djbon
New member
Great. Thanks for the information.
Nutrient Pulse Reactor (DIY DyMiCo filter)
- Thread starter dartier
- Start date
Great. Thanks for the information.
A bit of progress, but nothing worth posting about. I kind of got sidetracked with removing more wall and moving the doorway to my fishroom after deciding on a 10ft tank instead of 6ft one.
As for the filter, I have ordered a pipe cleaning brush online and am waiting for it to arrive before I do final assembly of the plumbing (inside the filter) just so I can test if it will be possible to clean the internal pipes with the planned pipe pattern.
For the media separator, I was going to source perforated PVC sheet like the original filter uses, but I think I am going to try using the white needlepoint mesh instead. Far easier to find locally (Walmart), cheaper, and it should mould itself nicely to the contours of the coarse layer it rests on. I was worried about the PVC sheet tilting and spilling the sanded into the coarse media over time. The fact the tank I am using for the body of the filter has a rim and cross brace, also influenced my decision to use the flexible needle point mesh. To give it a bit more structural support, I will zip tie in some fiberglass fence stretchers.
For the coarse media itself, I think I am going to have to make a trip to the border to take advantage of better pricing at BRS and free shipping inside US. Probably grab a big order of Reef Saver rock at the same time. Even dry rock is crazy expensive in Canada.
Dennis
Speak of the devil and the pipe cleaning brush from hell arrives. Not a very good brush at all, but it did highlight the issue that the 45 degree bends in the outside lower pipes will make cleaning them all but impossible once the filter is assembled and filled with media. This is what I was leaning towards as the possible outcome. So I will make a new set of pipes that is just straight across. I can install them with the slits pointed towards the outside of the tank to try and pull from area the dog legs were trying to account for.
Dennis
Dennis
If I might make a suggestion on your media separator...
Drain sleeve.
I buy a roll from time to time for multiple uses. It makes nice media bags but I have used it to separate media strata that I didn't want mixing up. It is marketed as a sleeve for drain pipes to prevent sand and dirt from clogging up the pipe. It is strong, flexible and doesn't break down, even after being in for years. Cut to length and open one end of the sleeve to make flat.
Aaron
Drain sleeve.
I buy a roll from time to time for multiple uses. It makes nice media bags but I have used it to separate media strata that I didn't want mixing up. It is marketed as a sleeve for drain pipes to prevent sand and dirt from clogging up the pipe. It is strong, flexible and doesn't break down, even after being in for years. Cut to length and open one end of the sleeve to make flat.
Aaron
Hmm, thanks for the suggestion Aaron. I looked it up on the HD site, and it kind of looks like soft mesh made of nylon. I will check it out this weekend when I am out and about.
The more that I have thought about the separator between layers, the more I have wondered how important it is to prevent the sand from making its way to the bottom area. The PVC sheet that the DyMiCo folks use, appears to have quite large perforations (5-6mm), so about 1/4". So some of the sandbed would escape over time. As the bottom of the filter is really intended to be a calcium reactor, these liberated sand particles are most likely just dissolved.
Hmm, in the manual, they do mention that you may need to replenish the sand bed after 1 year, whereas the coarse layer will last several years. That always seemed backwards to me, but I guess if the sand is being washed through the holes, then that would explain it.
Dennis
The more that I have thought about the separator between layers, the more I have wondered how important it is to prevent the sand from making its way to the bottom area. The PVC sheet that the DyMiCo folks use, appears to have quite large perforations (5-6mm), so about 1/4". So some of the sandbed would escape over time. As the bottom of the filter is really intended to be a calcium reactor, these liberated sand particles are most likely just dissolved.
Hmm, in the manual, they do mention that you may need to replenish the sand bed after 1 year, whereas the coarse layer will last several years. That always seemed backwards to me, but I guess if the sand is being washed through the holes, then that would explain it.
Dennis
Experiment
Experiment
I did a little experiment last night. I had previously built and set up a plastic box, about 8X8X20 in one of my tanks. The box contained four (4) 4X8X8 blocks of Marine Pure with two blocks placed back to back on each end of the box. An ORP probe was place in the center, along with a center exit pipe. I submerged the box below the water line in the sump, with the exit pipe slightly above the water line. This prevented water from flowing back into the box when the pump was turned off.
I used a small pump to push water into the box on both sides toward the middle, through the Marine Pure. The system ran for a couple of weeks in a small 60 gallon FOLR tank with four medium size fish.
Last night, I used an APEX DOS pump to apply organic carbon. I primed the system by first turning off the larger pump. I ran the DOS (CH2 500 ml @ 250 ml/min) in the same direction while dosing 6 ml of a homemade organic carbon solution (water, vodka and sugar). After that initial dose, I ran CH2 at 60ml/min for 120 ml per dose, about every 6 minutes to circulate the carbon slowly through the Marine Pure.
When the ORB on the Apex dropped to 150, I purged the dose out and started again to see if the reaction could easily be restarted. The second test, I let run all night and found it at 0 ORP the next morning.
I dropped the PH probe into exit tube and found the PH had dropped to about 7.4 from 8.0 I took a reading from the water in the center of the box and found the nitrate had dropped from 8.0 ppm to 4.0 ppm. The Phosphate had also dropped from 0.16 to 0.08.
I will try to see if I can automate this process so that it runs by itself. It seems that completely purging the box with fresh water may be too much because it takes longer than I expected to get the ORP back down.
Experiment
I did a little experiment last night. I had previously built and set up a plastic box, about 8X8X20 in one of my tanks. The box contained four (4) 4X8X8 blocks of Marine Pure with two blocks placed back to back on each end of the box. An ORP probe was place in the center, along with a center exit pipe. I submerged the box below the water line in the sump, with the exit pipe slightly above the water line. This prevented water from flowing back into the box when the pump was turned off.
I used a small pump to push water into the box on both sides toward the middle, through the Marine Pure. The system ran for a couple of weeks in a small 60 gallon FOLR tank with four medium size fish.
Last night, I used an APEX DOS pump to apply organic carbon. I primed the system by first turning off the larger pump. I ran the DOS (CH2 500 ml @ 250 ml/min) in the same direction while dosing 6 ml of a homemade organic carbon solution (water, vodka and sugar). After that initial dose, I ran CH2 at 60ml/min for 120 ml per dose, about every 6 minutes to circulate the carbon slowly through the Marine Pure.
When the ORB on the Apex dropped to 150, I purged the dose out and started again to see if the reaction could easily be restarted. The second test, I let run all night and found it at 0 ORP the next morning.
I dropped the PH probe into exit tube and found the PH had dropped to about 7.4 from 8.0 I took a reading from the water in the center of the box and found the nitrate had dropped from 8.0 ppm to 4.0 ppm. The Phosphate had also dropped from 0.16 to 0.08.
I will try to see if I can automate this process so that it runs by itself. It seems that completely purging the box with fresh water may be too much because it takes longer than I expected to get the ORP back down.
Attachments
Cool experiment. This is along the lines of how the DyMiCo filters operate, and how my DIY version will operate. I am not sure if you are running the DOS pump under ORP control or timed? The DyMiCo filters allow ORP to guide the process and derive the duty cycle from the ORP level and rate of descent (as far as I can speculate at this point).
When you noticed that it takes a long time to get the ORP level to drop if your flow is too high, this is likely due to the bacteria colony being insufficient to consume the oxygen quickly. Also if the ORP gets too high, you can also lose portions of the colony that are dependent on anoxic conditions and perish when the oxygen level becomes too high.
Since you are seeing reductions in PO4, your bacteria population is probably still becoming established.
Dennis
When you noticed that it takes a long time to get the ORP level to drop if your flow is too high, this is likely due to the bacteria colony being insufficient to consume the oxygen quickly. Also if the ORP gets too high, you can also lose portions of the colony that are dependent on anoxic conditions and perish when the oxygen level becomes too high.
Since you are seeing reductions in PO4, your bacteria population is probably still becoming established.
Dennis
Experiment
Experiment
The intent behind the experiment was to determine exactly how all of this might work and to see if I could get the CerMedia, housing the bacteria to perform the way this DyMiCo filter does, at least from a nitrate and phosphate removal process.
A few updates since this morning's post.
I found that the Apex does not know how to handle the lower millivolt ranges that the probe hits. The Apex bottoms out with a zero ORP reading and won't go negative. However, I can set the probe up as a pH probe and it will continue to read. I have a request in to Apex to get the raw milli-volt readings for both the pH and ORP values so I can convert to milli-volts.
I also saw what I believe is your "œknee" effect on the pHX (actually an ORP probe). Once the probe started to read about 10.5 it started to rise (or fall in reality if I were looking at the mv reading of the probe) rapidly. Check it the thumbnail below.
I also noticed that I don't need to dos the carbon that often. A small bump of water added will keep things from crashing. Also, the actual pH probe I moved into the filter is running a pH of around 7.2 to 7.3. I am not sure what that is a result of but I assume it is related to lower O2 levels.
Aaron
Experiment
The intent behind the experiment was to determine exactly how all of this might work and to see if I could get the CerMedia, housing the bacteria to perform the way this DyMiCo filter does, at least from a nitrate and phosphate removal process.
A few updates since this morning's post.
I found that the Apex does not know how to handle the lower millivolt ranges that the probe hits. The Apex bottoms out with a zero ORP reading and won't go negative. However, I can set the probe up as a pH probe and it will continue to read. I have a request in to Apex to get the raw milli-volt readings for both the pH and ORP values so I can convert to milli-volts.
I also saw what I believe is your "œknee" effect on the pHX (actually an ORP probe). Once the probe started to read about 10.5 it started to rise (or fall in reality if I were looking at the mv reading of the probe) rapidly. Check it the thumbnail below.
I also noticed that I don't need to dos the carbon that often. A small bump of water added will keep things from crashing. Also, the actual pH probe I moved into the filter is running a pH of around 7.2 to 7.3. I am not sure what that is a result of but I assume it is related to lower O2 levels.
Aaron
I was struggling to figure the chart out, but I think I get it now. This is generated from a single ORP probe and graphed as both ORP and PH using the single probe? So when the ORP flat lines at 0 the PH data set take over shown as a rising value?
Assuming my interpretation is correct, I agree it certainly looks like a nitrate knee just after 8:00 AM when the rate of change accelerates as the nitrate gets quickly consumed and the ORP reading plunges (rises on the PH graph).
I am not sure how the Apex is for custom programming, but if you can implement a rate of change detection function, that should get you a long way to making this a basic filter capable of dynamically adjusting to nitrate levels in the water column.
Dennis
Assuming my interpretation is correct, I agree it certainly looks like a nitrate knee just after 8:00 AM when the rate of change accelerates as the nitrate gets quickly consumed and the ORP reading plunges (rises on the PH graph).
I am not sure how the Apex is for custom programming, but if you can implement a rate of change detection function, that should get you a long way to making this a basic filter capable of dynamically adjusting to nitrate levels in the water column.
Dennis
As far as I know ORP is kept between 0 and -200mV in the dymico system where ORP is the sole measure regulating flow and dosing of carbon. The bacteria responsible for converting your NO3 into N2 (that you feed with the carbon source) also produce CO2 hence the lower pH in the reactor. The Dymico system only injects additional CO2 for the 'calcium reactor' function if needed to keep CA/KH at required levels in the aquarium. So it is not required if other means of supplementation are used or non calciumcarbonate media is used in the filter.
Experiment
Experiment
Dennis, you are correct. Unfortunately the Apex ORP bottoms out at zero, which is the value, zero mV. At -200 mV, the only way to get the Apex to read that low is to tell it that it actually has a pH probe connected to it, not the ORP probe that is actually connected to it. Also, the values are upside down, graphically. High pH is actually low or negative mV or ORP.
Apex did respond to me today and the 10.5 shown on the pH scale equates to -203 mV, so it seems this is the target number.
Also, pH continues to drop within the core of the filter. It is now down to 6.8. I now have the system set up to bump on the main pump for 7 seconds when the ORP is below 175 mV. This feeds a small amount of fresh water into the system while a small amount of nitrified water is pushed out the exit pipe. Without this activity, the ORP range will dip into the -300 to 400mV range.
I have not added carbon since earlier today when i manually purged the system too much and the reading got down into the 0 mV range.
Experiment
Dennis, you are correct. Unfortunately the Apex ORP bottoms out at zero, which is the value, zero mV. At -200 mV, the only way to get the Apex to read that low is to tell it that it actually has a pH probe connected to it, not the ORP probe that is actually connected to it. Also, the values are upside down, graphically. High pH is actually low or negative mV or ORP.
Apex did respond to me today and the 10.5 shown on the pH scale equates to -203 mV, so it seems this is the target number.
Also, pH continues to drop within the core of the filter. It is now down to 6.8. I now have the system set up to bump on the main pump for 7 seconds when the ORP is below 175 mV. This feeds a small amount of fresh water into the system while a small amount of nitrified water is pushed out the exit pipe. Without this activity, the ORP range will dip into the -300 to 400mV range.
I have not added carbon since earlier today when i manually purged the system too much and the reading got down into the 0 mV range.
Attachments
Experiment
Experiment
In regard to the programability of the Apex, it is pretty much unlimited. It can be annoying sometimes but there is a good user forum and you can usually find some programming examples for just about any task.
Now I just need to figure at what points to trigger what activities and make any fine adjustments as necessary. The graphical data is nice to have as well.
Aaron
Experiment
In regard to the programability of the Apex, it is pretty much unlimited. It can be annoying sometimes but there is a good user forum and you can usually find some programming examples for just about any task.
Now I just need to figure at what points to trigger what activities and make any fine adjustments as necessary. The graphical data is nice to have as well.
Aaron
Updates
Updates
A few updates to my test.
Since the nitrate knee seemed to be occurring at the same point (-200 mV) I decided to set a trigger point there. I actually set it at -175 mV, hoping to keep the reaction from getting out of control, although I am not really sure what control would look like.
So anyway, I increased the 125ml dose of fresh water to 325ml. I was hoping it would even things out. Based on what i had observed earlier, a larger fresh dose of water would pull the ORP value back up momentarily. I also used this set point to turn off the small carbon doses (1.5 ml) I had been administering while trying to drive the ORP values more negative.
I let the system run like this over night. As you can see from the graph, it sort of worked, but not really. Mostly the ORP just bottomed out (around -300 mV) and stayed there for hours. Once it did finally get enough fresh water to drive the ORP to around 0 mV for about an hour. The system dosed some carbon and slowed the water dose rate to 125 ml again and the ORP came back down, again to -300mv (which I found out is a limit of the Apex).
The really bad affect was that I found the tank water very cloudy. The pH in the filter had bottomed out to around 6.0 and there was a slimy coating in the sump. I am assuming this was a bacteria bloom since the water parameters in the tank never seemed to get too far out of range and the fish didn't seem to be distressed.
i decided to put the skimmer back in and see if I could clear out the cloudiness, assuming that if it is a bacteria bloom then the skimmer should remove some of it. I also stopped all dosing and control and just let the pump push water through the Marine Pure continuously. I will do a 50% water change if necessary tomorrow. I will try to start things back up in a couple of weeks.
Aaron
Updates
A few updates to my test.
Since the nitrate knee seemed to be occurring at the same point (-200 mV) I decided to set a trigger point there. I actually set it at -175 mV, hoping to keep the reaction from getting out of control, although I am not really sure what control would look like.
So anyway, I increased the 125ml dose of fresh water to 325ml. I was hoping it would even things out. Based on what i had observed earlier, a larger fresh dose of water would pull the ORP value back up momentarily. I also used this set point to turn off the small carbon doses (1.5 ml) I had been administering while trying to drive the ORP values more negative.
I let the system run like this over night. As you can see from the graph, it sort of worked, but not really. Mostly the ORP just bottomed out (around -300 mV) and stayed there for hours. Once it did finally get enough fresh water to drive the ORP to around 0 mV for about an hour. The system dosed some carbon and slowed the water dose rate to 125 ml again and the ORP came back down, again to -300mv (which I found out is a limit of the Apex).
The really bad affect was that I found the tank water very cloudy. The pH in the filter had bottomed out to around 6.0 and there was a slimy coating in the sump. I am assuming this was a bacteria bloom since the water parameters in the tank never seemed to get too far out of range and the fish didn't seem to be distressed.
i decided to put the skimmer back in and see if I could clear out the cloudiness, assuming that if it is a bacteria bloom then the skimmer should remove some of it. I also stopped all dosing and control and just let the pump push water through the Marine Pure continuously. I will do a 50% water change if necessary tomorrow. I will try to start things back up in a couple of weeks.
Aaron
Attachments
Yikes. Hopefully nothing was lost in this endeavour. 1.5ml is a lot of carbon (depending on the carbon source), and if you are injecting this often, then that will definitely contribute to a bacterial bloom.
I believe the DyMiCo filter doses carbon only once per hour, and I suspect the amount of the dose is varied based on the number of cycles per hour. Based on what their manual says, I think they are targetting 15 cycles per hour.
The other thing that I was able to glean, is that they control the flow (amount) of water entering the filter by having the return pump set to a known flow rate and then timing the on time to only add the desired amount of fresh water. As you are using a more precise pump, you could easily add a fixed amount. The DyMiCo folks seem to be adding only the amount of new water that will occupy the sand bed layer, and the water that was in the sand bed layer already gets pushed into the anaerobic layer. In your case, you only have the anaerobic area of the filter.
Dennis
I believe the DyMiCo filter doses carbon only once per hour, and I suspect the amount of the dose is varied based on the number of cycles per hour. Based on what their manual says, I think they are targetting 15 cycles per hour.
The other thing that I was able to glean, is that they control the flow (amount) of water entering the filter by having the return pump set to a known flow rate and then timing the on time to only add the desired amount of fresh water. As you are using a more precise pump, you could easily add a fixed amount. The DyMiCo folks seem to be adding only the amount of new water that will occupy the sand bed layer, and the water that was in the sand bed layer already gets pushed into the anaerobic layer. In your case, you only have the anaerobic area of the filter.
Dennis
Yeah, definitely not the result I had hoped for. Nothing lost, again it was only a temporary tank I had set up with 4 fish and some rock. They are fine. Cloudy water is clearing but I may still do the water changed tomorrow. I guess if this had been a reef tank, some of the bacteria would have been consumed by the tank inhabitants. Maybe this wasn't the best test.
For a carbon source, I water mine down significantly. 1 cup of sugar and 200ml of Bacardi in 2000 ml RO water. This is the same recipe I use when I dose my other tanks.
I can work on the set points and vary the amounts based on those. It was all just a test for now. The time it takes to restart the reaction after pulling it back from the brink sure seems to take a long time. I don't know how they get 15 cycles an hour. I can definitely see that once you have established the bacteria population, you really don't need to dose a lot of carbon but it sure seems like it takes the bacteria longer than a few minutes to start the cycle.
One other factor, the time it seems to take read the effect from the fresh water doses is an issue. You really can't see the effect until 4 or 5 doses. The fresh water has to work it's way to the probe. By then things are headed south, out of the target zone, and you are trying to restart things. I imagine this would be the same with the sand bed too.
Aaron
For a carbon source, I water mine down significantly. 1 cup of sugar and 200ml of Bacardi in 2000 ml RO water. This is the same recipe I use when I dose my other tanks.
I can work on the set points and vary the amounts based on those. It was all just a test for now. The time it takes to restart the reaction after pulling it back from the brink sure seems to take a long time. I don't know how they get 15 cycles an hour. I can definitely see that once you have established the bacteria population, you really don't need to dose a lot of carbon but it sure seems like it takes the bacteria longer than a few minutes to start the cycle.
One other factor, the time it seems to take read the effect from the fresh water doses is an issue. You really can't see the effect until 4 or 5 doses. The fresh water has to work it's way to the probe. By then things are headed south, out of the target zone, and you are trying to restart things. I imagine this would be the same with the sand bed too.
Aaron
To get the cycle started initially, they have the filter cycle once per hour, whether they have reached the target ORP level or not. Extra cycles are added as the filter matures, and are needed to prevent the filter from becoming anoxic. Also they are using a fairly large volume of water. The sand bed layer on my filter which is modeled after their 2000 version, has a sand bed that is 36 x 18 x 5. So several gallons of water. I am also not sure if you have a process pump in your test filter? This is the pump that mixes the carbon and draws the water through the lower substrate and this pump would help to expose the ORP probe to both existing water and water freshly added to the filter.
The other thing that may be impacting this, is that the DyMiCo version has 2 layers, and the lower anaerobic layer is really only be exposed to water that has already progressed through the sandbed layer. This water (where the ORP probe is located) is not really "fresh" water, but water that had entered the sandbed on the previous cycle. This would probably help to keep the ORP reading from too big of swings.
Dennis
The other thing that may be impacting this, is that the DyMiCo version has 2 layers, and the lower anaerobic layer is really only be exposed to water that has already progressed through the sandbed layer. This water (where the ORP probe is located) is not really "fresh" water, but water that had entered the sandbed on the previous cycle. This would probably help to keep the ORP reading from too big of swings.
Dennis
