8 Charts for Chilled Water Pipe Sizing | Augmintech

An undersized chilled water pipe does not fail. It just quietly costs money. Velocity climbs, friction loss rises with roughly the square of it, the pump works harder every hour of every year, and somewhere on the top floor a branch never gets its design flow. All of it is decided by a number chosen in an afternoon at design stage.

This guide puts the sizing criteria for steel, copper and cast iron in one place, explains where the numbers come from, and gives you a live chart and calculator so you can size a line rather than hunt for a row in a PDF.

TL;DR

The Two Sizing Criteria

Pipe sizing looks like a lookup and is actually a comparison. You apply two independent limits to the same pipe and take the more restrictive answer. For context on the wider system these pipes serve, see our guide to HVAC chilled water systems.

Criterion 1: Pressure Drop

Friction loss per unit length, because that is what the pump pays for continuously. Design practice is normally expressed as feet of water per 100 feet of pipe, and common values sit between 1 and 4.

Criterion 2: Velocity

Velocity limits exist at both ends, for different reasons.

• Minimum, around 0.6 to 1 m/s. Below this, entrained air is not swept along to the vent points and can collect at high points, forming an air lock that blocks flow entirely.
• Maximum, around 3 m/s for steel. Above this, noise becomes noticeable in occupied areas and erosion accelerates. The Carrier Systems Design criterion of 10 ft/s, which is 3.05 m/s, is the widely used ceiling.
• Copper is held lower, commonly 1.2 to 2.4 m/s, because copper suffers erosion corrosion at elevated velocity. The damage concentrates on the inside of elbows and at fitting entries where flow is turbulent.

The Chilled Water Pipe Sizing Chart

This is the friction chart in its usual form: flow rate along the bottom, pressure gradient up the side, one curve per pipe size, both axes logarithmic. Constant velocity lines run diagonally across it, and the shaded band marks the normal design zone.

How to Read the Chart in Three Steps

1. Find your flow rate on the bottom axis

   Read across the logarithmic scale in litres per second. Convert from m³/h by dividing by 3.6, or from US GPM by dividing by 15.85.

2. Go up to the design band

   Travel vertically until you reach the shaded 200 to 400 Pa/m zone. That is the normal target for commercial chilled water.

3. Read the pipe size curve you land on, then check velocity

   The curve passing through that region gives your size. Now check where the point sits relative to the diagonal velocity lines. If it is above the 3 m/s line, the velocity limit is binding and you must go up a size regardless of what the friction reading says.

Live Pipe Sizing Calculator

Enter a flow rate and the calculator applies both criteria, tells you which one governs, and marks your point on the chart above.

Steel Pipe, Schedule 40

Black steel to ASTM A53 or A106 in Schedule 40 is the default for chilled water mains and risers. It is economical at larger sizes, mechanically strong, available in every diameter, and welds or grooves readily.

The table below gives maximum flow for each size under both criteria at 4 ft per 100 ft and 3.05 m/s, with the governing one highlighted.

Amber rows are where the velocity limit governs rather than friction. Note the crossover at DN200: below it, friction is the binding constraint; above it, velocity is. Bores are nominal Schedule 40 internal diameters; confirm against the actual specification, since a heavier schedule reduces bore and pressure drop varies with roughly the fifth power of diameter.

Steel or Copper?

• Cost crossover. Copper is competitive at small sizes where labour dominates and jointing is quick. Steel becomes markedly cheaper as diameter grows, which is why you rarely see copper above about DN100 on chilled water.
• Size range. Steel covers the full range up to and beyond DN600. Copper availability and cost effectively cap it well below that.
• Corrosion. A closed chilled water loop with proper water treatment protects steel well, because the oxygen is consumed early and not replenished. Where treatment is unreliable, copper's inherent corrosion resistance is worth paying for. In coastal and humid GCC and Indian locations, the more likely corrosion problem is external, under damaged insulation, which affects both materials.
• Velocity headroom. Steel tolerates higher velocity than copper, which matters on space constrained risers.

Copper Pipe, Type L

Type L is the usual grade for chilled water and building services, heavier walled than Type M and lighter than Type K. Its lower velocity ceiling is the key sizing difference against steel.

Note that copper's velocity crossover arrives one size earlier than steel, at 6 in rather than 8 in, purely because of the lower velocity ceiling. Copper is hydraulically smoother than steel, so at the same bore it carries slightly more flow for the same friction, but the velocity limit takes that advantage back at larger sizes.

Cast Iron Pipe Data

Cast iron appears in this article's outline, and the most useful thing to say about it is a clarification: cast iron is not used for pressurised chilled water distribution. It belongs to a different part of the building.

Where you will meet it:

• Gravity drainage. Soil and waste stacks, particularly where acoustic performance matters, because cast iron is significantly quieter than plastic for falling water in a stack adjacent to occupied space.
• AHU condensate drainage in some specifications, for the same durability and acoustic reasons.
• Older retrofit systems, where existing buried or riser pipework may be cast iron and has to be surveyed, assessed and either retained or replaced.
• Buried municipal and fire mains, usually ductile iron rather than grey cast iron in modern work.

The Hazen-Williams Formula

For the Darcy-Weisbach approach used in detailed hydraulic design, see our companion guide on pressure drop piping calculations — it covers friction factor, Reynolds number, and equivalent length methods for fittings.

Hazen-Williams is an empirical formula from 1902 that estimates friction loss for water without needing viscosity, Reynolds number or a friction factor. That simplicity is why it survives in fire protection, municipal water and irrigation work.

Worked Example

1. State the duty

   42 L/s of chilled water through 100 m of DN100 Schedule 40 steel, bore 102.3 mm, taking C = 120 for design.

2. Convert to SI base units
   Q=0.042m³/sD=0.1023m
3. Substitute
   h_f=10.67×100×0.042^1.852120^1.852×0.1023^4.87
   h_f = 28.2 m over 100 m, or 2,763 Pa/m
4. Cross-check against Darcy-Weisbach

   The same duty solved properly, with a Colebrook-White friction factor at 7°C, gives 2,247 Pa/m. Hazen-Williams at C = 120 is 23% high here. That is conservative, so it will not undersize the pipe, but it will oversize the pump if carried through to a head calculation unchecked.

The practical conclusion: the C value you choose matters more than the choice of formula. At C = 130 the two methods agree within about 10%. At C = 100 they diverge by three quarters. If you use Hazen-Williams, be deliberate about C and know which direction the error runs.

Dynamic Losses Through Fittings and Valves

Straight pipe friction is only part of the total. Every bend, tee, valve and transition adds a dynamic loss, and in a compact plant room those losses commonly exceed the straight pipe entirely.

Fluid Properties That Change the Answer

Every table above assumes water. Three properties decide how far reality departs from that.

Glycol figures vary with concentration and supplier. Always use the manufacturer's published property data for the actual mixture rather than a generic value.

How Each Property Bites

• Viscosity raises friction loss directly, through a higher friction factor at the same velocity. Chilled water at 7°C is already about 40% more viscous than the 20°C reference most published tables assume, which is why this article's tables are computed at 7°C rather than 20°C.
• Density raises both the pressure loss for a given velocity and the mass the pump must move. Glycol at roughly 1045 kg/m³ is about 4.5% denser than water.
• Specific heat is the one people forget. Glycol carries less heat per kilogram, so you need more volumetric flow for the same cooling duty at the same delta T. That larger flow then has to pass through a pipe that is already more resistive because of the viscosity. The two effects compound.

Worked Example: Primary Secondary Network

A 250 TR plant, two chillers, serving four floors. Design delta T is 5°C, so total flow is 42 L/s. Here is where each sizing decision applies.

A Note on Insulation

Pipe sizing and insulation are specified together, because the insulated outside diameter is what actually has to fit through the building.

• Insulation thickness scales with pipe size, so a larger pipe costs void space twice over: once for the bore and again for the thicker jacket around it. Coordinate the insulated OD, not the nominal size.
• The vapour barrier matters more than the thickness on chilled water in humid climates. A cold surface below the ambient dew point will condense moisture, and if the barrier is broken that moisture reaches the pipe wall inside the insulation, where it corrodes steel from the outside in, invisibly, for years.
• Seal at every interruption. Supports, valves, flanges and hanger penetrations are where vapour barriers fail. A perfectly insulated straight run with an unsealed valve box is not insulated.
• External corrosion under insulation is the dominant failure mode for chilled water pipework in coastal Indian and GCC locations, more than internal corrosion in a properly treated closed loop.

Where This Skill Leads

Pipe sizing is one of the first genuinely engineering tasks a junior MEP engineer is handed, and it is a good test of whether someone understands a system or only a table. Our dedicated pipe sizing calculation guide covers the full step-by-step method beyond the chart lookup. Reading a chart is straightforward. Knowing which criterion is binding, what a C value assumption costs you, why the fittings on a short plant room run outweigh a long straight riser, and how glycol changes all of it, is the part that turns into a defensible design.

Frequently Asked Questions

Sources and Further Reading

• ASHRAE Handbook: Fundamentals, Pipe Sizing and Fluid Flow chapters, ASHRAE. Source of the friction chart basis, fitting loss data and fluid property tables for water and glycol solutions.
• ASPE Data Book, American Society of Plumbing Engineers. Velocity and pressure drop criteria for copper and steel water piping.
• Carrier System Design Manual, Part 3, Piping Design. Source of the widely used 10 ft/s velocity ceiling and the ft per 100 ft pressure drop criteria.
• Crane Technical Paper No. 410, Flow of Fluids Through Valves, Fittings and Pipe. The standard industry reference for loss coefficients, valve data and worked hydraulic examples.
• ASME B36.10M, Welded and Seamless Wrought Steel Pipe, for Schedule 40 dimensional data, and ASTM B88 for copper water tube Types K, L and M.
• Copper Development Association guidance on velocity limits and erosion corrosion in copper tube systems.

Standards are revised on fixed cycles. Confirm the edition adopted by your project before issuing a specification. This article was last verified against the sources above on 1 August 2026.