Why initial discharge increase?

Hello

Can you please exsplain to me why initial discharge increase along the rivercourse,in spite of no lateral inflow. I use option 12. I attach the the input file.

Yours sincerely
Morten Skoglund
Morten.Skoglund@civil.sintef.no

Dear Morten Skoglund :

I have looked at your model and have determined why the initial discharge differs over the length of the model.

1. At cross-section 95.1 km, you have a lateral inflow of 421 cms. The next upstream cross-section is at 90.7 km. DAMBRK takes the lateral inflow that you have defined and then uniformly distributes it along this reach of river between these two cross-sections. In reviewing the initial conditions output, you will see the discharge gradually changes at each interpolated cross-section between these two user-defined cross-sections. This is done so that DAMBRK does not experience a shock to the system. If DAMBRK did not do this and the lateral flow was entered all at one cross-section, the model could become unstable and not converge to an solution.

2. You have a dam structure at cross-section 97.4 km. You have a spillway rating curve at this dam that defines the amount of flow allowed out of the dam structure. So, the flow downstream of this cross-section is dependent of what the initial water surface elevation has been defined for the dam structure, not what is the value defined as the inflow hydrograph at the upstream end of the model.

Note that DAMBRK appears to be operating correctly for the way it has been defined. However, you should be concerned about the following items that I saw defined in your model.

1. The initial inflow hydrograph runs out of data (it endds at hour 27.941) before the entire inflow hydrograph has passed. This will cause the model to end when it runs out of inflow data. The flood hydrograph may not completely route to the downstream-most cross-section by this time.

2. The dam structure spillway discharge rating curves that are defined, stop prematurely (below the defined dam crest elevation). What can happen in this case is DAMBRK will extrapolate from this curve data whenever the reservoir water surface elevation is greater than the head values defined in these rating curves. This could cause incorrect outflow values to be computed at these spillways during high flows.

3. The downstream outlet stage hydrograph that you have defined as a boundary condition has a water surface elevation of 0.0 m. This is well below the channel bottom elevation of 61.5 m MSL at the downstream-most cross-section 95.1 km.

If you have any additional questions or need any further information, please contact me again.

Sincerely,

Chris E. Maeder, M.S.
Senior Technical Engineer
chris.maeder@bossintl.com

Hello

I refer to earlier correspondance and your answer. I will say that i can't understand why initial flow increase along the river.

In the file I attach til message there is no lateral inflow.

If the flow over the dams begins with a high waterlevel, wich I see now, se will give a high startingflow. This flow should be lower and lower as time goes on because the waterlevel will decrease since outflow is greater than inflow. The flow did not decrease, it is constant during time at many crossections.

I will ask: In the input file I have earlier written inflow hydrograph. Now I have take this hydrograph away in the channel cross-section table. Is this the reason why initial waterflow increase? Did the program use old parameters?

Morten Skoglund
Morten.Skoglund@civil.sintef.no

Dear Morten Skoglund :

I have reviewed your email, and I believe that DAMBRK might be "manufacturing" flow in your model. Note that I have copied this from our DAMBRK User Manual, which explains this:

THETA WEIGHTING FACTOR
Theta weighting factor in the finite difference solution. If a value of 0.0 is entered then a value of 0.60 is used by the program. If a value of 0.5 is used, theta is set to 0.60 internally by the program, the water surface elevation safety net feature is turned off (allowing the flow to drop below the initial conditions), and the model is then capable of allowing negative flows to occur (flow up the river valley). If a value of 0.51 is used, theta again is set internally to 0.60 and the model routing is performed by the diffusion method (the first two inertial terms in momentum Eq. (9) in Chapter 6 are omitted) iinstead of by dynamic routing. [36-F1I]

Note that the Theta Weighting Factor is defined in the DAMBRK Boundary Conditions screen. I believe that this is most likely what you are seeing happen with your model.

If you have any additional questions or need any further information, please contact me again.

Sincerely,

Chris E. Maeder, M.S.
Senior Technical Engineer
chris.maeder@bossintl.com