Darcy–Weisbach is exact and wants a friction factor, a roughness and a viscosity. A water utility modelling four hundred kilometres of main wants none of that, so it uses Hazen–Williams: $h_f = \dfrac{10.67\,L\,Q^{1.852}}{C^{1.852} D^{4.8704}}$, read aloud h-f equals ten point six seven L Q to the one-eight-five-two, over C to the one-eight-five-two D to the four-eight-seven.

The letters: $h_f$ is head loss in metres, $L$ the length in metres, $Q$ the flow in cubic metres per second, $D$ the inside diameter in metres, and $C$ the Hazen–Williams C factor — the pipe's whole character in one number. High is smooth: 150 for new plastic, 130 for steel with some years on it, 100 for unlined cast iron that has been in the ground since before you were born. The exponents are not derived; they are fitted, to a century of water flowing through real mains, and the constant 10.67 belongs to those SI units alone. In feet and cubic feet per second it becomes 4.727.

The velocity form of the same empiric is $v = 0.849\,C\,R^{0.63} S^{0.54}$, where $v$ is the mean velocity in m/s, $R$ is the hydraulic radius — flow area divided by wetted perimeter, which for a pipe running full is $D/4$, not $D/2$ — and $S$ is the hydraulic slope, $h_f/L$, metres of head lost per metre of pipe travelled, a bare ratio.

Know its edges, because an empiric has them. Hazen–Williams was fitted for cold water only, at ordinary velocities. It has no viscosity term at all, so it knows nothing about glycol, nothing about oil, and nothing about hot water. Take it near any of those and it will still hand you a confident number — which is the dangerous part.