Long pipe runs, thermal expansion and the movement you must allow
How far a run grows, why plastics move far more than steel, and when expansion loops stop being optional.
The engineering question
A long pipe run is anchored at both ends. What movement must be allowed, and at what point does the restraint itself become the problem?
Worked example
The figures below were produced by running the Pipe Thermal Expansion Calculator with the same inputs — during the build, from its authoritative engine.
| Expansion over the run | 21.78 mm |
|---|---|
| Expansion per 100 m | 72.6 mm/100 m |
| Material | Carbon steel |
| Compensation needed | Large expansion — provide expansion loops, bellows, or guide/slide anchors; PVC/HDPE need allowance for creep and UV. |
| Assessment | 30 m of Carbon steel grows 21.8 mm for a 60 °C rise (≈ 72.6 mm per 100 m). Large expansion — provide expansion loops, bellows, or guide/slide anchors; PVC/HDPE need allowance for creep and UV. |
- Movement is the coefficient of expansion times length times temperature rise.
- Steel moves roughly 12 mm per 100 m per 10 °C; uPVC moves several times that.
- Beyond about 10 mm of movement at a terminal, loops or bellows are needed.
Sanity check
If a plastic line is being restrained like a steel one, the material selection is wrong before the expansion design is.
Tools for this problem
These are the EngiMetric calculators that compute the numbers in this topic. Each one shows its formula, its units and its limits rather than returning a bare result.
Engineering review required
This topic supports professional engineering work; it does not replace a sealed design. The result depends on site data, the governing standard and engineering judgement, and should be confirmed by a competent engineer before it is built.
References & standards
- ASME B31.3 — piping flexibility
- EN 13445 — pressure piping systems
- ASME PCC-2 — support design
Related problems
Where the pressure in a pipe line actually goes
Separating frictional loss from fitting loss, and why small bore runs lose far more than intuition suggests.
Sizing a heat exchanger from the log mean temperature difference
Why the arithmetic temperature difference gives the wrong area, and how the log mean fixes it.
Also: Chemical & Process