[leman_opc]

BTW I changed my pump for a new one this weeked and I am pretty sure the flow issues I had previously (normal injection couple of seconds, then failsafe activation; or even no flow at all) were due to dying pump and not due to injection mixture freezing. As the pump is activated only when the injection actually starts it is highly important that it spools up fast enough to keep the pressure in the lines; this seems to be a problem when the pump motor shaft bearing starts to fail (i.e. it is not seized fully but the motor spool time is increased).

Interestingly, I had my old pump disassembled and cleared the debris from the bottom shaft bearing (it was completely seized by then) so that it could work at least somehow until I replace the bearing. The spool time was still bad - but I arranged the feed from the battery directly (i.e. the pump was on all the time, whether there was any injection or not) and this setup worked OK for me for a couple of evenings. So I'm pretty confident it's the spool time being significant enought for the failsafe to kick in which was my problem previously. Probably, if I was injecting enough for the pump capacity to become critical this would be another problem I'd face with a failing shaft bearings...

With a new pump there is no problem anymore even with the pump powered up as should (i.e. only when the injection starts).