If a PED had interfered with the high-impedance parts of a circuit that controlled a relay, then I could see it, but by the fact that we are talking about relays acting on their own I assume that has been ruled out
Could that possibly mean high-impedance parts or connections in the wiring close to the relays somewhere between the relays and ELMS?
If there were high-impedance points in the connections then the relay would probably fail to operate at all. After all it will require something in the region of 100mA to operate. ie around 3 watts.
Just toying with ball-park figures, but imagine a relay with a 28 volt coil. Number of turns required would result in around 1mH inductance and a self-resonant frequency around 1MHz.
What sort of field strength at 1 MHz would be needed to induce 3 watts in the coil??
P (dBm) = E (dBuV/m) - 20log F + Antenna gain - 77,2
Putting the relay coil and associated wiring as a 0 gain antenna and Power as 34 dBm the required field is 111 dBuV/m, or less than 1 V/m.
Do medium wave transmitters still exist?