How to size a gas distribution network

From customer demands to commercial diameter selection: the complete calculation path, with the flow law, the per-class checks and a numerical example worked to the last digit.

1. The input data

Sizing a gas distribution network starts from four families of data:

2. The flow law

The heart of the calculation is the relation between flow and pressure drop for each pipe. The general form, valid from low pressure to transmission mains, is the steady isothermal flow equation for compressible gas:

P₁² − P₂² = R(L, D, s, T) · f · z · Q²

where R is the pipe's base resistance (length L, internal diameter D — which weighs with the fifth power —, relative density s, temperature T), f is the Darcy friction factor from the Colebrook-White equation on the pipe's real roughness, z the gas compressibility and Q the standard flow. Italian and Spanish practice long used the Renouard formula, an empirical approximation valid in a precise domain; the general law carries no such constraint. Why pressure appears «squared» is explained in the guide on pressure drop.

On a branched network the pipe flows follow from the nodal balance; on a meshed network the split must be solved iteratively (Newton-Raphson nodal method) because each pipe's friction depends on the flow it carries.

3. The checks

A sizing is acceptable when every pipe and every customer respect two families of limits, tied to the pipeline class:

Maximum velocity

Italian distribution practice (UNI 9165 tradition) adopts a ladder by operating pressure band — details in the guide on gas velocity:

Operating pressure (gauge)Maximum velocity
above 3.5 bar25 m/s
1.5 – 3.5 bar20 m/s
1.0 – 1.5 bar15 m/s
0.04 – 1.0 bar10 m/s
up to 0.04 bar (low pressure)5 m/s

Minimum pressure at the customers

Every delivery point must receive the gas above a minimum pressure that depends on the network's class: in low pressure (7th class) Italian practice requires at least 18 mbar at delivery, because the service regulator and the appliances downstream still need to work.

4. Choosing the diameters

The criterion is economic and regulatory at once: for each pipe, choose the smallest commercial diameter that passes every check. «Commercial» is the key word: not a theoretical diameter, but a catalogue pipe (PE 100 SDR 11, steel…) with its real internal diameter and its roughness. The procedure is iterative by nature: changing one diameter changes the drops, which change the pressures, which can flip the checks on other pipes — by hand you proceed by trial, while an automatic solver repeats the calculate → check → substitute cycle to convergence.

5. A worked example

The simplest possible case, with real numbers (it is Case 1 of the validation dossier, developed digit by digit):

Data: source at 25 mbar gauge; one customer drawing 30 Sm³/h; a 100 m pipe in PE 100 DE63 SDR 11 (internal diameter 51.4 mm, roughness 0.007 mm); typical grid natural gas (relative density 0.591).

  1. Reynolds number: Re ≈ 13,568 → turbulent flow.
  2. Colebrook-White friction at convergence: f = 0.0288.
  3. Drop: P₁² − P₂² = R·f·z·Q² → 3.19 mbar lost along the pipe.
  4. Pressure at the customer: 25 − 3.19 = 21.81 mbar → above the required 18 mbar ✓
  5. Velocity in the pipe: 3.92 m/s → below the 5 m/s low-pressure limit ✓

The DE63 passes. Had the velocity exceeded 5 m/s, or the customer dropped below 18 mbar, the next catalogue diameter (DE75) would be tried and the calculation repeated.

6. From the calculation to the report

Sizing ends with the calculation report: input data, methodology with references, full results, check outcomes and the bill of quantities. Gasnetics automates the whole cycle — map drawing, calculation, per-class checks, diameter selection from the catalogue and the PDF report — and every report includes an appendix that recomputes the critical path by hand, so every number stays verifiable.

Calculation results, whatever the software, support the design work: the design choice and the final check remain the professional's responsibility.

Frequently asked questions

What data is needed to size a gas network?

Four families of data: the customers' demands (in Sm³/h, or in kW/kg/h to convert), the supply pressure of the source, the layout with pipe lengths and ground elevations, and the distributed gas (the mixture's composition determines density, viscosity and compressibility).

What criterion selects the diameters?

For each pipe you pick the smallest commercial diameter from the catalogue (PE, steel) that passes every check: maximum velocity for the pressure band and minimum pressure to guarantee at the customers per pipeline class. A larger diameter than necessary is a cost; a smaller one is a non-conformity.

Why does the calculation use squared pressure?

Because gas is compressible: density changes with pressure along the pipe, and integrating the flow law the drop is naturally expressed as the difference of the squared absolute pressures (P₁² − P₂²), valid from low pressure to transmission mains.

Size your network with Gasnetics — free beta