All tools
Open Hazen-Williams calculator →

Hazen-Williams Equation — Forms, Constants & Units

Hazen-Williams is an empirical friction-loss relation for water in pressurized pipe. Its appeal is that roughness enters as a single coefficient C that does not depend on velocity or Reynolds number, so head loss solves in closed form. Its cost is that the equation is dimensionally inhomogeneous — the lead constant changes with every unit set, and most wrong answers come from pairing the right constant with the wrong units.

Velocity Form

UnitsEquationWhere
US customaryV = 1.318·C·R0.63·S0.54V in ft/s, R in ft
SIV = 0.849·C·R0.63·S0.54V in m/s, R in m

R is the hydraulic radius (D/4 for a full circular pipe) and S is the slope of the energy grade line, hf/L — dimensionless in both systems.

Head Loss Form — the One You Actually Use

UnitsEquationQDhf, L
US, cfs & fthf = 4.73·L·Q1.852 / (C1.852·D4.87)cfsftft
US, gpm & inhf = 10.44·L·Q1.852 / (C1.852·D4.8655)gpminft
SIhf = 10.67·L·Q1.852 / (C1.852·D4.87)m³/smm
NFPA 13 (sprinkler)p = 4.52·Q1.852 / (C1.852·d4.87)gpminpsi per ft
C is the same number in every unit system. Only the lead constant changes. If a spreadsheet is producing head losses off by a factor of several hundred, the cause is almost always diameter in inches driving a constant written for feet — D4.87 makes a 12× unit error into roughly a 300,000× error.
Sizing a whole distribution network, not one reach? HydroComplete carries friction loss through every reach of the conveyance run and ties it back to the contributing watershed.

Flow Form (Solved for Capacity)

UnitsEquationQDGradient term
US, cfs & ftQ = 0.432·C·D2.63·S0.54cfsftS = hf/L, ft/ft
US, gpm & inQ = 0.282·C·D2.63·S0.54gpminS = hf/L, ft/ft
US, gpm & in, pressureQ = 0.442·C·D2.63·(Δp/L)0.54gpminΔp/L, psi per ft
SIQ = 0.278·C·D2.63·S0.54m³/smS = hf/L, m/m
0.442 versus 0.282 — this one bites regularly. Both are published as "the" gpm-and-inches Hazen-Williams flow constant, and they differ only in what the gradient term means. 0.442 expects pressure gradient in psi per foot; 0.282 expects head slope in feet per foot. Using 0.442 with a dimensionless head slope over-predicts capacity by about 57 percent. If you inherit a spreadsheet with 0.442 in it, confirm the S column is psi/ft before you trust the output.

These follow from the velocity form via Q = VA with R = D/4, and are the exact algebraic inverses of the head loss constants above — 0.432 → 4.73, 0.282 → 10.44, 0.278 → 10.67. Rounding of the exponent 4.87 versus 4.8655 accounts for the small differences between published values; any of them is well within the empirical accuracy of the method itself.

The Exponents, and Why They Matter

TermExponentConsequence
Flow, Q1.852Head loss is not quadratic in Q — doubling flow raises loss by 21.852 ≈ 3.61×, not 4×
Roughness, C−1.852Dropping C from 130 to 100 raises head loss by about 63 percent
Diameter, D−4.87Going one nominal size up is by far the cheapest way to kill head loss
Slope, S0.54Reciprocal of 1.852; the two are the same relation rearranged

Validity Limits

ConditionValid rangeOutside it
Fluidwater onlyUse Darcy-Weisbach — C carries no viscosity term
Temperature~40–75°FError grows at both extremes; hot-water and chilled systems need D-W
Flow regimefully turbulentInvalid for laminar or transitional flow (Re < 4000)
Diameter2 in and largerSmall-bore service tubing is outside the calibration set
Velocitybelow ~10 ft/sAccuracy degrades at high velocity
Pressure conditionfull, pressurizedPartial-flow gravity pipe is Manning's, not Hazen-Williams
Hazen-Williams is a distribution-system tool, not a general pipe-flow tool. It was calibrated on municipal water mains at ordinary temperatures and it does that job well. EPANET, WaterCAD and most modern network solvers offer it for compatibility but compute Darcy-Weisbach internally when asked. On the PE exam, if the fluid is not water at room temperature, the problem wants Darcy-Weisbach.

Sources: Williams, G.S. & Hazen, A. (1920). Hydraulic Tables. AWWA M11 (steel pipe), M22 (sizing water service lines), M32 (distribution system modeling). NFPA 13, Standard for the Installation of Sprinkler Systems — friction loss formula. Mays, L.W. (2010). Water Distribution Systems Handbook. Hwang & Houghtalen, Fundamentals of Hydraulic Engineering Systems.

Need C values for a specific pipe material? Open the C coefficient reference → · Run the numbers with the calculator · Or compare against Darcy-Weisbach.

Related cheat sheets and tools

The companion to this page is the Hazen-Williams C coefficient table — material-by-material, new versus aged. For the physically-based alternative see absolute roughness ε for Darcy-Weisbach and confirm the regime with Reynolds number. Fittings and valves are handled separately — see minor loss K values and equivalent lengths. If the pipe runs partial-flow as gravity sewer, switch to Manning's n. For whole-network pressurized and gravity modeling, see HydroComplete, the SaaS sister product to PE-Calc.

Get future cheat sheets in your inbox

One reference card or worked-example case study per month, delivered the moment it's published. Free.

Free. Unsubscribe anytime. Privacy.