There is a deep underlying reason why the two "best" ways of maintaining clacium and alkalinity in reef tanks have such drastically different pH's. Both of these methods (limewater and CaCO3/CO2 reactors) achive balanced inputs of calcium and alkalinity by relying on the natural balance of calcium and alkalinity in calcium carbonate. While it is possbile to get limited amoutns of calcium carbonate equivalents into solution by allowing aragonite to dissolve in deep sand beds, or by running topoff water through Halimdea hash _a la_ Adey, calcium carbonate is at best a sparingly soluble substance, so you can't get much of it into solution those ways. It makes sense that corals would use some sparingly soluble substance for skeletal material. Organisms that try to make skeletons out of soluble substances stand the test of evolutionary pressure. You can imagine them having the same problem as the wicked witch of the west? in the Wizard of Oz.... pour a little water on them and they are melting....

So we need to do something to coax calcium carbonate to dissolve in water. Calcium carbonate by itself gives a pH similar to seawater (when you equilibrate it against the atmosphere in seawater... so we can either go higher in pH or lower in pH. It turns out that both work. The high pH side is of course limewater, and the low pH side is the kalkreactor.

Underlying this is a very simple piece of chemistry... the concept of a solublility product. The ion concentrations relevant to the solubility of calcium carbonate are calcium and carbonate, curiously enough. Note that bicarbonate is not directly relevant, and hydroxide ion isn't relevant (except in the high pH limit of limewater.) Granted, in a given solution with various forms of inorganic carbon present, all of these values are going to be interrelated, but calcium and carbonate ions are where it is at for this discussion.

So, aside from chelating the calcium ions, there isn't much way to affect their activity. However, by adding carbon dioxide to a solution, you can drop the pH to the point that the carbonate ion concentration is very low. That is how kalkreactors work. You can also imagine a solution that limits the carbonate ion concentration by starting out with essentially zero dissolved inorganic carbon. That is how limewater works.

So, it isn't terribly surprising that there are two major ways of coaxing calcium and alkalinity into solution, and that they have dramatically different pH values. Fundamentally, there are really two ways to do this... get rid of inorganic carbon (limewater) or add so damned much inorganic carbon that the carbonate ion concentration goes down (kalkreactor.)

You should never directly mix the output of a kalkreactor with limewater, because you will make an extraordinarily highly supersaturated solution with respect to calcium carobnate at that point. Remember that the two gambits for getting calcium carbonate in soluble form work by moving to the very high CO2 limit and the zero CO2 limit. When you mix them, you wind up right back in the middle, and calcium carbonate is going to fall out of solution. If you mix them separately in your system, then they both have a chance to do what they are supposed to before they spontaneously flop out of solution.

You should never mix limewater with seawater, because you will absolutely precipitate the magnesium as magnesium hydroxide when you do that. This is why there are no Nilsen-style reactors that are closed-loop systems.

No, Nilsen-style reactors are not necessarily volume limited, becasue you can dispense milky limewater (a suspension of calcium hydroxide) with them.

Are Nilsen reactors necessarily better or worse than kalkreactors? No, they aren't. The only reason why kalkreactors are more favored in this country is.. well, there are two or three.

1. they aren't commercially available.
2. you need to add calcium hydroxide more often than you need to replenish the calcium carbonate in a kalkreactor. Since a desire to find a system that demands little attention is what often drives people to go with one of these systems, kalkreactors are a little gentler on the lazier aquarists.
3. Nilsen reactors may be somewhat more difficult to adjust than CaCO3/CO2 reactors, although once you get that down, they are in principle no more difficult to run than a kalkreactor.
4. you need one more technical gadget with the Nilsen-style reactor, a pH controller.

The Nilsen system does have advantages that the calcium carbonate/carbon dioxide reactor doesn't have... or perhaps we should call it a property because it isn't free. If you use a pH controller, you have a really great way to keep the system pH reasonably high 24 hours per day. In fact, you could clamp it at just about any value you like. That isn't true for kalkreactors, and when people try to modulate pH with a controller and a solenoid, they find that they are more difficult to adjust.

I have a followup question regarding the following statement "You should never mix limewater with seawater, because you will absolutely precipitate the magnesium as magnesium hydroxide when you do that. This is why there are no Nilsen-style reactors that are closed-loop systems."

Yeah, it is a bit frightening when I re-read what I've written. I guess it was clear that what I meant is that you should never use sea water to prepare limewater.

Why don't we have localized precipitation of magnesium hydroxinde in our tanks when we drip in limewater? My guess is that we do, but that if we have sufficient mixing the microscopic bits of precipitate go right back into solution. Is this correct? (I gather that this is the case for the calcium carbonate that forms when limewater is dripped in as well?).

Yes, there is probably localized precipitation, but the material formed (brucite is the name for magnesium hydroxide the mineral) is not stable at seawater pH. The partial pressure of carbon dioxide needs to be orders of magnitude lower than NSW for brucite to be stable. That condition is never satisfied in your tank outside of the tiny area where a droplet of limewater hits the tank water. It is of course satisfied very well in limewater solution.

This is also the reason why the "special" limewater mix on the market that contains both calcium hydroxide and magnesium hydroxide is... well, a bit silly. Magnesium hydroxide is rock insoluble at limewater pH values.

Regarding other questions... about Nilsen reactor plans. There is a call for such things occasionally, but I'm personally unaware of extremely well debugged plans laying around on the web somewhere. Regarding the questions asked, my personal preference would be to use an overhead mixer rather than a stirbar, and gravity output from the lime container. Obviously some sort of metered input.

Some ages ago, I remember some chap giving a talk at the Brooklyn Aquarium society... he was quite produ of his Nilsen-esque limewater additions. I think that he used a modified fish feeder to slowly meter lime into the mixing container.

And.. again, these devices are more trouble than a simple CaCO3/CO2 reactor. There are some advantages, though. I remember when Alf put a CaCO3/CO2 reactor on his system, that had been managed (very successfully) for many years with his own limewater reactor. The results were not great, and I believe that he broke the tank down some time after that. There was a massive proliferation of algae and also those pesky anemones. The algae was probably related, the anemones... maybe yes, maybe no.