How does the Implicit dynamic solver calculate HEC-22 2nd edition structure losses and what assumptions are involved?
Why is the headloss through my catchbasin or manhole smaller than expected or zero, when using the HEC-22 2nd edition energy loss method?
In general, when using the Implicit dynamic solver in SewerGEMS or CivilStorm, HEC-22 headlosses are calculated the same way as the StormCAD and SewerCAD solvers; based on equation 7-10 through 7-16 in the HEC-22 manual.
However, there are a few assumptions that need to be considered when using the Implicit numerical solver with SewerGEMS or CivilStorm:
1) HEC-22 2nd edition is intended to be used for junction loss between inflow and outflow pipes through a structure. Therefore headloss is currently (as of version 08.11.04.55) not calculated in cases where there is no upstream pipe and the headloss will be reported as zero. This applies to the other headloss methods as well (with the Implicit numerical solver in SewerGEMS and CivilStorm). Future versions may reconsider this and allow headloss calculations in upstream-most nodes.
The individual values for loss due to benching, plunging, and angled inflow are reported as N/A, because they cannot be reported accurately for this solver.
2) To avoid stability issues, if the velocity of the downstream pipe is greater than 5 ft/s, a special "filter" is applied, by modifying the velocity as follows:
Vm = 5.0 + 0.1 *(V-5.0)
Where V = calculated pipe velocity Vm = modified velocity used by structure loss calculations
So, in cases where the velocity is high, the calculated headloss can be smaller than expected.
3) In SELECTseries 4 (08.11.04.54) and below, the Implicit numerical solver in SewerGEMS and CivilStorm does not have awareness of the angle of the incoming pipe, so an assumption is made in calculating the "Initial headloss coefficient" (K0) component of the HEC-22 2nd edition structure losses. So for a pipe with a relatively large bend angle, this can sometimes lead to a smaller headloss than expected.
An enhancement was made in the SELECTseries 5 release to account for the bend angle. This fix is also available in the latest cumulative patch set for SELECTseries 4.
Note: as of version 10.00.00.40, the Standard and HEC-22 Third Edition structure headlosses are supported with the Explicit (SWMM) solver. HEC-22 2nd Edition structure losses are currently not supported as of that version.