The March/April 2008 issue of Stormwater had several great articles. I was particularly interested in your “Editor’s Comments” on first flush and the article it referenced.
I believe that one of the reasons for the “phantom first flush,” in addition to varying TOCs in large watersheds, is the existence of what I call mixing basins. These mixing basins can be small potholes in the street, catch basin sumps, or baffle boxes without a high flow bypass. What is happening is that during low flow rates, these mixing basin work as settling basins and remove a percentage of the pollutants in the “first flush.” Later in the storm event, the flow rates increase to a point where the previously settled pollutants are re-suspended, thus obliterating the first-flush phenomenon. I encountered this when I was trying to design a first-flush diversion system with an upstream watershed area of about 35 acres of commercial and residential property. The result was that the diversion system would not be effective unless the upstream mixing basins could be eliminated. The conclusions are:
1) Effective suspended solid removal BMPs must have a high-flow bypass to avoid the resuspension of the pollutants.
2) Suspended solid removal BMPs should not be used in series (see “Treatment Trains: Don’t Get Run Over” by Gary Minton in the July/August 2006 issue of Stormwater).
3) Careful consideration must be given to locating a suspended solid removal BMP where it can optimize its removal performance relative to the TOC and volume of the first flush.