Tidal model

Dear all:

I am modeling a system that intended to minimize the effect of tidal water into farm/residential area and also to create a wetland with time. The system consists of four ponds or reservoirs connected between each other by channels. We have data as water surface elevation for tidal end and have decided to consider runoff as negligible because the system will not collect runoff other than falling rain inside the ponds.

My question is: How can I input the tidal watersurface elevations in a continuous way so I can model the water coming in and out of the ponds? when tides are high ponds will be filling up and when tides are low then ponds will be draining out.

Thanks for your advises,

William Medina
wmedina@WAM.UMD.EDU

Hi William,

I have modelled the tidal effects in XP-SWMM. In XP-SWMM you could select the node as an outfall node. Under the outlet control you can assign your astronomical tidal parameters appropriately (The parameters can be selected based from your tidal elevation data). Although the parameters entered are for one tidal cycle, the cycles will be repeated until your run ends.

Regards

K.Umakhanthan, PhD.,MASCE, MIEAust.
k.umakhanthan@cpgcorp.com.sg

Bill,

If you use the the "6" th option on the J1 line of SWMM 4.4g to SWMM 4.4h as in this example:

J1 6

you can input a stage or tide boundary condition at multiple points in Extran. The tide or stage data is referenced as the 5th scratch file or NSCRAT(5) in the JCL of SWMM. Here is an example of how the data may look in the ASCII NSCRAT(5) file:

2 1
'outlet1' 'outlet2'
'outlet3' 'outlet2'
2001 8 1 0 633.550 711.015
2001 8 1 1 632.090 710.980

which means there are two boundary conditons in the file ( 'outlet1' 'outlet2') but a third boundary conditon 'outlet3' is the same as 'outlet2' - see the Extran DOC file for more information. This was added to SWMM by Chuck Moore of CDM and it is a feature that we use extensively for complex continuous modeling.

Best Regards,

Robert Dickinson
DickinsonRE@CDM.COM

Thanks for your advises. I am using option 5 which allows me to copy historic data but the limit is 500 data. Can I have more than this included in one run or do I have to use a hot start...

Bill Medina
wmedina@WAM.UMD.EDU

Robert:

It seems that my alternative using option 5 is not feasible (no replies from the group ... yet). Can you help in setting up option 6? I have one outfall controlled by tidal therefore I have set up the ascii file as follows:

1 1
2
2
2003 1 1 0 0.256
2003 1 1 0.25 0.275
..... (9098 data lines)


I have named it "tidal.bcf" extension is suggest by EXTRAN.DOC.

I have changed MM line to:

* NITCH NSCRAT(1) NSCRAT(2) NSCRAT(3) NSCRAT(4) NSCRAT(5)

MM 5 1 2 3 4
5

But I get an error message saying that I am not allowed to use nscrat
(5).

===> ERROR !! PROGRAM EXECUTION STOPPED FOR THE FOLLOWING REASON.

A JIN, JOUT, OR NSCRAT FILE uses file numbers 5, 6, or 55. This is not allowed.

What am I doing wrong?

Thanks for your help,

William Medina
wmedina@WAM.UMD.EDU <mailto:wmedina@WAM.UMD.EDU>

Bill,

I don't think this is still a problem but this error message is something I added in the 1980's. A JIN, JOUT, OR NSCRAT FILE uses file numbers 5, 6, or 55. This is not allowed. Change your NSCRAT(5) file number to 20 or 30. I don't like how your example data looks.

Here is another example of how this file should look:

2 1
'out-2' 'out-1'
'out-3' 'out-1'
2001 8 1 0 633.550 711.015
2001 8 1 1 632.090 710.980
etc...

or there are three boundaries, out-1 and out-2 are in the BC file and out-3 uses the same data as out-1. It eventually makes sense.

Robert Dickinson
DickinsonRE@CDM.COM <mailto:DickinsonRE@CDM.COM>

Robert:

My system has one outlet and that outlet is controlled by tide. The identifier I used for it is 2 as number. I have setup the first lines of the ascii file as

1 0
2
2003 1 1 17. 0 -0.1
2003 1 1 17.25 0.0
2003 1 1 17.5 0.01
...
etc...

Because following what I understand from EXTRAN DOC line 1 is 1 out let with 0 other outlets affected by tidal data, line 2 is the name of the outlet 2, and I don't need the third line shown in your example so I go directly to tide data as year, month, day, hour in decimal, tidal elevation.

The model runs but I get 0 flows, but if I use the option 5, same data I get flows into and out of the system.Is there something wrong with the set up of the ascii file?

Thanks for your help,

William Medina
wmedina@WAM.UMD.EDU

William,

Perhaps the reason you are getting zero flow is that it appears that your tidal elevation fluctuates from -0.01 to 0.01. This very small difference may give zero flow. A recent model I completed has tides which fluctuate from 0.0' to 3.2' for a 2-yr storm event, and 5' to 8' for the 100-yr storm event. Check the tide tables in your area to get a feel for the proper range of fluctuation.

Andrew B. Scott, P.E.
ascott@RKKENGINEERS.COM

HI Andrew:

In fact the maximum tidal elevation is above 2.6 ft, but my system is only 4.75 ft. One question would be the time step, can I have 15 minute intervals instead of hourly.

Thanks,

William Medina
wmedina@WAM.UMD.EDU

William -

The main problem is that your tide file starts Jan 31, but your simulation started Jan 1 and ended before the start of the tide file. I shifted the simulation forward, and the model runs. I also set NEQUAL=1 to let the program lengthen the conduits for stability. I'd also recommend setting the junction inverts to the minimum invert of the connecting conduits. Odd things may happen if you just leave them at zero. You were also missing Y0 data from the end of the D1 lines. Take a look at the attached files (Rt33 your original; RT34 with my modifications). I did a few other things such as add random XY coordinates and strip out the comments (I recommend you add them back in though).

If you don't have a visual interface for your model, consider that your's is small enough to use the demo version of MikeSWMM. You can download the software from www.dhisoftware.com <http://www.dhisoftware.com>and view the result file I set it up to generate.

Regards,

Mitch Heineman, P.E. DEE
HeinemanM@CDM.COM

William,

My tidal fluctuations are set up like this for a 10-yr event; and I'm using a time step of 1 second. A time step of 15 minutes should be fine if your system can remain stable.

I1 'Chisman' 1
* NTIDE
J1 5
* KO NI NCHTID DELTA
J3 0 8 0 0.005
* Time Elev Time Elev Time Elev Time Elev Time Elev
J4 0.0 4.0 6.3 1.0 12.6 4.0 18.9 1.0 23.2
4.0
J4 29.5 1.0 35.8 4.0 42.1 1.0


Andrew B. Scott, P.E.
ascott@RKKENGINEERS.COM

Mitch,

Thanks a lot, model is working now thanks to you all.

William,

P.D. Is there a converter from SWMM44 to XP-SWMM? Can I export this model into PC-SWMM using the same historic data file? I could do it with the data written into the code but it only allows 500 data series and I have more than that.

XP-SWMM allows the direct import of EPA SWMM 4.X files and XP Software has also created a converter to import the most recent versions of 4.4 by creating an XPX (ASCII) file as an intermediate step. This XPX file is a standard import file for XP products.

Anthony Kuch
XP Software Inc.
anthony@xpsoftware.com <mailto:anthony@xpsoftware.com>

In addition to Tony Kuch's answer regarding import into XP-SWMM, you should consider the following options:

1) MikeSWMM, which I suggested previously, uses the EPA SWMM4.4 engine, so no conversion is necessary. Your small model works with the free demo version of MikeSWMM. You would actually only need to use the MikeView post-processor, not the MikeSWMM pre-processor.

2) PC-SWMM, the lowest-cost commercial interface, also uses the EPA engine, so no conversion is necessary.

3) SWMM5 is free and has its own GUI. You'd need to run the conversion program to import your SWMM4-format input file.

Both PCSWMM and MikeSWMM thus read the external tide file you've already created. I haven't used SWMM5 much yet, but presume that it reads a tide file with an analogous format.

Regards, Mitch Heineman CDM

HeinemanM@CDM.COM

William,

Do you mean that you need more than 500 data points for your boundary condition or that you have more than 500 BC nodes? If you do need more than 500 data point - how many data points would be enough?

Regards,
Bob Dickinson
DickinsonRE@CDM.COM