Ok so that would indicate that those stored procedures with schema lock waits are trying to modify the schema in some way? This sounds strange because the SPs that are waiting for those locks do not make any schema modifications at all. They are just simple select statements. Does this sound right to you?

I would like to add one more thing to all of this. At the beginning when I said that each time this happens there is one particular stored procedure in here that is listed about 70 times? Well, if I use my load testing tool and load it up with about 20-30 concurrent queries running this SP then I can reliably recreate a load of RESOURCE_SEMAPHORE waits. These wait types indicate that this SP is chewing up the memory. If the memory is being chewed up then am I right in saying that SQL Server will start ditching cached plans from the plan cache as it starts running out of memory. If this is correct then would this in turn cause all of these Compilations?

Just a thought?