 1 acre disturbed area typically don't trigger NCDEQ post-construction permits. Anything beyond that is full SWMP territory.

For a project that needs the full hydrograph + pond routing + BMP credit + state submittal package, see HydroComplete — the SaaS sister product to PE-Calc that handles those workflows end-to-end.

Tools used in this example

Every step above is reproducible in the live PE-Calc tools: time of concentration · Rational Method · NRCS curve number · culvert hydraulics · Manning's equation · interpolator. Reference values came from the Manning's n, runoff coefficient, and Tc methods cheat sheets.

Buying the lot? Check the drainage before you buy.

If you're considering a rural lot like the one in this example, the question "what culvert do I need?" comes after "is this site even buildable?" Wetlands, FEMA flood zones, soil suitability for a septic, slope stability, and crossing permits all change a project's scope and cost dramatically. SitePrior runs the federal-data property screening report (FEMA, NWI, NRCS, USGS) for $29 in about 60 seconds. Run it before the offer, not after the survey.