Has anyone ever had problems with volume sources not producing the desired energy?

I have a heating element that I defined as a 1000W volume source (along with several other volume sources in this simulation).  For some reason two of the elements (same geometry as each other, but greatly different from the others) don't solve correctly.

When I probe surface parameters the other elements have Integral Heat Transfer rates equal to the specified Source conditions.  However, these 2 elements are WAY off - i.e. 1000W volume source that only transfer 93W.

This is however the first time I have used non-radiating surface definitions in proximity.  (They are heating elements that are protected by a quartz sleeve to prevent interaction with process gases.)  If I replace the quartz with a high emissivity radiator (graphite, etc) the temperature goes up and the volume source works correctly.  The fact that it has happened both times I used non-radiating nearby makes me suspicious...

There is no heat generation under Volume Goals (likely because it SHOULD define that as a requirement, not a goal) - do I need to add a Surface Goal for Heat Transfer Rate instead?  I've had this happen in 2 simulations recently (the same element failed to heat correctly), but never before (with a different configuration of the same simulation).  My first inclination is to refine the mesh more to make sure it has accurately resolved the body and distinguishes radiators from non-radiating bodies, but I would think if it was confused it would say "it generates 1000W and doesn't radiate therefore it goes super-nova and temperature skyrockets".

Thoughts?......