What is the difference between these moment connections?

It really comes down to how thick your end plate is and how stiff you need it to be. I agree that the flange plates are probably stiffer...but also consider you'll have a bit of bolt slip before the flange plates engage.

Not only is the extended end plate cheaper to fabricate, it is also less time consuming to install on site - fewer bolts, easier placement of the beam, less sensitive to over-roll/under-roll on the beam. Lots of reasons why the end plate is better.

I'll use the extended end plate for lighter capacity cranes and smaller frames but on the larger frames with higher capacity cranes I prefer the bolted flange seat connection for better rigidity. I'm asked to go out of my comfort range to save money. Curious what others opinions are on the extended end plate.

I don't think there is anything wrong with it, I've used it successfully in the past. It is a recognized standard moment connection. Check out AISC Design Guide 16 for more information. You can always put a stiffener above and below the beam to make the connection more rigid.

May also want to check out AISC Design Guide 4 and/or AISC 358-16 (prequalified seismic connections) - both have design examples for end plate moment connections. Former has a discussion on pros/cons of this type of joint in chapter 1.

Extended end plate. Love it. Better control for fabrication and easier for fit-up.

I've only ever used the one with the red arrow.

It is a good rigid connection.

We once had deflection problems with it after they failed to tighten them up properly and clamp the plates shut. It took an inspection in a boom to see what had happened. There was a slight gap, which translated into excess deflection at the ridge.

I have several opposing view points, and offer them to see what response comes back..
1. Why use headers at all? - if the concrete is adequate, it can resist the moment via larger columns
(in the range of 18 x 60)with heavy base plate & anchors.
2. Assuming the concrete is inadequate, and hence headers are used to control sway, I question the need for moment connections, The lateral loads you are trying to control are like very light.. in the range of 2000# (ballpark 11,000# x .2 = 2200# minus distribution to next column) If the loads are light, moment connections not needed?
3. This takes me to the 12x65 header - seems way oversize considering the light loads.
I suspect it would need to be in the realm of a light W16 to resist deflection though.
4. If you wish to consider the "free standing column" approach (no header) - I remind you that each column sees only 1/2 of the 2200# load.
I'm going into all this because I run thru these systems very often, and I'm trying to better understand several aspects of the design

I've questioned several PE's regarding the connections in this type of frame and I keep hearing that Moment connection is not needed, - we are only trying to control column rotation - and not necessarily transmit flange strength (which is what the moment connections do)

I look forward to hearing you folks response to this