12 replies to this topic

[Art Montemayor]

Sara:

I don’t know what year of study you are in your ChE curriculum, but your questions are good ones and the type that I consider valuable and important for any student – the earlier in their school years, the better.

First, I want to make sure you know about these web sites:

1. http://www.mcnallyinstitute.com/

2. http://www.gouldspum...20040903_1.html

Download as much of the technical literature that McNally gives you; cut and paste it into Word for Windows and form a Pump Handbook of useful information. Read and study it religiously.

Try your best to get the Goulds CD-ROM offered by them which contains their GPM pump catalog. It is full of practical, useful, and illustrated pump information in their technical and engineering section.

I’m not at home at present, but if you show interest and continue this thread, I will research all the internet sites and information I have downloaded in the past and this may be of great help to you in understanding pumps and resolving their problems. Now on to your 5 questions – which, by the way, you didn’t number:

1. A piston-type pump is a basic positive displacement pump. Positive displacement pumps are radically different from a dynamic type of pump – like the centrifugal type. If you increase the stroke length in a piston pump you will increase the pump’s displacement directly and the immediate effect will be one of increased discharge flow rate of fluid.
2. If you increase the number of strokes per unit of time you will also increase the discharge flow rate of fluid. This is very easy to visualize when you study and learn how a piston pump “pumps”.
3. The priming of a pump means that the entire suction line and the pump’s volute (or pump chamber) are entirely filled with liquid fluid (or in other words, that there are no gases, vapors, or compressible fluids in the suction line or the pump proper. A pump moves liquids; it cannot move, transport, or “pump” compressible fluids because it isn't designed to do so. NEVER FORGET THIS IMPORTANT STATEMENT. If you don’t ensure that the system is 100% liquid-full, the pump will not pump or it will have a difficult time doing so. All pumps require priming.
4. In spite of advertising and popular belief, there is no such thing as a “self-priming pump”. Simply stated, I repeat what I stated above: All pumps require priming – whether initially, or every time they are started up. What is popularly called a “self-priming” pump still requires priming at the initial start-up. Once these pumps are initially primed, their over-sized casings employ an excess inventory of liquid stored inside to displace any air or compressible fluid trapped in the system after it shuts down. There are many ways to prime a pump: by gravity flow, by trapping the liquid in the system through the use of a check valve (“foot” valve) in the suction line, by using a jet ejector to pull a vaccum and induce liquid into the pump’s suction system and casing, or simply by using another pump.
5. If you startup a pump (presumably a centrifugal type) with its discharge piping blocked by a valve, you will force the pump (& motor driver) to initially do more work. You will have a rush of electrical current into the motor and the pump will immediately go to the “top” of its performance curve and reach what is called a “dead-head” condition – where the discharge pressure is at its maximum value and the flow through the pump is zero. Under these conditions the impeller continues to turn and converts the work input into friction and mixing, raising the temperature of the fluid within the pump. If you persist under this condition, you will overheat the fluid and, more importantly, the pump seals. This is hard and nasty wear on the seals and will eventually cause them to leak or fail outright. Starting a pump under these conditions is not necessarily bad if the duration of the dead-head is not long (several minutes). I like to start my centrifugals with a minimum of discharge pressure and usually have a recycle stream back to the fluid’s source in order to clear out all air or gases and then slowly bring up the pump’s discharge flow rate by closing down on the recycle stream. However, in some processes this is not possible and the pump must be started up immediately and against a set discharge pressure.

I think that answers your questions and I hope you follow up on what I recommended. You will be rewarded in the future if you continue on with your engineering career.

[Art Montemayor]

Sara:

You will note that my prior comments were dedicated primarily to a centrifugal type of pump, since that was what I suspected you were referring to using. In the future, always identify the type of pump you are alluding to. There are many different classes of pumps in the industries and they all behave in different manners and are used for different reasons and in different applications. The original class of pump ever devised or employed by humans was probably the positive displacement class – of which the piston type is probably the first example. There are different types within the positive displacement class as there are different types within the dynamic class – including the centrifugal types.

Since you are still a student, you should concentrate on the basics before you get too deep or immersed into the subject of pumps. The first thing to conquer is to understand the different types and the manner in which they move liquids. After you understand the different types, then you can begin to differentiate within the types by studying the different models of each type. There are specific reasons for the different types and the different models – and these reasons play a very important part in selecting and applying them in the field. The basic importance for all students is to definitely understand the difference between a positive displacement pump and a centrifugal pump – and their respective characteristics. You cannot allow yourself to discuss pumps without knowing this basic knowledge because you will absolutely get mixed up and design, select, and apply them wrongly. For lack of any further information, I have to assume that you know these facts (even though I may be wrong) and go forward on this important subject. It’s up to you to tell me if (and what) you don’t know yet.

1) A piston pump is a type within the positive displacement class of pumps. The main characteristic of this class is that the pumps move a definite amount of liquid volume with each stroke (or revolution) by positively displacing the liquid – hence the name. Note that the liquid is positively displaced. What is meant by this terminology is that the liquid is discharged from the pump without regards to the pressure existing at the discharge port. This class of pump typically continues to pump liquid with the discharge closed off until a discharge relief valve (PRV) is activated, there is internal leakage, the driver trips out (i.e., an electric motor would have its circuit breaker trip), the coupling of the driver breaks off, or the pump (or its discharge piping) bursts apart from excessive pressure. Usually (and hopefully) there is a pressure relief valve always installed adjacent to the discharge port. This PRV is piped to discharge to the suction side of the pump (bad feature), to the source tank of the liquid, or to the atmosphere (worse feature). The prime rule in applying positive displacement pumps is that they should always have a PRV installed in their discharge port or exit.

2) You select a centrifugal pump from information found in its respective characteristic (or “performance”) curve(s). You do this only after you have defined or calculated the following:

- the Net Positive Suction Head Available (NPSHa) within your actual physical pump system;
- The Total Dynamic Head (TDH) – a.k.a. “Total System Head” or “Total Developed Head” - required by your system;
- The speed (RPM) of your motor or other driver;
- The System Curve. A centrifugal pump always operates at the intersection of its characteristic curve and the system curve which shows the head required to make the liquid flow through the system of piping, valves, fittings, etc.
- The sensitivity that you can tolerate in the capacity-TDH relationship; this has to do with the shape of the pump’s characteristic curve – some curves are flatter than others and their TDH is very sensitive to the pumped flowrate.

3) Actual power is not defined by you. Do you mean the power delivered to the electric motor drive? (Note that you haven’t defined how you intend to drive the pump – this is another flaw in your questions) Or do you mean the power delivered to the coupling at the pump’s shaft? Or do you mean the power required by the pump itself? As you can see, it makes a big difference as to what you intend to say and how you say it. As an engineer you are expected to be specific and accurate – unlike a poet who is accepted as being general, non-specific, and relaying only feelings and emotions. You must relay FACTS.

The work performed in pumping or moving a liquid depends on the weight of the liquid being handled in a given time against the total head (in feet of liquid) or differential pressure (in psi) being developed.

Since one horsepower equals 33,000 ft-lb per minute, the useful or theoretical horsepower (usually called the hydraulic horsepower – hyd hp) will equal:

Hyd hp = (lb of liquid per minute) (total head, feet) / 33,000

The actual - or brake horsepower (bhp) of a pump will be greater than the hyd hp by the amount of losses incurred within the pump through friction, leakage, etc. The pump efficiency will, therefore, be equal to:

Pump efficieny = hyd hp / bhp

or

Pump Brake hp = hyd hp / pump efficieny

Since the above expressions apply to both centrifugal and reciprocating type of pumps, horsepower calculations can be simplified if the weight of liquid being handled (capacity) is expressed in terms of gpm and the differential pressure (TDH) in terms of head in feet of liquid for centrifugal pumps, and psi for reciprocating pumps as follows:

For centrifugal pumps -
Brake hp = (gpm) (TDH, in feet of liq.) (specific gravity) / (3960) (efficiency)

For reciprocating pumps –
Brake hp = (gpm) (psi) / (1714) (efficiency)


I hope the above clears up your dilemma. You haven’t furnished your email address, so I’m unable to send you the additional information I previously offered.

If you continue to have interest or need in centrifugal pumps, I strongly recommend you obtain a copy of "Cameron Hydraulic Data", a book that is considered a basic need in studying pumps and their applications. You can buy this book through the Internet - probably at Amazon.com or other technical book sites.

[gvdlans]

You can buy the Cameron Hydraulic Data book directly at https://www84.ssldom...t2/merchant.mvc?
I bought my copy (19th edition) a few months ago. I explicitly asked them to send the book via normal mail. The book reached me (in The Netherlands, Europe) within 2 or 3 days for a much lower shipping & handling cost than what they originally stated...

[Art Montemayor]

AFD:

You are correct; there are many pumps which have an integral PSV incorporated in the pump which relieve any over-pressure discharge liquid back into the suction chamber of the pump. In effect, what is happening is that the liquid is being re-circulated – and, as a result of a dead-head, is also being re-heated. This over heating of the liquid and the pump is a potential hazard for many pump seals and fluids. Additionally, you will discover that manufacturers of positive displacement pumps such as gear pumps and sliding vane pumps do not have any calculations or documentation for their integrated PRVs. I can assure you from personal experience that major processing companies such as DuPont, Huntsman, etc., will not tolerate this type of situation without some solution. The solution is often to recirculate with an external, engineered PRV back to the fluid’s original source. I have always removed or obstructed the integral PRV in such pumps and installed an external, calculated and documented PRV discharging back to the original suction tank. I have found this to be a more controllable situation. I can spot PRV leakage or failure much easier.

I haven’t pumped xylene with a PD pump and I don’t know the reason why a manufacturer would not accept the integral PRV except that I highly suspect that they have taken the same stance as I have described in the previous paragraph.

I don’t know what make or model of Piston pump you have on your peroxide and I also can’t guess what internal design your pump has. However, I am very familiar with piston pumps – especially since I’ve used them on cryogenic services, pumping liquid Oxygen and Liquid Nitrogen from 5 psig suction pressure to 3,500 psig discharge. That’s quite a load to put on a piston and two check valves – which defines all the moving parts inside the pump. You can imagine the sensitivity of the NPSH. However, I frequently varied speed and stroke at will and never had problems. Theoretically there is no difference in varying the capacity of such a pump by either method. But you have to depend on the 100% sealing action of your valves – both suction and discharge. If they are not operating 100%, you will get leakage and a loss of capacity. This capacity loss is more noticeable at the lower speeds than at the higher ones. I would suspect that your valves are the culprits that are causing you discrepancies in operation. When is the last time you checked them out for sealing action? The only way a plunger (or rod) can cause trouble for you is by leaking fluid out through the rear packing gland or by generating too much heat due to an operator over tightening the packing gland. Both these effects are very observable and can be quickly spotted or felt.

[Art Montemayor]

AFD:

A lot of people fear or hate the piston type of positive displacement pumps. Most of the bias comes from the belief that they are “old-fashioned” and “clunkers” that characteristically have pulsating flow and pressure. I have learned through the years to appreciate their simplicity and ruggedness. (Here, I’m assuming you have a simple plunger – “ram” – type of piston, without any piston rings or double-action) If you study your pump carefully and you have no external leaks or suction restrictions, you will note that there are only two basic things that can affect your flow capacity: your valves are leaking or discharge PRV is leaking back into your suction. If your PRV is sealing properly, then it has to be your valves that are giving you problems.

When you are pumping a super-cooled liquid with a positive suction head like the one you have, you should not see any capacity difference between the varying of stroke and the varying of speed. They both have the same positive effect --- if your valves are working properly.

I kind of doubt your pumps are pumping with a discharge of 31,900 psig, but I’ll let you confirm that. I am accustomed to using XXH pipe schedule, but I’ve never heard of anyone working at that pressure level on an industrial level. I am familiar with UHDE. They make good process equipment and have a good reputation.