I look further and realize Grundfos uses what they described as "mains borne signalling" to get the commands to the pump over the legacy 3 wire cable running to the well. I scope the 240 line coming out of the pump controller back to the electrical panel and see data pulses near the zero crossinf very similar to X10. It looks like they do not bother to filter the signals to stop them at the controller. Damn them!
Interesting! Are these data pulses constant or random? If you're tech-savvy enough to own and use a scope, you may want to assemble an:
X10 Oscilloscope Adapter for a better view.
They reccomend (at my cost) Corcom 10VB6 EMI filter for about $25. I have one on order but just to be sure I also order a pair of the X10 Pro XPF 20A filters.
Filters are what first came to my mind too, but since this GRUNDFOS signal seems to be so similar to X10's signal, I wondering about "just how much" filtering you're going to need. It would be interesting to see how much signal strength the GRUNDFOS generates on the ELK X10 Signal Strength Meter. If it's pumping out something like 10V right on the X10 "sweet spot", then you may need more attenuation.
GOOGLE NG:comp.home.automation for ACT, Phil Kingery and filters - IIRC, I remember reading about 30:1 and 60:1 numbers somewhere. Personally, I'd perfer installing one 60:1 filter instead of two 30:1 filters.
Best of luck and keep us all updated with your progress!