Cruising Info

General Cruising Info
Pacific Ocean Cruising
Indian Ocean Cruising
Provisioning Food & Fuel
Cruising Recipes

Equipment Pages

KAP Files
Sail Drives
Epoxy Work
Solar Panels
3G Modems


When we cruised in the '80s, watermakers (or desalinators) on private boats were extremely rare, and usually found only on larger, expensive yachts.  But now many watermaker companies cater to smaller boats.  The units are still fairly expensive to buy and to run so they're still a bit of a luxury item, but they're much more affordable than they used to be, and several good options are available depending on your needs.  Subjects discussed in this article include:

Do you really need a Watermaker
How Reverse Osmosis works
How much water do I need?
Watermaker Types
Selection Criteria
A crazy alternative
Installation Tips
Testing water Quality
Watermaker Care
Watermaker Costs
What we do on Ocelot

Do you really need a watermaker?  (or Why buy a cow when milk is so cheap? )
There are actually very few cruising grounds in the world where you really need a watermaker.  Just about anywhere in the world with a local population will have potable water somewhere ashore, be in from a tap or sold in big jugs.  It takes a lot of bought water to come up to the thousands of dollars you'll spend on a watermaker system.  So why spend the money?

Abby from Estrela collects non-potable water at the well in Chagos
Getting non-potable water from a well in Chagos

One thing a watermaker does is to extend the amount of time you can spend in places that don't have easily accessible water ashore -- places like the San Blas Islands of Panama, the outer islands of the Bahamas, the Tuamotus of French Polynesia, isolated atolls in the Pacific, Chagos Archipelago in the southern Indian Ocean, and many other beautiful areas that are thinly inhabited or even totally uninhabited.  These are exactly the sorts of places that we seek out, and we like to spend as much time in them as possible.

Another advantage of a watermaker is that it increases the level of personal comfort on board: Dishes can be washed in fresh water, clothes can be washed as often as you wish (and boats with watermakers can often support having a washing machine), showers can be taken as often as you want -- including each time you come out of the ocean after swimming.

Watermaker water is guaranteed healthy (assuming you've kept up to industry standards), a great advantage when cruising in places with potentially iffy water, or where heavy rains can cause run-off to pollute the local water supply.

Watermakers save you the wear and tear on muscles when the alternative is to schlep 20 liter jugs of water from shore to the dinghy and then up to the boat deck.  We have a friend who, after wrenching his back and having to spend a long time lying down to recuperate, wrote on all his water jugs: "Warning: Carrying water jugs can be hazardous to your Back!"

12v DC motor driving a high-pressure pump
12v DC motor directly driving a high-pressure pump
This system delivers 60 L/hr but draws 30 amps

How much water do I really need?
This is another number that varies tremendously between boats.  On Ocelot, with 4 adults all taking fresh water showers (but from a sun-shower, not a pumped system) and washing our dishes in fresh water, and doing a load of laundry every few days, we'd go through 400 liters a week, or about 50 liters a day.  Note that nobody goes down below salty - we always rinse off after swimming, so the upholstery doesn't get clammy.  Washing dishes and/or bodies in salt water, or catching rainwater (which we generally don't do) would certainly bring that number down.  Washing your decks down in fresh water after a sail, showering with pressure water, doing lots of laundry, or even doing a lot of scuba diving (and having to rinse the gear afterwards) will push it up.  But 12 liters/person/day (3 gallons/person/day) is probably a good starting point.

How Reverse Osmosis Watermakers Work
Virtually all commercial watermakers use a "Reverse Osmosis" (RO) semi‑permeable membrane to remove the salt from seawater.  This bit of magic is typically a hollow tube about a meter (39") long and 60 mm (2.5") in diameter, encased in a high‑pressure housing.  Such a membrane is called a 25 40 or even a 2540 (for 2.5" diameter by 40" long).  When clean seawater is forced over the outside of this membrane at a pressure of 40‑60 bar (600‑1000 psi), about 10% of it will go through the membrane and emerge as drinking water.  Magic!

The reason it's called "Reverse Osmosis" is that the salt in seawater has a natural tendency to flow towards water that has no salt.  The "osmotic pressure" of seawater is about 24 bar (350 psi).  By applying a pressure much higher than the natural osmotic pressure, this tendency is reversed, and the semi‑permeable membrane allows only the fresh water to pass.  If you want more than this layman's explanation, Wikipedia does a pretty good job.

Watermaker Types
Getting seawater up to the pressures necessary for the RO membrane to work requires a fair amount of energy.  Boats with large AC generators can use an AC motor to drive a high‑pressure pump.  Alternatively, the high‑pressure pump can be belt‑driven off a generator, or even the main engine.  This adds a bit of vibration but if the hoses are long enough this shouldn't present a problem, and it's one of the more efficient designs.  These designs can produce a lot of water fairly quickly (hundreds of liters an hour).  However, the AC watermakers don't care as much about energy efficiency.  They're far more concerned with money efficiency, producing a reliable unit cheaply.  Therefore, some of them consume a lot of power.  Make sure you check the specs and make sure that your generator has the extra capacity.  We've seen several reports of folks who've installed AC RO systems only to find that their generators weren't big enough to handle the motors.  (Remember that starting a motor will require 10 times its normal running current.)

Boats that have limited AC power generation capabilities will probably want to go with a 12v design.  The quick-and-dirty approach here is to connect a really big 12v motor to a high‑pressure pump.  Several companies produce this type of desalinator.  It's a relatively cheap approach but the low liters/amp‑hr ratio mean it's relatively inefficient, so you end up paying (in energy) in the long run.

The pressure amplifier uses no electricity at all
The pressure amplifier uses no electricity at all

A more efficient (but higher initial cost) design uses fairly standard 12v water pump(s) to produce relatively large volumes of water at relatively low pressures (4‑8 bar or 60‑120 psi).  This water is fed into a "pressure-amplifier" which converts it to much lower volume but much higher pressure - typically by a factor of 10:1 each way.  This produces the 40‑60 bar (600‑1000 psi) needed to make the RO membrane work.  The efficiency improves because the high pressure waste water in the membrane housing (the water that did not go through the membrane) can be used to help power the next stroke of the amplifier, making the whole system much more efficient, typically producing twice as much water for the same number of amp‑hours.  The "pressure-amplifier" is strictly a mechanical device and uses no electricity at all.  Spectra Watermakers are probably the best known company to use this design approach, and they're considered the Rolls Royce of 12v watermakers.  They have a cute animated graphic of their "Clark Pump" amplifier for those who're interested.

There are (or were) units available that use no electricity at all, but are towed behind your boat and are powered by the boat's motion through the water.  Reviews seem to be generally positive, but at least one company's website doesn't seem to work anymore, so perhaps they're no longer in business.  We have no direct knowledge of these systems, but given the number of fishing lures we lose, we'd be a bit hesitant to tow very expensive gear behind Ocelot - it might be eaten!

In the interests of completeness we should mention that some big ships use the excess heat and vacuum from their engines to boil seawater, and then they use cool seawater to condense that steam to produce absolutely pure distilled water.  This technique is also used by big land-based multi-stage "flash plants", especially in the Middle East, where they're usually located next to big power-plants so they can use the excess heat from the generators.  However, this is an extremely energy intensive method, suited only to large installations with lots of excess heat.

Selection Criteria
The first issue is how do you want to power your watermaker?  If you've got a generator then you almost certainly want to use it to drive a high-pressure "Cat" pump, either directly (via a belt from the engine) or via an AC motor.  Driving your pump directly from the engine is cheaper (since it dispenses with the AC motor) but it also limits flexibility somewhat.  When your generator breaks down, how will you power your pump?  With the more expensive AC motor option, you can use another AC source if necessary.

There are many more AC watermaker systems available than DC systems, so you'll have a much larger selection of units to buy.  With an AC system, your selection criteria is probably based on how much water your unit can produce for a given amount of money.  That is, you'll want to maximize (liters per hour)/cost.  Running costs are pretty much the same for most units, as long as your generator has sufficient capacity to drive the pump as well as perform its other duties.

Another 12v direct-drive installation
Another 12v direct-drive installation
Membrane housing in foreground

For either AC or DC installations, another decision is to buy a monolithic system in its own enclosure, or to buy a system that uses separate components that you can tuck in where they'll fit better.  Most cruising sailboats have limited space for such add-ons, and watermakers can be pretty big.  Distributed systems are certainly more work to install, but the installation isn't difficult and they can fit in places a monolithic unit can't.  The components you'll have to mount and plumb together include a seawater strainer, several water valves, 1-3 standard domestic water filters, the membrane in its pressure housing, and at least one water pump.  DC systems will also have to run fairly large cable from your electrical distribution panel to your motor(s).  Possible other equipment includes a pressure amplifier (if your system uses one), flow meters, pressure gauges, control panel, pressure accumulator, etc.  This may sound complicated but it actually isn't, and it's a good way to become more familiar with the components of your watermaker.

For DC systems, there are fewer buying options so fewer decisions.  If you're going to be cruising long-term then you might want to consider the extra up-front expense to go with a more efficient system using a pressure amplifier.  But if money is tight then perhaps a cheaper direct drive system like the Katadyne (who bought Pūr) is better for you.  Katadyne also makes units that can be powered by hand, which sounds like a useful unit to have in a life raft or a ditch-bag.

RO Membranes don't last forever, so make sure you ask your manufacturer if their system uses standard sized membranes (like a 2540), or if they use a custom membrane that can only be purchased from the manufacture.  Standard sized membranes are manufactured by several companies and are available from a variety of sources for about $150.  Custom membranes often cost 2‑3 times that much.  Our first membrane lasted 6 years but we've heard of them lasting 10 years.

A fairly crazy alternative
We've heard several people talk about building their own watermaker but I've never actually seen anyone try.  There are several companies selling RO membranes for less than $200.  High-pressure membrane housings can also be found on the web.  Then you need a high pressure (Cat) pump and some way to drive it, a pressure regulating valve and perhaps some instrumentation, like a pressure gauge and perhaps a flow meter.  Total cost so far is still less than $1,000, much less than the $3,000‑$9,000 that you'll spend on a commercial unit.

Installation tips
These will vary with each boat so we can only give generalities here.  DC watermakers use a lot of power and their performance tends to be fairly voltage sensitive.  Make sure your electrical connections can't corrode and that your wires are big enough that you don't get any significant voltage drops.  We put ours under Chris' bunk, a good 10m (30') from our power panel, so we ran huge wires and soldered all connections we could.  Even going through 3 circuit breakers before it gets to the motors, we still only lose about 0.3v.  We also tend to run the system when the batteries are at their highest voltage - we watch our Solar Controller and when it starts limiting current to the batteries we run the watermaker for an hour or 2.

The components should be as low as possible in your boat - ideally below water line.  This will help keep air from getting into the system and will also mean that less pressure is lost to lifting the water up.  High pressure pumps generally can't pump if there's air in the system, and getting air out can be difficult.  If your primary pump is too high, you may need a "lift pump" which costs more in terms of money, electricity, and complexity.

Make sure you install your watermaker somewhere that can take getting wet.  With such high pressures, it's likely that some connections will leak a bit.  You also won't want to put the first 5 minutes of water produced into your tanks (it often smells a bit of hydrogen sulfide).  We save the first few minutes of water produced in a bucket and usually use it for washing dishes.

Any watermaker makes enough noise that you probably shouldn't run it while folks are trying to sleep nearby.  Ours is fairly silent, but it's right under a bunk.  The motors vibrate a bit and the pressure amplifier clicks every few seconds as it shifts.  But it's only an issue in that cabin.

Testing the water quality
Watermakers generally produce very good quality water, but you should test the water periodically to make sure that everything's working OK.  Do NOT depend on tasting the water.  If you can taste salt, the water is 2‑3 times saltier than is good for you.  Seawater usually starts out at something like 30,000 parts per million (PPM) of Total Dissolved Solids (TDS: salt, etc).  A watermaker should bring that down to 200-500 PPM.  More than 500 PPM is not good for you long term, and can lead to kidney stones and other unpleasantness.  But water doesn't taste salty until it gets up over 1,000 PPM.  A small battery powered TDS meter generally only costs $10‑30 and is well worth the investment.  Many watermaker suppliers will include a TDS meter when they sell you a unit.

Watermaker installed under a bunk, with the membrane in the bilge
Our watermaker was installed under a bunk,
with the membrane in the bilge and the
pressure amplifier under the step to the bunk.

Care of your watermaker and components
Watermakers like to be run everyday, if possible.  In cooler climates you can get away with running it less often, but in the tropics you need to at least flush the system with (un‑chlorinated) fresh water every other day to prevent growth on the membrane.  But if you can't run it for a week or so, you should "pickle" your system with special preservative.  Most units can be pickled with simple Sodium Metabisulfite, which is commonly available and much cheaper to buy from a chemical supply house than from the watermaker manufacturer.  But some units, like the Spectra, need special pickling solution.  Un‑pickling your system usually involves running it unpressurized for about 30 minutes, then pressurizing the system but not using the water for drinking for another 30 minutes or so.

Growth on the membrane can sometimes be cured with either an alkaline or an acid wash.  Both chemicals can be purchased from your manufacturer.  Mix the chemical up in a bucket of (ideally) hot water, de‑pressurize the membrane, and place both the input hose and the brine-discharge hose into your bucket, to create a closed system.  Then run the system (unpressurized) for 30‑60 minutes, circulating the chemical solution through the system.  Forcing the chemicals through the membrane will damage it, but running the system unpressurized will clean the outside of the membrane.

Another maintenance item is the input pre-filters.  At the very least, you should have a 5 micron filter on your water input line.  We have an additional 20 micron filter in front of the 5 micron filter, and it's the 20 micron filter that clogs up the fastest.  How fast it clogs depends on the waters you run your watermaker in.  In the open ocean, we can go months without changing the filters, but coastal cruising usually causes the filters to get dirty somewhat faster.  As a general rule, we like to be able to see into the water at least 10' (3m) before we'll run our watermaker.  If we can see where our bridle connects to our anchor chain, that's usually good enough.

RO membranes are very sensitive to chlorine.  Make sure that no chlorinated water gets near the membrane.  If you flush your system after each use (as you should in the tropics) then put an activated charcoal (carbon) filter in your flush circuit and make sure that you replace that filter at least every 6 months, no matter how little you use it.  Better yet, flush only from the water you produce, and make sure that only watermaker water goes into that tank.  Our watermaker feeds into our port-side tank, so domestic water never goes into that tank, only into the starboard tank.

What does a watermaker cost?
This is another number that varies tremendously, even with time.  Luckily, the internet allows you to do cost comparison research quickly and easily.  The cheapest commercial systems we've seen cost about $3,000 and higher output or more efficient systems can cost 2‑3 times as much.  We can say that our Spectra 200C cost us about $5,000 from Emerald Harbor Marine in 2001.  It produced about 30 liters (8 US gallons) per hour in its initial configuration, before we added the second pump to bring it up to 50 liters/hour.  Spectra is one of the highest quality and best respected manufacturers of 12vDC watermakers, and their prices reflect that.  In 8 years we've probably spent another Boat Buck ($1,000) for a new membrane, new pump heads, pre-filters, pickling and cleaning solutions, and bringing our pressure-amplifier pump in for routine service once.

Installation costs depend on local labor rates, but unless money is no object we'd recommend doing the job yourself.  It's not difficult and it's a good way to get to know your system.

Besides energy, running costs are fairly low.  AC systems are usually run when the generator is on to charge batteries so the incremental costs to run the watermaker aren't a lot.  Our unit consumes about 15 amp‑hours/day, which is about 40% of the output of one of our 120W solar panels.

Pre-filters have to be changed when they get dirty.  We usually use the kind that can be cleaned and reused, although they cost a bit more.  The filter housings can be bought from any hardware store for much less than from a marine store.

If you won't be using your watermaker for more than a week or so then you'll need to pickle it, which usually costs about $10 in chemicals.

Our first membrane lasted 6 years, but we've heard of membranes lasting up to about 10 years.  If your membrane is a standard size like a 2540, they're usually available on the internet for under $500.  But some companies use special custom membranes, forcing you to go to the manufacturer and pay 2‑3 times that much for a replacement membrane.

We don't know how long high-pressure "Cat" pumps last, but we usually have to replace one of our ShurFlo pump heads every year or so.

What we have on Ocelot
Ocelot is a 12v "battery" boat without a generator set so we needed a 12v solution.  We started off with 4 adults on board, 2 with long hair that needed frequent washing.  We also like to go scuba diving, and scuba gear needs to be rinsed in fresh water every time it's used.  Being an engineer with (at the time) a paying job, I decided to go with the most efficient design I could find - efficiency in this case being defined as the most liters of water produced per amp-hour of electricity, not liters/hour/dollar spent.

In 2001, the hands-down leader in this category was the Spectra 200C.  This unit uses a single 12v pump (actually produced by ShurFlo - their model 8008‑943‑839) working into a "Clark Pump" pressure amplifier.  The specs say that in ideal conditions it produces 8 US gallons (30 liters) per hour while consuming only 8 amps, thus yielding an efficiency of 1 gallon (3.7 liters) per amp-hour of electricity.  Watermakers don't produce as well in warm water so we found that reality in the tropics didn't quite get to this level, but we're usually within 80‑90%.  The 200C could also be expanded by adding a second ShurFlo 8008‑943‑839 pump in parallel with the first, which we did by buying a secondhand pump from another cruiser in Grenada.  The second pump doubles the power consumed (of course) and almost doubles the water output.  So it's not quite as efficient as the single pump design, but the higher pressure also means that water quality is improved, because the RO membrane produces better quality water with higher pressure.  Unfortunately, the higher pressures of running both pumps together also causes the pumps to wear out faster.  On Ocelot we have switches that allow us to run either pump singly or both together.

Ocelot's watermaker installation
The 4 valve panel is to the left with one pump next to it.  The 2 filters
are next to that, with the accumulator to the right of the bilge pump.
The second pump & the carbon flush filter are mounted on the near wall.

Our installation has not been trouble free, but no watermaker installations ever are - depend on it!  In 2003 the support from Spectra was excellent.  Our first problem occurred in French Polynesia, after we'd been using the system for 2 years.  Our pressure amplifier pump started shifting erratically, much faster on one side than the other, and output plummeted.  We emailed Spectra and based on our symptoms they knew exactly what the problem was - a small "annular ring" inside our pressure amplifier had cracked.  They promptly emailed a quick-fix: remove the annular ring, reverse it, and replace it.  Only the high-pressure side is prone to cracking so reversing the ring would get us working again.  Of course, this solution worked beautifully, but we weren't happy having a cracked ring still in the system.  So Spectra air freighted us 4 new annular rings, 2 special removal tools, and a large fist-full of spare O‑rings, all at their expense.  We'd also mentioned that our output had been slowly dropping over the last year or so, so they included a pair of brand new pump-heads in their shipment, also at no cost to us.  Unfortunately, Glenn, who was responsible for that excellent service, no longer works at Spectra, so their service has dropped off markedly.

Our first membrane lasted about 6 years but eventually died in the Maldives, of all difficult places to try to get spares.  The symptom was that the salinity went up.  It still removed most of the salt, but instead of 350 ppm, the water was more like 1,300 ppm, good enough for dishes but not safe to drink long term.  We eventually had a new membrane shipped into the Seychelles.  Replacing the membrane was fairly straightforward: disconnect and remove the membrane housing, unscrew one end, remove the old membrane, install the new one (lubricating the seals lightly), screw the end back on, and re-install where it belongs.  The manual that comes with the Spectra details all this (and more).

Our only real issue with our Spectra is that our (older) model tends to chew through pump-heads.  After about 700 hours (say, 2 years) the pressure and output drop low enough that we need to replace the pump heads.  We've gone around and around with Spectra on this, and they tend to deny that they have a problem at all.  They used to give us a special deal on new pump-heads but they no longer do.

But Spectra has changed their pumps recently (without admitting that their previous pumps had any problems) replacing the ShurFlo pumps with fancier vane-pumps.  These are more expensive and they consume more electricity so they're less efficient but apparently they last better.  They also apparently can't be run dry, so Spectra has placed elaborate sensors and control systems around the pumps in their newer designs.  The net result is that their systems are somewhat more complicated and expensive, but should have better reliability.

Equipment Pages: KAP Files | OpenCPN | Turnbuckles | LEDs | Anchors | Sail Drives | Epoxy Work | AIS | Solar Panels | Watermakers | 3G Modems

Cruising Info: Up | General Cruising Info | Pacific Ocean Cruising | Indian Ocean Cruising | Provisioning Food & Fuel | Cruising Recipes

Top Level: Home | Destinations | Cruising Info | Underwater | Boat Guests | Ocelot | Sue | Jon | Amanda | Chris | Site Map | Make a Comment

The Triton - Nautical News for Captains and Crews
If our information is useful,
you can help by making a donation

Copyright  2000‑2017  Contact: Jon and Sue Hacking --, svOcelot.comAll rights reserved.