Which Pipe Material Should You Use? PVC, Copper, Steel & PEX Through the Numbers

PVC, copper, steel, and PEX compared the way the equations see them — friction C-factors, inside diameters, thermal movement, and surge stiffness.


Updated August 20, 2026

The four numbers that separate pipe materials

Strip away the brand names and the aisle signage and a pipe material is, to the equations that size a water system, four numbers: a Hazen-Williams C-factor that sets friction loss, an inside diameter that the nominal size only hints at, a thermal expansion coefficient that decides how much the run grows when it warms, and an elastic modulus that controls how hard the pipe fights back during a pressure surge. PVC, copper, steel, and PEX each land in a different corner of that four-dimensional space, which is why no single material wins every job. Temperature ratings draw a first hard line before any arithmetic: rigid PVC is typically rated to about 60 °C (140 °F), while CPVC reaches roughly 93 °C and PEX about 95 °C — so ordinary PVC stays out of hot-water service entirely.

To keep the comparison honest, this guide threads one running example through every section: a 100-ft (30.5 m), nominal 1-inch supply run carrying 8 gpm (0.505 L/s), priced out in three materials — PVC Schedule 40, copper Type L, and galvanized steel Schedule 40 — with the thermal and surge behavior of the same run examined along the way. By the end, the same pipe run will have been worked through friction, geometry, expansion, and hammer, and the pattern of trade-offs becomes hard to unsee.

Friction: what a C-factor of 150 vs 100 costs you

The Hazen-Williams equation, h_f = 10.67·L·Q^1.852/(C^1.852·D^4.87), condenses a material’s wall smoothness into the single coefficient C — higher is smoother. The verified new-pipe values: PVC and other smooth plastics (PEX included) run about 150, copper 140, galvanized steel 120, and old unlined cast iron falls to about 100. Because C enters at the 1.852 power, the ratio matters more than the difference: swapping C 150 for C 100 multiplies friction loss by (150/100)^1.852 ≈ 2.12 — the aged main loses more than twice the head of the plastic line at identical flow and bore.

The second friction lesson is about time, not catalogs. Plastic and copper hold their smoothness for decades, but unlined iron and steel roughen as they age — the C = 100 figure represents old tuberculated cast iron, not a factory product. Designers of long-lived systems therefore often run the numbers with an aged C rather than the new-pipe value, effectively pricing in the material’s future. A material choice is also a maintenance forecast: the pipe you install at C 120 may be a C 100 pipe by the time the mortgage is paid.

Nominal size is not inside diameter

Every material names its sizes differently, and none of the names is a bore measurement. Steel and PVC share the iron-pipe-size system of ASME B36.10M and ASTM D1785: a fixed outside diameter per nominal size (1-inch pipe is always 1.315 in OD) with the schedule’s wall thickness eating inward, so 1-inch Schedule 40 measures 1.049 in inside and Schedule 80 less still. Copper tube plays by its own rules — the OD is always the nominal size plus 1/8 in, and the type letter sets the wall, K heaviest through M lightest — so 1-inch copper Type L (1.125 in OD, 0.050 in wall) has a 1.025 in bore. PEX is a third system again: it follows neither table, and its inside diameters come from the manufacturer’s own datasheet.

These fractions of an inch are not pedantry, because the diameter enters the friction equation at the 4.87 power: cutting the bore by 10% raises head loss by about 59%. The bore also sets what the pipe holds — 100 ft of 1-inch Schedule 40 contains 4.49 gallons — which matters for drain-downs, dosing, and glycol fills. In the running example, copper’s 1.025 in bore against PVC’s 1.049 in is a 2.3% deficit that, compounded with C 140 versus 150, will cost copper about 27% more friction on the same run. Same shelf label, different pipe.

Thermal movement: why plastic runs need slack

Warm a pipe and it grows by ΔL = α·L·ΔT, and the coefficient α is where materials diverge dramatically: carbon steel runs about 12×10⁻⁶ per °C, copper 17×10⁻⁶, and PVC around 70×10⁻⁶ — nearly six times steel and four times copper. Put the running example through a 40 °C (72 °F) warm-up, installation temperature to hot service: the 100-ft steel run grows 0.58 in (14.6 mm), copper 0.82 in (20.7 mm), and PVC a full 3.36 in (85.3 mm). Three-plus inches of movement has to go somewhere, which is why plastic runs are installed with slack, offsets, and room to snake, while short metal runs often get away with none.

Restrain the pipe instead of letting it move and the growth converts to axial stress, σ = E·α·ΔT — and here plastic’s floppiness becomes a mercy. For that same 40 °C rise, fully anchored steel develops about 96 MPa of axial stress (E = 200,000 MPa), copper about 80 MPa, but PVC only about 8 MPa, because its modulus of roughly 3,000 MPa is tiny. Metal runs shrug off the movement per foot but punish rigid anchoring with enormous thrust; plastic moves far more but pushes gently. Either way the design answer is the same discipline: give the run a place to flex — a loop, an offset, an elbow with travel — sized to the growth you just calculated.

Water hammer: stiffness cuts both ways

Slam a valve shut and the moving water column piles into a pressure wave whose severity depends on how fast that wave travels — and the pipe wall is part of the spring. The Korteweg celerity a = sqrt((K/ρ)/(1 + (K/E)(D/e))) says a stretchy wall slows the wave: in the classic textbook geometry (0.5 m bore, 10 mm wall, water), a steel line (E = 200 GPa) carries the wave at about 1,191 m/s, while the same geometry in rigid PVC (E = 3 GPa) drops it to roughly 242 m/s. Since the instantaneous-closure surge is the Joukowsky head ΔH = a·v/g, stopping a 2 m/s flow dead produces about 243 m of surge head (~345 psi) in the steel line but only about 49 m (~70 psi) in the PVC one — a nearly five-fold difference from wall stiffness alone.

That is the both-ways cut: the flexible plastics that demand expansion slack repay you with inherently gentler transients, while stiff metal lines transmit brutal surges but barely notice thermal anchoring. Material choice is only half the hammer story, though — closure time is the other half. If the valve closes slower than the critical period 2L/a, the reflected relief wave arrives in time to shave the peak roughly in proportion to (2L/a)/t_c, which is why slow-closing valves, surge vessels, and gentle pump ramps are fixes that work in any material. The 100-ft residential example rarely sees textbook surges, but the physics explains the banging pipes behind fast solenoid valves on washing machines.

A 100-ft supply run, worked in three materials

Now the full friction bill for the running example — 100 ft, nominal 1-inch, 8 gpm (0.505 L/s), a mean velocity near 3.0 ft/s (0.90 m/s), comfortably inside Hazen-Williams’ calibrated range. PVC Schedule 40 (C = 150, ID 1.049 in): about 3.6 ft of head loss (1.11 m), or 1.6 psi. Copper Type L (C = 140, ID 1.025 in): about 4.6 ft (1.41 m), or 2.0 psi — the 27% penalty promised earlier, part smaller bore, part lower C. Galvanized steel Schedule 40 (C = 120, ID 1.049 in): about 5.5 ft (1.67 m), or 2.4 psi — same bore as the PVC, so this gap is pure surface roughness, (150/120)^1.852 ≈ 1.51. And if that steel ages toward C = 100, the run drifts to roughly 7.7 ft (3.3 psi), more than double where PVC started.

Two honest caveats before shopping. First, these are straight-pipe numbers; fittings, valves, and elevation are added separately, and a real house run has plenty of each. Second, a couple of psi on one 100-ft branch may not decide anything by itself — friction compounds across a system, and it is the cumulative gradient on long runs, well feeds, and irrigation mains where the C-factor difference turns into pump horsepower. The Hazen-Williams calculator runs this whole comparison in seconds — length, bore, flow, and a verified C table in, head loss, pressure drop, and a velocity sanity check out — so testing a fourth material or a second size costs nothing but a keystroke.

Choosing in one pass

Recap the run: 100 ft of nominal 1-inch pipe at 8 gpm. PVC gave up 1.6 psi to friction, copper 2.0, galvanized steel 2.4 — but PVC grows 3.36 in over a 40 °C rise where steel moves barely half an inch, PVC is barred from hot water by its ~60 °C rating while PEX runs to about 95 °C, and the stiff steel line would carry a water-hammer wave five times faster than the plastic one. That is the whole decision in miniature: smooth-and-flexible (PVC, PEX) wins on friction and surge but demands expansion slack and respect for temperature limits; stiff-and-tough (steel, copper) wins on temperature and rigidity but pays in friction — increasingly so as unlined steel ages — and transmits harder transients. Bore honesty underlies it all: compare actual inside diameters, never nominal names. For the spec-sheet version of this decision — pressure classes, dimensions, and use-case verdicts side by side — a dedicated PEX vs copper vs CPVC comparison page is the natural next read; this guide’s job was the physics behind the columns.

Frequently Asked Questions

Why does copper pipe have different dimensions from steel and PVC of the same size?

They are different sizing systems. Steel and PVC share iron-pipe-size outside diameters (a 1-inch pipe is 1.315 in OD) with schedule walls, per ASME B36.10M and ASTM D1785. Copper water tube instead fixes its OD at the nominal size plus 1/8 in — 1.125 in for 1-inch — with the wall set by the type letter, K heaviest through M lightest. The result: 1-inch copper Type L has a 1.025 in bore versus 1.049 in for 1-inch Schedule 40, and fittings are never interchangeable between the two systems.

How much more does plastic pipe expand than metal pipe?

Roughly four to six times. PVC’s thermal expansion coefficient is about 70×10⁻⁶ per °C against copper’s 17×10⁻⁶ and carbon steel’s 12×10⁻⁶, so a 100-ft PVC run warming 40 °C grows about 3.4 inches while the same run in steel grows barely 0.6 in. The compensation is that plastic’s low elastic modulus makes its restrained stress far smaller — it moves a lot but pushes gently, whereas anchored metal barely moves yet develops enormous thrust.

Is PEX or copper quieter during water hammer?

The flexible pipe is the gentler one. Surge pressure follows the Joukowsky relation ΔP = ρ·a·Δv, and the wave speed a falls sharply as the pipe wall’s elastic modulus drops — a rigid PVC line carries the wave at a fraction of steel’s ~1,191 m/s in the same geometry, cutting the instantaneous-closure surge several-fold. Flexible plastics like PEX benefit from the same physics. Closure time still matters in every material: valves that close slower than the pipe’s critical period 2L/a shave the peak regardless of what the pipe is made of.

Does a higher Hazen-Williams C always mean the better pipe?

Only on the friction axis. C ≈ 150 plastics lose the least head, and the (C₁/C₂)^1.852 scaling makes that worth real pump energy on long runs — but the same materials carry temperature ceilings (about 60 °C for rigid PVC) and the largest thermal growth, while lower-C metals handle heat, pressure, and mechanical abuse. C also degrades with age in unlined ferrous pipe, so the honest comparison uses a design C for the system’s life, not the catalog number on day one.

Try the Calculators

Sources & Further Reading

  • Hazen-Williams C coefficients verified against published tables (Engineers Edge; CECALC): PVC 150, copper 140, galvanized steel 120, old unlined cast iron 100; SI-form equation per standard hydraulics references
  • ASME B36.10M (steel pipe schedules), ASTM D1785 (PVC Sch 40/80), and ASTM B88 (copper tube Types K/L/M) — outside diameters and walls, with inside diameters computed as ID = OD − 2·wall
  • ASME B31.1/B31.3 flexibility practice — thermal expansion coefficients and moduli (carbon steel α = 12×10⁻⁶/K, copper 17×10⁻⁶, PVC 70×10⁻⁶) and restrained-stress relation σ = E·α·ΔT
  • Joukowsky (1898) surge equation and Korteweg celerity relation, with representative pipe moduli per Streeter & Wylie, Fluid Transients, and AWWA M11
  • Pipe material service ratings (PVC ~60 °C, CPVC ~93 °C, PEX ~95 °C; Hazen-Williams C 150 for smooth plastics) per ASHRAE Handbook — Fundamentals, Crane TP-410, and manufacturer data as encoded in the pipe & duct sizing material library