DIY Watermaker in another way

[Lars_L]

I have got an idée how to make an efficient watermaker in another way.
First I get the idée.
Second I make a principle sketch of the construction. The image is attach.
Third I talked with friends if they could see any fatal error with the idée. They didn’t.
Next I did a spreadsheet to se what performance I could get. (The spreadsheet is attach.) It shows that I can make a watermaker that produces 30 l/h while it uses 110 W or 3.7 Wh/l. This is performance near Spectra Venture 300 that costs $8000.
Next step in development is to let the CF community check if I have made any big mistakes. If the idée passes this test it will just remain to build a unit.

I know that US is great (and getting greater) but it will always be a small part of the world. The spreadsheet is in metric.
The values I have used in the spreadsheet are sometimes wild guesses, sometimes qualified guesses, and sometimes wished values.
Fields that are market yellow are input fields (not locked).

The piston pump works in two stages.

Stage 1: (Blue valves)
The piston moves to the right.
It suction in saltwater through the pre-filters and one-way valve 1 to the left side of the piston.
At the same time the piston presses out used water through the governed valve 4 back to the sea.
The piston-pump will only operate against the pressure drop over the pre-filters and therefore take little power. There is not any need for a booster-pump since the piston-pump will have enough power to take in the seawater.
The current that the system takes is a god indicator of the clogged stage of the pre-filters.
The one-way valve 2 and the governed valve 3 are locked in this stage.

Stage 2: (Red valves)
The piston moves to the left.
The water on the left side of the piston is pressed through the one-way valve 2 in to the membrane.
At the same time water in pressed out from the membrane trough the governed valve 3 to the right side of the piston.
The right side of the piston has a smaller volume (it has an piston rode) then the left side. This results in that the pressure in the membrane will be high and let out fresh water.
It’s during this stage that the piston-pump will take its maximum power.
By measuring the current, the computer will have a god control over the overall stage of the system.
The governed valve 4 and one-way valve 1 will be closed in this stage.

Since the piston pump wont make an enough flow over the membrane for it to be happy, I have a circulation-pump that increases that flow. Even though it is in the high-pressure part of the system it is not a high-pressure pump. It will only feel the pressure-drop over the membrane. Now that the membrane is happy I can have a low input flow of seawater to the piston-pump and thereby an easy job for the pre-filters.

The easiest solution for the governed valves 3 and 4 are to have electro magnetic valves. This valves will feel a pressure of 5.5 MPa and therefore easy to find.

Using a crankshaft will probably be the easiest way to move the piston. But with a crankshaft the side force on the piston rode will be quite high (see spreadsheet) and therefore need some form of governing.
The piston in the piston-pump will feel the same pressure as the circulation-pump, the pressure drop over the membrane. But the piston rode sealing in the end of the cylinder will get 5.5 MPa. The cylinder itself will also feel 5.5 MPa.

Put a electrical maundered two-way valve between the pre-filter and the one-way valve 1 that either, in normal operation, takes seawater, or during flush, takes freshwater. This valve will not sense any high-pressure.

After that the system have made the wanted amount of freshwater it will do a
Automatic Freshwater Flush (AFF) of the system:
The piston-pump is temporary stopped.
The circulation-pump is continued operating.
The two-way valve in the inlet is placed in freshwater position.
The magnetic valves 3 and 4 are activated (opened) and depressurise the system.
The piston-pump makes 20 strokes (47 s) and takes thereby in 3.5 l fresh water in the system.
Wait 133 s while the circulation-pump flush the system.
The piston-pump makes 6 strokes (14 s) and takes in 1.0 l fresh water.
Wait 46 s while the circulation-pump flush the system.
The piston-pump makes 6 strokes (14 s) and takes in 1.0 l fresh water.
Wait 46 s while the circulation-pump flush the system.
The circulation-pump is stopped.
The magnetic valves 3 and 4 are deactivated (closed).
The two-way valve in the inlet is placed in seawater position.
Now is the membrane flushed, the whole system is filled with fresh water, all pumps and valves are deactivated and do not take any power. It’s only the control computer, that’s is in standby mode, that takes power.
In total the flush have taken 5 minutes, used 5.5 l freshwater and 2.7 Wh.

The motor that drives the piston-pump could be a 220 W electrical motor with a spoon gearbox and an output of 200 rpm. On the shaft sits a 32 mm spoon belt wheal. This is connected via a 19 mm spoon belt to a 250 mm spoon belt wheal that’s sits on the crankshaft. This belt gear reduces the speed to 25.6 rpm. The spoon belt will feel a maximum force of 860 N.

There could be a need for some type of pressure equalizing system in the high-pressure part of the system. During stage 2 the pressure could maybe so high that the security valve release, and then would an equalizing system prevent that.
It’s preferred that the high-pressure part of the system have a low volume, it will make a flush more efficient.
Will a peace of rubber in a tank do the job as pressure equalizer?

It is possible to build a system that in emergency situation can be operated by hand. In this situation there is not any need to operate the circulation-pump. If the governed valve 3 and 4 are operated mechanically when the piston comes to the end points, it will be easier to build. Five strokes will make .1 litres freshwater.

It is possible to built a system with a cylinder diameter of 20 mm and a piston rode of 4 mm and this way get much smaller forces on the piston pump. Unfortunately will the circulation pump then still take the same amount of power and the total performance of the system will be much lower.

[Captain Bill]

With the exception of the circulation pump and the fact that you intend to use electric valves instead of spring loaded valves this design is essentially the same as my Katadyn 160 and it is not nearly as efficient as you predict. One thing to consider is that when you use a single piston pump the pump is not applying pressure to the membrane on the intake stroke so the membrane produces little to no water during that part of the cycle. My Katadyn produces a pulsed output and not a steady flow. The membrane it uses is spec'd at 300 gallons per day yet the unit produces only half that amount. This is because the piston is only applying pressure 50% of the time. I really don't see how the circulation pump is going to improve it as it only increases the speed of the water over the membrane, it doesn't increase the pressure or the time the pressure is applied to the membrane. The real test is to build one and see what it does. My katadyn use about 10 Wh/liter, so see if yours does any better.

[Lars_L]

Captain Bill

Thats correct that the piston pump just increases the pressure during stage 2, but the volume of water that I take in to the membrane is higher than the volume I take out. The difference is the produced amount of fresh water, and that is 30 l/h. If the water comes out from the membrane only during the stage 2 or over a longer time, that I cant tell.
With a lager diameter on the piston I would get enough flow over the membrane for that to be happy and don’t need any circulation-pump. But it puts much higher demand on the piston-pump and the flow though the pre-filters will double. With a circulation-pump I get an easy way to be sure that I have enough flow over the membrane. But as you say, it will not give any pressure.
With a smaller diameter on the piston I would get a lower demand on the piston-pump, a lower demand on the pre-filters, but a little higher demand on the circulation-pump. Totally I will get a little better performance, but how small piston can I have? Since I get a higher concentration of salt over the membrane I will probably need higher pressure, and thereby lose in performance.
As I have said, this I only a theoretical test of the idée. If this works, next step is to build one.
As a compare, a standard watermaker needs 35 Wh/liter.