If you define an arbitrary plane in free space, then the mass, momentum, and energy of virtual particles crossing the surface will balance (statistically over a given time period). If this plane is parallel to and very close to an event horizon, then an imbalance occurs as some of the virtual particles are lost to the event horizon, making re-combination impossible and thus creating a surplus. The net surplus at the other side of the surface in terms of the mass, momentum and energy of the orphaned particles is then balanced by the mass, momentum and energy of the particles that strayed far enough from the original plane to cross the nearby event horizon. This implies that if an event horizon occurs, it will leak energy (mass and momentum from orphaned virtual particles) into the "real particle" universe.

My problem is that this model leaks from the "vacuum energy", and creates Hawking Radiation in equal, statistically interchangeable, forms on both sides of the gap between the arbitrary plane, and the event horizon. Everything you loose on one side, you have therefore gained, as an orphan, on the other. You get opposite charge, but still the same individual mass and overall momentum. So the total vacuum energy erodes (very slowly) but the black hole expands (also very slowly).

Even if you invoke negative energy particles, they are either repelled by the gravitational field (making things much worse) or at best they cancel the in-falling Hawking Radiation, and produce some very strange external emissions, but no net loss of mass from the black hole. I can’t create a bias, as it implies a bias in free space. Besides, I can’t tell in advance which virtual pairs will become separated, or which individuals orphaned, so there is no biasing mechanism available.

If this was the case then black holes would remain stable. Can someone please explain what is missing from this model? How does it differ from the accepted model of Hawking Radiation and the decay of Black Holes?