Here is a guide.
Max.

I was not able to display the pdf but did some research on snubbers. One variation seemed to be a series cap/resistor combo in parallel with the motor. I do not have complete motor data so cannot calculate the exactly correct values, but one rule of thumb suggestion was .1 mfd/120 ohms.
Any thoughts? Is this as good as the varistor solution?
I had made an earlier suggestion to connect both sides of a 30a rated DPDT in parallel (the circuit requires only a SPDT.)
Assuming that contacts probably won't close at exactly the same instant (thus effectively raising the 30a capacity to 60) I would still think this arrangement would effectively double the time until the relay unit would need attention since initial contact closure would randomize between the two sides. Does this make sense?

 

I have no specific data on this but I would expect it not to be random. A single moving piece in the relay begins to move to open/close the contacts. Because the mechanical parts will not be precisely identical measurements, one set of contacts will open/close first.

 

I was not able to display the pdf

Strange, I have no problem seeing it from the posted link, you may get the PDF from the Electrocube site direct, as this is where it originated.
http://www.electrocube.com/resistor-capacitor-networks
Max.