Product(s):	HAMMER
Version(s):	08.11.XX.XX, CONNECT Edition
Area:	Calculations

Problem

When looking at a Time History (graph) or animating a profile in the Transient Results Viewer, changes are seen (for example a change in HGL or flow) in the very beginning of the transient simulation, before the transient event occurs.

Solution

This is referred to as an "Initial Surge" or "Spurious surge" and is often an indication of a discrepancy between the initial conditions (steady state) and the first transient timestep. Meaning, HAMMER sees that a change in momentum has occurred in the very first timestep, and graphs and profiles display the resulting transition (albeit typically small). In some cases this manifests itself as a small "wiggle" or "jumpiness" in the HGL before the real surge event occurs (such as a pump shutdown or valve closure), while in other cases you may see a noticeable drop or spike in HGL in the first timestep. Addressing this problem is important, to ensure that the starting conditions are in a true steady state condition and that the transient simulation results are new skewed.

Note: the latest versions of HAMMER will include the following user notification to alert you:

1) First, to properly check for an initial surge, make sure that your "real" transient event is configured to begin a few seconds after the start of the simulation. For example if you are simulating an emergency pump shutdown, use a "time (delay until shutdown)" of five seconds (not zero). This way, you can check to make sure that the first five seconds of the transient simulation remains steady, before the pump shutdown occurs. This can give you confidence that there are no discrepancies between the initial conditions and the first transient timestep which would otherwise skew the transient results.

2) If you see an initial surge occur before your "real" transient event, check for any obvious problems with the model configuration in the general area where you see the initial surge occur. For example pump status that does not match with the transient pump type, a non-zero value in an air valve's "air volume (initial)" field while initial pressure is positive, a valve's initial calculated "relative closure" not matching the "initial relative closure" of its assigned pattern, or any of the issues described in the articles in the "See also" at the bottom of this page. Also check your User Notifications list for any clues and make sure the "specify initial conditions?" transient calculation option is not set to "true" (which would mean the initial conditions are manually typed into each element instead of being taken from the initial conditions calculation.)

3) If the source of the initial surge is not clear, you should next utilize the details of the User Notification (latest versions of HAMMER) to identify the source(s) of the initial surge:

"Possible initial surge detected: Initial head of Flow difference at this location exceeds the Initial Consistency setting specified in calculation options ...."

For older versions of HAMMER, use the Transient Output Log to identify any specific elements to focus on. Navigate to Report > Transient Reports > Analysis Output Log (referred to as the "Transient Analysis Output Log" in older versions). This will open a text-based report in WordPad. Press CTRL+F (Find) , type in "initial surge", and it will jump to the relevant section of the report, as seen in the example screenshot below:

Note: in the Transient Calculation Options, the Initial Flow Consistency and Initial head Consistency determine the thresholds for displaying an endpoint under the above section in the report. Meaning, a change in flow or head between the initial conditions and the first transient timestep, that is above the values entered in these calculation options, will cause the pipe endpoint to appear in the list as the cause of an initial surge. The reason is because there can sometimes be a small movement in the first timestep that is expected, such as from a change in tank level for a tank that was filling or draining in the initial conditions.

4) If the Output Log does not clarify where the initial surge is coming from, try making a few shorter profiles, re-compute the transient simulation and review those profiles in the Transient Results Viewer, as it will be easier to identify the specific element that the initial surge starts at. Then, carefully review the element configuration.

5) If the cause of the initial surge is still not clear, ensure that you are using the latest available version of HAMMER (see article on downloading). If you are using the latest version, contact technical support (forum or Service Request) for the latest cumulative patch set (which could potentially contain a fix) or to troubleshoot further.

Using a Top-fill tank

Using a Top-fill tank will cause an initial surge because at the point the water enters the tank, it is no longer pressurized, and the Engine creates a hydraulic "jump" to try and match the HGL in the tank.

We can work around this by creating a Child Scenario as a Steady State run. You will need to set the "top fill / inlet valve" tank property to false in the new Steady State scenario.

Then, capturing the conditions that exist at a time of a "stable timestep" into the EPS simulation, place a Discharge to Atmosphere element before the tank and use those values as the "typical flow" and "typical pressure drop" fields in the D2A element properties. See the section titled "Top-Fill Tanks in a Transient Simulation in HAMMER" in this Knowledge Base article.

These initial settings should also include things like the observed tank levels, whether a tank is filling or full, water level elevations, and anything else that could change during an EPS simulation. The "typical flow" and "typical pressure drop" fields in the D2A element properties can be derived from the orifice equation if needed. This Knowledge Base article contains more details - Modeling Reference - Discharge To Atmosphere in HAMMER

Using this approach, water leaves the system during the transient at the D2A and does not go into the tank, but that is likely acceptable since the tank level would be unlikely to increase significantly during the short period of time simulated with a transient.

Troubleshooting an unexpected Initial Surge during a transient simulation

Problem

Solution

Using a Top-fill tank

See Also