The battle of Trent's Reach (the long E-W area around the I295 crossing) ... from a previous discussion.
http://www.historynet.com/air_sea/naval_battles/3025591.html
The three confederate ironclads were:
Richmond: 16'
Virginia II: 13'
Fredericksburg: 11'
They were able to make it down the river at high tide, but Virginia II and Richmond were stranded under fire when the tide went out. Since the tide was about 4' here, I surmise the historic channel at high tide was about 18', and CSS Va2 was slightly out of the channel.
Earlier episodes in the war:
The USS Pawnee threatens Richmond.
The Virginia is scuttled, because even completely unloaded, her draft (18-19') is too deep to pass through the area.
If modelling shore batteries that were significantly downstream (Drewry's, Chaffin's, Signal Hill, etc.) is getting in the way of adjusting the region, then the border of the region needs to be adjusted. Drewry's & Chaffin's Bluffs are about 3.5 miles upstream from Trent's Reach, I would put the border of the region just north of that.
As I see it, the problem is not with a massive landing on Richmond from transports, that can be done no matter what depth is used. It is not with the blockade of Richmond, which was accomplished by a combination of deep water ships near Fort Monroe and the invasions of the NC Sounds. That leaves naval fire support as the problem. The ability to support a direct amphibious assault on Richmond with Steam Frigates and Armored Frigates is gamey. I think everybody knows my opinions about balancing fire support, but I would like to see this issue fixed, just in case naval fire support ever gets fixed for vanilla.
I think what really throws off any discussion of this area is CSS Virginia. She was an ironclad, just like an Eads city-class (in game). In RL CSS Virginia had a 21' draft, as compared to 6-7' for a city-class ironclad. She was built for local service in Hampton Roads, and would've had great difficulty in getting anywhere else.