I've had a longer look at your picture.
(Just as an aside, it's a lot neater than many of the rat's nests one comes across. I find I have greater confidence in the capability of someone who produces tidy work, despite knowing that the two don't necessarily go hand in hand!)
The 24v supply does seem to be isolated from the Pi's supply, and the GPIO wiring seems to be kept well away from the switched wiring, so I can't see a way for the one to affect the other.
The only weak point (if it is one) is the 5v feeding the relay coils. Are you taking that from the Pi? I suppose it's possible that, when a relay is switched on, the current surge could affect the Pi's power if the Pi's supply is on edge.
It's possible that the inductive load of the valves could be causing interference to radiate across to the Pi. However, that should only happen when a valve is switched off, not on.
At switch-on, the current rises slowly, but at switch-off, the current falls quickly, creating a quite large voltage to appear across the relay contacts (and maybe arcing).
Since it's DC, you could try putting a capacitor across the relay contacts. I've no idea about the size of capacitor - that will depend on the inductance of the valve. A diode to suppress the spike would have to be placed across the valve, not easy with the way you've wired them (though it may be possible to put one inside the valve, if there isn't one already there)
If you're convinced that it's the magnetic field from the valves that's causing the problem, mu-metal or permalloy shielding could be an answer (though I've never had much luck with the stuff - it seems more trouble than it's worth. YMMV)
One thing you could try:
With the Pi just idling, use a pair of needle-nosed pliers to short across the relay contacts.
That will switch a valve on, without any associated software input, and without the relay coil being activated.
If it really is the valve current causing the problem, the Pi should crash.