Storm Duration

Compressing a 24-hour storm to one hour will increase the relative peak by 24x, so this is probably not what you want to do.

Please check the regulations carefully. If they specify the SCS/NRCS runoff method, they should also tell you which rainfall to use. For a 1-hour rainfall, they need to specify the exact rainfall distribution. For example, NJ has a 2-hour "water quality" distribution.

If you can't identify the distribution (and you're sure they're not talking about the Rational method!) then you could select the "Constant Intensity" storm in HydroCAD.

I guess I have two goals here. One is to demonstrate that with the proposed basins, I have enough capacity to store the 100-yr 1-hr event as determined by the rational method, as dictated by the state. I don't use hydrocad for that process, it is a pretty simple set of calculations and spits out a required volume at the end.

What I am using hydrocad for is to demonstrate that the various pipes, ditches, etc have the capacity to handle the flows, and specifically the basins. There are four basins, tied to each other with pipes, so there's a lot of interaction, with dynamic tailwater conditions on the basin pipes. For that reason I am using the dynamic SI method. For consistency, I am wanting to use a 100-yr 1-hr event but that doesn't sound correct for reasons listed above. There is no specified 1-hr distribution of which I am aware.
(The local jurisdiction is not sophisticated with stormwater regs and defers to the state standard).

Any comments would be appreciated.

Sizing a pond based on rational method really isn't recommended. As discussed in many other threads, rational method is designed to give peak flows, but is marginal (at best) for estimating runoff volume. But if that's what your regulations require...

If you want to run an additional SCS analysis, then I suggest you use the appropriate 24-hour SCS rainfall distribution for your area, (Type II, III,etc) together with your 24-hour 100-year rainfall depth. Since the 24-hour rainfall distribution includes all events up to 24 hour duration, you don't need to run a separate analysis for the 1 hour event.

If you want to stick with a 1 hour duration, then you could use the constant intensity distribution with a duration of one hour, but I don't really recommend this approach.