axial piston soft start pump 2

[oceanman]

We are using a small axial piston pump to pump several liter per minute quantities of water against a pressure head of about 600 psi or so. If the system has to start up against the full 600 psi, it seems to just go into a dry running mode of operation. If we gradually increase the back pressure, we do not seem to have this problem. However, under our use conditions (deep under under the ocean), I cannot really provide for a gradual ramp up in pressure. The problem may stem from cavitating, which I sort of doubt due to the fairly high inlet pressure of hte pump, or it might inadvertantly be degassing the water, causing the pump to run unproductively. There could be another reason that we had not thought of yet. The water inlet pressure to the pump can be controlled fairly well, from just a few psi up to a hundred psi or so. In any case, the pump will not productively pump against a 600 psi pressure head on start up. It does appear to pump at least a little bit of water, but just a bit. I have three questions related to this:

1. Are any centrifical or positive displacement pumps more tolerant of starting against an initial 600 psi head as compared to an axial piston pump? Can any of these pump air laden water against such an initial pressure head?

2. Is there a way to get an axial piston pump to pump small amounts of air mixed with water on start up, or is this a lost cause? The ones we use just utilize water as the sealing fluid.

3. If air is the problem, we can probably separate it out, but it would then the air would need to be pumped out into the ocean a 600 psi head. Is there any sort of micro sized air compressor that would allow us to pump just a tiny amount of air out in this manner?

Any sort of comments or suggestions would be welcomed.

Oceanman
Thanks.

 

[Montemayor]

Ocean:

This is not only a unique and interesting post, it seems to be a very important one and one that will draw a lot of attention from experts. However, we need more basic data!

1) The pump in question, just to make sure we understand, is “wobble-plate” type – as opposed to a radial piston type – and is a positive displacement type. What is the make & model of the pump? What is the driver?
2) The pumped fluid is seawater. Or is it sweet, potable water? Here, I think you lose a lot of readers because you’re pumping “water” and you’re down at the sea bottom. But which “water” – drinking water or seawater?
3) The pump starts up pumping against a 600 psi pressure head on initial start up and remains doing this until stopped. The operation is stop-and-start as far as controls are concerned. What other controls are there available? By-pass? speed control?
4) You state the inlet pressure of the pump is high. What exactly is the inlet pressure? And what is the suction temperature? Yet you state that you can reduce the inlet pressure. How and why is this available?
5) You say the pump “just utilize water as the sealing fluid”; is this the shaft seal sealing fluid – like an API seal system? What is the atmosphere around the pump? Is it compressed air? Or is the pump submerged into the depths of the sea bottom?

More details on what you are doing at the sea bottom would also help explain how or why you are getting gasification in the suction or loss of NPSH on the pump.

We’ll await further input in the shape of basic data.

 

[oceanman]

Thank you for the post. I will try to respond to your questions as follows:

1) The pump in question, just to make sure we understand, is “wobble-plate” type – as opposed to a radial piston type – and is a positive displacement type. What is the make & model of the pump? What is the driver?

I believe that the pump is a wobble plate type of pump. It has a swash plate, which I think is the same thing. The pump is a Danfoss APP type. It is driven by an AC motor with variable speed drive. The controlled parameter for the VFD drive is the inlet pressure to the pump.



2) The pumped fluid is seawater. Or is it sweet, potable water? Here, I think you lose a lot of readers because you’re pumping “water” and you’re down at the sea bottom. But which “water” – drinking water or seawater?

It is potable water which is at 10 degrees Celsius. It does contain naturally occurring dissolved gasses. Some of these come out of solution when exposed to 20 degree C air temperatures.

3) The pump starts up pumping against a 600 psi pressure head on initial start up and remains doing this until stopped. The operation is stop-and-start as far as controls are concerned. What other controls are there available? By-pass? speed control?

We are controlling with a VFD drive. The VFD tries to keep the inlet pressure to the pump constant. We have tried settings of from 10 psi to about 80 psi inlet pressure. So far, we have had the best luck at about 20 psi inlet pressure control. When the pump fails to pump water, then the VFD sends a signal for maximum speed, which has no effect on pumping out more water. The pump mostly, but not totally, just spins at high RPM without moving water.


4) You state the inlet pressure of the pump is high. What exactly is the inlet pressure? And what is the suction temperature? Yet you state that you can reduce the inlet pressure. How and why is this available?

Again, the inlet pressure has been set at 10 psi, 20 psi, and 80 psi. We thought that 80 psi would keep any dissolved gasses in solution and would prevent the pump from cavitating. The inlet temperature is always about 10 degrees C. The inlet pressure, in an uncontrolled state could go up to about 200 psi or so. The source of potable water is limited in terms of flow, so the pump is supposed to keep the pressure in check. If the pump cannot keep it in check, which is cannot for more than a couple of minutes when it is mostly running dry, then a shut off valve isolates the potable water supply from the inlet of the pump. At that point the system is shut down.

5) You say the pump “just utilize water as the sealing fluid”; is this the shaft seal sealing fluid – like an API seal system? What is the atmosphere around the pump? Is it compressed air? Or is the pump submerged into the depths of the sea bottom?

The Danfoss pumps use water as the lubricant between the piston and the cylinder. I do not know what an API is, so I cannot comment on this. The atmosphere around the pump is normal air at atmospheric pressure. The pump is contained in a pressure vessel and it not exposed to the outside seawater.



More details on what you are doing at the sea bottom would also help explain how or why you are getting gasification in the suction or loss of NPSH on the pump.

I do not wish to reveal exactly what we are doing at the sea bottom, as we wish to preserve any future patent rights for our particular application. We are happy to provide additional details as may be necessary. I am not sure what NPSH means. Please excuse my lack of knowledge.



Thank you.


Oceanman

 

[Montemayor]

Ocean:

Thanks for the timely and detailed data. With a response and data like this, you will be rewarded with some serious and professional responses from our members - of that, I'm sure. This promises to be an interesting thread.

The first comments I have are:

1. I now know the exact type of pump and driver you have - as well as controls. This type of pump is stalwart and does yeoman service in thousands of hydraulic oil applications around the world. I have personally used it in many hydraulic oil applications - up to 2,500 psig. I have never seen them in potable water applications and I have some doubts as to the lubricity of the water in the application. My concern is with the heat generation of the piston parts.

2. I certainly don't like the part where you allow the pump to literally pump itself dry. This, plus the fact that the fluid pumped is pure water leaves me suspecting a lot of undesired heat buildup is taking place in the pump - with no lubricant or cooling taking place at times.

3. I would not operate this pump (or any other pump) without keeping the suction flooded in a positive manner by maintaining a level control on the source of the water. Then I would apply the VFD control and resort to possibly recycling the discharge back to source if the VFD still could not keep a positive suction level in the source. These pumps generate a lot of heat mainly because of the friction caused by the surface-to-surface contact – unlike a free-wheeling centrifugal type. That’s why they excel in hydraulic oil applications. Water has next to nil lubrication to offer this type of pump and unless you are constantly moving the water in and out of the pump, it could cause suction problems by raising the vapor pressure in the suction water.

NPSH is an acronym for Net Positive Suction Head. All pumps have a NPSH required that can’t be violated. The result is loss of prime or pumping gasification. Cavitation (which is the creation or existence of vapor in the suction with subsequent instantaneous vapor collapse by condensation) is also a culprit that can appear and disappear in your application. Keeping the pump’s suction totally flooded with a positive head of water above the suction port is one way to ensure that the pump sees and has the liquid it needs to pump, cool, and lubricate itself. You must move the water in and out of the pump – regardless of whether it is to pump to the target or simply to recirculate. The pump needs cooling and lubrication and moving the water through the pump is the only way you can supply these needs.

I’ll close for now and let some of the real experts get in their valued opinions.

Art Montemayor

 

[rzrbk]

A lot of if’s, but something to consider:

I cannot really think of a way that varying discharge pressure could affect the suction in a way to cause gassing up the pump unless there is in fact some gas coming out of solution before the pump is started. (or coming out of solution due to pressure drop via fluid acceleration on startup) If conditions are right, this gas could be easily pushed through the pump against a low backpressure, but not against a high backpressure.

If flow is established at a lower backpressure, then the water movement could provide the continual cooling necessary to stabilize pressure, temperature, and keep the gasses dissolved as discharge pressure is ramped up.

If pumping is begun against high pressure, this gas may not be able to be pushed out of the pump. The friction heat generated from lack of cooling would make the situation worse by forcing more gas out of solution and expanding the existing gas, and could result in a dry running pump. (although a different type of pump, a similar situation can be seen in self priming pumps if an air bleed is not provided on the pump side of the check valve)

The easiest way to provide a ramping discharge pressure at start up would probably be a recycle line from the discharge back to the suction vessel as mentioned above; although this would have to be designed carefully as it could provide a route for 600psi water back to the suction system.

As a side note, the control of the VFD may need to be kept in manual till actual operating experience is seen. This seems like a situation where the speed could be controlled in manual easily, but would require involved logic to run automatically. (eg. On this fixed volume pump, raising the speed when more flow is required would work, but raising the speed when flow suddenly drops off will probably make the situation worse.)

 

[ApC2Kp]

To oceanman,
The pump seems to be air bound. Priming the pump full of liquid is necessary for starting. The 600 psi backpressure would suggest adding a discharge vent valve for start up. If an actuated valve could be added for bleeding small flow off of the discharge (before the checkvalve), the starting flow through pump could eliminate air as well as give a soft start.
Once the system has been installed at the 30 m depth, and the pump stays full after shutdown, then it should stay primed, if I understand the operation. The discharge vent valve might not be needed again, if the pump keeps primed. Otherwise, the discharge vent valve needs to be interlocked for pump start up.

 

[oceanman]

Thanks for all of the comments, and please keep them coming. I did wish to clarify a couple of things. The pump is in a pressure vessel at a depth of about 1300 feet below the surface. This makes venting airout a bit tricky. The low inlet pressures are possible only because the pump is in a pressure vessel at atmospheric pressure. Automating the start up is key, and the comments given so far might work. In the past, we went out of our way to fully prine the system. It seems to quickly unprime quickly. There may well be some air outgassing from the water, or there might be some residual air in the plumbing prior to the pump. If we vent air out, we will have to push the air against the 600 psi pressure head, which is possible, but difficult. We thought about the issue of the pump accelerating quickly, inducing cavitation. This might be part of the problem.

Oceanman