HEC-RAS Mass Balance Output Error

[Anonymous]

Not sure if anyone else has ever run into this problem…let me explain the best i can. I have run multiple tests and the results are comparable however i want to focus on one thing here and thats the computation log for the representation of the mass balance output.

Take a small section of a natural channel and contain it within a 2-D mesh. Add a hydrograph to the US end (small flows that stay within banks) and create 3 different DS boundary conditions within a different geometry:

#1 ND for one Boundary Condition
#2 Stage for another
#3 and the last create a 1D storage area DS of the 2D mesh and connect it with a SA connector at the exact level of the terrain. And set the elevation of the 1D storage area to 100-ft below the connecting terrain and give it an infinite volume of storage.

All 3-models produce similar results internally within the mesh. However in the last 2 tests the mass balance is absurdly high with 99% error. I have done multiple scenarios trying to fix this by changing tolerances and htab parameters and dt and weir coefficients and different stages…the list goes on.

At this point i believe the error is occuring at the face of the stage BC and at the face of the weir for the other model. In RAS Mapper all 3 scenarios show water moving in and ultimately out.

[cameron]

for the run with the stage set, how does the stage compare to the normal depth boundary?

For the storage area run, does the flow leaving the 2D model match which is going into the storage area?

[Anonymous]

Hi Cameron,

The normal depth % error =0.002723 Qleaving=24.02 cfs VolumeOut=3.06 ac-ft

The Storage Area run % error = 65.14 Qleaving=24.01 cfs VolumeOut=5.065 ac-ft

Inflow for both are 3.075 ac-ft of volume

I have done numerous testing with multiple scenarios. I do not believe the storage area geometry set up is incorrect. I believe the calculation of the mass balance is incorrect when a storage area and or stage is used as your DS boundary condition. Or the error is occurring exactly at the weir but does not impact the interior portion of the 2D model or its results

[Anonymous]

I am able to minimize the error to almost zero by changing the weir coefficient to 0.2 which is outlined in the 1D/2D connection tables listed in the 2D users manual however the stage test still produces 99% error

[cameron]

Is there water flowing into the model from the stage boundary condition?

What is your time step?

What are your initial conditions?

[Anonymous]

When i set my stage at the lowest elevation of the terrain under neath the stage BC, i do not get water flowing back into the model (i.e negative flow): actually i get water flowing out however the mass balance states that 0.00009 ac-ft gets out but the hydrograph shows differently.

As i raise that stage elevation the error becomes less and less however as you pointed out, thats when backflow begins to occur. My dt is 1 second and initial conditions have no parameters as the small 2D area is runnning fully dry until the rising limb of the inflow hydrograph enters.

Currently the HEC-RAS users manual on page 3-50 contains a “lateral structure weir coefficients” table and explanation. The exact statement within says ” In, general Lateral Structure weir coefficients should be lower than typical values used for inline weirs.”

This statement is incorrect. Both inline and lateral structure weirs which are used as an outflow mechanism within a hydraulic model and at some point becomes fully submerged, should ALSO be much lower than typical values. Inundation limits, culvert hydraulics and depths will all be the same but the mass balance will show upwards of 65% error if the coefficient is not changed

[cameron]

do you get really high velocities at the boundary with the low stage?

[Anonymous]

Within RAS mapper, with velocity set at “max” yes. Velocities that are not even possible within this model show. However if you move the slider to view velocity mapping over time those ridiculous velocities dont show

[cameron]

My guess is that those velocities are occurring either at the beginning of the simulation or when water reaches the downstream end and causing the model to freak out and give volume errors.

[Anonymous]

I agree. And i think this i a large bug within the software. I dont believe its really happening. I dont believe there is a 99% error and a velocity at one point in time of 23,414 fps. I think there is a bug in the software

[cameron]

Is this with the full momentum solver or diffusion wave solver?

[Anonymous]

I have run both scenarios. Both accumulating 99% error. One trick ive discovered is i can put a BC of normal depth and use a high slope, for example 0.5, and it represents results as runoff being able to freely leave the system without resistance because of the high slope.

However, currently if you use the stage as DS BC the errors are simply unnacceptble.

When using a 1D storage area connected via a weir which simulates an infinite outflow storage area and the weir represents the ground profile a few things must be adjusted to obtain acceptable errors. The first is that the weir submergence curve variable within Calc/tolerances in the 1D tab must be changed from its default (1) to 3. The second thing is that a weir coefficient of 0.2-0.5 must be used to represent a submerged weir. This is for both inline and lateral flow.

I truly think there is a bug or i am not fully understanding whats going on but my experience is telling me that its not the later

[Author]

Posts