Are you an engineer? If you are loading only one of the two beams and attaching them with bolts through the webs I would not count on them to share loads equally (when I say equally I assume they have equal stiffness, if they don't then I would not count on them to share load based on their stiffness). They will share load, but not necessarily equally. Additionally, you will have some loacalized bolt stresses in the web as one beam tries to deflect and the other is trying to hold it back by the bolt rotating as a rigid body which is going to bend the web out-of-plane.
How are the beams loaded? What are the sizes of the beams?
How are the beams loaded? web, top flange bearing, etc.
Is only one beam loaded?
What is the detail of how you connect them together?
What is the spacing of the connectors?
How large are the beams?
I am an engineer and my belief goes with StructuralEIT's response,I understand situations are different but the general principal is that it is largely unacceptable regardless of whichever situation we are in, the theory could be O.K. in some rare situations only.
No answer with out more details: Where are the loads applied and in which direction? How many and what size bolts connect the webs? What is the spacing between the webs and between the bolts?
In general, I would guess uneven loading of the beams if the loads were not applied equally and symmetrically.
If I had to do this, I would not connect the beams with thru the web bolts. If you really need both beams for strenght and the load is not symmetrical you will have a problem, because I´m not so convinced that they will share the load. As has been stated above, it depends on many things.
I would have tought about connection plates on the flanges. If the load is not symmetrical, you will end up with torsion on the built up section. I think this arrangement is more adequate in that respect.....
