Towards a ranking of improvements for SWMM: Water Balance

John: re stormwater-GW infilt. interest.

Mass balance, plus Hantush approach may be most suitable (which we already have most of). Detailed and accurate GW routing, especially in urban area, would be a pipe-dream.

Bob

To all users but especially to Wayne:

First of all thanks to Ching Len Chen for his posting and reply. I take this opportunity to reopen the discussion of ranking potential upgrades to the SWMM engine.

It seems to me that current stormwater management strategies are, or ought to be, much more oriented towards infiltration BMPs. At least in this part of the world. Of course the major concern is the fate of pollutants in the ground.

SWMM RUNOFF does not attempt to route groundwater or interflows from one compartment to another. Better coding is clearly required. I have mentioned this requirement quite often both on this list and in personal notes to Wayne. My purpose here is to seek support from you all.

Put simply we need better groundwater compartment routing, and I know that this is a tough assignment, involving a redesign of the link between surface/ground hydrology, and also TRANSPORT. Ipso facto, it should be done first. After that, we could consider integrating planview graphics, and generalize input for physical drainage (RUNOFF, TRANSPORT, EXTRAN and STORAGE all currently require different forms of similar info.) because only then will we know what the fundamental code will be.

If you agree perhaps we could discuss minimum specs for routing groundwater - e.g. does urban groundwater often follow predictable paths? Is much urban groundwater in fact somewhat managed (flows to basement drains)? Or towards ponds placed on drainage systems? Could urban grounwater routines initially be along a preset direction IOW as a dendritic network (which allows very much simpler coding)? Can urban groundwater be modeled explicitly, as is done in most surface water models, rather than implicitly as is done in most groundwater models? What % of urban applications can be covered by dendritic groundwater networks?

If so perhaps SWMM5 could easily be made to be more useful but not require a major rewrite. Personally I would like to see in SWMM some sort of start towards comprehensive groundwater/surface water interaction, not more accurate than one subcatchment (iow not specifying local hillslopes).

Or is this a pipe-dream? Thanks for your time.
/bill

from Dr Chen:
The issue is not the continuity check, but the soundness of simulation of physics.
--snip--
There are a lot of other issues can be discussed in the groundwater module. We may discuss that latter if you are interested.
--snip--

I believe I have heard this message earlier, quite likely from the same person!

Regarding improving the groundwater routine in the Runoff Block, there are a couple of improvements that can be made, I hope without too much trouble. 1. Groundwater can be routed from the saturated zone of one subcatchment groundwater storage to another groundwater storage, on the basis of elevation difference. This implies that an absolute datum be introduced into the Runoff Block. 2. Bank storage can be simulated by moving water out of an adjacent channel into the subcatchment groundwater storage. Both of these changes would be based on elevation differences at the *beginning* of a time step, and cannot be iterative (probably, not positively) or dynamic in the sense of Extran.

However, SWMM cannot incorporate a technique as sophisticated as, say, MODFLOW, for 2- and 3-D groundwater routing using a finite difference scheme. If you need something that advanced, please use MODFLOW, or whatever the model of choice is these days (I confess, I'm out of date here). What would be useful would be an easy linkage from SWMM to MODFLOW, as well as to several other potentially useful models, e.g., HSPF, WASP, etc. I hope we can agree on a general interfacing scheme for some of these models this year, that is, establish a "standard" for interfacing of time series and state variables for models of this kind.

I do not envision quality routing through the subsurface regime in SWMM in the near future. This is something so complex, especially in the vadose zone, that I simply don't have time to devote a year of research to it. If you need this feature, please use HSPF or other alternative.

Finally, I agree with Bob Pitt that "Detailed and accurate GW routing, especially in urban area, would be a pipe-dream." At least not with SWMM, and unlikely with any model, given the highly disturbed nature of soils in an urban area.

Wayne

> However, SWMM cannot incorporate a technique as sophisticated as, say, MODFLOW, for 2- and 3-D groundwater routing using a finite difference scheme. If you need something that advanced, please use MODFLOW, or whatever the model of choice is these days (I confess, I'm out of date here). What would be useful would be an easy linkage from SWMM to MODFLOW, as well as to several other potentially useful models, e.g., HSPF, WASP, etc. I hope we can agree on a general interfacing scheme for some of these models this year, that is, establish a "standard" for interfacing of time series and state variables for models of this kind.

The linkage of groundwater and surface water models is not a trivial task. In 99% of the cases there is an interaction of groundwater and surface water. To adequately "compute" (that's for you Bill) movement of groundwater and surface water the models would have to be dynamically linked. In other words both models running at the same time on the same or near the same timestep. In WASP there are ways to introduce base flow conditions from groundwater as well as any associated mass. There is no formal linkage for carrying out the task but it is pretty straight forward.

As for development of "standards" for time series files, these is about 10 years late.

Tim Wool

A small thing to add to the wishlist for SWMM 4.5 (or should i say SWMM 5, or even SWMM99):

Add an option that prints an output summary file (in addition to the big output file). I know it's easy to do this in Visual Basic or an EXCEL macro or whatever, but I want a small text file that I can quickly load into my spreadsheet or to present without too much explanation. No input echo, no intermediate goodies, and certainly no page breaks or repeated column headings. Just a list of all warnings, the continuity error, the timing and peak flood stage at each node, the velocity and flowrate in each conduit, the worst node and conduit change percentages, and input/ output file names. Oh yeah, and total run time, to quantify complaints that I need a faster computer :(

Michael A. Gregory, P.E.
gregoryma@cdm.com

We (chi) have created a new tool for the PCSWMM environment that creates html pages from SWMM output files and displays them in the browser of you choice. The html files are based on template files of different complexities (which you can modify) to suit the level of detail you wish to present. For example output data from a single swmm run can be presented in different ways for modelers, politicians, citizen groups, your boss etc... all with colorful graphics, links to other PCSWMM tools (like the Dynamic HGL tool, etc.) and, if you wish, links to other html pages on your project. The pages are created and displayed instantaneously with the click of one button in the PCSWMM shell. Look for it in the '98 release (February Retreat in Toronto).

Sincerely,
Rob James
CHI