One of the most basic problems
that arise in our hobby is the difficulty
in assuring a constant source of
pure fresh water. We need fresh
water for mixing our saltwater and
for our daily replenishment of the
water that has evaporated from our
tanks. When I started keeping reef
tanks, more years ago than I'd
like to remember, we knew that we
needed clean water for
our systems. We weren't aware,
though, that we actually required
pure water, free of chlorine,
nitrates, phosphates, ammonia, etc.
The result was that, even after
aquarists had figured out that proper
light, sufficient flow, and stable
levels of things like Alk, Ca, and
Mg, and pH were necessary to sustain
SPS and many LPS corals, we still
had problems. Our fresh water just
wasn't pure enough.
The great majority of aquarists
used tap water when mixing up their
salts or topping off their tanks.
The tap water in some homes comes
from municipal sources, others access
residential wells. Each source has
it's own problems when it comes
to pure water. Municipal water is
usually dosed with chlorine and
chloramines. Depending on the type
of pipes installed in the home,
metals such as copper came become
a problem. Silica, chromium, nitrate,
and phosphate are other possible "contaminants".
Depending on the geographic region,
well water can contain even larger
amounts of nitrate and phosphate
than city water due to heavy agricultural
based fertilization, etc.
With the event of home sized
reverse osmosis and deionization
(RO/DI) systems, it has become a
simple exercise for everyone to
have pure fresh water for their
What is reverse
We all learned in school that
osmosis is the diffusion of molecules
through a semipermeable membrane
from a place of higher concentration
to a place of lower concentration.
Sounds good, huh? The point, of
course, is that osmotic pressure
tends to equalize the molecular
concentration of two liquids which
are separated by a semipermeable
membrane. In our RO systems the
membrane has been designed with
holes that are only large enough
to let water molecules through.
Anything larger (the bad stuff)
can't get through the membrane.
So we have a situation where "dirty"
water from the tap passes through
a membrane and all the gunk stays
on one side and only pure water
gets to the other side. Something's
not right, though. We said osmosis
is where a liquid of lower concentration
(pure water) passes through to a
solution of higher concentration
(gunky water). What we need to achieve
is the opposite - uh, the reverse.
Oh yeah, reverse osmosis. How do
we reverse osmosis? By applying
pressure to the gunky water and
pushing it through the
membrane! The pressure, of course,
comes from the pressure in your
water lines. If we can apply enough
pressure, we get nice pure water
on one side of the membrane and
really gunky water, on the other
side, which gets eliminated down
So, that was my totally unscientific
attempt at defining reverse osmosis.
The pressure thing, by the way,
is very important. If your water
lines don't carry enough pressure
(usually defined as a 40 PSI minimum)
your reverse osmosis system will
be very slow at supplying water
and the quality of the water that
is supplied will not be optimal.
In cases of insufficient pressure
there are booster pumps available
which can raise the pressure of
the water reaching the membrane
up to 80 PSI.
Before the tap water can be allowed
to reach the RO membrane, we need
to take care of a couple of things.
First of all - sediment. Tap water,
and especially well water, often
carries significant amounts of sediment.
Not good for our RO membrane! The
answer is a sediment filter. Sediment
filters come in various pore sizes.
Normally a size between 1mm
is used. The sediment filter is
normally placed as the first pre-filter
in an RO system.
I get all my water from a well.
Chemically the water is fairly pure,
but it is loaded with sediment.
For this reason I have a total of
three sediment filters in my system.
and a 0.5mm.
Activated carbon (block) filters
are placed next in line. Among other
things, the activated carbon removes
chlorine and chloramine from the
water. Both these substances can
damage the RO membrane. Depending
on the chlorine and chloramine levels
in the water to be treated, a second
carbon filter may be advisable.
The RO membrane
Not all membranes are created
equal. Aside from the material from
which membranes are manufactured,
there are also differences in the
rated daily amount of pure water
production (in gallons per day).
With membranes suited for our needs,
the range goes from 24gpd to 150gpd.
Experience has shown that for reefkeeping
purposes, the Dow FilmTec 75gpd
membrane delivers the optimal results.
It should be noted that the amount
(and quality of the product produced)
depends on the temperature of the
water being treated and the source
water pressure. A water temperature,
as it comes from the tap, of at
least 70 degrees is advantageous.
As for water pressure, most RO system
manufacturers recommend that water
pressure no be allowed to fall below
40 PSI. If you find that you are
not able to supply this pressure,
it's possible to buy booster
pumps to increase the pressure to
around 80 PSI.
Under optimal conditions of temperature
and pressure, the Dow FilmTec 75gpd
membrane can remove 94% to 98% of
impurities from your tap water.
We judge our success in removing
impurities from the water by measuring
We try to shoot for the lowest TDS
value possible. To use my water
as an example - the TDS of water
at my tap is around 190. After the
water has passed through my pre-filters
and RO membrane, the TDS has dropped
to 3 or 4. That means I'm getting
optimal results from my unit. Later,
we'll discuss how to bring that
3 or 4 down to zero.
shut-off valve (ASOV)
The automatic shut-off valve
enables your RO unit to
turn off the source water supply.
This is accomplished using
pressure from the pure water side
of the unit. The ASOV turns off the
raw water supply, whenever the pressure on the
pure water side of your system
reaches approx. 60% of the raw
water pressure. This happens,
for example, when a float valve
shuts off the flow to a storage
tank as the water level reaches
the top of the tank. In sort,
anything that raises the
pressure sufficiently, on the
pure water side of the RO unit,
will cause the ASOV to block raw
water from entering the RO unit.
The best known, and probably
most used booster pump is the Aquatec
This pump raises the incoming
water pressure to around 80 PSI.
If needed, the pressure can be adjusted
using a small flat head screwdriver.
The pump comes with it's own
power adapter and, if ordered in
a kit, a pressure switch is available.
The pump is installed between the
tap and the RO unit. The switch
is installed between the RO unit
and the DI cartridge(s). The pressure
switch has a built-in switching
point of approx. 40 PSI. If pressure
falls below that, the switch allows
power to reach the pump - the pressure
is raised to approx. 80 PSI. When
the flow of water is stopped, for
instance by a float valve closing
as an RO storage container fills
up, the switch stops the flow of
electricity and the booster pump
We said that RO, by itself, will
only bring down TDS to a value of
3 or 4. But, optimally, we need
to reach zero. That can be accomplished
by allowing the product water, leaving
the RO membrane, to flow through
a deionization resin. These resins
have microscopic pores that serve
to trap any charged (negative=anion,
that come into contact with them.
Molecules/atoms with no charge pass
right through and are not held back.
DI resins will not work indefinitely.
When the resin reaches a point were
it is saturated with ions it becomes
exhausted. It is possible to regenerate
DI resins. The regeneration process
essentially cleans the resin beads
of the charged ions
it has collected.
There are resins designed to
catch cations (plus + charge) as
well as resins designed to catch
anions (minus - charge). In this
country most aquarists use mixed-bed
deionization cartridges where both
types of resins are mixed together.
This has its negative and positive
(no pun intended) aspects. If you
intend to regenerate your resin
after it has become exhausted, mixed-bed
cartridges cause problems. Each
type (cation or anion) resin must
be regenerated separately. This
means, in the case of a mixed-bed,
the resins must be separated before
they can be regenerated. This can
be done, but is a time consuming
process. Better is to use the resins
separately, if you intend to regenerate.
DI resins can be purchased with
a color indicator that changes color
when the resin has been exhausted.
The price is a little higher than
resins without the indicator. I
don't find it particularly useful
as I use the TDS value of the product
water to indicate when it's
time to change resin.
DI resin tends to exhaust completely,
all at once. One minute your product
water has a TDS of zero and next
it's already 3 or 4. For this
reason, I use two DI cartridges
in my system, run in series. For
the sake of this discussion, let's
call them DI position one (just
after the RO membrane) and DI position
two (just after the cartridge in
position one). As soon as TDS drops
below zero, I know that the cartridge
in position one is exhausted, so
I remove both cartridges and refill
the first (totally exhausted) one
with fresh resin. I place the cartridge
from position two (which isn't
exhausted yet, position one always
exhausts itself first) in position
one. Then I place the newly filled
cartridge in position two. In this
way my product water is always at
There are a few items that can
enhance the functionality of a RO?DI
- booster pumps - as already
mentioned above, booster pumps
can greatly increase the amount
of product water produced in
a given time in situations where
water pressure, at the tap,
is below 40 PSI.
- pressure gage - placed just
before the RO membrane, a pressure
gage can aid in determining
when one or more of the pre-filter
are clogging. This is indicated
if the pressure drops from the
- dual TDS meters are available
which allow permanent mounting
to an RO/DI system. Measurements
can then be taken directly after
the RO membrane and after the
DI cartridge(s). This tells
us when the RO membrane is no
longer producing RO product
water in the 3 to 5 TDS range.
We can also see when the DI
product water raises above zeroTDS.
The nuts and bolts
So let's take a look at some
of the components that go into making
up an RO/DI system. Here's a
pic of my current installation:
The system is mounted on a 24"
X 48" piece of plywood that
I covered with some old Formica
I had laying around. Starting from
the left and moving right, I'll
describe the individual parts:
sediment filter. This is places
before the booster pump to eliminate
any danger of sediment (read
sand) damaging the pump.
- the Aquatec 8800 booster
pump. This pump turns on whenever
the the pressure falls below
around 45PSI, which is almost
always. The pump turns off automatically
when the water movement stops
due to a float valve closing
when the water level in a storage
tank has reached the top.
- the power adapter for the
- a pressure gage. The gage
is connected to the input side
of the RO membrane, which is
output side of the pre-filters.
In this position the gage can
tell us when pressure falls,
indicating that one, or all,
of the pre-filters is clogged
- first activated carbon filter.
Gets rid of chlorine chloramine
and some organic chemicals
- 2nd activated carbon filter.
This pic is a few months old.
I have since decided my system
doesn't really require two
carbon filters so I replaced
it with a
- the 1st of two DI filters.
- I have two 75gpd FilmTec
RO membranes installed in my
system. Two membranes greatly
increases the amount of product
water produced every 24 hours.
You can see one membrane cartridge.
The other is below the one that
- the 2nd of two DI filters.
- a DIY RO/DI ATO (automatic
top-off) unit. It takes care
of controlling when the RO/DI
turns on so that the fresh water
storage tank is filled.
- the dual TDS meter. The
meter reads TDS at two locations:
after the RO membrane - indicates
the condition of the membrane,
and after the first DI filter
- lets me know when the resin