Here's the mod, posted with Jorge's permission.
NOTE: The new reactors don't need this mod as the new design is being shipped
Piping assembly modification
The following modification will minimize the issues with flow slowing down over time, while maintaining the reverse flow and its benefits:
Move the effluent output from the top line (suction of the pump) to the bottom line Outlet of the pump). The pump will maintain a more stable pressure but will require that the effluent control valve be installed also at the output and not at the input of the reactor. As this literally will run the reactor pressurized at whatever the inlet pressure of the inflow water is, the overall higher operating pressure will have the advantage to make the small variations in the loop pressure negligible so the effluent flow will be more stable. In addition, higher operating pressure will help for faster and better dissolution of CO2. At a higher partial pressure the water is able to hold more CO2 increasing the overall performance. The only disadvantage with this modification is that if too much pressure is applied to the inlet, like when feeding directly from the return line, the excess pressure will make the reactor leak at the flanges if not potentially crack it open. To prevent this, a feed system with a bleed line must be used to limit the inlet pressure. (See attached diagram)
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/0.jpg>
Needed parts and Instructions
For reactors without uni-seals:
a) 1 Ã"šÃ‚½" Male adapter
b) 1 Ã"šÃ‚½" Threaded Tee (SxSxFips)
c) 1 Slip * Thread Bushing 1/2" * 1/4"
d) 1 John Guest male pipe connector 3/8" Tube x Ã"šÃ‚¼" NPT
e) 1 Ã"šÃ‚¼" or 3/8" John Guest X Ã"šÃ‚¼” MPT valve*
f) 3" of Ã"šÃ‚½" PVC Pipe
g) 3/8" OD tubing, enough to go back from the reactor to the sump
*If you have already bought the needle valve I sell, you won't need to buy another valve since you can reuse the ball valve it replaced.
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/1.jpg>
1 Begin by emptying your reactor.
2 Disconnect the effluent lines
3 Unscrew the pump output from the elbow in the lower part of reaction chamber
4 Remove the effluent valve on the circulation assembly and the JG fitting from the effluent output
5 Screw the new 3/8" JG (d) where the effluent valve was
6 Screw the ball valve (e) in the output
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/2.jpg>
7 Glue the pipe (f) and the bushing (c) to the T (b) and screw the effluent valve to the bushing
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/3.jpg>
8 Screw the male adapter (a) to the elbow in the reaction chamber.
9 Glue the assembly you just made to the male adapter (a) making sure the pump output can be screwed into the T without much stress on the flex tube.
10 Screw the pump output to the T (b) (you might have a blue ribbed tube instead of the tube pictured)
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/4.jpg>
11 Connect the input to the new 3/8" JG fitting (d).
12 Connect the output to the effluent valve using the new piece of 3/8" tubing (f).
13 The new valve on top of the circulation assembly (e) will work as a purge valve. It will need to be opened to let the air escape while filling the reactor. You can use the old effluent output tubing (green) to take the valve output back to the sump to avoid spilling water.
For reactors with uni-seals:
a) 1 Ã"šÃ‚½" Threaded Tee (SxSxFips)
b) 1 Slip * Thread Bushing 1/2" * 1/4"
c) 1 John Guest (or similar) male pipe connector 3/8" Tube x Ã"šÃ‚¼" NPT
d) 1 Ã"šÃ‚¼" or 3/8" John Guest X Ã"šÃ‚¼" MPT valve*
e) 2" of Ã"šÃ‚½" PVC Pipe
f) 3/8" OD tubing, enough to go back from the reactor to the sump
*If you have already bought the needle valve I sell, you won't need to buy another valve since you can reuse the ball valve it replaced.
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/5.jpg>
1 Begin by emptying your reactor.
2 Disconnect the effluent lines
3 Unscrew the pump output from the elbow in the lower part of reaction chamber
4 Pull the elbow out of the uniseal
5 Remove the effluent valve on the circulation assembly and the JG fitting from the effluent output
6 Screw the new 3/8" JG fitting (c) where the effluent valve was
7 Screw the ball valve (d) in the output
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/2.jpg>
8 Glue the pipe (e) and the bushing (b) to the T (a) and screw the effluent valve to the bushing
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/7.jpg>
9 Insert the new assembly into the uni-seal in the reaction chamber.
10 Screw the circulation pump output to the T (a) (you might have a blue ribbed tube instead of the tube pictured)
<img src=http://img32.photobucket.com/albums/v97/saltyseaman/Reeftek/8.jpg>
11 Connect the input to the new 3/8" JG fitting (c).
12 Connect the output to the effluent valve using the new piece of 3/8" tubing (f).
13 The new valve on top of the circulation assembly (d) will work as a purge valve. It will need to be opened to let the air escape while filling the reactor. You can use the old effluent output tubing (green) to take the valve output back to the sump to avoid spilling water.
Dial in directions
When setting the effluent rate, set it to your initial value and let it run for 12 hrs. At this point, if you are using a very slow flow it might have stopped or slowed down. Don't panic. It takes some time for the pressure to equalize in the reactor. Do not close the valve. Just open it enough for the effluent to begin flowing again and set it back to the rate you need. The flow will stay constant. It also helps if you don't use a flow slower than 20ml/m. Reduce the CO2 count rather than the effluent rate beyond that point to adjust the output of the reactor.
). I guess the Kalkwasser and water changes will have to do for now. I'll send a pic of what I had Jorge do, when I get the reactor tomorrow. I can still run the reactor without media and CO2 addition I suppose just to see if everything is a go... but it probably wouldn't give me the same result. But at least this way, maybe we can see if it's the pressure that's causing the slow down.
