Modeling other fluids or fluid temperatures

  Applies To 
  Version(s): V8i, CONNECT Edition
  Area:  Modeling
  Original Author: Mark Pachlhofer, Bentley Technical Support Group


Is it possible to model fluids other than Water, or water at a different temperature or specific gravity?

Solution - WaterGEMS, WaterCAD, HAMMER

Other liquids can indeed be modeled, and the steps below describe how to do it.


  • Only Newtonian liquids are valid
  • You might need to adjust your pump definitions if the viscosity of your fluid differs from water (see more below).
  • As the Hazen-Williams friction method is an empirically based formula, its calculated friction losses are only applicable to Water @ 20 deg C so it should not be used for other liquid types. You will not see a change in headloss when changing the specific gravity for example (just a change in calculated pressure). When modeling other liquids, consider using the Darcy-Weisbach method.= instead.

1) Go to Analysis > Calculation Options and click on the calculation option that you'll be using for this analysis to access the properties

2) Go to the Hydraulics section towards the bottom of the properties window and click the ellipsis button on the right side of the liquid label cell as seen below.

3) This opens the Liquid Libraries for the Engineering Library and you can select the liquid that you'd like to use from there. Note that on the right hand side of the window the liquid properties like Kinematic Viscocity, temperature, Vapor Pressure, and the bulk modulus of elasticity are displayed.

4) After you create or decide on one of these use the select button to choose that liquid.

5) If you'd like to create a new liquid with different characteristics go to Components > Engineering Libraries and expand the liquid library. Left click on the 'Liquid library.xml' label as shown in the screen shot below and then click the 'Add Item' option. Finally, enter your data and name the new liquid. Click close when done. See more here.

Pump Definition Implications

You will need to change the pump curve itself to represent the flow and head for the fluid you are modelling and you might want to use the Darcy Weisbach fiction method. See more on this here: When the user changes the fluid, why doesn't the calculated pump head change?

Transient simulation implications in HAMMER

HAMMER has been used successfully in the past for analyzing certain mining slurries. In addition to changing the fluid specific gravity and viscosity, you would also use a friction factor that is appropriate for the slurry. If it isn’t Newtonian then HAMMER’s standard friction models won't be appropriate and won't work very well. This is more of an issue after the initial transient occurs and the resulting pressure waves are being dampened by friction (although it is also a factor to consider when computing the steady state).

You may also need to adjust the vapor pressure limit for your liquid. To do this, click on the Vapor Pressure field in the properties of the Transient calculation options, then click the ellipsis button (...) next to it. Expand the liquid library, select the appropriate liquid, then click the Select button. You will notice that the Vapor Pressure field will change accordingly. For more information on the impact of this, see the article linked to at the bottom of this page.

HAMMER was developed with water in mind. The more you deviate from a specific gravity of 1.0, the more you may need to you engineering judgment to assess the results of the transient analysis.

A Note on Temperature Change in HAMMER

Note that only one liquid temperature can be set for a given simulation run. This is a consideration in a case where you have a significant change in temperature in the network being analyzed, such as for a heat supply network / district heating/cooling network. From the HAMMER Help topic “Bentley HAMMER V8i Theory and Practice”,


With Bentley HAMMER V8i, you can analyze drinking water systems, sewage forcemains, fire protection systems, well pumps, and raw-water transmission lines. You can change the specific gravity of the fluid to model oil or slurries, for example. Bentley HAMMER V8i assumes that changes in other fluid properties, such as temperature, are negligible. It does not currently model fluids with significant thermal variations, such as can occur in cogeneration or industrial systems.


So, unless you can make a valid, conservative assumption, then the transient results may be skewed if you have a significant change in temperature. One approach is to try a sensitivity analysis: do several model runs, varying the liquid properties for multiple temperatures. If there’s no notable difference in transient response, then you might conclude that HAMMER can model their situation OK.

Be sure to account for differences in vapor pressure limit for the different temperatures, which can be very significant for "downsurge" pressure waves. For example the vapor pressure difference between 95C and 4C is about 9 meters, so you would want to be mindful of that.


Or, if there is a clear separation between the segments of the network at two different temperatures, you may be able to do two separate model runs.

Solution - SewerGEMS, SewerCAD, CivilStorm, StormCAD

For pressure pipes (force mains) in the GVF-Convex solver (available in SewerCAD and SewerGEMS), friction methods such as Hazen-Williams and Darcy-Weisbach can be used and the liquid properties can be changed in the Calculation Options just like with WaterGEMS and WaterCAD (described further above).

For the Implicit, Explicit (SewerGEMS and CivilStorm) and GVF-Rational solvers (StormCAD and CivilStorm), non-Mannings friction calculations for gravity flow are not supported. The Mannings equation was developed for water and therefore since gravity conduits only support the Mannings friction method, only water is supported. If you need to model gravity flow of fluids other than water (with a different density / viscosity), consider performing an offline conversion to determine the estimated equivalent Mannings n, then update the Mannings n in the conduit elements. Consider a conservative assumption for modeling purposes.

See also

Related Discussion on crude oil

What is the difference between pressure head and pressure?

When the user changes the fluid, why doesn't the calculated pump head change?

How can I change the vapor pressure limit based on fluid properties, for fluids other than water?