Why EPANET Is the First Choice for Water Network Modeling

EPANET is the first choice for water network modeling because it is free, it comes from the U.S. Environmental Protection Agency, and it answers the questions every water system faces before pipe goes in the ground: what pressure reaches each node, how each tank fills and drains, and how old the water is at the last house. It does not design the system. It tests the design you give it.

Why do engineers choose EPANET for water network modeling?

Because it is free, it models hydraulics and water quality together over time, and anyone can open the model file.

EPA's EPANET software page calls it public domain software that can be freely copied and distributed, so the designer, the utility and the reviewer can all run the same model. EPA's account of the EPANET 2.2 release says EPA developed it in the 1990s and that an open source project begun in 2015 now carries it forward.

The EPANET 2.2 User Manual describes the .INP export as plain text, "a human readable version of your data." A reviewer can check a pump curve in a text editor and rerun your case.

What does an EPANET model tell you about a water system?

Flow in every pipe, pressure at every node, the level in every tank and the water quality throughout the network, step by step over a simulation period.

The manual calls this extended period simulation, typically in one-hour steps. Along the way, EPANET:

• computes friction loss with Hazen-Williams, Darcy-Weisbach or Chezy-Manning, plus minor losses;
• runs constant or variable speed pumps and computes their energy use;
• models pressure reducing, pressure sustaining, flow control and other valves;
• gives each node several demand categories, each with its own time pattern, and runs pumps and valves from tank level, timer or rule-based controls.

Does EPANET size your pipes and pumps?

Not by itself. EPANET checks a design; the engineer picks the sizes.

The engineer sets a diameter, runs the model, reads pressures and velocities, and changes the pipe until the results pass. Pumps work the same way: the manual's FAQ sets up a run that reports the head a pump must deliver at the required flow, and the engineer picks the pump. The same loop compares options, such as one pump station or two; the manual suggests side-by-side sessions for that.

A typical run set covers average day, maximum day, peak hour, and maximum day plus fire flow, today and at buildout.

Bigger pipe is not a free margin. EPA's fact sheet on water age management says water age can become excessive when piping or storage tanks are larger than needed, and older water brings more disinfection byproducts, lost chlorine residual and microbial growth.

How does a model show a system meets state pressure rules?

The rule sets the pressure. The model shows where the system comes closest to missing it.

In Texas, 30 TAC §290.44(d) requires at least 35 psi at all points in the distribution network at flow rates of at least 1.5 gallons per minute per connection, and 20 psi under combined fire and drinking water flow where the system is meant to fight fires. EPA's fact sheet on distribution system pressure management notes that most states use 20 psi as a lower limit, and states with a maximum usually set it between 60 and 150 psi.

For fire flow, the manual adds the fire demand to a node's normal demand and reads the pressure, or uses an emitter to read the flow a node can deliver at a set pressure. Watch for negative pressures. EPANET's default demand driven analysis, the manual warns, can meet demands at nodes with negative pressure, a physical impossibility; pressure driven analysis lets the delivered flow drop with pressure instead.

How does EPANET handle steep terrain and pressure zones?

It models every pressure zone, PRV and booster pump, so you can see where elevation pushes pressure out of range.

The manual converts pressure to head at about 2.3 feet per psi. A main that drops 230 feet gains roughly 100 psi of static pressure. Steep service areas, like Colorado's foothills, are split into pressure zones, with PRVs stepping pressure down and boosters lifting it. For distribution system design in Colorado, the model sets the zone boundaries before anyone sizes a valve.

Texas adds a rule here. Under §290.44(d)(2), service to more than one pressure plane gets its added pressure from booster pumps taking suction from storage tanks. A booster drawing straight from the mains is an exception that goes to the executive director for review, and its suction line must hold at least 20 psi at all times.

EPA's pressure fact sheet recommends monitors at the highest and lowest points in each zone; check the model against those readings. A rural water system model and fire flow memo shows these steps on a real project.

Can EPANET find water age and chlorine problems?

Yes. EPA first developed EPANET to study how water quality moves through distribution systems.

It tracks water age, follows water from each source, decays chlorine and grows disinfection byproducts, with reactions in the bulk water and at the pipe wall, and mixes each tank as a complete mix, plug flow or two-compartment reactor. Wall reaction rates cannot be measured directly, so the manual says to fit them against field data. And water age problems show up when demand is low. A resort or lake community that designs for its busy season also needs a quiet-season run to see which dead ends and tanks hold water too long. EPA's water age fact sheet lists fixes: tank mixers, looped dead ends, changed tank cycles and gentle flushing.

Can EPANET test main breaks and pump failures?

Yes, one scenario at a time. Close the pipe or shut off the pump, rerun, and read which nodes fall below the pressure floor.

For dozens of failure cases at once, EPA and Sandia National Laboratories publish WNTR, the Water Network Tool for Resilience, an EPANET compatible Python package that adds pipe leaks, repair strategies and pressure dependent demand. For utilities and municipalities, the result is a ranked list of the mains and pumps whose loss hurts most.

Where does EPANET fall short?

It is not a surge model, and it is only as good as its data.

The manual assumes all pipes are full, and each step balances flow and head with no term for pressure waves. A pump power failure or a valve closed too fast creates water hammer, which EPA's pressure fact sheet describes as alternating very high and very low pressures that can cause main breaks. A surge can last only a few seconds, far shorter than an hourly step, so it takes a separate transient analysis. And while EPANET can print a calibration report against field measurements, closing the gap between model and field is still the engineer's job.

Is your model ready to support a design decision?

It is if you can answer yes to every question below. Any no means the results are still a draft.

1. Do the demands come from meter or billing records, with the peaking factors written down?
2. Do the tanks, pump curves and PRV settings match the field, not the original drawings?
3. Has the model been checked against field pressures or a hydrant flow test at the high and low points of each zone?
4. Were the fire flow and outage runs checked for negative pressures, with pressure driven analysis used where they appear?
5. Is there a low-demand run showing water age at tanks and dead ends?
6. Is there an .INP file the reviewer can open and rerun?

A note on limits: this is general education, current as of October 2026. Pressure rules differ by state and utility, and the engineer of record and the reviewing agency decide what applies.

Frequently asked questions

Is EPANET free?

Yes. EPA distributes it as public domain software, and continued development happens in an open source project on GitHub.

What version of EPANET does EPA offer?

EPA's download page lists EPANET 2.2, dated July 2020. The first improvement its manual lists is pressure dependent demand.

Can EPANET model water hammer?

No. Surge from a pump trip or a fast valve closure needs a transient analysis program.

I run public works for a small Colorado mountain town, and our summer demand is several times our winter demand. What should our EPANET model include?

Run both seasons: summer for peak hour and fire flow, winter for water age at tanks and dead ends, with tank levels set the way your operators run them.

I'm developing a subdivision in Texas and the city wants a hydraulic analysis of my water system. Does it have to be built in EPANET?

Not under the state rule: 30 TAC §290.44(d) sets pressures, not software. Ask the city which program its reviewers accept and for the pressure at your connection point, often from a hydrant flow test. If the city runs EPANET, send the .INP file.

Related resources

• EPANET Water Modeling
• How to Read a Hydraulic Grade Line
• How to Conduct a Fire Flow Analysis for a Texas Development or Subdivision
• Water Storage and Pressure Zone Design for McKinney's Rolling Terrain

Not sure your water system holds pressure at buildout?

We build and check EPANET models for new developments and existing systems, from the first demand estimate to the fire flow memo the reviewer reads.

Modern Engineering Solutions, McKinney, Texas, with offices in Golden, Colorado and San Jose, California. 525+ projects completed.

Talk with an engineer about your water model or call (214) 833-6748.