EPANET alternatives: the pipe-network tools, compared
EPANET is the US EPA's free, public-domain water-distribution modeller. It solves steady-state and extended-period hydraulics, tracks water quality and age, and has been the reference implementation for pipe-network analysis for three decades. If you are reading a page about alternatives to it, you have almost certainly hit one of its edges rather than a fault: it is a Windows desktop program, it models water, and its interface was built for municipal distribution modelling rather than for the plant, process or building-services network on your desk this afternoon.
This guide covers the tools engineers actually move to, what each is genuinely for, and where each one stops. It is written by the people who build one of them, Fluid Network Studio, so read our own entry with that in mind. Every claim about another product is drawn from that vendor's own public material, and pricing moves, so check the vendor before you buy anything.
First, decide what you are actually short of
Most "EPANET alternative" searches come down to one of five gaps. Naming yours narrows the field fast.
Access. You need it on macOS or Linux, on a locked-down work machine, or on site from a tablet. This rules out most of the field, which is Windows desktop software.
Fluids other than water. Compressed air, process gas, a hot or cold line where temperature changes along the run, a sludge or a polymer solution. EPANET models incompressible water only, and this is the single most common reason people leave it.
Time behaviour. Tanks filling and draining, pumps cycling on level, valves on a schedule. EPANET does this well, so if it is all you need, the honest answer is to stay.
Fast transients. Water hammer, surge, a pump trip, a valve slam. Neither EPANET nor any quasi-steady tool models fluid inertia. You need a dedicated surge package, and no amount of extended-period simulation substitutes for one.
Proof you can hand to someone. A report that names its methods, states its assumptions and shows the results were checked against something independent.
The tools
EPANET itself
Free, public domain, and still the right answer for a large class of work: municipal water distribution, water-quality and age tracking, chlorine decay, and long extended-period runs on big models. It is a Windows program, it models incompressible water, and it does not do gas, heat transfer or non-Newtonian fluids. Its solver is the Global Gradient Algorithm of Todini and Pilati, which is also what most of the field, including us, uses. If your only complaint is the interface, look at the EPANET-based front ends before you leave the ecosystem entirely.
Fluid Network Studio
Ours. A browser-based network solver: draw the network in a tab, press solve, read flows, pressures, velocities and temperatures. It covers incompressible liquids, compressible gas, liquids and gases with heat transfer through pipe walls and insulation, and homogeneous non-Newtonian fluids (power-law and Bingham plastic). Extended-period simulation, with tanks, valve schedules and pump control, is included on every plan rather than sold as a module. Water and liquid networks are free with no account needed to solve; gas, heat, non-Newtonian flow and the CFD-coupling export are on Advanced at A$39 a month.
It reads and writes EPANET .inp files, so an existing model comes across rather than being retyped. Methods are named openly - Global Gradient Algorithm, Darcy-Weisbach with the Churchill friction factor, Crane TP-410 fitting coefficients, least-squares pump curves with affinity scaling - and results are checked against an independent reference problem set that includes cross-checks against OWA-EPANET 2.2 and NASA's GFSSP, with mass and energy conservation residuals reported on every solve.
What it will not do: water hammer, surge or any fast transient, because its time simulation is quasi-steady and carries no fluid inertia; settling or heterogeneous slurries; water-quality, age or tracer tracking; and relief-valve sizing to a code. Those are boundaries, not roadmap items.
Pipe Flow Expert
Windows desktop software with openly published annual licences, from a small Lite edition up to full and multi-user versions, with lower renewal fees after the first year. In spirit it is the closest desktop tool to what a practising engineer wants day to day: focused network sizing, no municipal-modelling scaffolding. If you want desktop software with a price on the website and no sales call, start here.
AFT Fathom
A deep incompressible pipe-flow suite from Applied Flow Technology, now part of Datacor, widely used in process and power engineering. Pricing is by quote through the vendor and its resellers. Its stablemates cover compressible flow and surge, so the family spans more than Fathom alone does. It is a serious engineering package with a learning curve and a licence to match, and it is a sound choice when pipe-flow analysis is a large part of the job rather than an occasional task.
PIPENET
From Sunrise Systems, the specialist for oil and gas, power and process plants. Its Transient module does surge and water hammer properly, and its Spray/Sprinkler module targets NFPA fire-protection design. Pricing is discussed with the vendor. If your problem is a surge study or a code-compliant fire system, this class of tool is where that work belongs, and nothing in the browser-based category replaces it.
PipelineStudio
Emerson's tool for long-distance gas and liquid transmission pipelines, covering steady-state and transient analysis, priced by quote. This is transmission engineering rather than facility or utility networks. If you are sizing a plant, a compressed-air ring main or a pumped liquid network, it is more tool than the job needs; if you are running a transmission line with transients, it is the right one.
Bentley OpenFlows WaterCAD and WaterGEMS
The municipal water-distribution standard, sold as annual practitioner licences through Bentley's Virtuosity store with pricing tiered by pipe count. WaterGEMS adds GIS and CAD integration and model-management tooling on top of WaterCAD. For thousands of nodes, water-quality tracking, fire-flow studies and a model that lives for years, this is the category EPANET users step up into.
Which one for which job
| If the job is | The right home |
|---|---|
| Municipal water distribution with quality tracking | EPANET, or WaterCAD/WaterGEMS at scale |
| Everyday sizing, balancing and pump-versus-system checks | Fluid Network Studio, or Pipe Flow Expert on the desktop |
| Compressed air, process gas or a fan and duct system | Fluid Network Studio, or an AFT compressible product |
| Heat loss or gain along a line, insulation, delivered temperature | Fluid Network Studio |
| Sludges, polymer solutions and other homogeneous non-Newtonian fluids | Fluid Network Studio |
| Water hammer, surge or a pump-trip pressure wave | PIPENET Transient, or another dedicated surge package |
| Code-compliant fire-protection design | PIPENET Spray/Sprinkler, or a fire-specific tool |
| Long-distance transmission pipelines | PipelineStudio, or a pipeline-specific suite |
| Settling slurries and deposition | A specialist slurry package |
If you are leaving EPANET, what comes with you
An .inp file is the closest thing this field has to an interchange format, and it travels better than people expect. Importing one into Fluid Network Studio brings across the junctions, reservoirs, tanks, pipes, pumps and valves, their geometry and coordinates, and the unit system the file was written in. Control valves keep their control law rather than degrading to plain throttles.
What does not come across, and we would rather say so before you try it: [QUALITY] and [REACTIONS], because water-quality modelling is out of scope here; demand patterns, controls and rules, which are not evaluated on import, so a patterned model arrives as a steady one; and pressure-dependent emitters. The import reports every element it skipped and why, so you get a list rather than a silent difference. Export runs the other way too, so a network built here can go back to EPANET for a water-quality run.
Frequently asked questions
What is the best free alternative to EPANET?
For water networks in a browser, Fluid Network Studio is free with no account needed to solve, and it cross-checks its incompressible results against OWA-EPANET 2.2. For water-quality and age tracking, nothing free matches EPANET itself, which is why we do not claim to replace it. Most engineers who switch do so because they need gas, heat transfer or non-Newtonian fluids, or because they are not on Windows.
Is there an EPANET alternative that runs in the browser?
Yes. Fluid Network Studio runs in any modern browser on any operating system, with nothing to install and no licence server. That also means it works on a locked-down work machine and on a tablet on site.
Can I open my EPANET .inp file in something else?
Yes. Fluid Network Studio imports EPANET .inp files, keeping the network, its geometry and its control-valve behaviour, and exports back to .inp. Water-quality sections, demand patterns and rule-based controls are not carried across, and the import lists exactly what it skipped.
Does any EPANET alternative handle gas or heat transfer?
EPANET models incompressible water only. Fluid Network Studio solves compressible gas in absolute pressure with mass-flow balances, and models heat loss or gain through pipe walls and insulation with named film correlations. In the desktop category, the AFT family and PIPENET both cover compressible flow, and PipelineStudio covers gas transmission.
What can EPANET do that a browser tool cannot?
Water-quality, age and tracer tracking, and very large municipal models that live for years and are managed as an asset. Those are genuine strengths, and a focused browser solver is not the right home for them.
Which alternatives model water hammer?
Of the tools here, PIPENET Transient and PipelineStudio do transient analysis; dedicated surge packages exist for exactly this. Neither EPANET nor Fluid Network Studio models fluid inertia, so neither can size a surge vessel or predict a valve-slam pressure. Extended-period simulation is not the same thing, and no tool should let you believe it is.
How to choose without wasting a fortnight
Take a network you already know the answer to - one you have hand-calculated, or one from a textbook, or an existing model you trust - and build it in whatever you are evaluating. If the tool cannot reproduce a result you already trust, nothing else about it matters. If it can, then judge it on the things you will live with daily: how long it takes to get from a question to an answer, whether it tells you what it did, and whether the output is something you would put your name on.
That last part is why our verification cases are public and solved live rather than quoted from a document. See how it works for the methods end to end, the verification cases for the numbers, and the full comparison against desktop packages for where each one wins.
Launch the Studio - free for water and liquid networks, with 20 worked examples built in.