Booster pumping in series questions 2

[jacky89]

I have two theoretical questions:

1) What would hapen if an existing water line is flowing (say 5,000 gpm), and I place a centrifugal pump in the line with a pump curve that does not dexceed 1,000 gpm?

2) Can someone enlighten me why when you place two pumps in series, the outlet TDH and pressure is doubled but the flow rate is not? I cannot see how the equation works since if you increase the pressure, shouldn't flow rate also increase since nothing else changes in the piping system?

Thanks.

 

[bimr]

What is incorrect is when a pumping system is designed to operate with 2 pumps in series, presumably, the pumping system will not work at all with a single pump. The flow from a single pump will not have enough pressure, but your diagram shows point 1 is where the system operates with one pump running.

For example, if you needed two 1,000 gpm pumps in series to pump to the top of the hill where the total pressure required is H_total, a single pump with 1/2 the pressure will not produce any flow.

 

[Artisi]

Lets all step back a bit and look at what is actually being pointed out. LittleInch 10Dec., this is a correct case if the flow thru a particular system need to be increased, adding a pump will increase flow with a corresponding increase in head.
Bimr 15 Dec. is also a correct case, if the TH is greater than 1 or 2 pumps can achieve in series, it will require additional pump/s.

Horse for course, depends what, where and why there is a need for series pumping.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

Artisi - thank you. The other time you may need to run two in series is when the system curve changes, e.g. the end point of the system has a floating pressure. Sometimes one pump will suffice but at other times you need two in series to maintain the same flow or indeed any flow at all if the end pressure rises higher than the first pump head. Many many possibilities.

Pipeline booster pumps when the pumps are a long way apart can operate in both cases, e.g. you have a large hill in the middle, one pump generates enough head to get arrive at the base of the hill (but not enough to get over the hill) with a low pressure, enough to feed a larger head pump to generate enough head to get over the hill.

A different profile would mean that one pump at the start would be able to pump some liquid from one end to the other, but to get a higher flow you need a second pump half way along the line.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[BigInch]

With an extreme case, my favourite example, pumping hot, heavy oil, two pumps in series might be needed, as a cold pipeline starts up with hot oil being injected at the inlet. As the hot oil makes its way down the pipeline, its lower viscosity begins to have a major effect in reducing system head losses even while flow through the system is actually increasing. That effect may continue to the point where the pump configuration might eventually need to be changed to parallel arrangement.

 

[LittleInch]

Of course if you use an insulated heated pipeline you don't have that effect (in joke)

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[BigInch]

Who would want one of those???

 

[dcasto]

We run two pumps in series, the second pump on a VFD for flow control, works great.

 

[Artisi]

And so it should

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[BigInch]

Actually flow control VFD shouldn't be used for pumps. That's for gas systems. How do you control flow in a liquid piping system. Open or close an inlet valve, or open or close an outlet valve. VFD flow control only works for pump systems when there are no valves in the system. How often is that. Pressure control VFD works for pumps with, or without valves. You know what NPSHR you have to maintain to keep the pumps running properly, so use pressure control to VFD to keep suction pressure >= NPSHR. Speed up pumps when suction pressure is greater than NPSHR, and slow down (to a stop) when NPSHA drops below NPSHR. Combine with a maximum pressure trip, for when pressures rise due to outlet valve closure. If you get too much flow at the discharge close down on the outlet valve and back the pump up on the curve anywhere, at least until it reaches the high pressure trip. You could do that by an outlet flow control valve, but why try to control the pump with that? The pressure control settings are already doing that for you.

 

[itdepends]

I respectfully disagree BigInch. Flow control VFD works very well for pumps and is particularly useful when pumping abrasive solids which will destroy control valves and (potentially) downstream piping in short order.

As a chem eng/metallurgist the first part of any answer I give starts with "It Depends"

 

[BigInch]

I don't think you disagree at all. I have no problem with VFD, only the part about the "flow control" method of controlling it. Put it on pressure control and any other problems can also be resolved one way or another.

 

[itdepends]

But pressure control still requires a valve to be closed on the discharge to regulate the flow- unless I'm mis-reading your post. So instead of using a pump with VFD + Flowmeter- I now need two additional control elements (pressure transmitter and flow control valve).

Why install the additional equipment when I don't need it? (and in the case of slurries- the control valve becomes a wear point).

As a chem eng/metallurgist the first part of any answer I give starts with "It Depends"

 

[BigInch]

No. Back pressure at the end of a pipe is often supplied by static head alone. You don't always need a valve. In the case of a pressure controlled vfd without a pipeline end valve, the pumps will run at full speed, or whatever lower speed becomes might become the limiting speed such that suction pressure remains above NPSHR.

If you absolutely must control FLOW from the end of the pipeline (be sure you really do have to control flow), you might be able to do that with a pump somewhere within the system (this applies to flow or pressure control of that pump), but only if the system hydraulic characteristics allow it. If the pump's downstream pipeline goes over a high hill, then down again to a lower elevation, you will probably have to use a valve at the end of the pipeline, or your pipeline might not run full in the downhill segment. In such circumstances high flow rates might have enough head loss that the pressure loss per unit length acts the same as an end valve (kind of a distributed backpressure) and the design flow rate can flow without a valve holding backpressure at the pipeline's end, although at other lower flow rates some back pressure might be required at the end of the pipeline to keep the pipeline flowing full at any lower flow rate. If you don't mind pipelines not flowing full, then you wouldn't need a valve at the end of the pipeline.

In chemical plant work flow control is often required. That is not true in liquid pipelines where the business objective of the pipeline is to move as much product as possible in the fastest time possible. You have to first realize that flow control is inherently contrary to the business objectives of liquid pipelines. It is actually the same for gas pipelines, but due to the compressibility of gas, there is wide pressure variation with flow which can vary constantly with temperature and line pack so customer requirements are usually based on meeting specific flow rates at specific minimum delivery pressures. It is rarely necessary that liquid pipelines actually have a specific required delivery flow rate from the end of a pipeline. Most liquid transportation contracts are based on moving a specific volume (10,000,000 bbls of your oil and maybe 20,000,000 of somebody elses) within a given month. We are usually not interested in doing that at anything but the fastest flow rate possible within a given system. For that example a design flow rate might be 1,000,000 bbls/day, but if we could do it at 1,100,000, we just might want to do it. That flexibility to flow at higher capacity should not be limited by somebody's arbitrary flow rate setting. It should only be limited by the maximum flow rate when the pumps start hitting NPSHR. That is a pressure setting, not flow rate.

 

[itdepends]

BigInch your latest posts have made it clear to me that we are talking from very different backgrounds. My background is in metal extraction and refining plants where flow control is an inherent requirement for most pumping systems. This is obviously different to your background in moving bulk volumes of product from A to B.

Where I have experienced the use of booster pumps has been where high pressure is required to inject liquids/slurries into chemical reactors or to pump viscous slurries longer distances to tailings/waste disposal lines. Nearly all of these systems are two phase (liquid/solids) with minimal levels of ullage/storage between stages. Running at full flow and then stopping results in bogged/plugged lines.

As a chem eng/metallurgist the first part of any answer I give starts with "It Depends"

 

[BigInch]

Certainly there are requirements for control of flow, I'm just saying that controlling the pump to achive that objective is not always the most suitable method, especially in transportation pipelines. Control of pumps is best accomplished based on pressure, because it keeps the system within operating limits, no matter what the flow and pressure relationship might be at any given time. IMO pressure control is better with two phase gas/liquid system flows. I think more often than not in two phase flow you will be lucky if flow control works at all, as flow and pressure relationships can vary wildly depending on if liquid is running up a hill and gas down the other side, or vice versa, or other liquid hold up and slugging characteristics of the system. It might work at certain flow rates where liquids are not held up, but if flow drops to the point where liquid holdup becomes dominant, flow control may not work at all. VFD-Pressure control also works for slurries, as the pump can be made to slow down, or speed up based on suction and discharge pressure limits. If you can't maintain minimum suction pressure, or if you exceed discharge pressure, you have no alternative but to slow down the pump, no matter what you are pumping, then try to solve the problem some other way.

I suspect that you are not quite as interested in actual flow control as you are in simply varying the flow rate by any method that might work. In processes where you must inject a specific amount of one chemical into another, I can readily see that flow control (by mass or volumetric rate) is useful. Heavy crude oil pipelines benefit by injecting a diluent oil into a heavy oil at a controlled rate such that a lower resulting viscosity is maintained, or from injecting a specific amount of corrosion inhibitor, quantity varying directly with flow rate. As those processes are independent of pressure, flow control, normally in addition to control of suction and discharge pressures to keep the pumps and piping system within pressure operating limits, is highly useful.

If we look at heating/cooling systems where two flows are mixed to maintain a certain temperature, you could do that with flow control, or pressure control, but depending on entering water temperature variations, those relationships may not always require mixing in constant proportions. In that case neither flow or pressure control would be appropriate to control the process, however pressure control might still be included to keep pumps and system within operating limits. Temperature control should be used to control the process itself.

I still think that controlling pumps and system safety is one thing and that can usually be based on pressure, sometimes including temperture as well. Controlling process can be (and usually is) another thing entirely and the two can only be combined where there are supplemental and direct predictable relationships between them. Otherwise provide pressure control for system and equipment safety and provide separate controls based on whatever other parameter you might need to monitor and control, such as flow, temperture, color, viscosity, or salt content as necessary, but not necessairily wired to VFD pump control.