vGreetings:

I am involved in a CSO storage facility design project. The facility is
being designed to ensure that 90% of WWF volume system wide is captured
and treated. The sizing is based on continuous simulation of the
"typical year" (determined by statistical analysis of 30 years of local
rainfall data) using a trunk system model. The system is very large,
covering some 7,000 ha, and involving 15 flow monitoring sites for model
calibration. Typically there were 3 to 5 decent events for calibration
at each site, but insufficient data was available to do much
verification. Measured and modelled volumes were within a 10% envelope
for 41 of 52 events, and a 20% envelope for 48 of 52 events. Generally
the flow monitoring data is quite reliable (especially for the typical
events that we are most interested in) and involved calibrated primary
devices and depth-velocity sensors at each sites. In a few cases the
local hydraulics were poor and there is less confidence in the data.

My question is with respect to a reasonable approach (considering
limitations in the remaining study time and budget) to assessing the
overall uncertainty in the resulting tunnel volume calculation, so that
an appropriate contingency volume can be added to ensure that the
performance objective will be met on average over the long term.

Thanks in advance for any advice.
Chris Rogers
Christopher.Rogers@OTTAWA.CA