12 replies to this topic

[CS Kang]

Why is it essential to half open V5(discharge valve) and fully open V1 before starting the pump? Is it because of safety reasons?

Attached Files

•  practical4fluidmechanics.JPG   237.27KB   85 downloads

[Padmakar Katre]

Dear
Centrifugal Pumps never started (should not start) with the partail opening of the suction and full open discharge valves.In the start-up sequence the suction valve is fully open (oherwise you will not be able to maintain the NPSH required),discharge valve fully closed (Its crack open in case of multistage centrifugal pumps) while starting and then start opening discharge valve by monitoring the discharge pressure, current drawn by motor and discharge flow, vibrations motor winding temperature and so on.
As you stated that before start-up of the pump this is the situation like partail open suction and fully open discharge valves but I am doubtfull because as such there is no specific reason to do so may some body senior will elobrate some more on this.Waiting for the right comment by the other Jedies.

[Art Montemayor]

CS Kang:

Please don’t take my first comment as being negative or critical, but you are asking a question with an assertion inserted into it as if it were a fact. As Padmakar states, the opposite is actually true.

Many times a question is posted on our Forums based on an erroneous assertion. The point here is that we can help you via our experiences if you reveal to us where and how you obtained this erroneous idea. That way, we can further help you by disqualifying or clarifying the source of your information.

I have started pumps with the discharge line both opened or closed – all depending on the process, the type of pump, starter, and instrumentation allowed. But I have never – and would never - started a centrifugal pump with the suction line closed or even throttled. This is a totally wrong and bad practice to get into. You can damage a pump and/or its seals this way. It is very important to have a centrifugal pump fully primed prior to attempting a start up. When you close or throttle the suction line, you are starving the pump and depriving it of a need to be 100% full of liquid prior to starting the pumping action. As I said, this is a bad practice; it is also very impractical and makes no sense.

The block valve installed on the suction line of a centrifugal pump is never meant for throttling purposes. The sole reason for using a block valve in the suction line is for isolation purposes when it is necessary to do maintenance or to remove the pump. That’s all it is there for. The discharge line also has a block valve (normally a gate or ball valve type) for the same reason as the suction line; additionally, a globe type valve is used on the discharge line to control the developed head (or throttle the discharge flow).

Before you attempt to start any centrifugal pump it is critical that you ensure that the pump casing and the connected suction pipe be 100% primed (or "flooded" - totally filled with liquid fluid with no air or gas included or in pockets). If you want to control the pump’s flow rate, you should throttle the discharge line with the globe valve mentioned above. That is the operational technique that a centrifugal pump was designed for.

Please tell us how you got the idea that it is essential to half open the suction valve. I’d like to know how these types of half-baked ideas get circulated.

[CS Kang]

Dear all, it's a question in one of my practical, flow through venturi meter and orifice meter. The procedures stated that check that valve V1 is fully open. (V1 is what i meant by the discharge valve leading to the venturi and orifice.) Half open rotameter valve V5. Start the pump. I consider V5 as the suction valve in this case. After that, we are suppose to take a range of rotameter settings and for each reading, record the difference in manometer heights associated with pressure drops across the Venturi and orifice meters.

[Adriaan]

Dear respected Art and Padmakar, hello CS,

as far as I can see CS Kang nowhere states what type of pump (s)he is talking about, nor the medium the pump is used for (liquid or gas).

I fully agree with Art that one should never start a pump with a closed or throtteled suction valve, irrespective of the type of pump. To me that's a good way to ruin a liquids pump by producing cavitation!

For pumping liquids I assume that opening the discharge valve and partly opening the suction valve could be a method used to make sure the pump and system are filled (especially when the system upline is pressurized) before it is started? Would also be a way to check for leaks etcetera and possibly to pre-warm the pump. The actual starting should not be done with the valves in those respective positions however.

So CS, next time please provide more (significant) information so you can receive an answer you can use!

[CS Kang]

Thanks Adriaan and all, I will take note. it's a centrifugal pump used to pump liquid. in this case, the liquid is water.

[Padmakar Katre]

Dear CS,
Could you make a drawing of your practical problem in excel workbook and upload here for us so that we can get the clear idea about your system.And firther we can comment.

[CS Kang]

Dear CS,
Could you make a drawing of your practical problem in excel workbook and upload here for us so that we can get the clear idea about your system.And firther we can comment.

Okie, i have attached the drawing under my modified post. Unforturnately, I cant upload the video because of the large file size.

[Adriaan]

The valve V6 in your drawing is the suction valve, V7 is a drain, V5 is the discharge valve.

As valves V1 - V4 are check valves and you state that you have to do measurements with the venturi and orifice plate you'd fully open V1 and close V2, V3 and V4. Check valves shouldn't really be used to control flow volume, they are either open or closed. If somebody tells you to half open a check valve tell them to replace the damn valve

Close V7, fully open V6 and open V5 so that the system is filled with water. Close V5 (could keep it open a tad) then start the pump.

If you leave V5 open too far you might damage the equipment when the pump is started (lab equipment usually isn't built to last including the seal on the discharge side of pumps) so just leaving V5 open a little bit when the pump is started might be a good idea .... but .... well under normal conditions I'd start the pump with V5 fully closed.


Oh and its a bit odd that there seems to be no drain valve other than V7 in that system, would have expected at least one more (under the vertical pipe after that globe valve).

[Padmakar Katre]

Dear CS,
Here everything is right i.e. your procedure to measure flow through the orifice and venturi meter.V5 is the pump discharge valve whose opening you have to adjust as per the flow you required to maintain through the rotameter.Keep open V1/V2/V3/V4 valves simulteneously i.e. one at a time and other three closed and find the delta P in the respective flow measuring device with the manometer and subsequent flow as per the delta P obtained.
The main confusion was that you were termed the V5 as suction valve which is pump discharge valve and your V6 is the pump suction valve which will be fully open before you will start the pump. And V7 is to drain the entire system.

Dear Adriaan,
All these V1 to V4 should not be check valves because if you see the drawing the U-loop is provided which has elevation equal to the HLL of water reservior.And V5 purpose is to vary flow through the rotameter and as per the different flow we will measure the delta P in orifice and venturimeter and subsequently flow through them.As well if we keep V5 open fully nothing is going to get damage only thing to be considered is the rotameter range should be adequate to measure the flow in this case(V5 fully open)

I hope it will help you.

[CS Kang]

Thanks adriaan and Padmakar for pointing out my mistake. I feel sorry that I make this stupid mistake which cause quite a trouble for everyone. The purpose of the partially open V5 is to prevent cavitation?

[Art Montemayor]

C.S. Kang:

There some important points that we are all missing on this thread up to now and I take this opportunity to bring them out into the light:

1. The hydraulic circuit, or pumping system, that you show is nothing more than a simple, closed system. What goes around, stays around. In other words, the pump will never lose prime or stop pumping at any rate on its performance curve as long as the suction tank has a positive level, the suction block valve (and I stress BLOCK valve – not throttle valve) is kept open, and there are no losses or leaks in the system. What goes into the suction tank comes out – continuously. Therefore, it doesn’t make any difference as to whether the discharge valve (V5) is open or closed at start up. The discharged fluid becomes the suction fluid soon enough. I think that everyone reading this has still not noticed that I strictly emphasize the difference between BLOCK valves and THROTTLE valves. There will always be a practical need for block valves – maintenance and safety deem it so. There isn’t necessarily a requirement for throttle valves on the discharge of a centrifugal pump UNLESS you need or want to control the flow rate (and/or the head) of the pump. In your system, you definitely need a GLOBE type throttle valve. A gate or a ball type simply will not do for good flow control. Your schematic drawing shows what normally is depicted as a gate valve – not a globe valve. This is misleading but since you probably got this through your university lab I can understand the mix up since academics don’t place much importance on the valve differences that are mandatory and very important in real life applications. You need block valves for the same reason you need drain valves – to do inspection, maintenance, and repairs.
   
2. There is an important reason for reading and studying about cavitation and not giving into temptation and using the word loosely as if one knows exactly what one is talking about. Research and learn exactly what “pump cavitation” means. Do not resort to using the term loosely because you project and transfer wrong or misleading information to other engineers who do know exactly just what cavitation is and what it does. Briefly stated, cavitation is the vaporization of a normally liquid fluid into bubbles combined with the subsequent, rapid collapse or “implosion” of the same bubbles downstream. This sometimes causes violent and traumatic noise and shocks in the system – resulting in equipment movement and damage.
   
3. When a pump is starved of liquid feed it does not necessarily experience a fluid cavitation! This is very important to note by all reading this. Engineers – especially younger ones – live in fear of cavitation and its effects (& rightly so); as a result, some resort to using the term loosely and conveying the wrong message or impression. I have run countless centrifugal pumps – at one time or another – starved of fluid due to operational mistakes or instrument failure and NO cavitation took place. This is simply because the vapor pressure of the fluid in question was not reached – in other words, the fluid’s vapor pressure at the pumping temperature was so low that it remained a liquid. This is another way of stating that a super-cooled liquid is difficult to cavitate. Now, what does happen when a pump is starved of super-cooled liquid is that it loses prime and, as a result, stops pumping. This, of itself, does not cause cavitation – primarily because liquid centrifugal pumps don’t have the capability of pulling a partial vacuum. If they did, they would cause cavitation frequently under starved conditions. Believe me, when you experience cavitation, you will quickly become aware of it!

I am not trying to “preach” in this thread, but rather want to make sure that we all are on the same track or subject when we mention such basic things as pumps losing prime or cavitating. It makes a BIG difference. This is also why I asked you to explain where you got the idea of opening the suction block valve half-way - which you haven't answered yet.

[CS Kang]

Thanks Art for your clear and concise reply. The half opening suction valve was all because I mistook V5 which is a discharge valve as the suction valve.