Calcium Reactor CAP Mk. II: User's Manual

Chris Paris

November 3, 1997

This calcium reactor meets my design goals (I think), but there are some implementation problems that you should know about. You could say that this is a CRAZEy project, because of all the crazing that happened. Crazing is the small cracks that you see all over everything. Those were caused by the acrylic glue. The crazing on the main reactor tube is particularly bad (and also shows where I was sloppy with the glue). That tube came from a surplus place, and dates from the early '60s. The age might have something to do with the degree of crazing. It was $15 (including shipping to me) instead of AIN Plastics' price of $130 for the same piece, and the crazing shouldn't affect performance nor reliability.

The flexible tubing

I've included a five foot and a ten foot length of 1/4'' inside diameter flexible PVC tubing. I hope that's enough for you. This stuff is cheap, and I would have included more, but I'm running low and I hate to place a tiny order with US Plastic Corp.

The bubble counter

The bubble counter also suffered some crazing. The hose barbs on the top and bottom are homemade from extruded acrylic rod. That stuff isn't very strong, and it is brittle, so you need to be very careful when attaching or removing tubing. I've already attached a check valve to the top of the bubble counter (where the CO2 enters). You shouldn't ever need to separate the check valve from the bubble counter, so there should be little risk of breaking the top hose barb. The connection at the entrance to the check valve is some sort of semirigid plastic, and you don't have to worry about breaking that.

The bottom of the bubble counter, where the water enters, is a very delicate hose barb. It's also cut for an aggressive grip on the tubing. My suggestion is to put tubing on it once and plan never to remove it. Cut the right length of tubing to connect your feed pump in your sump to the bubble counter. Cut it oversize in case you change your configuration around. Heat one end of the flexible tubing in boiling water to make it soft, then press it straight onto the bubble counter's hose barb. If you ever have to remove that tubing, I suggest carefully cutting it off rather than pulling it off. If you cut it off, be careful not to score the hose barb with your knife, or else you may create a path for water leakage. If you break the hose barb on the bubble counter, I can repair the bubble counter, but let's try to avoid that.

The other hose barbs--the one on the output of the water needle valve and the two on the lid of the reactor--are commercial units made from nylon, and there's little risk of breaking them.

As an experiment, I used a very small hole to release the CO2 gas into the bubble counter. This means that you get much smaller bubbles than in the bubble counters that most other people are using. I thought that might make it easier to gauge fine changes in the CO2rate. This also means that it's meaningless to compare your bubble rate with other people's bubble rates. If the Dwyer flowmeter that you have registers the CO2 flow, then you won't need to pay attention to the bubble rate anyway. Even if that's the case, the bubble counter still serves its purpose of premixing the aquarium water and the CO2, so that calcium carbonate is highly soluble in the water that passes through the water needle valve.

The output in the side of the bubble counter connects via a 1/4'' nipple to the needle valve. I've already made that connection, and you shouldn't have to undo it. The output of the needle valve connects to the input of the calcium reactor. The input of the reactor is the black hose barb that's positioned over the circulation pump intake.

The reactor input

The input is pretty straightforward. I placed it directly over the pump intake, though I don't think that's a critical requirement. Some people think that the CO2 and maybe the input water should be injected into the pump impeller. With this reactor's overall design, I don't think that it matters. But I had to put the input somewhere, and it was just as easy to put it over the pump intake as anywhere else.

The reactor output

The reactor really has two outputs. One is controlled by a ball valve. The output that is controlled by the ball valve draws water from the very top of the reactor. The other output draws water from about half an inch below the top of the reactor. The two outputs are connected on the lid by a tee fitting and a short length of tubing, so you need only one piece of tubing to take water from the reactor back to the aquarium (or sump).

The reason for having one output draw from below the top of the reactor is so that you can trap CO2 gas at the top of the reactor, where it can dissolve over time. I believe this is better than the conventional design where CO2 that doesn't react immediately just exits the reactor unused.

The purpose of the valved output that draws from the very top of the reactor is so that you can bleed off the air that will be at the top of the reactor each time you close it. The idea is that you fill the reactor with water with the valve open, and then close the valve once all of the air is out of the reactor. This causes all further output from the reactor to come from the other output, which draws from half an inch below the top of the reactor, and thus all of that output should be liquid.

The closing screws on the lid

The top of the reactor is held in place by twelve hex head nylon screws. You need a 7/16'' wrench to operate the screws. I could have provided thumb screws for the lid, but on my own reactor I found that I like to be able to tighten the lid more tightly than I can comfortably achieve with thumb screws. Unlike the flanges on the ETS-type protein skimmers, this flange has to be water and gas tight, not just tight enough to keep nonpressurized foam from escaping. Be careful not to overtighten the screws. I think--I hope--that the nylon screws would strip before the acrylic threads in the top of the reactor, but I've not tested this idea. In any case, it would be easy to apply too much torque with a wrench when we're used to tightening metal threads. Perhaps a good technique is to get the screws as tight as you comfortably can with your fingers, and then tighten them a little bit more with a wrench.

Because of my simple methods for laying out and drilling the holes for the lid flange, the lid goes on only one way. Some day I will be rich enough to buy that $500 rotary table or dividing head that would let me do it right. On the perimeter of the lid, there is a mark with a permanent black marker, and a matching mark on the perimeter of the lower half of the flange. These marks show you which way the lid goes on. The neoprene gasket is also homemade and also goes on only one way. It has a similar marking, but on the top face rather than the edge. Be sure that the mark on the gasket is facing up.

The alignment between the holes in the lid and the holes in the flange of the reactor isn't superb, and as a result it can sometimes be difficult to get all of the screws to fit into their holes. Here's how I've been doing this. After making sure the holes in the lid are lined up with the correct holes in the reactor, I start each of the twelve screws by hand. I don't screw them in all the way yet. Once all screws are started, I screw each one all the way in by hand (unless it's too stiff). Then I make a third pass, gently tightening each screw with a wrench.

The power cord

In order to get the power cord for the circulation pump through the lid, I cut the cord, passed it through a small hole in the lid, spliced the cord back together, and sealed the hole around the cord with Weld-On #40 glue. I'm worried that the seal around the cord will start leaking. It will depend on how much it gets messed with. It's watertight now. If it starts leaking, you can try sealing it with silicone, probably from the inside more so than from the outside. In the meantime, treat the cord delicately.

The circulation loop

The circulation pump is a Maxijet 1000. With a rated throughput of 1000 liters/hour, and a typical reactor throughput of 1-2 liters/hour, the water is turned over within the reactor on the order of 1000 times the rate at which it flows through the reactor. So the pump is strong enough. Got that!?

Because of the way the cord passes through the lid of the reactor, replacing the pump would be fairly major work. However, you can replace the impeller assembly easily, and that's usually what goes bad, not the body of the pump.

The intake side of the pump has a CPVC elbow glued into it with PVC glue (so it's permanent). The elbow has a short length of 1/2'' CPVC pipe placed in it, but it isn't glued, so you can change its length if you wish. The purpose of the elbow and the pipe is to make the pump draw water from near the top of the reactor. And if CO2 gas accumulates at the top of the reactor, it doesn't have to build up much before it starts being sucked through the pump and pushed to the bottom of the reactor, which will help it dissolve and react. When the CO2 rate and water input rate are properly matched, essentially all of the CO2 that is injected should be used, and you should not have an ever-growing volume of CO2 at the top of the reactor. There will be some gas bubbles there, to be sure, but the volume of trapped CO2 should reach a fairly steady state and then stop increasing.

The output of the pump is a slip fit into a 1/2'' inside diameter acrylic tube, which brings the water to the bottom of the reactor. This acrylic tube has some slots cut in the bottom so that the water can escape. It is glued to the bottom of the reactor, but is not firmly attached to the pump. This way, you can remove the lid of the reactor (with the attached pump) while the reactor is full of sand, and then reclose the reactor. If the tube were connected to the pump rather than to the bottom of the reactor, it would be difficult, at best, to push it down through a reactor full of sand.

The tube is attached to the bottom of the reactor via a Weld-On #40 glue joint. This glue forms a strong bond, but you still must be very careful not to apply side pressure to the tube. Because of the leverage present at the top of the tube relative to the joint at the bottom, a small force at the top becomes a large force at the bottom, and it would be easy to snap the tube or the glue joint at the bottom. I can't overstate how delicate this joint is. I probably should have provided some horizongal bracing between the top of the central tube and the sides of the main reactor tube, but that would have made the reactor slightly less neat looking during operation. The only risk of breaking the central tube is when you have the reactor apart and are messing with it, so just be careful then.

When you replace the lid, remember that it goes on only one way. The pump output should easily slide into the output shaft as you gently set the lid down. If it doesn't, then one of two things has probably happened. The first possibility is that you may not have the mark on the edge of the lid lined up with the corresponding mark on the top of the reactor. The other possibility is that you have rotated the impeller housing of the pump. Be careful of this if you've removed the impeller housing to inspect the condition of the impeller. If this is the problem, just gently fiddle with the orientation of the impeller housing until the pump output slips easily into the acrylic output tube.

Filling the reactor with aragonite

I tested the reactor with CaribSea Geo-Marine, but any of the common aragonite products should be fine. With Geo-Marine in place, I didn't see any shifting of the material when the recirculation pump was operating. Keep that in mind if you see a lot of shifting with a finer material.

You fill the reactor by dumping in your aragonite, but you have to take some precautions.

First, the twelve screw holes in the top of the reactor don't go all the way through the plastic. Actually they do, but I found that water leaked up around the screws, so I put a dab of silicone sealant over the bottom of each of the holes. If you let sand fall into the holes, it'll be hard to get it out, since you'll have to pick it out from the top. In the worst case, you can tear out the silicone in the holes, and redo it. Perhaps a better solution to this would be to tear out the silicone at the bottom of the screw holes, and provide the necessary seal by putting a tiny O-ring around each screw, pushing it up against the screw head, and screwing in the screws with the O-rings in place. I'm making this up as I write, so take it for what it's worth.

To avoid getting sand into the screw holes, I suggest adding aragonite with the gasket placed on the top of the reactor. Rotate the gasket slightly so that the holes in the gasket don't line up with the screw holes in the reactor flange. When you're done filling, clean any spilled aragonite off of the gasket before rotating it back into the proper position for closing the reactor. Any aragonite left on the gasket, or on either of the two acrylic surfaces that the gasket seals against, will potentially interfere with the seal.

Try not to dump sand down the central acrylic tube. If you do get sand down there, don't worry about it, and don't try to remove it for fear of breaking the tube off. (Acceptable ways to remove it would be to try to blow it out the bottom using the pump, and to remove the aragonite and turn the whole reactor upside down.) I've included a rubber stopper with the reactor. Put the rubber stopper into the tube before dumping the aragonite in, and remove it after all the aragonite is in place. That should keep the sand out of the tube. Don't let the aragonite level go past the top of the tube. If you do, you won't be able to place the pump output into that acrylic tube--it'll be buried in sand.

``What! I can only fill my new reactor 2/3 full of sand?'' I hear you cry. Sorry. This reactor has an inside diameter of about 8-1/4'', and the length of the central acrylic tube is about 7-1/2''. That allows for a volume of aragonite equivalent to a reactor with 4'' inside diameter and a length of 32''. You shouldn't need to refill this very often.

Starting or restarting the reactor

The procedure for starting or restarting the reactor is as follows.

Close the reactor and tighten each of the twelve nylon screws.
Put the output valve on the lid of the reactor in the open position. The valve is open when the end of the handle is pointing down.
Open the water needle valve so that water can enter relatively quickly. You don't want to wait all day for the initial fill.
When the reactor is completely full, close the output valve. Water will continue to exit through the unvalved output.
The circulation pump may be started now or at any time henceforth. Never run the circulation pump before the reactor is full of water to the top.
Close the water needle valve until water exits the reactor at a drip rate.
Open the CO2 needle valve to allow gas to enter the bubble counter, if you have not done so already. (There's nothing wrong with leaving the CO2 running through all this. It ``wastes'' CO2, but when you have it running continuously forever, what's five minutes of waste?)

Tips on operation

The text of this section and the remainder of this manual is taken from my article on calcium reactors that appeared in the November 1997 edition of the ORAS newsletter.

Opinions differ on how to operate a calcium reactor. The most common point of disagreement is over what the pH of the output should be. The most common recommendations are in the 6.5-6.8 range. I've personally had good results with the output pH between 6.5 and 7.1. In a well designed reactor, there's a relationship between the pH of the output and the calcium content and alkalinity of the output. A lower pH output means that the output has more calcium carbonate dissolved in it, so you need less of that water each day to meet your tank's demand. A higher pH output means the output is less concentrated, and you will need a higher water drip rate to meet your demand.

You do not need a pH controller to safely use a calcium reactor. Most users run their CO2 24 hours a day, 7 days a week, and the pH of the tank never goes outside acceptable limits. However, you most definitely should have a pH monitor to help you establish the correct water flow and CO2 flow rates. The difference between a pH monitor and a pH controller is that the monitor simply tells you what the pH is, whereas a pH controller has not only a readout of the pH, but also the ability to turn your CO2 supply on and off, depending on how the tank's pH compares with preset pH limits. Once you have the reactor working well for your tank, you won't need the pH monitor regularly, but I would not consider beginning to use a new reactor without owning a pH monitor.

The most important setting on a calcium reactor is the rate of CO2addition. As long as the water flow rate is within a wide acceptable range, the CO2 rate determines how much aragonite you will dissolve. The correct CO2 flow rate depends on your own tank's calcification rate. No one can tell you, absolutely, how much CO2 you will need. But you need some ball park estimate of where to start. Many bubble counters release the CO2 through a piece of common rigid airline tubing (outside diameter: 3/16", inside diameter: 1/8"). If your bubble counter uses this size tubing, then a figure of 30 bubbles per minute is a good place to start operating your reactor. Small tanks and/or tanks with few fast growing corals will need less, and large SPS-coral dominated tanks may need more CO2 than this. But 30 bubbles per minute should be in the right order of magnitude for most home tanks. Since you'll be watching the effect of the reactor closely in the first few days, it will be okay if you have to make significant adjustments from your initial CO2 rate. Since your bubble counter produces finer bubbles than the hypothetical bubble counter described in this paragraph, you will need a higher bubble rate.

As an initial water rate, try a moderate to fast drip. In order to see the drip, you have to have the end of the reactor's output tube positioned above the water level in the sump. A more accurate way to measure the water flow rate is to time how long it takes to fill a graduated cylinder (or even a kitchen measuring cup). For the purpose of adjusting your reactor for your tank, such precision in the water measurement isn't required, though for estimating the calcification in your aquarium based on the output of the reactor, such accuracy is necessary. But we won't look into that topic now.

Whenever you make a change to the flow rates into your reactor, and especially when you first fill it, it will take several hours for the pH, calcium concentration, and alkalinity of the output to stabilize. After your reactor has been running for several hours, you can measure the pH of the output by collecting some in a cup and placing your pH probe in the cup. If your reactor output's pH is in the 6.5-7.0 range, you're doing fine.

The pH of the reactor output is just a rough sanity check to make sure your CO2 and water flow rates are in a reasonable ratio. Note that the pH of the output, considered alone, is not enough to determine how fast you're adding calcium and alkalinity the tank. To illustrate this, consider that at any output pH, you could increase both the CO2input and water input rates in a ratio that would leave the output pH unchanged, but would increase the total calcium carbonate being dissolved and added to the aquarium.

For long term adjustment of the reactor, use the calcium and alkalinity of your aquarium water as your guide. If the tank levels fall over time, turn up your CO2. If the tank levels go too high, turn down your CO2. If you make large changes in the CO2 input rate, you may have to adjust the water input rate to keep the output pH in the 6.5-7.0 range. When making small changes in the CO2 rate, I don't change the water input rate. I strive for a particular alkalinity level in the tank, not for a particular reactor output pH or reactor output alkalinity.

Since a calcium reactor is a balanced calcium and alkalinity supplementation discipline, it will tend to increase both parameters together. Thus, if your calcium and alkalinity are in the correct seawater ratio (Calcium 420 ppm, alkalinity 2.4 meq/l, or thereabouts), they should stay in that ratio no matter how you adjust your reactor. Likewise, if your calcium and alkalinity are in the wrong ratio, there's nothing you can do with your calcium reactor to change that situation. For example, if you've been maintaining your calcium level with calcium chloride, and not adding a buffer, you may have a calcium level of 450 ppm but an alkalinity of only 1 meq/l. You could run the reactor in ``overdrive'' to raise the alkalinity to something reasonable, but in doing so you would be driving the calcium level above what is reasonable. You have to get your calcium and alkalinity into a reasonable ratio by other means, water changes probably being the best of them. (I say that water changes are probably the best fix to ionic balance because whatever you did to mess up your Ca:alk ratio probably messed up the general ionic balance in the process). For a more detailed treatment of this topic, see Craig Bingman's article, ``Repairing Calcium and Alkalinity Imbalances,'' in the May/June 1997 issue of Aquarium Frontiers.

Since the calcium concentration and alkalinity in the aquarium should always move together, you don't have to test both parameters. I test alkalinity, because I find alkalinity test kits more pleasant to use, and because among the tests available to the hobbyist budget, alkalinity test kits are a more sensitive way to measure the dissolved calcium carbonate (this statement holds only because I use a balanced supplement). I have tested my calcium level about twice in the past year (it was right where I wanted it), but I test alkalinity regularly.

Another point of differing opinions among calcium reactor users is on what particular calcium carbonate sand product to use in the reactor. I personally use CaribSea ``Reef Sand.'' Some aragonite materials contain some phosphorus, which is released as phosphate when the material dissolves. I personally suspect that the level is small enough that a decent protein skimmer should be able to remove it as fast as it's added. I've seen no increase in unwanted algae since installing my calcium reactor. As a matter of fact, I saw a decrease in unwanted algae, which I attribute to the coincident addition of more snails rather than to the addition of the reactor. Nevertheless, some people are in a virtual panic about the phosphate released by calcium reactors, and are willing to pay more money for calcium carbonate materials that are low in phosphate. Two such products are Koralith and Super Calc Gold (which is also the title of my high school calculus textbook). I have no personal experience with those products. You can read a chemical comparison of them in Craig Bingman's article, ``Calcium Carbonate for CaCO3/CO2 Reactors: More Than Meets the Eye,'' in the August issue of Aquarium Frontiers Online, which is available at