How to Size a Process Pipe: A Step-by-Step Guide

A practical method every process engineer should know — with a full worked example.

Line sizing is one of the most common calculations in process engineering, and one of the easiest to get subtly wrong. Undersize and you pay forever in pumping energy and noise; oversize and you waste steel and risk settling in two-phase lines. This guide walks through the method the right way — requirements in, diameter out.

The golden rule: size, don't rate

A sizing calculation should take your process conditions (flow, fluid properties, limits) as inputs and return the pipe diameter as the output. If a "calculator" asks you to input the diameter first, it's a rating tool wearing a sizing hat. Decide the size from the constraints, then verify.

Step 1 — Gather your inputs

• Flow rate (volumetric or mass)
• Fluid density and viscosity at operating conditions
• Service type — continuous or intermittent (sets the erosional constant)
• Your design limits: maximum velocity and maximum allowable pressure drop per 100 m

Step 2 — Set the velocity limit

Two limits apply. First, a practical maximum from guidance (e.g. Crane suggests ≤ 3 m/s for typical liquid lines). Second, the erosional velocity from API RP 14E:

Ve = C / √ρ

where C ≈ 100 for continuous service and 125 for intermittent, and ρ is density. The governing velocity limit is the lower of the two.

Step 3 — Compute pressure drop for candidate sizes

For each standard pipe size, work out the velocity (v = Q/A), the Reynolds number, and the Darcy friction factor from Colebrook-White. Then the pressure drop:

ΔP = f · (L/D) · (ρ·v²/2)

Normalise to ΔP per 100 m so you can compare against your allowable gradient.

Step 4 — Pick the smallest size that passes both limits

Start small and step up. The first standard size where velocity ≤ your limit and ΔP/100 m ≤ your allowable is your answer. Note which limit governs — it tells you whether you're velocity-bound or pressure-bound.

Step 5 — Verify and document

Record the governing constraint, the final velocity and ΔP, and the assumptions (roughness, service factor). Future-you — and the reviewer — will thank you.