[Delphinus]

There's already a shutoff valve, so the pump maintenance angle is covered.

In the case of a power failure .. there may be some air introduced .. I'm not sure how to completely eliminate that variable. I had also thought of using a separate feed line from one of the overflow chambers but it's basically still the same problem, in a power failure it will siphon until it sucks air.

There are a couple mitigating factors though, which I thought might have been enough: (I dunno, maybe not? I realize it's not foolproof but this is why I decided to chance it...)
- the reactor sits lower than the sump wall, so the reactor itself doesn't drain. There's always water in the 2 chambers.
- With an upflow and the output of the 1st chamber fully open, and on top, any air will escape out right away. It might make 1 pass through the media but hopefully no more.
- They say when the reactor's cycled, that you can increase the flow-through, so obviously the media can handle some O2, just not a lot. (??)

Anyhow, this is the only way I can see to slow down the flowthrough considerably from what I had before. Since you can increase the flowthrough once it's cycled I was hoping to change the feed once again to something else (after all I want my drip acclimation feedline back sooner or later).

But if it's no good then it's no good.

I'll try seeing if the recirc pump is enough to draw water from the sump, before going online. If that will work then it's not too hard to ensure that the input is always submerged. Thing is there are a lot of bubbles in the sump so the odds of sucking in the stray air bubble are also nonzero in this situation either. Although probably less and also don't have to worry about the power failure scenario.