| Article Index |
|---|
| Dido's Watermaker |
| Fresh Water Production |
| Build your own! |
| Experiences, Operation, and Repairs |
| 2011 Modifications |
| The Real Truth |
| The New pump in Action |
| Conclusion |
| All Pages |
Why and how we have built a DIY watermaker
We need water.
During summer we spend 6 to 8 weeks on the Adriatic. Anyone who has three kids knows that 65 liter (17 US gall.) of fresh water is practically nothing. It lasts for 24 hours, if you take care maybe 48 hours. When you need water assumingly you would simply go alongside somewhere and top up your water tank. Not so on the Adriatic: The islands there are short of water anyway, due to the climate. Furthermore, there are water taps not in every harbor at the pier. If there are any, you need the right adaptor and an endless hose. If you have all this, it is not yet sure that the water is on when you are there or that there is any vacant berth close enough to the very few taps or that the existing tap is unlocked and you need to find someone with a key. And finally, if you have this all, you will be charged ridiculous high prices. In marinas the situation is of course better - and they charge you even more. This however is totally understandable as there are thousands of boats during summer, and all of them need water ...
Usually you get water more easily when you are at the gas station bunkering diesel, but we only need diesel about every 10 days.
Thus our strategy for water supply was: take a 25 liter canister (6 gall.) when going out for dinner and carry it back to the dinghy, with your arms getting longer and longer. To cut it short: Water turned out to be a major problem for showering and doing minor laundry.
We do not stay in marinas for overnight but prefer remote and picturesque anchor bays. Lying at anchor we talked to other boaters how they managed this challenge. There solutions were straightforward: much bigger tanks (no option for us), no kids (ha-ha!), or – a watermaker on board!
Watermaker turned out to be a magic word. I had heard about it before but no clear idea how it works exactly, nor how much this would cost. Those boaters who had one told us: take into consideration some 3 to 5 thousand Euro (4 to 6 thousand USD). Before the dream was really dreamt it started to vanish.
Back home by the end of summer I started to google around. After a few days I understood that one could build a DIY model for around 2000 EUR / 2500 USD. This sounded also very much, it was, however, within what we were willing to invest for fresh water available when and where we'd like. At the end it turned out to be even less.
How does it work - in two sentences?
Reverse Osmosis: Seawater is pumped through a membrane at high pressure (55 bar / 800 psi). You get 20% fresh water, 80% goes overboard again as brine.
A few more details about our DIY installation
(Many technical details and instructions that I found when searching the internet are listed at the end of this article. So here I'll rather describe my experiences and challenges during the project.)
Energy, Energy
Most important is the energy-question: You need a pump that makes 55 bar / 800 psi pressure and has a flow rate of about 5 liters per minute / 1.3 gpm or more. Typically this will be a plunger pump that you might know from the high pressure washer at your garage or at home for pressure washing your car etc. Such a pump needs about 1500 Watt or about 2 hp. You may use an AC-motor and a DC-AC-inverter, or you may use a 12V DC motor, or you may use the boat's main engine and a belt drive with a clutch, similar as the compressor of the A/C in your car works. All three methods are similar in terms of costs, the main engine driven version is certainly the most attractive one: no additional equipment necessary, no heavy electric loads (1500 Watt from 12 Volt DC means 125 Amps and wires as thick as a finger!).
Dido's engine compartment, however, makes a main engine driven pump very complicated due to space restrictions. So I decided to install the DC-AC-inverter. Rough calculations convinced me that the 50 Amps alternator of our Solé Diesel had to be replaced by something bigger. I found a 110 Amp alternator that would fit. Furthermore - as I wanted to top up the batteries quickly (the alternator would only provide 2/3 of the energy necessary during water making, the rest would have to be taken from the batteries) - I installed a so called "alternator-to-battery-charger" (A2B), which charges the batteries really incredibly fast (there is an article about this available here also).
Under optimum conditions the membrane itself is capable of producing almost 110 liter / 30 gall. fresh water per hour if you provide a pump with a flow rate of 5 to 6 times as high. A pump of such a flow rate (11 lpm / 3 gpm) needs about 5 hp, it is big and heavy, and it is horribly expensive. Together with an AC-engine or an electro-magnetic clutch and a belt drive a seawater resistant pump will cost about 1200 Euro / 1500 USD, maybe more.
Pipes and threads and valves and hoses and BSP and NTF and inch and millimeter ...
The membrane is within a tube (called high pressure vessel), the vessel has 3 openings with threads to screw fittings into. Then you need a high pressure needle valve to adjust the pressure, a relieve valve for security reasons, 3-way-valves, hose connectors, bows, filters with connectors and threads, and and and. All of them should be seawater resistant.
When I looked for them in the internet I found items for 10 USD and for 100 USD doing the same thing. I wanted to understand the differences (which turned out to be a time consuming and anyway useless hobby. Some parts are the same but without a stamp that says: "medical proof" or "may be used in nuclear power plants" whereas others are intended to be used in agricultural machines or pressure washers). Putting together the most reasonable parts turned out to drive me mad: one part had a 1/4" NPT thread and should be connected to a 1/4" BSP thread. Then there is a 1/4" NPTF that should fit onto 10 mm. some of them are self-sealing, some of them need a sealant, some of them for hydraulics, others for gas, others for marmalade or so.
One inch equals 25.4 mm, but if you measure a 1" pipe you find that it is 25.4 mm neither inside nor outside. Same is true for 1/4 inch, 1/2 inch, or 3/4 inch. You may measure any way, you will never find one inch when you measure a 1" pipe or hose or thread. I felt confused to a certain extent.
I studied anthropology at university not any technical sciences, and I suddenly understood why the NASA once lost a space probe on Mars due to the fact that in a computer landing software they mixed up meter and feet. Thus, I thought, if this happens to NASA engineers why should I do better. So I ordered many parts at least twice. My workshop is full of brand new parts now that might make a small rocket or at least a chemical laboratory.
But finally the foot bone was connected to the ankle bone, the ankle bone connected to the shinbone, the shinbone connected to the knee bone and them bones walked.
The High-Pressure Pump
The critical question still was the pump: I already had a few offers for plunger pumps of around 1200 USD when I occasionally read in a sailing newsgroup (I have become member in dozens of newsgroups during this research) that someone used a standard low-cost ordinary high-pressure washer from Kärcher for 70 USD without problems. The idea attracted me immediately: 70 instead of 1500 motor included! So I tried to find out why not everyone is doing this and what would be the disadvantages.
Very obviously the ordinary high-pressure washers use the same plunger pumps, however the material of the pumps is different: The watermaker pumps are made from stainless steel and / or seawater resistant bronze, whereas the "normal" pumps are made from brass and aluminum, to explain it the simple way.
Now, the sailor who experimented with the Kärcher wrote that especially the very cheap washer pumps are made mainly from plastic, so the corrosion problem should not be one. I did not disassemble one to find out but I thought I should give it a try. If such a pump only would last for 6 - 8 weeks holidays I could buy a new one every season for ten years and still save half the money of a CAT pump or something similar. Further, I said to myself, as I am not bluewater sailing the worst case scenario is that the pump refuses to work after a certain number of days and I will be where we were before: canisters and long, long arms ...
Another limitation of such a pump is the flow rate. The maximum pressure (100 bar or 1450 psi) is far beyond what we need and will cause no trouble, but the maximum flow rate is stated as 340 liter per hour or 90 gph. This equals 5,5 liter per minute or 1.5 gpm. When actually running I estimated that it probably would be rather 4 liter or 1 gpm.
With these figures the calculation goes like this:
Nominal pars of the membrane:
Maximum feed rate: 23 lpm / 6 gpm
a) Maximum permeate rate at water temperature of 25°C / 77°F 108 lph / 28.5 gph
b) Maximum permeate rate at a water temperature of 20°C / 68°F 90 lph / 24.5 gph
c) Maximum permeate rate at a water temperature of 15°C / 59°F 76 lph / 20 gph
With the Kaercher Pump installed the results should be:
Maximum feed rate of the Kärcher pump: 4 lpm / 1 gpm
a) 19 lph / 5 gph
b) 16 lph / 4 gph
c) 13 lph / 3.5 gph
This means: for our “spring trip” in April/May (sea still cold) we should expect about 13 liter or 3.5 gph, for our summer trip (water warmer) we should expect about 19 liter or 5 gph fresh water with the Kärcher pump. The watermakers “from the shelf” with their expensive plunger pumps and the same membrane produce twice as much water the brochures say. But anyhow, still not bad for that money.
So I went to the Home Depot, bought one of those 59.95 Euro Kärchers, opened it and took off the pump/motor-unit. I connected it to the referring hoses (again: various different threads), turned on the water and electricity – and them bones walked!
The “test drive” at home with homemade seawater proved that the system worked. The first glass of “permeate” as the fresh water coming from the membrane is called tested like – water! Now the final question was: would the system work on board under real life conditions?
