# How to do water balance check and how to calculate thermal water fraction - Home - Delft3D

## Message Boards

### How to do water balance check and how to calculate thermal water fraction

AF

##### Amaal Fathy, modified 2 Years ago.

#### How to do water balance check and how to calculate thermal water fraction

Youngling Posts: 6 Join Date: 3/26/16 Recent Posts 00

Dear all,

I'm simulating WQ of a low flow part of a river,and I have a few questions.

I need to perform a water balance check before the simulation to ensure the validity of the model. I've read that to do this check, both continuty and boundary concentrations should be equal to 1 and the resulting solution must be equal to 1 as well. I've no idea how to perform this check so any help is appreciated.

Another inqury, a thermal discharge effluent is simulated in the hydrodynamic model, is it possible to calculate thermal water fraction between the boundaries? if yes, how to do this?

I've searched the manual but reached no answer.

Thanks in advance

I'm simulating WQ of a low flow part of a river,and I have a few questions.

I need to perform a water balance check before the simulation to ensure the validity of the model. I've read that to do this check, both continuty and boundary concentrations should be equal to 1 and the resulting solution must be equal to 1 as well. I've no idea how to perform this check so any help is appreciated.

Another inqury, a thermal discharge effluent is simulated in the hydrodynamic model, is it possible to calculate thermal water fraction between the boundaries? if yes, how to do this?

I've searched the manual but reached no answer.

Thanks in advance

AM

##### Arjen Markus, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti

Jedi Knight Posts: 223 Join Date: 1/26/11 Recent Posts 10

Regarding your first question:

The basic idea is that a conservative substance which has a concentration of 1 (or any other value) all over the area and which flows in over the boundaries with a concentration of 1 (or that same other value) must keep that value. If you see deviations from that value, then the flow field is not mass/volume conserving. Small deviations (let us say concentrations that range from 0.999 to 1.001 after several days or weeks) are not bad - they cannot even be avoided as there is always some inaccuracy in the calculations.

That said, we normally use a substance called "continuity" and define the boundary concentrations as 1 and the initial condition as 1. If you do a calculation with this set-up and inspect the result, using a tool like QuickPlot, then you should see that the concentration remains near to 1 all over the model and from the start to the stop time.

Regarding your second question:

Depending on what you want exactly, you could use QuickPlot to show the area where the temperature is above a certain threshold or you could use the statistical output facilities of DELWAQ to get numerical information. For the latter:

The basic idea is that a conservative substance which has a concentration of 1 (or any other value) all over the area and which flows in over the boundaries with a concentration of 1 (or that same other value) must keep that value. If you see deviations from that value, then the flow field is not mass/volume conserving. Small deviations (let us say concentrations that range from 0.999 to 1.001 after several days or weeks) are not bad - they cannot even be avoided as there is always some inaccuracy in the calculations.

That said, we normally use a substance called "continuity" and define the boundary concentrations as 1 and the initial condition as 1. If you do a calculation with this set-up and inspect the result, using a tool like QuickPlot, then you should see that the concentration remains near to 1 all over the model and from the start to the stop time.

Regarding your second question:

Depending on what you want exactly, you could use QuickPlot to show the area where the temperature is above a certain threshold or you could use the statistical output facilities of DELWAQ to get numerical information. For the latter:

- Define a monitoring area (use DIDO for this) over which the concentration or temperature is to be averaged
- Import the temperature as a segment function - it is exported via the coupling procedure
- Define the statistical output you need

AF

##### Amaal Fathy, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti

Youngling Posts: 6 Join Date: 3/26/16 Recent Posts 00

Thank you Arjen for your reply,

I used the continuty substance as you mentioned but it seems there's a problem. Data for discharge points are not being read. I attached a photo for what appears to me.

Model is 4 layers only and Both time step for communication file and WQ file is 6 hours.(I know it's too large but I'm only checking.

I used the integration method 12.

What might be the problem here ?

I used the continuty substance as you mentioned but it seems there's a problem. Data for discharge points are not being read. I attached a photo for what appears to me.

Model is 4 layers only and Both time step for communication file and WQ file is 6 hours.(I know it's too large but I'm only checking.

I used the integration method 12.

What might be the problem here ?

### Attachments:

MJ

##### Michelle Jeuken, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti

Jedi Knight Posts: 156 Join Date: 1/21/13 Recent Posts 10

Deaar Amaal,

I think your model has become unstable. A communication time step of 6 hours is indeed very large. You are also using an explicit scheme, (12), so your calculation time step in Delwaq is limited by the smallest residence time. You should lower your Delwaq calculation time step, because the automatic choice in the user interface is the coupling time step of 6 hours, which is probably (much) to big...

Michelle

I think your model has become unstable. A communication time step of 6 hours is indeed very large. You are also using an explicit scheme, (12), so your calculation time step in Delwaq is limited by the smallest residence time. You should lower your Delwaq calculation time step, because the automatic choice in the user interface is the coupling time step of 6 hours, which is probably (much) to big...

Michelle

AF

##### Amaal Fathy, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti

Youngling Posts: 6 Join Date: 3/26/16 Recent Posts 00

Thanks Michelle for your help,

I repeated the hydro-dynamic run to reproduce the communication file with time step 1 hour but I still have the same problem. The Define input window appears as I attached although the discharge value is 10.2 m3/s(uniform over depth) . Why is the model unstable?

- Hydro-dynamic time step is 6 sec, com. and WQ time step is 1 hour.

- Explicit scheme 12 is used.

Thanks in advance

I repeated the hydro-dynamic run to reproduce the communication file with time step 1 hour but I still have the same problem. The Define input window appears as I attached although the discharge value is 10.2 m3/s(uniform over depth) . Why is the model unstable?

- Hydro-dynamic time step is 6 sec, com. and WQ time step is 1 hour.

- Explicit scheme 12 is used.

Thanks in advance

### Attachments:

AF

##### Amaal Fathy, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti

Youngling Posts: 6 Join Date: 3/26/16 Recent Posts 00

I realized something here, After a few time steps, the unstability starts exactly from the discharge points so the problem is definitely here. I will reduce the time step to be 6 seconds and try the WQ simulation.

Thanks.

Thanks.

### Attachments:

AF

##### Amaal Fathy, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti

Youngling Posts: 6 Join Date: 3/26/16 Recent Posts 00

Dear Michelle,

I reduced the time step to 1 minute and the model is still unstable. Where could be the problem?!!

I reduced the time step to 1 minute and the model is still unstable. Where could be the problem?!!

MJ

##### Michelle Jeuken, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti (Answer)

Jedi Knight Posts: 156 Join Date: 1/21/13 Recent Posts 00

Dear Amaal,

There might be a lot of reasons why the model still is unstable.

First of all, check the discharges panel. If you select the hydrodynamics a second time in an existing scenario, you might have the discharges twice in the model. Selecting hydrodynamics always adds discharges, because existing discharge specifications might be modified. The time resolution of the discharges should match the time resolution of the other hydrodynamics files. You can't match your old 6h discharges with 1h hydrodynamics.

You might want to make a test run without the discharges, so you can be more sure that they are the problem.

Your grid is very fine (5x5m?), and I don't know the depth of your system. But I guess the depth near the shores is lower. With explicit scheme 12, your WAQ time step is limited to the residence time which is V divided by Q. Don't know how big the thermal discharges are, and the volumes of the segments they are discharged to, but you could make an estimate of V and Q.

It might be a good idea to move the location of the discharges a bit further of the shore into somewhat deeper waters.

You could also try to use scheme 15 (implicit) or scheme 21 (flexible explicit/implicit), which might be more stable.

Also, you could make your grid less fine, or aggregate the results for the water quality calculation to bigger segments. This will make the V's per segment bigger, thus giving you higher residence times (=V/Q).

I think you have

Michelle

There might be a lot of reasons why the model still is unstable.

First of all, check the discharges panel. If you select the hydrodynamics a second time in an existing scenario, you might have the discharges twice in the model. Selecting hydrodynamics always adds discharges, because existing discharge specifications might be modified. The time resolution of the discharges should match the time resolution of the other hydrodynamics files. You can't match your old 6h discharges with 1h hydrodynamics.

You might want to make a test run without the discharges, so you can be more sure that they are the problem.

Your grid is very fine (5x5m?), and I don't know the depth of your system. But I guess the depth near the shores is lower. With explicit scheme 12, your WAQ time step is limited to the residence time which is V divided by Q. Don't know how big the thermal discharges are, and the volumes of the segments they are discharged to, but you could make an estimate of V and Q.

It might be a good idea to move the location of the discharges a bit further of the shore into somewhat deeper waters.

You could also try to use scheme 15 (implicit) or scheme 21 (flexible explicit/implicit), which might be more stable.

Also, you could make your grid less fine, or aggregate the results for the water quality calculation to bigger segments. This will make the V's per segment bigger, thus giving you higher residence times (=V/Q).

I think you have

Michelle

AF

##### Amaal Fathy, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti (Answer)

Youngling Posts: 6 Join Date: 3/26/16 Recent Posts 00

Thanks Michelle, that was really helpful. I moved the discharges and it worked proberly.

MJ

##### Michelle Jeuken, modified 2 Years ago.

#### RE: How to do water balance check and how to calculate thermal water fracti

Jedi Knight Posts: 156 Join Date: 1/21/13 Recent Posts 00

You're welcome. And to get back to one of your first questions:

>Another inqury, a thermal discharge effluent is simulated in the hydrodynamic model, is it possible to calculate thermal water fraction between the boundaries? if yes, how to do this?

I've searched the manual but reached no answer.

You can do that using simple tracers. Attached is a sample 3 tracers file.

Use tracer one for the initial: initial concentration one, and zero at all boundaries

Use tracer two for the river: initial concentration zero, and one at all river boundaries

Use tracer three for the thermal waste loads: initial zero, and one at all waste loads

You can edit the file with a text editor to give the tracers more sensible names (initial, river, thermal), or add other tracers to distinguish between the two thermal discharges by copying one of the definitions.

Michelle

>Another inqury, a thermal discharge effluent is simulated in the hydrodynamic model, is it possible to calculate thermal water fraction between the boundaries? if yes, how to do this?

I've searched the manual but reached no answer.

You can do that using simple tracers. Attached is a sample 3 tracers file.

Use tracer one for the initial: initial concentration one, and zero at all boundaries

Use tracer two for the river: initial concentration zero, and one at all river boundaries

Use tracer three for the thermal waste loads: initial zero, and one at all waste loads

You can edit the file with a text editor to give the tracers more sensible names (initial, river, thermal), or add other tracers to distinguish between the two thermal discharges by copying one of the definitions.

Michelle