The ETS Foam Fractitioner
- [M][R] Description of a Novel Skimmer Design (Long)
by cb77-at-inibara.cc.columbia.edu (Craig Bingman) (20 Feb 1995)
by cb77-at-inibara.cc.columbia.edu (Craig Bingman)
Date: 20 Feb 1995
The ETS Foam Fractionator
I mentioned here some time ago that I had seen a foam
fractionator of a novel design, and one that caused me to
reevaluate my paradigm of what is important for a foam fractionator
to function optimally. I typically don't initiate threads here,
but I'll break from my usual pattern of simply answering questions
with a description of this skimmer, its principles of design and
operation, and take my first cut at explaining why it achieves the
remarkable results that it does.
The design is unconventional, and it is actually derived from a
device that was supposed to strip ammonia directly from the water.
The old timers here will remember the claims made several years ago
that trickle filters removed ammonia directly from the water by
outgassing. This was tested, and for freshwater systems, trickle
filters just don't cut it, it takes a lot more than water gently
flowing over some DLS to cause ammonia loss to the air, especially
when the ammonia is almost 100% ammonium ions. Ammonia
fractionation towers can be made to work, they just have to be
about 20 feet tall and have throughput like Niagara falls, and they
work best at high pH. Gary was trying to build a home version of
an ammonia stripping tower for marine aquaria, and noticed that it
produced a lot of foam. The notion that this side effect might be
exploited was the first step in the development of the ETS skimmer.
The unit works by a novel "downdraft" design, where the output from
a high volume, high pressure water pump over a 5+' acrylic tower
filled with bioballs. Anyone who has stood next to a waterfall
appreciates that the water carries a substantial downdraft of air
with it, and this is how the air is introduced into the bioball
tower. There is no venturi valve or air pump associated with it,
the water drags the air into the bioball tower. There is
something like a high pressure water "gun" at the top of the tower,
it narrows down to maybe a quarter inch outlet in some designs.
The water emerges from this "gun" with substantial pressure and
muzzle velocity. When it hits the bioballs, it starts breaking up
into a mixture of water and air bubbles. By the time the water is
a foot or so down the column, the bubbles are becoming very fine,
and this fast moving froth continues down a tortuous path through
the bioballs, mixing and remixing the water with the air, keeping
the bubbles very fine.
The output of the bioball tower then goes into a rectangular tank,
where the process of separating air from water begins. There are
some tricky baffles in that box that I don't fully understand, it
is a bit difficult to see exactly what is there when the skimmer is
in operation, and I've never had occasion to look at one when it
was dry. This seems to be the subtle part of the design, in my
estimation, and I'm told that relatively small changes in the
layout of this component change the results substantially.
In any event, the foam emerges from this chamber into the foam
riser. It is of large diameter acrylic tubing (in the 6-8" range)
and is about 24" tall. In all of the operational skimmers I have
seen, the foam riser has been about 3/4 full of foam, of which the
top 12 inches are very dry, thick foam. The neck of the riser has
a substantial ID, according to Sprung and Delbeek's book,
relatively large OD tubing at this point gives the best results.
I first saw the design at Westchester Pet center, where it was
being tested on the tropical marine fish section. This prototype
was huge, and it had an impressive head of foam on it. The next
time I saw it was at the same store, this time on the reef aquaria.
Tony Vargas (co-proprietor at the time) and I were having a chat
about the skimmer, what do you make of it, that sort of thing.
There was some sentiment that it might be a superior design, and
being my usual "self" suggested that there were few ways of
quantifying skimmer performance (there isn't a foam fractionator
equivalent of an iodine or molasses number as there is for
activated carbon.) So I opined that the only way to straighten
this out was to run two skimmers against each other on the same
system. If one was markedly better than the other, it would then
be apparent in the quantity and quality of skimmate from each. It
also takes out variations in the organic load, presence of anti-
foaming surfactants (stick your hands in your aquarium and see what
happens to your skimmer) and other confounding and difficult to
I was surprised (and pleased) that such tests were actually
conducted, and the ETS was tested against several other skimmers,
including a very upscale, well respected, professionally
constructed and designed $700 venturi unit which was previously on
top of my "if I had the money" list of skimmers. All the skimmers
against which the ETS was run were effectively "shut down" meaning
that they were not able to make enough of a foam head to produce
skimmate. This was a substantially more "black and white" result
than I had expected, and in my opinion the only reasonable
interpretation of this is that the dissolved organic concentration
in the water had been reduced by the ETS so far that the other
skimmers failed. The bubbles in the units being run against the
ETS were not very persistent at the surface of the water, actually
it looked more like what I expect to see when fine air bubbles are
blown through fresh water. However, the ETS was maintaining an
approximately 18" head of foam, and actively producing a dark,
Why does it work as well as it does? Well, my best answer is that
it is a Brute Force design. A phenomenal amount of air is mixed
into the water, and that mixing is done with considerable force and
violence. This is unlike most venturi designs, where there is
substantial turbulence as the air/water mixture leaves the venturi,
but then goes into a relatively quiet fractionation column. CC
units also tend to be quiescent. While this should theoretically
allow concentration gradients to be set up in the skimmer, in
practice, both venturi and airstone driven CC units are limited in
the amount of air they mix with the water. Because air only rises
so fast through the countercurrent section of the skimmer, they are
also restricted as to how much water one can put through them, and
they are also limited in the amount of shear they deliver to the
solution. Mixing is sufficiently violent inside the bioball-filled
column of the ETS skimmer that it is plausible that proteins in the
water are being denatured and rendered into a more skimmable state.
Other molecular aggregates may similarly be broken up and find
their way to the air/water interface. There is simply no way that
the skimming process is diffusion limited in the ETS. All of that
water and whatever is dissolved in it is being intimately mingled
with the air/water interfaces as the water passes through the
Probably the concentration gradients which are essential for foam
fractionation are set up in the foam riser. There is about 18"
of foam in the riser, and this prodigious column of foam is
probably protecting itself from turbulence and draining
quiescently. When I looked at it, it seemed that turbulence was
damping out to essentially nothing before the dry foam layer was
reached. There is a LOT of air moving though this skimmer, and
it embodies a unique mixture of brute force and grace, from the
violent mixing tube to the draining foam in the riser.
I've seen the skimmer in operation six times now, and was present
for some of the testing (not all of it.) So I am relying on
individuals who I feel are reputable sources for a description of
the rest of the testing, and their notes of what has happened in
What else does it do besides act as a foam fractionator?
Well, all of the effects that I am describing below are attributes
of all foam fractionators to some extent, this design just seems to
have More of them because it is a very high throughput system.
1. In all cases that I am aware of, there was a substantial
increase in redox potential (~50 mV) of the water. Note that this
was observed on functioning reef aquaria that would have been
considered well skimmed and well maintained by anyone. Part of
the increase in ORP is probably due to enhanced oxygen saturation,
part of it is no doubt because this is an excellent skimmer, and
organics are removed efficiently.
2. Gas exchange in general seems to be improved, and this also
manifests itself in enhanced pH stability. Large reef aquariums
with high intensity illumination and even moderate bioloads have a
tendency to be a bit "strangled" when it comes to either picking up
or losing CO2 to the atmosphere. The higher the throughput of the
skimmer (air and water volume) the more this situation is going to
be improved. The ETS directly helps pH stability in this way.
3. The air flow increases the evaporation from the aquarium, and
that allows one to add more limewater. Also, and this is very
important, since the water is in better contact with the
atmospheric reserve of CO2, one is less likely to get into the
regime where calcium and alkalinity are being lost from the system
(which can occur when the pH of the aquarium is very high, the most
dangerous time is toward the end of the photoperiod.) For all the
aquaria running this skimmer about which I have inquired, the
alkalinity has improved. Good gas exchange is crucial for
limewater to be effectively used as a source of calcium and
alkalinity replenishment. By promoting excellent gas exchange and
increased evaporation the ETS promotes enhanced stability in
soluble calcium and alkalinity.
4. Some have indicated that the concentrations of problematic
inorganics have also decreased. An indirect mechanism is called
for to explain this, as nitrate and inorganic phosphate are not
directly skimmable. Terry Siegel is reporting decreases in both
on his "new" system, which is a converted fish aquarium, and this
observation may be explained by a couple of effects, all of which
A. Nitrate and inorganic phosphate are often derived from
organic precursors (nitrogen in uneaten food, organic phosphates)
and based on reasons given above, this skimmer is probably doing a
better job of intercepting organics before they are mineralized
than anything I've previously seen.
B. Skimmers are pretty damned good mechanical filters, they
take a lot of particulates out of the water. Those particulates
may be bacteria, phytoplankton, or organic debris. All of them are
going to contain some phosphate and organic nitrogen.
C. Additionally, because the amount of limewater that may be
added is increased, one may get more inorganic precipitation of
phosphate, and perhaps removal of this from the system in
It should be remembered that this report is from a converted fish
to reef aquarium, a lot of changes were made at the same time, the
most prominent being the replacement of a very good venturi skimmer
with the ETS, the addition of high intensity lighting (which will
tend to promote algal growth and uptake of both nitrate and
phosphate from the system) and the addition of limewater.
If anyone cares to dismiss the reduction in inorganic pollutants out
of hand because they are not directly removed by the foam fractionation
process, I invite them to disconnect their skimmer for an extended
period of time and report what happens to the concentration of inorganic
pollutants. They will go up. It isn't necessarily a one way street,
if you reconnect your skimmer they will probably go down to some extent,
as sources of these nutrients are intercepted before mineralization,
and nutrients incorporated from that existing soluble pool are utilized,
converted into organic form and rendered eligible for removal by a foam
5. Most of the people who have these skimmers so far have felt
more at ease with feeding the inhabitants more than previously, so
there should be benefits to the organisms from this as well.
Again, all foam fractionators contribute to the above effects.
This skimmer is higher performance, so it just gives more of these
benefits than is typical.
What are the other considerations related to this skimmer?
Well, you need a competent water pump to drive any ETS, and a mail-
order little giant pump with enough "humph" to drive the smallest
will set you back another hundred dollars or so, more if you decide
to go with an Iwaki pump. There will be some incremental heat load
on the system because of this. It will probably be offset to some
extent by increased evaporation. The need for a strong water pump
is shared with venturi skimmers as well, so this is hardly a unique
consideration. You will be getting a slightly higher power bill
as well. Again, not a unique disadvantage.
Everyone's first concern about the skimmer is "do the bioballs
cause nitrate to accumulate in the system?" "Do they become
colonized by nitrifying bacteria?" The flow is really
exceptionally fast and turbulent throughout the bioball section of
this skimmer, and I think the bacteria are going to need mountain
climber's gear to hang onto the bioballs. No one has reported an
increase in nitrate concentration on any system which is running an
ETS, and I have asked very pointed questions to both system owners
and the designer on this topic. I will pass along any reports of
problems along this line.
The skimmer apparently "glitches" in the beginning. It doesn't do
much for a couple of days, but when it starts producing foam:
Watch Out. People have found a few gallons of water on their
floor. After that, it seems quite stable. The present version of
the skimmer will has an optional collection chamber with a floating
check valve which will block the formation of additional skimmate,
preventing overflows. Something along these lines is a good option
for any skimmer. For your peace of mind, your carpet, and your
marriage, it is probably a wise to get this. Why doesn't it skim
for a day or so after it is installed? Who knows? Probably
because the plastic surfaces in the skimmer need to be primed with
organics before it starts to work.
Earlier designs returned a small amount of fine air bubbles to the
system, for this reason, most people are running them out of their
sumps. This problem may have been solved, I saw prototypes. A few
bubbles in the sump aren't a big aesthetic issue, and in any event,
there wasn't a problematic amount of air being returned. It isn't
going to generate noticeable salt spray.
There are two possible adjustments on the skimmer, one of which is
its height relative to the sump. The second is a water valve that
you may choose to supply yourself to control the amount of water
going through the skimmer, although most people are running the
correctly sized pumps wide open. I don't think it is a good idea
to tee off an existing pump to feed the skimmer, and most people
don't have enough reserve pumping capacity to make that practical.
In practice, this skimmer seems to require a dedicated, correctly
All of the units I have seen so far were plumbed into the system
with flexible (large ID) tubing. This preserves flexibility in
height adjustment, and you can keep the turns sweeping and low
resistance with flexible tubing, if you do it correctly. As with
anything you plumb into your system, you need to do it correctly,
with an eye to reducing the possibility of various failure modes.
The unit can handle a lot of flow, indeed, up to its design limit,
the skimmer works better with increasing water flow. You generally
don't want to lose flow or pressure before the water hits the gun
at the top of the tower.
The ETS will not fit under the cabinet of Anyone's aquarium. And
you will probably need to feather the height of the skimmer
relative to your sump to get optimal results. Given the amount of
time people devote to tuning their skimmers, the initial height
adjustment doesn't seem like a major drawback to this design.
I've recently spoken with Tony Vargas, who is marketing the skimmer
for Gary Loehr, the inventor. They now seem satisfied with the
results of the testing and are now ready to distribute the product.
The phone number for inquiries is (914) 654-8002, if he is not
there, there is a combination answering machine/fax on the line.
There are to be exclusive licenses granted to about five stores in
five states on the east coast. In other states, skimmers will be
mailed out directly by Tony. Pricing and some product literature
is now available. The "owner's manual" is being written by Terry
Siegel. Terry is now running ETS skimmers on two of his large reef
aquaria, and is presently looking to replace his last remaining
venturi unit with a third ETS.
Greg Schiemer was also closely involved in the testing and
evaluation of these skimmers, and his review of this product will
appear in the next issue of Aquarium Frontiers. I have a great
respect for Greg's skills as an aquarist, having seen his aquaria
now on three occasions, it is clear that he is extraordinarily
successful with these systems. Greg's most recent articles include
a piece on herbivorous hermit crabs and a review of calcium test
kits in Aquarium Frontiers.
The skimmer doesn't look like anything else I've seen before, and
it disrupted some of my cherished ideas of what is truly important
for foam fractionation. However, I am only too happy to see
paradigms overturned (even when they are mine) if the result is a
superior aquarium environment. Based on what I have seen, the ETS
skimmer is an example of such a paradigm overturning invention.
There is a patent pending on this design. It is well constructed
from cast cell acrylic, the seams look strong, and attention has
been paid in the design to convenient maintenance (there are
flanges in the right places.) The construction quality is on par
with the very best commercially available units (for example, the
MTC venturi skimmers) and is far, far better executed than most.
If you are considering the purchase of a quality skimmer, either
for a new system or to upgrade your existing one, I urge you to at
least investigate the ETS, and wish you and your corals only the
best fortune with whatever design you chose.