4 replies to this topic

[Helena_Ali]

Hi..

I have been having a bit of trouble on deciding how to calculate the total discharge head via multiple lines..I have different flows for each line..Basically I have a main line leaving a pump and then it splits into two lines which lead to different vessels.

So am having problems on how should I do the calculations for this case.. Do I calculate the velocity of the fluid in each line then find reynold's number and further calculate the head due to frictional losses for each pipe then sum them together?

Please look at my diagram and tell me what i should do.. Remember my flows are different for each split line...Thank you for your time and consideration.

Attached Files

•  pump_diagram.doc   151.5KB   72 downloads

[Helena_Ali]

I forgot to mention the flows leading to the quencher and filter which are 3,066 kg/h and 1,022 kg/h respectively..Total flow in main line leading from pump is 4,088 kg/h...

Attached Files

•  SPECIFICATION_SHEET.doc   45.5KB   48 downloads

[Art Montemayor]

Helena:

Welcome to our Forum. It's nice to have Trinidadians joining our Forum and sharing their knowledge and queries with us. If we get a few more members from Trinidad, I'm going to request that Chris Haslego, the Administrator, to get us some Steel Band music to play in the background.

What you have is a nice, practical centrifugal pump application problem - very typical of what you will find out in industry when you graduate. I like this type of problem - not only because of its practicality, but also because it requires a lot of the basic, real-life engineering that you will need out in industry: common, God-given horse sense. Let me describe what you have to do to resolve this problem:

• First, you have to address the practical nature of the problem. You have one pump with one source and two targets. That means the pump must be able to overcome the total head required to reach both targets simultaneously. To do that, the pump must have enough head to overcome the highest head among both targets - no matter which one it is.
• Do not worry too much about how the heads are split at the outset. Concentrate on visualizing that your pump has to have enough head and horsepower to get liquid into the targets. And you must also understand that you have to regulate the flow that goes to each of the targets in the quantities that you have specified. This is a very important point. The pump is basically a very stupid machine. It doesn't know what flow rate has to go where. You must engineer a method whereby you can ensure that the proper, pre-determined flowrates go where they are supposed to. To do this, you employ manual globe valves that can throttle each of the two target streams. These are shown on your diagram, but what is just as important and is missing is one or two flowmeters - each on the line leading to their respective globe valves. You must have this basic instrumentation in order to succeed in having the proper flows go into the respective targets.
• The next thing you must realize is the basic principle of fluid flow: You must have (or create) a driving force in order to induce fluid flow. That driving force is a pressure drop. The pump will develop the head ("pressure") that you require to overcome the frictional resistance offered by the piping, valves, instruments, fittings, entrance losses and exit losses - as well as the pressure existing at the target(s). Here is where I have to stop and let you know that you are missing vital Basic Data for this application's solution: the pressures exising at the source tank and the two targets. You must have this information because otherwise you can't calculate the amount of driving force ("total developed head") you have to generate in order to overcome all the resistances + the target pressures.
• Next, you calculate the expected pressure drop in each of the two discharge loops, taking into consideration that you have to allow for the globe valves to throttle the discharge of the pump and build up a discharge head big enough to overcome the pressure drop downstream and allow the desired flfow to each target. In other words, the pump is going to have to develop excess head over the theoretical required. How much depends on the type of valve and the flow rate through the valve. You may have to allow an empirical, contingency pressure drop to make sure that the pump works OK.
• You next back-calculate the total pump flow pressure drop from the pump's discharge to the junction where the flow is split off.
• Next, you calculate the pressure drop in the suction side of the pump.
• Now you are in position to calculate the total developed head required of the pump.
• With all the pressure drops identified, you next apply Bernoulli's equation to find the pump work required to carry out the flow system.

I recommend that you always resort to using a Spreadsheet to draw your system and also do your calculations. And draw your piping schematic drawing as an isometric sketch so that you can show all the off sets and fittings as well as the instrumentations and elevations. A word processing program is a tool for Liberal Arts Students, not engineers. You will find that Excel can do all you need plus more -- if you take the time to study it and learn how to communicate with it. Search out some of my posts and threads on these Forums and you will find out what I mean about communicating and calculating with Excel.

I hope I have been of help in getting you started. We've only begun to define the problem. Please don't fail to come back or continue to ask questions and inquire when you have need.

[Helena_Ali]

Hi..I got the reply and I apologize for the missing data..The pressure of the quencher and filter is 2500kPa and 101.3kPa respectively...The pressure leaving the tank is 810.6kPa...So from what you mentioned before I dont have to consider flow rates,so I jus find the friction factor for the main pipeline then use that along with the total length for all the discharge pipes to calculate the head loss due to friction in the pipes..Well I know that I have to consider the losses for the valves and entrance and exit losses too..So for total discharge head I need to find the total pressure head at the surface,the static discharge head and the discharge line losses,right?

[Helena_Ali]

oh yea,am glad that you are thinking bout adding the pan music to make us feel at home but i recommend some soca instead...lol..