Watermaker production amount

[Norbert1]

Hello,

I hope Tellie will answer but I am happy about any input.

At first I have to say, I got lots of help from spectra allready. They would answer every single mail.


At the moment the options for repair are just to wide spread.
I could spend lots of money for a whole kit, then for other pressure tubes.
I would like to point it down to a more single problem.


Here is the story


I found a second hand clark pump 10%

I tried to rebuild as good as I could 1 year ago.
Change annular rings and many O rings.

I have it connected to two feedpumps, shurflow and 2540 membran.

It should be the same like cape horn with 2 pumps.

The problem is, the product flow only 35 liter/hour
should be 50+

So I did a flow test with the following results.

I used a 18.8 liter bucket as the base.

I did already convert min to decimal.



Results

A. 2 Feed pumps:

5,92 liter/min brine

0,58 liter/min product


9.8%

B. Older feedpump

4,00 liter/min

0,416 liter/min

10,4 %

C new feedpump

3,6 liter/min

0,4 liter/min

11%



So for me it looks like the clark pump works with 10 %

The other problem is, I have connected two high pressure hydraulik tubes to the membrane. They are 50cm and 150cm long.

But the inside diameter of the fittings are only 4mm. But spectra use much bigger pipes.

I still waiting for a calculation from my friend, he is ingenieur to tell me how that would reduce the flow.

So I am not shure how to begin with.

The Sytem works on the other hand well.
The sound of the pump is normal, spectra say.
No leaks.

The ppm is not perfekt. 550 now. But I am 100% shure thats part of the same problem.

Thanks for help

[Tellie]

My first recommendation is to add a 0-200psi pressure gauge and a 0-20gph flow meter to the system. These two items are cheap enough and will tell you a lot more of what is actually going on with the Clark pump and the feed pumps. Also 4mm ID would certainly restrict the flow, 12mm would be closer to what is used. If you need a diagram for the flow meter and gauge let me know.

[GordMay]

Greetings and welcome aboard the CF, Norbert.

[Norbert1]

Hello Tellie,

thanks a lot for reply. The gauge and flowmeter I will try to get.
I am just thinking how I can get bigger high pressure hoses.
Spectra is using these pipes, but of course they are not flexible. And on standard membran housing there are these small, guess 1/4 connectors.

[Norbert1]

Today there was a leak on the clark pump.
So I have taken the chance and disambeld the clark pump..
I realized that the annular rings looking good.

But what I found is, the cylinders looking not good at all.
Of course I have to hone and replace the piston.


But my question would be, if the clark pump make 10 % and sounds ok so far.
Could the reason for less production and less over all flow be the bad cylinder?

[Tellie]

Originally Posted by Norbert1

Today there was a leak on the clark pump.
So I have taken the chance and disambeld the clark pump..
I realized that the annular rings looking good.

But what I found is, the cylinders looking not good at all.
Of course I have to hone and replace the piston.


But my question would be, if the clark pump make 10 % and sounds ok so far.
Could the reason for less production and less over all flow be the bad cylinder?

Yes of course it would. Usually if your cylinders and pistons are scored, so are a few other parts internal to the pump. I would suggest to strip the pump down completely to inspect all parts.

[Norbert1]

Thanks Tellie for reply.

I can't get the flowmeter and and the gauge now I will do that later.

But I managed to hone the cylinder absolutly perfect.
But there is no effect at all. Big disappointment.

I have questioned thismany times but never got an answer.

On the clarkpump there is high pressure outlet and inlet.

My understanding is that from outlet the water would flow towards the membrane
Than from the membrane to the inlet.

If I look to my pressure vessel, the inlet is on the side where a single connection is.
The outlet is together with the production outlet.
There is a flow symbol, so no mistake possible.

If I look to spectra watermakers it looks not like that.

The side where is only a single connection on the pressure vessel is connected to the inlet from the clark pump.

And opposite , where the production water came out and the second connection, is connected to the outlet of the clark pump.

So, either pressure vessels are not all the same, or somthing is wrong.

[Tellie]

Originally Posted by Norbert1

Thanks Tellie for reply.

I can't get the flowmeter and and the gauge now I will do that later. This is going to be very important especially with building from scratch. These will save you countless hours and frustration trying to guess without them. Even harder for me without being there.

But I managed to hone the cylinder absolutly perfect.
But there is no effect at all. Big disappointment. Did you get two new pistons as well? What did the piston rod and lip seals look like,What did the annual rings and spool valve look like? Also what kind of feed pump are you using?

I have questioned thismany times but never got an answer.

On the clarkpump there is high pressure outlet and inlet.

My understanding is that from outlet the water would flow towards the membrane
Than from the membrane to the inlet.

If I look to my pressure vessel, the inlet is on the side where a single connection is.
The outlet is together with the production outlet.
There is a flow symbol, so no mistake possible.

If I look to spectra watermakers it looks not like that.

The side where is only a single connection on the pressure vessel is connected to the inlet from the clark pump.

And opposite , where the production water came out and the second connection, is connected to the outlet of the clark pump.

So, either pressure vessels are not all the same, or somthing is wrong.

All pressure vessels are the same bi directional regardless of the sticker on the vessel tube. The membrane vessels are non directional. It all depends on how you plumb it. But in your case it is quite simple. On the Clark pump itself at the two HP 90 degree fittings. The bottom fitting closet to the base is the "Output" and the fitting above it is the Return". Follow the "Output" line to the end of the membrane vessel. It doesn't matter which side it goes to on the membrane vessel. All you have to remember is that where ever the bottom 90 degree fitting is attached to the membrane vessel that the membranes single brine seal is also at that same end. This will give you the proper flow. On Spectra's membrane vessels you have two 1/4" product out puts on each membrane vessels end cap. Regardless of the orientation of the brine seal you can tap into any one of the four you'd like. I assume you have the black Spectra vessel. You will soon learn that these are the best membrane vessels in the world because they are the easiest to work with.

[billknny]

10% of the total flow as output is exactly how a Clark pump works. That is FIXED in the design it can never make more. Anyone who doesn’t understand this is giving bad advice.

You really need to learn how a Clark pump works before you tear it apart again!

If your pressures are right, and your membranes match the pump, you just need more capable feed pumps.

[Tellie]

Originally Posted by billknny

10% of the total flow as output is exactly how a Clark pump works. That is FIXED in the design it can never make more. Anyone who doesn’t understand this is giving bad advice.

You really need to learn how a Clark pump works before you tear it apart again!

If your pressures are right, and your membranes match the pump, you just need more capable feed pumps.

With all due respect, this may be misleading. (There are four different sizes of displacements on Clark pumps, 7%, 10%, 15%, and 20%) 10% pumps are used on four different range models of Spectras that produce 8,14,15, and 16 gallons per hour. The flow and pressures from the feed pumps make the difference in the Clark pumps total production with an identical sized membrane. A 10% pump with one good feed pump (like Norbert1's) and a 21" membrane and he will produce 8gph. Just change the membrane to a 40" membrane and production will drop to 7gph, same feed pump and Clark pump just a different membrane size. The percentage of production is measured by the input flow and pressures. (again the real need for the flow meter and pressure gauge) A 10% can make anywhere from 6gph to 20gph depending on flow and the right pressures which of course is dependent on the feed pumps capabilities. But in Norbert1's case he has already started out with a severely damaged pump with heavily scored cylinders. Just honing and polishing the cylinders and replacing the pistons is most likely not enough. We need to know symmetry, (the reason for the flow meter and pressure gauge) and the other issues scoring can do to other parts of a Clark pump. It is a guessing game at this point (without the pressure gauge) to address or condemn the feed pumps until we know exactly the condition of the Clark pump. Also, the pressure gauge is essential in determining the condition of the these two type of feed pumps as well. I would invite Norbert1 to tear this pump completely down and take pictures part by part along the way and post them here for all to see. One little trick that helps reveal a problem, I should have suggested this to Norbert1 as well, is watching the product flow from the 1/4" product tube (even at his 9gph). The Clark pump only makes pressure on one side at a time. Watching the flow, an asymmetrical Clark pump will flow stronger on one stroke and weaker on the other. Also, the time between strokes should be even, so if it shifts on one stroke in five seconds the second stroke should shift in five seconds as well. A disparity of these numbers also indicates a problem with the Clark pump. The great thing about Spectras is that there are really only three main components, the feed pump/s, the Clark pump and the membrane. If we can determine any two of these components are working properly it is almost always the third that is the problem. Making sure the Clark pump is operating properly is the first thing to check for a faster diagnoses. In fairness it can certainly be a combination of all three things not working properly in this particular case. This is after all an older watermaker where the sins of the previous owner are not known.

[Norbert1]

Hi Tellie,

Thanks a lot again for answer.
At least the direction question is no issue anymore.

I will have a closer look to the product flow as you have suggested.
At the moment we are in Tobago, but we will go to Greneda soon.
I have to think about order Items to greneda.
First the gauge and the flowmeter.
Hope to find a way to order.
Second, maybe parts for clark pump.

Also I can send pictures to you.
But I need a email adress.

My feedpumps are more or less the same shurflo like spectra use. Only they are for chemicals ,agriculture. The flow is 1.8gpm and 100psi
Shurflo
8000-543-238
But after ask pumpvendor, the psi would go up to 150 psi, if the switch is bypassed, wich i did.

I did not change the pistons. Once I worked as a lathe operator, so I know about surfaces etc. and for me it looks as if the pistons after a short threatment with 1000 grid paper are just perfect.
The annular rings are perfect, under magnifying glass. ( I know how they look if bad) Also the spool valve.
I also checked and overhauled the checkvalves.
One had a little leak, but thats solved.
The piston rod looks good.

The only thing I have not inspected is the pilot spool because thats difficult.

I managed to open the testport- no water at all

But as you say, the gauge and flowmeter are important.

There are some issus arround but they are not importend yet for the problem.

1-I need new J tubes, maybe the fittings as well.

2- I like to connect the membrane with the flexible tubes spectra use. Because they a bigger inside.
Of course with the fittings.

3- some O rings, ( maby the oversea kit)

[Tellie]

My E-mail is HaldenMarineService@yahoo.com