My experience is that pH probes constantly get out of calibration - and air intrusion happens sometimes, media gets melted - which creates confusing situations. Here are a few situations that I deal with every six months or so with my GEO 618.
1. Air intrusion. This happened when I used a peristaltic pump to "pull" water through the reactor. Especially, since the reactor was on a shelf above the sump. The reactor pH might be ok (6.6 or something) or might be high (7.0), but the effluent ALK is too low (15 dKH for example), or flow may have stopped due to air damming. Running the reactor with a little positive pressure will solve this problem. To test it... make sure water is flowing through, shut off the CO2 gas... and see if the air bubbles disappear. If they don't after 24 hours... you have air intrusion. To run at positive pressure, use a pump to push the water into the reactor and something to limit the flow out of it. A peristaltic pump or pinch valve like the GEO instructions say can control the outflow. Just don't use a big pump to feed the reactor or your might crack your reactor.
2. CO2 saturation with out of calibration probe. So... the reactor is gassed out and my apex says the pH is not yet at target (6.8 or something). I am running with a little pressure now (see situation 1 is corrected), the flow is good, the effluent ALK is 38 or 40 dKH. In this situation... the probe is wrong. The real pH is probably 6.4. Re-calibrate the probe and decrease the bubble rate or increase the flow rate depending on the needs of the tank. If you are trying to use pH to automatically control your reactor within a narrow range... then things are off the rails at this point and God help you if your tank ALK is now 15 dKH or something. Also.. you might have melted your media and set yourself up for the next confusing situation... so look closely to see if it is full of dust. See situation 3.
I do NOT rely on the pH probe to shut the CO2 on AND off constantly within a narrow pH range. It is more important to find the steady bubble rate from your CO2 which will run continuously without much need for automatic control. This way... I will be less likely to spike my ALK if other things get out of whack.
If the pH goes up and down due to flow changes... but the CO2 bubble rate is stable... then Alk delivery will stay stable. BRS did some nice experiments to show this. Again.. if CO2 delivery is stable... then modest changes in flow and therefore pH don't matter much.
I use the pH probe and controller only to shut the gas OFF if the pH in the reactor gets really low - like 6.5. At this point there is probably a clog in the tubing.
... and don't program the controller to let gas back in until the pH is back up to 6.9 or 7). If the pH goes back up to 7... then you have flow.
3. Bad media. So... the reactor is gassed out and my apex says the pH is not yet at target (6.6 or something). I have calibrated the probe. I am running at pressure (so there is no air intrusion), the flow is good, but the effluent ALK is just lame... like 15 or 16 dkH. I this situation, the media is probably old, melted, dirty, covered with algae and needs to be rinsed and at least 1/2 or more replaced.
My next reactor will be a saturation reactor, than runs constantly with bubbles and at pressure, no need to monitor pH. So the gassing out situations I described above should no longer give me fits.
