Reducing KH with acid

O.K. here goes again the thread that won't die.

The discussion was whether it would be possible to save on CO2 by
lowering KH, assuming that the pH is adjusted so that the levels of
CO2 are unchanged.

I speculated that higher bicarbonate concentrations possibly could
affect the ammount of CO2 needed to achieve a certain concentration of
dissolved CO2. I started thinking about this since the CO2 consumption
is very high in my 140 gal tank with open top and W/D filter, KH10,
fairly strong pump. Despite a very effective reactor system (Dupla
stuff), I was having difficulties to keep the CO2 levels at/above 18
mg/l. A pH of 7.2 was about as low as practically possible.

Here's what I did:
540 l (150x60x60 cm) FW, fairly densly planted tank, very light fish
load, temperature 25 C (77 F), 2x150W MH (HQI) lights 10 h/d,
Wednesday 'rainy day', with only four hours of light. One corner of
the tank works as a trickle filter as an overflow chamber (volume
about 17 l or 4.5 gal., filled with densly stacked Sera Biocubes).
Large biofilter underneath the tank with four filtration chambers and
a sump. This filter holds about 150 l, so the total water volume is
about 600 l or 158 gal, accounting for gravel, rocks and decoration.
Pump is an Eheim 1060 rated at 38 l/min. It has to pump against a
water column of 130 cm. No flow reduction valve. The water enters the
aquarium 8 cm below the water surface to minimize surface motion.

GH 14 dGH, KH 10 dKH. Without CO2 the pH is at 8.2 after one day of
circulation and light (I tried this when I first set it up). No
phosphates, nitrate of the tap water is around 10 mg/l, it's zero in
tank. The electronic dupla pH meter is set for 7.2, shooting for a CO2
concentration of about 18 mg/l. The reactor is a Dupla 'S' type driven
by a small Eheim 1046 pump rated at 5 l/min. Flow is in parallel to
the last two biofilter chambers.

A KH of 10 dKH is 179 ppm or 3.6 meq/l. To half KH, about 1.8 meq/l of
ions would be needed. That's in the range of 123 ml of HCl 32% (8.78
mol/l) or 260 ml of acetic acid (vinegar  or CH3COOH) 25% (4.16 mol/l)
for 600 l of
water. Assuming acetic acid is completely dissociated at a pH around
7, which it is _not_!
I started with acetic acid 25%, prediluted it some more and added
small ammounts to the tank under close monitoring of the pH. After the
acetic acid proved unpracticable, I continued with HCl and finally
reached a KH of 5. Set the pH to between 6.90 and 6.95. That should
give a CO2 concentration of around 18 and should correspond well to
the KH10-pH7.2 state. Nothing else in the tank was changed during the
experimental period. The tank had been running for one month prior to
the experiment.

CO2 consumption was estimated by the water level in the Dupla S
reactor. Once a certain pH is stabilized, I found the water level
(below the CO2) to be quite stable.  For those who don't know the
reactor: it's a cylinder shape and the more CO2 is injected, the lower
water level. Water goes in at the top and trickles through the CO2
atmosphere, leaves at the bottom.

Here's what happened:
Adding small ammounts of acid causes instant drops in pH by definition
and as expected. The ammounts I added never caused the pH to go down
more than 0.15 units per single addition and the pH during the whole
game was
never below 6.80.
After this instant decrease, pH _SLOWLY_ goes back up to the initial
value (as long as there is enough buffering capacity left). This can
be explained by the reaction  H+ + HCO3- --> H2CO3 --> H2O + CO2.
The last step appears to be very slow in aquarium water. CO2
then leaves the water to the atmosphere. 
I found out quickly that acetic acid is not very good for decreasing
KH, because one needs a lot of it! It's not dissociated completely at
a pH around 7. And who knows what such large ammounts of acetate do to
the living stuff in the tank. One would assume it gets metabolized,
since acetate is such a basic compound of every cell metabolism, but
who knows?
After 300 ml of acetic acid (over three days) I went on with
HCl. One can already tell that acid titration is
not a practical way to decrease KH. It just plain takes too long. You
can only add about 10 ml of HCl 32% (prepreprepre-diluted!!!) per hour
to a 600 l volume, and that's already pushing it. pH goes to below
6.9. But
you need more than 100 ml. So by the time you're at KH of 5, it's
already time for the next water change. Messing with RO water can't be
much more work. I needed a total of 300 ml (1.25 mol) of CH3COOH and
100 ml (0.88 mol) of HCl. That's nicely in the range of 1080 meq/l of
H+ which I had calculated.

The CO2 consumption of the tank during the titration period was
markedly reduced. No wonder, pH was lower than the value set at the pH
meter most of the time, so the valve was shut. CO2 for the plants came
from splitting H2CO3 (did it really?). After the acid titration was
stopped, KH remained stable at 5 and pH was set to 6.90 to 6.95 (see
above). The CO2 consumption - as judged by the water level in the
reactor - was pretty much exactly where it had been before the
experiment, when KH was 10 and pH was 7.2. Plants didn't grow that
much during the test period and judging from the day/night difference
I observe in my tank (not much difference), CO2 consumption by the
plants is not the main factor in CO2 loss in my aquarium.

As a side experiment, I reduced the flow of my filter pump. This was
after all the other experiments. It actually happened by accident when
the small prefilter sponge in the Eheim pump got suddenly clogged up
with some slimy stuff. CO2 consumption dropped significantly and
instantly! It went back up after cleaning the sponge and restoring
normal flow.

I also have a valve, where I can adjust the flow from the TF to the
biofilter. It allows very fine adjustment and thus allows me to adjust
the water level in the TF. I can 'flood' the TF completely if I want.
I haven't determined yet whether that makes a difference or not, but
will try later today or tomorrow.


- titrating down the bicarbonate concentration in aquarium water is
not practical. It takes too long.

- acetic acid is difficult to use, because the ammount needed is
difficult to calculate.

- HCl is effective, completely dissociated and therefore easy to
calculate, but again, the proccess is time consuming.
- the whole mess is needless if one wants to save CO2, because it
doesn't work. CO2 consumption to achieve 18 mg/l of CO2 is independent
of pH and buffering capacity.

- the whole mess is an effective way to lower the pH in an aquarium
with a  pH - controlled CO2 injection system. Here from 7.2 to 6.9.
it's a real pain.

Important note:
My fish didn't show any side effects from this experiment. However,
some of the plants didn't like it at all. While it is difficult to
blame either acetate or Cl- as the causative agent, the
time-correlation is striking. Among the species that got hurt were
Cardamine lyrata, Echinodorus griesebachii, one Valisneria species and
Shinnersia rivularis. Most of the plants, including some sensitive
species like Rotala wallichii were unaffected. What happened was that
these previously fast growing species stopped growing completely and
seemed to loose the chlorophyll from their leaves. The leaves turned
either 'glassy' (C. lyrata and S. rivularis) or yellow (Val. and E.
griesebachii). Luckily, they are already recovering rapidly after a
30% water change. I don't think the pH swings caused this, the swings
weren't that big (maximal from 7.2 to 6.8). Acetate or Cl- could be
the reason. Acetate, I don't know, Cl- unlikely as I only added 1.5
mmol/l, which is close to nothing, equivalent to adding 0.088 g of
NaCl per liter. 
Another possibility would be that the plants were
actually starving CO2 during the experiment. Maybe the dissociation of
H2CO3 to CO2 and H2O was too slow to hold sufficient CO2 levels? A
hint in this direction would be the primary effect on the fast growing
species. I did not measure CO2 directly, unfortunately.

Oh, almost forgot: nearly all of my algae are gone now.

Any comments welcome and my apologies for pestering everybody with
this monstermessage, I will shut up now.


The next thread will be:
"How much should I reduce the flow of the pump to still get optimal
filtration, but reduce surface motion?"