Random runaway power usage in circuitry with transformers

Connecting power transformers in this obscure manner may highlight some sort of rounding error. The power used by the last transformer connected into the circuit started using power incrementally slow, then later maxed out at 5Kw. This creep up to 5Kw occured when there were potential loads on the circuit. After severing the lines to the loads, and reconnecting just the transformers as they are, the power consumption jumped to 5kw on one transformer immediately.

I don't know what the intended behaviour of such a circuit is supposed to be - save to act as a sort of two-way bypass or zener diode. The bug here is that it's not actually using power from the batteries, but reporting a high power consumption on the line. - there's some intense power looping going on between the transformers causing line damage everywhere. Please note this was just an experiment.