Jessica,
Even if the program runs the whole way through, you still get error messages and warning messages. You can view them by clicking the second to last button from the right (immediately to the left of the button with the letters DSS on it).
Your error/warning messages will tell you if there was no subcritical answer (in which case it would default to critical depth and indicate a dashed red line on your cross-section plot). You want to start with subcritical flow, as this is the highest water surface elevation.
If your error/warning messages indicate that critical depth was used extensively in your model, you still might not need to run subcritical flow (or preferably, mixed flow). I had this problem once with a small stream that had a well-defined channel, but the channel walls were heavily vegetated. By using my channel bottom Manning's coefficient on the channel walls too, I was overestimating the velocity in the channel. When I changed my Manning's coefficient on the channel walls, I no longer had the critical depth warning messages.
Of course, if your problem is high water surface elevations, then it's unlikely that you've underestimated your Manning's coefficients, but perhaps overestimated them.
Describe the area where you're having trouble and tell me what Manning's coefficients you've used. (Did you get them from the Hydraulic Reference Manual or another source?)
You're right that you can't specify boundary conditions for junctions, but in the Geometry Editor you can specify a junction width and an entry angle. An error I made initially with my junctions was stationing my minor stream channel right up to where it meets the major stream channel. In fact, under flood conditions, the water will be flowing on the overbanks and your entering stream should be stationed with zero where its waters meet with the water flowing on the overbanks of the main channel. You then enter a junction width to account for the width of the main channel.
In the image above, Stream 1-2 is the north-south stream; Stream 2-3 is the east-west stream. If you use the Energy equation you don't need to specify the entry angle, but if you use the Momentum equation you do.
The HEC-RAS help screen says the following about editing junctions:
Reach lengths across the junction are entered in the junction editor, rather than in the cross section data. This allows for the lengths across very complicated confluences (i.e., flow splits) to be accommodated. In the cross section data, the reach lengths for the last cross section of each reach should be left blank or set to zero.
In this example the energy equation will be used to compute the water surface profile through the junction. If the momentum equation is selected, then an angle can be entered for one or more of the reaches flowing into or out of a junction. The momentum equation is set up to account for the angle of the flow entering the junction.
Good luck,
Francesca
