initials PPT is used in the sense of parts per trillion in spite of the 

 fact that the established usage of PPT is parts per thousand, a difference 

 of 6 orders in magnitude. A better choice might be PPTr to indicate parts 

 per trillion. The EPA has used standardizing reporting units as parts per 

 billion (PPB) . 



Table 4.21 

 Comparison of Units Used in the Qrganotin Literature. 



Parts per thousand [PPT] grams/liter 10~° g/1 



Parts per million [PPM] milligrams/liter 10~ 3 mg/1 



Parts per billion [PPB] micrograms/liter 10 -6 M/l or ug/1 



. . . — Q 



Parts per trillion [PPTr] nanograms/liter 10 9 ng/1 



4.9 NUMERICAL MODELING 



The Dynamic Estuary Model was used to mathematically simulate the hydro- 

 dynamics of San Diego Bay to predict the increased TBT loadings that would 

 be expected if the U.S. Navy implemented the plan to paint the entire 

 600-ship fleet with TBT antifouling paint (Waltin et al., 1986). The 

 model was used to evaluate the environmental impact of continuous releases 

 from docked ships and pulsed releases from drydock activities. 



The simulated concentrations in San Diego Bay were derived in two steps. 

 First the hydrodynamics of the Bay were modeled using a dynamic, pseudo 

 two-dimensional link node model. The governing equations for the 

 hydrodynamic model are the momentum equation, mass transport, dispersion 

 and the continuity equations. Losses of organotin from the water column 

 were estimated to occur as an exponential decay function. Simulated 

 concentrations were generated using sets of assumptions of continuous 

 releases, pulsed releases of different magnitudes and finally a set of 

 decay rates derived from laboratory microcosm studies ranging from 4%/day 

 to 11%/day. Values generated from the model have shown that pulse 

 loadings are rapidly dispersed in the estuary, and that concentrations of 

 TBT in the southern end of San Diego Bay where flushing is poorest are 

 d e pendant on the assumptions of degradation rate used. Good correlation 

 was obtained between the model and environmental concentrations where 

 samples have been analyzed. The predictive ability of the model cannot be 

 established until better information is available on the leach rates from 

 underwater surfaces and the extent of underwater surfaces coated with TBT. 



rv-50 



