Weird pH reading in my calc reactor

[dangros]

Late last night (12:20am), I was too tired to sleep so I decided to check my apex. I noticed that the pH in my calc reactor was way too high. Looking at the graphs I could see that this issue started 24hr before - slowly at first, drifting, then a steady upward climb in pH. I initially suspected the cleaning lady must have hit the regulator and somehow jarred the power cord out. That would have stopped the CO2 from getting to the reactor. But the timing wasnt right and didnt make sense b/c the reactor ph would have spiked much faster vs drifted slowly upwards. I unplugged, reset the reg, and cranked it up a bit after noticing the frequency of bubbles in the bubble counter to be a bit less than normal. I then waited a few min and saw the pH falling quickly. Happy with my "fix" and puzzled by the issue, I went to bed.
This am, I woke up and noticed that the pH had come down drastically but not all the way and then started to drift back up again! This made no sense! I didnt see a leak and the CO2 canister was full. I'm attaching the apex graph b/c it puzzles me. Notice on the left side, the apex is kicking in perfectly and keeping the pH in a very tight range... then the drift and spike, the correction after I notice, and the subsequent drift. WEIRD! Maybe it's a leak or blockage that I havent found... yet. Any ideas?
here's the image of the graph: https://flic.kr/p/AVuCYn

 

[reef life]

You need to program apex to turn on co2 (soleniod)when ph reaches desired level and shut off when it gets to your low ph level. How fast is your effluent drip?

 

[gcrawford]

Possible that the media has broken down limiting flow through the reactor. Mine did this and the media was mush when I opened it up and checked it. I had the ph set too low on my apex.

 

[dangros]

I'm pretty sure I just found it. When I purchased my reactor from Aquarium Plants, they recommend that I use 2 check valves in case one failed. They are insanely expensive so they must be good right?! NOPE! One failed by sticking and creating too much resistance for the C02 to get through. Once I pulled it out of the loop, I saw WAY more bubbles in the bubble counter and even had to turn down the pressure. After only a few minutes, I can already see the pH plummeting towards normal.
@reef life, the apex is programmed in the exact way you mentioned. You can see it in the graph I provided in the first link.
This does bring up a good idea: I will program an email alert if the pH gets above or below a certain threshold in the reactor.