Dissolved oxygen surprises

Spaced Cowboy

South Bay Reefer
Maybe this is well-known to everyone out there but it was news to me, thought it was worth a write-up.

I have one of the Dissolved Oxygen probes from Neptune Systems which I've recently put on my system and I was looking at the readings and trying to figure out what was going on in the graph below:

My first thought was that it might be something to do with the lights - it always starts at ~21:45pm or so, and ends at 3:00am. I couldn't think of any reason why the end might match, but I couldn't think of anything *else* that was correlated with the start of the period. Looking at my light timing graphs ...

... however, revealed only a *very* rough correlation with the lights going off and the DO spiking. That's not sufficient to explain the delta-function-like spike in Dissolved Oxygen, and certainly doesn't have any bearing on the slump in DO after 3:00am.

The DO sensor is placed in the 3rd section of the sump (the one which the Reeflo Dart return pump pulls from). It's after the skimmer and bubble-trap and generally in very calm water but with reasonable flow (~3000 gph),

The only contents of this section are the sensors (DO, Ph, Temp, Salinity), the floating algae-scrubbers, and to keep the bottom relatively clean, there's a pump that is turned on for 3 minutes in every 5 that just stops any detritus from falling to the bottom and building up there.

I next thought that it might be the algae-scrubbers affecting the water in that sump-section, but looking at their timings they had a turn-off at 1:00am, turn on at 3:00am. These are the SURF8X scrubbers from Santa Monica Filtration, and require air to be bubbled up into them, so there was a time and an oxygen connection, but for the life of me I couldn't see how they could fit the 9:45 -> 3:00am profile of the "top-hats" in the graph.

So I went through every outlet on the Apex. Eventually we got to the "A-ha" moment...

Ding ding ding. Ladies and gentlemen we have a winner.

The UV (an Emperor Aquatics Smart HO UV, 50W) is plumbed in right *after* the sump exit. You can just about see it as the upright black column by the stand leg in the below:

... so it's as far removed from the DO sensor as it could possibly be. Even then, only 6g/minute goes through the UV filter according to the flow meter, and yet this is the culprit. It hadn't occurred to me that the UV filter would have such an effect on the dissolved oxygen, but it certainly does for me. If I turn off the UV filter briefly (just before 10:00 am, below), I can see the DO spike back up to the 92,93 or so level for the quarter-hour graph reading; with the UV switched back on, we flip right back down to the 86-88 level.

So maybe if you're investigating Carbon dosing or some other technique which has the potential to reduce the dissolved oxygen in your tank quickly, it might be a good idea to give yourself a bit more buffer and turn off the UV filter, especially if you run high-flow through your UV filter, I tend to run it low-flow and that already makes a measurable difference...