Centrifugal Pump with Different Capacity

What will happen? Pump 2 will take over. Depending on dozens of things you haven't told us, you may have uncontrolled back flow and catastrophic failure of the entire system.

Or just reduced flow.

No cavitation though.

Understanding the phenomenon will require a knowledge of hydraulics you do not posses.

Is this your design?

Lyn is correct. Here are some more things to consider. If your system is running at maximum capacity when the change from Pump 1 to Pump 2 occurs, there will be back-pressure on pump two that it cannot handle as all of the pump two output pressure, at any flow rate, lies below the output curve for pump 1, otherwise your flow diagram would be drawn differently. You must include some mechanism for handling the subsequent increased fluid volume in the tank. If the tank is sealed, it has to vent - to something - or it could possibly overpressure or explode. If the tank is open, it might overflow, depending on the volume of fluid in the system downstream of pump 1 that will flow back into the tank.

Sorry but I disagree with you. Pump Capacity is flow and pressure not volume on its own.

Assume this is a closed loop system with a total volume(tank, pump and system) of 1000 liters. Pump 1 pumps 100 liters per minute, Pump 2 pumps 50 liters per minute, at the operation point.

The returns to the tank match the pump rate out of the tank (100 liters a minute for pump 1, 50 liters a minute for pump 2)

Tank level doesn't change.

When is your homework due?

Read This (and your notes, if you have any or weren't sleeping during class)

http://www.engineeringtoolbox.com/pump-system-curves-d_635.html

Do your own work and post the answer and we'll mark it for you. Don't expect us to do the work.

Without the NPSH_R curve for pump 2, the possibility of cavitation is unknown.

If assumed that you can effect the changeover instantaneously without any bad effect and that the system is delivering to an open discharge and the system curve consist of static and friction.

The flow rate and pressure will oscillate until the system finds its own stable condition.

At the moment of changeover the pressure at pump 2 will be at the intersection of P1 and system curve.

You now have to draw a horizontal line to P1 and find Qmin2. At that flow the friction will be much less and the pump head will drop causing the flow rate to increase (possibly overloading the power source).

The increased flow will once again increase the friction to above the P2 duty point but it should be less than P1. etc.

The word should be fluctuate .

Can we ask why such non-identical pumps are installed in this system ? I dont understand why anyone would design a system this way....

In my opinion, this kind of system configuration is just begging for a disaster.

Please explain also why the system cannot be brought to a stop when there is a switch-over of pumps. Why must this be done "suddenly" as you stated ?

Oh, wait, I understand....you are a new engineer and your boss told you "that is just the way it is."

From what i experience, if one wants to expend the effort why, there usually is an explanation, a poor one, but still an explanation, i.e., it happens.

Whatever happens, and that is determined by a number of things that the original posting has not stated, the operating point will remain on the System Curve, somewhere, at all times.