Alan, I guess I'm not making myself clear. There is no need for a "DC bias" of the power input.
The "wire trick" (simply running a complete second circuit with both conductors hidden in a single "wire"), uses only the normal A/C voltage supplied by the mains. It isn't the voltage that is rigged, it is the current. The report states that the "phase 3" "hot" wire appeared to have no current flow. If that wire were rigged so it provided the same amount of power as the other two legs, then the actual staady-state power input would be completely normal A/C, but varying between 1200 Watts and 400 Watts, instead of the measured 800 Watts and 0 Watts. This provides an apparent COP of 2.5, which is exactly what they measured. Your model shows that there is no apparent "excess" power until the E-Cat reaches operating temperature. This could be easily accomplished with the addition of a thermostatically-controlled switch for the 3rd phase line (thus being totally automatic in operation). John
