Engineering context

Compressible gas networks behave differently from liquid systems because gas density changes continuously with pressure and temperature. As gas flows and pressure drops, density falls, velocity rises, and the actual volumetric flow increases — only mass flow (and standard volumetric flow) is conserved at junctions. Because of this coupling, results must be read across the whole connected network, and gas flow is always solved with real-gas equation-of-state methods rather than simplified ideal-gas, purely isothermal, or purely adiabatic assumptions.

A practical workflow starts by defining the analysis objective and operating cases — whether the question is capacity (how much the network can deliver) or distribution (how a set throughput splits across branches) — which sets the inlet boundary condition (Known Pressure for capacity, Known Flow for distribution), taking care not to over-constrain the model where choking can occur. The gas is selected from FluidFlow’s fluid database, where its properties are modelled and calculated using an equation of state (such as Peng-Robinson, Lee-Kesler, or BWRHS); operating conditions are set at each boundary (inlet pressure and temperature or flow and temperature; outlet pressure or flow), choosing the pressure model correctly — stagnation for vessels, static for pipes or wall measurements. After building and connecting the network and setting one standard-volume reference (STP or NTP) for volumetric-flow inputs, the model is inspected for choking issues (gas models can fail to fully converge where multiple chokes appear; review Mach numbers and choked-flow warnings for endpoint, restriction, or expansion choking), then analysed as a normal pipe system for pressure drop, velocity and noise, flow distribution, and along-pipe property behaviour.

Engineering workflow

1. Define the analysis objective and operating cases — decide whether you are asking a capacity question (how much the network can deliver) or a distribution question (how a set throughput splits across branches), and list the operating cases.
2. Choose the inlet boundary condition to match the analysis objective — Known Pressure for capacity, Known Flow for distribution. Where choking can occur, avoid over-constraining the model (for example, demanding a fixed flow through a restriction that would choke).
3. Define the boundary conditions — add the boundaries, then select the gas from FluidFlow’s fluid database, where its properties are modelled and calculated using an equation of state (such as Peng-Robinson, Lee-Kesler, or BWRHS); build a mixture with the database’s mixture function where needed. Set the operating conditions at each boundary: at the inlet, pressure and temperature or flow and temperature; at the outlet, pressure or flow. When defining a pressure, choose the pressure model correctly — stagnation for vessels, and static for pipes or where the measurement is taken at the pipe wall.
4. Build and connect the network — add the equipment (compressors, blowers, fans, valves, restrictions, and fittings) and connect it with pipes to establish topology.
5. Set the calculation basis — choose one standard-volume reference (STP or NTP) and keep it consistent across the model; this reference is the basis on which gas volumetric flows are defined as input data (the active reference shows on the status bar).
6. Inspect the model for choking issues — gas models can fail to fully converge, particularly where multiple chokes appear in the network. After solving, review Mach numbers and any choked-flow warnings (endpoint, restriction, or expansion choking) to locate where sonic conditions are being reached, and confirm whether any convergence problems trace back to choking.
7. Analyse the model as a normal pipe system — once choking is understood, work through the usual pipe-system review: pressure drop, velocity (including noise-velocity screening using the 60 m/s rule of thumb or the density-based limit), flow distribution across branches, and along-pipe property behaviour (pressure, temperature, and density fall while velocity rises; mass balance holds at junctions).

How FluidFlow helps

FluidFlow analyses compressible gas networks within its steady-state pipe network solver, always using equation-of-state property methods and the Duxbury method to capture changing density, velocity, and friction along each pipe. It reports mass, actual, and standard volumetric flow, along-pipe properties, Mach numbers, and choked-flow warnings, and scales from single lines to branched headers and parallel paths. It is not CFD and does not perform transient or surge analysis.

Go deeper