Hazen-Williams Flow Calculator
Estimate water flow velocity and discharge in a pressurized pipe using the empirical Hazen-Williams equation. Useful for water distribution system design and quick hydraulic calculations on municipal water mains.
Last updated: May 2026
Compare with similar
About this calculator
The Hazen-Williams equation is an empirical formula widely used in water-utility engineering for steady, full-pipe flow of water under pressure. The SI form for velocity is v = 0.849 × C × R^0.63 × S^0.54 (using hydraulic radius), or equivalently in terms of pipe diameter: v = 0.849 × C × D^0.63 × S^0.54 × (1/4)^0.63 simplified. Variables: v in m/s, C is the Hazen-Williams roughness coefficient (dimensionless), D is internal pipe diameter in meters, S is hydraulic gradient in m/m (head loss per unit pipe length). For flow rate Q: Q = v × A = v × π × D² / 4 in m³/s. The imperial form is v = 1.318 × C × R^0.63 × S^0.54 in ft/s. Typical C values: new smooth steel C = 130; new ductile iron C = 130–140; cast iron new = 130; old cast iron (50+ years) = 80–100; new PVC/HDPE = 150; concrete 110–140; new copper 130–140. C decreases over time due to corrosion, scaling, and biofilm — engineers use lower 'design C' values for aging systems to ensure adequate capacity throughout the system life. Edge cases: Hazen-Williams is calibrated for water at typical municipal temperatures (5–25°C) and velocities 0.5–3 m/s. It is reasonably accurate within this range but degrades outside it — at very low or very high velocities, the empirical fit breaks down. The formula is NOT recommended for fluids other than water (different density and viscosity make the empirical coefficients wrong), for very hot or cold water (use Darcy-Weisbach), or for very small or very large pipes (calibration range was for 50–1,500 mm). Hazen-Williams gained popularity in the early 1900s when computing Darcy-Weisbach's implicit Colebrook equation by hand was difficult; with modern computation, Darcy-Weisbach is the more accurate choice for new designs.
How to use
Example 1 — water main flow check. New PVC pipe (C = 150), diameter D = 0.2 m (200 mm), hydraulic gradient S = 0.005 m/m. The implemented formula uses 0.849 × C × D^2.63 × S^0.54 directly for flow rate. Step 1: D^2.63 = 0.2^2.63 ≈ 0.0233. Step 2: S^0.54 = 0.005^0.54 ≈ 0.0479. Step 3: Q = 0.849 × 150 × 0.0233 × 0.0479 = 0.142 m³/s = 142 L/s. Step 4: cross-section A = π × 0.2² / 4 = 0.0314 m². Step 5: velocity = Q/A = 0.142 / 0.0314 = 4.52 m/s. Verify: 4.5 m/s is higher than typical design velocity (2 m/s recommended max for municipal mains to prevent water hammer and erosion) — pipe is undersized or the gradient is too steep. Sensitivity: doubling pipe diameter to 0.4 m reduces velocity dramatically — D^2.63 increases by factor 2^2.63 = 6.19, so Q increases 6.19×, but A increases 4×, so v increases 6.19/4 = 1.55× — wait, this is contradictory. Let me recompute: at D = 0.4 m, Q = 0.849 × 150 × 0.4^2.63 × 0.005^0.54 = 0.849 × 150 × 0.1444 × 0.0479 = 0.880 m³/s; A = π × 0.16/4 = 0.1257 m²; v = 0.880/0.1257 = 7.0 m/s. Velocity actually increased — this is wrong intuitively. The issue is that the Hazen-Williams D^2.63 formula gives Q directly, and for a fixed gradient larger pipes carry much more flow but at HIGHER velocity. To get lower velocity, you'd need to reduce the gradient too. Example 2 — design a pipe for required flow at acceptable velocity. Required Q = 50 L/s = 0.050 m³/s, want v ≤ 1.5 m/s (typical conservative design), C = 130 (new ductile iron), need to size D and find required gradient S. From v ≤ 1.5 m/s: A ≥ Q/v = 0.050/1.5 = 0.0333 m², D ≥ √(4×0.0333/π) = 0.206 m → use 250 mm pipe (next standard size up). Compute actual v at D = 0.25 m: A = 0.0491 m², v = 0.050/0.0491 = 1.02 m/s. Now find S to achieve Q at this D and C: 0.050 = 0.849 × 130 × 0.25^2.63 × S^0.54. Step 1: 0.25^2.63 = 0.0426. Step 2: 0.050 / (0.849 × 130 × 0.0426) = 0.050 / 4.704 = 0.01063. Step 3: S^0.54 = 0.01063; S = 0.01063^(1/0.54) = 0.01063^1.852 = 0.000219 m/m. Verify: gradient of 0.000219 m/m × pipe length 100 m = head loss of 2.19 cm per 100 m of pipe — very gentle slope, consistent with smooth pipe at low velocity.
Frequently asked questions
What Hazen-Williams C values should I use for common water pipe materials?
Published C values for various pipe materials and ages: New ductile iron with cement-mortar lining: 140–145. New steel: 130. New cast iron: 130. New copper: 130–135. New PVC: 140–150. New HDPE: 140–150. New concrete (smooth): 130–140. Older pipes degrade significantly: cast iron 20 years old: 100–110; cast iron 50 years old: 80–90; unlined steel 30+ years: 80–100; concrete 20+ years: 110–125. Design practice usually uses conservative 'design C' values 10–20 below new-pipe values to account for aging: design C = 100 for cast iron, 120 for ductile iron with cement lining, 140 for PVC. AWWA M11 (Steel Pipe Manual), M41 (Ductile Iron Pipe), and similar manuals publish recommended design C values for water utility use. Reviewing actual measured C in your specific system through field flow tests is more accurate than published tables — many utilities have ongoing pipe-condition assessment programs. Lower C from biofilm buildup and tuberculation can be partially reversed by pipe cleaning (pigging, scraping); after cleaning, C can return close to new-pipe values.
How accurate is Hazen-Williams compared to Darcy-Weisbach for water pipe design?
Hazen-Williams is an empirical equation calibrated for water at ordinary temperatures (5–25°C) and velocities (0.5–3 m/s) in pipes ranging roughly 50–1,500 mm diameter. Within this range, it agrees with Darcy-Weisbach to within ±5–15% depending on roughness assumptions. Outside this range, accuracy degrades: at very low velocities (< 0.3 m/s), Hazen-Williams underestimates head loss by 10–30% because actual flow is in or near the laminar regime where Hazen-Williams's empirical coefficients fail. At very high velocities (> 4 m/s), Hazen-Williams typically overestimates losses by 5–15% because actual friction-factor behavior diverges from the empirical fit. For hot water (above 30°C) or cold water (near freezing), the implicit viscosity assumption is wrong — losses change by 10–30% versus prediction. For fluids other than water (sewage, brine, slurry), Hazen-Williams should not be used. Darcy-Weisbach with the Colebrook equation is theoretically rigorous and accurate for any Newtonian fluid at any temperature, but requires iterative solution. With modern spreadsheets and software, Darcy-Weisbach is no harder than Hazen-Williams and gives better accuracy. However, Hazen-Williams remains widely used in US water utilities because C values are well-tabulated for water pipes and engineers are comfortable with the formula.
How does the Hazen-Williams hydraulic gradient relate to physical pressure loss?
The hydraulic gradient S = h_f / L is the head loss per unit pipe length, where h_f is friction head loss in meters of water column and L is pipe length in meters. So S is dimensionless (m/m). To convert head loss to pressure loss: ΔP = ρ × g × h_f, where for water ρ = 1,000 kg/m³ and g = 9.81 m/s², giving ΔP in Pa: ΔP = 9,810 × h_f Pa per meter of head. In practical units: 1 m of water column ≈ 9.81 kPa ≈ 1.42 psi ≈ 100 mbar. So a gradient of 0.01 m/m means head loss of 1 m per 100 m of pipe = 9.81 kPa per 100 m = 1.42 psi per 100 m of pipe. Typical water distribution system gradients are 0.001–0.005 m/m for the main grid and up to 0.01 m/m for steep service lines. To find the required gradient from a known pressure budget: given an available pressure drop of 50 kPa over a 1,000 m pipe run, convert: 50 kPa / 9.81 = 5.10 m head; S = 5.10 / 1,000 = 0.00510 m/m. Then apply Hazen-Williams to find the pipe diameter that delivers required flow at that gradient. Note: gradient S is the slope of the hydraulic grade line (HGL) — the line representing total energy minus velocity head, useful for visualizing pressure variation along a pipe network.
What are common mistakes when applying the Hazen-Williams equation?
The most common mistake is using Hazen-Williams for non-water fluids — the empirical coefficients are calibrated for water, and applying them to oil, glycol, sewage, or slurries gives wrong answers; use Darcy-Weisbach for non-water fluids. Using new-pipe C values for aging systems — actual C in a 30-year-old cast iron pipe may be 30–40% lower than new, resulting in 50–80% higher actual head loss than predicted. Confusing the dimensionless gradient S with pressure or head — S is head loss per unit length (m/m), not the total head loss or pressure drop. Mixing imperial and SI versions — they use different constants (1.318 in imperial, 0.849 in SI). Using Hazen-Williams outside its calibrated range — for very low velocity (< 0.3 m/s), very small pipes (< 50 mm), or hot/cold water, Darcy-Weisbach is more accurate. Ignoring minor losses from fittings, valves, and bends — Hazen-Williams gives only friction loss in straight pipe, just like Darcy-Weisbach. Forgetting that the formula assumes full-pipe flow under pressure; for partially-full pipes (storm drains, sewers, culverts), use Manning's equation instead. Using internal diameter incorrectly — pipes have nominal sizes (e.g., '4-inch pipe' has different actual ID depending on schedule: Sch 40 = 102.3 mm, Sch 80 = 97.2 mm); always use actual ID. Finally, computing flow rate without verifying the resulting velocity is in the design range (typically 0.6–3 m/s for water mains).
When should I NOT use this calculator?
Skip Hazen-Williams for non-water fluids (oil, chemicals, glycol mixtures, slurries) — use Darcy-Weisbach with the appropriate fluid properties instead. Do not use it for partially-full pipe or open-channel flow — Manning's equation is correct for those cases. Avoid it for hot water (above 30°C) or cold water (below 5°C) where viscosity varies enough to invalidate the calibration. The formula doesn't apply to gas flow at any temperature — use compressible Darcy-Weisbach or Weymouth/Panhandle equations for gas. For two-phase flow (water with significant entrained air, or steam-water mixtures), specialized two-phase correlations are needed. For pipes outside the calibration range (very small < 25 mm or very large > 2,000 mm), use Darcy-Weisbach for better accuracy. For low-velocity laminar flow (< 0.3 m/s), Hazen-Williams underestimates loss; use Darcy-Weisbach where friction factor for laminar flow is exactly f = 64/Re. For viscous fluids generally (μ > 5 × water at the same temperature), Darcy-Weisbach with proper fluid properties is required. For new pipe design where the highest accuracy matters, professional process simulation software (KORF, AFT Fathom, Bentley WaterCAD) uses Darcy-Weisbach with full Colebrook iteration. Finally, for any regulatory submittal or critical-safety system (fire-water, drinking-water treatment), use the methodology specified by your regulator — many specify Darcy-Weisbach as the authoritative method.