How this instrument works
Friction head loss is the pressure a pipe run steals from water moving through it, expressed the way hydraulic engineers actually think about it: not in pascals, but in metres of water column — the extra height a pump or elevated tank would have to supply just to make up for wall friction over that length of pipe. Flowing water drags against the pipe wall, and rougher walls or faster flow drag harder, so head loss climbs with both.
The exponents look odd — 1.852 on flow, 4.8704 on diameter — because Hazen-Williams was never derived from first principles the way Darcy-Weisbach was. Allen Hazen and Gardner Williams fit these numbers directly to measured pipe-flow data in the early 1900s, trading the generality of a dimensionless friction factor for an equation solvable in a single pass, with no need to iterate toward a Reynolds-number-dependent value on a Moody chart.
That convenience has a price. The formula is calibrated for water near ordinary temperatures, moving in the fully turbulent range engineers actually design pipe runs for, roughly 0.9 to 3 metres per second. Push it outside that band, or ask it about oil, air, or slurry, and the C coefficient stops meaning anything — that is Darcy-Weisbach's job. C itself is not fixed for a given pipe either: it falls as scale, corrosion, or biofilm build up inside the bore, so a main rated C=130 when new is routinely re-checked closer to C=100 after years of service.
- Enter Flow rate — litres per second by default; switch the unit menu to gpm if that is how your data arrives.
- Set Pipe inside diameter to the pipe's actual bore, not its nominal size; use mm or switch to inches.
- Enter Pipe length as the straight-line pipe run in metres or feet.
- Set Hazen-Williams roughness coefficient C (130 = new PVC) — lower it for older or corroded pipe, raise it for very smooth new plastic or copper.
- Read Friction head loss in metres or feet of water — the pressure a pump or supply must overcome beyond elevation change and fittings.
Worked example — 10 L/s through 100 m of new PVC
Take a 100 mm (0.1 m) inside-diameter PVC line, C=130 for new pipe, carrying 10 L/s (0.01 m³/s) over a 100 m run — a typical service line from a water main into a building. Converting flow to 0.01 m³/s and diameter to 0.1 m and applying the formula gives h_f = 10.67 × 100 × 0.01^1.852 ⁄ (130^1.852 × 0.1^4.8704) = 1.9034 m.
That head is what a source or pump must supply above elevation change and fitting losses, purely to overcome friction — about 1.9 m of water column over 100 m of pipe, close to a 1.9 percent grade pictured as a slope. At that flow, water moves about 1.27 m/s through the bore (flow divided by cross-sectional area), squarely inside the 0.9–3 m/s band Hazen-Williams coefficients were originally measured against — exactly why this answer can be trusted.
Questions
Why doesn't Hazen-Williams need an iterative friction factor like Darcy-Weisbach?
Because the roughness coefficient C is a fixed, looked-up number baked directly into the equation instead of something solved for. Darcy-Weisbach's friction factor depends on both Reynolds number and relative roughness, which usually means iterating with the Colebrook equation or reading a Moody chart. Hazen-Williams trades that generality for a single algebraic pass — fast, but valid only for water in ordinary turbulent flow.
What does the roughness coefficient C actually represent?
An empirical smoothness rating for the inside of the pipe, not a measured physical roughness height. Higher C means a smoother bore and less friction for the same flow — new PVC or copper commonly rates around 130 to 150, while old, tuberculated cast iron can drop below 100. Because C is looked up from tables rather than derived, it folds material, age, and scaling into one convenient number.
Can this formula be used for fluids other than water?
No. The constants in Hazen-Williams were fit to water near ordinary temperatures, and the result stops being reliable for oils, gases, slurries, or anything with meaningfully different viscosity. For those fluids, Darcy-Weisbach, which accounts for viscosity through the Reynolds number, is the correct tool. Treating C as if it applied to a viscous fluid quietly produces a wrong head-loss figure with no warning.
Why does head loss grow faster than flow rate?
Because flow rate is raised to the power 1.852, not 1. Doubling the flow through the same pipe more than triples the friction loss, since 2^1.852 is about 3.6 — which is why a modest increase in demand can produce a surprisingly large pressure penalty, and why engineers oversize pipe diameter up front rather than lean on higher pump pressure to push more water through later.
Does a pipe's C value change as it ages?
Yes, and significantly. A new smooth pipe might rate C=130 to 150; the same pipe after years of scale buildup, corrosion, or biofilm growth can fall to C=100 or lower, which roughly doubles the friction loss at a given flow. Utilities often design against a lower C than a pipe's as-installed value specifically to leave room for this predictable degradation over its service life.
Who actually uses the Hazen-Williams equation in practice?
Water utility engineers sizing transmission mains and distribution pipes, fire-protection engineers checking sprinkler branch-line pressures against NFPA requirements, and irrigation designers laying out drip or sprinkler mainlines all reach for it, because it turns a pipe run and a flow target into a single, non-iterative pressure-loss figure without needing viscosity data most of them don't have on hand.