We would like to run long term (~15 days) experiments in a pioreactor cluster consisting of a leader and two workers, both in turbidostat or morbidostat mode. As a probe, I ran an experiment in morbidostat mode, but:
-One of the workers stopped running after 4 days when I removed the flask to see if there was biofilm formation.
-Another of the workers stopped after six days, also while checking for biofilm formation. I tried to restart it (reboot the pi) but it did not respond.
-The leader stopped working at day 8 for no reason.
I realised that removing the vial caused a spike in the voltage detected by the photodetectors and the consequent system crash. I had experimented the same problem with our custom diy morbidostat. I think this consideration (temporarily stopping OD readings when sampling) should be added to the manual. Also, I noticed abrupt changes in OD when reinserting the vial, and I opened a thread about the idea of installing a poka-yoke solution in the vial/holder to reinsert the vial in the same position. However, I then realised that even rotating the vial does not achieve that the OD come back to similar values as after I removed the vial. Do you have any idea why?
I would like to know which is the best solution for long term experiments in the pioreactor. I suposse than the crashes are related to a long cycle of writing/querying to the database. Could reducing the OD sampling frequency solve this problem? Maybe every 30 or 60 seconds instead of 5 seconds.
Hm, this much crashing is odd. I’ve seen clusters run for weeks without issue.
One known thing that can cause a crash is if the vial is removed, and the heating surface is active and hot, a thermal runaway may occur. This is mitigated with the Pioreactor 20ml v1.1, but still could occur. The logs will say something like “surface exceed temperature X, shutting down”. You can “pause” the temperature-automation job to keep this from happening, too.
It’s not possible that a spike in OD or or too much writing to the database cause crashes.
Hm, a few things to consider:
check power supplies, looking for at least 12.5W (Volts * Amps) in your PSU.
How are you connecting Pioreactors? Via a local-access-point or some other network?
It might be useful to send us the logs for this experiment (Export data → select this experiment → Pioreactor Logs → Export), so we can see if they tell us anything specific about the crashes. You can share here or email us at info@pioreactor.com.
However, I then realised that even rotating the vial does not achieve that the OD come back to similar values as after I removed the vial. Do you have any idea why?
Possibly a temperature artifact? I’ll have to think more about causes here.