# What happens if we operate a positive-displacement pump in parallel with a centrifugal pump?

I have a system that consists of a centrifugal pump and a positive-displacement piston pump, and both pumps get water from one reservoir and then discharge this water back in the same reservoir. The two pumps are connected in parallel.

The specs of each pump are as follows:

• Centrifugal: $Q_{max} = 33 \frac{m^3}{h}$ , $H_{max} = 9.5m$ , Power= 1.5hp at 3000 rpm

• Piston: $Q_{max} = 1.5 \frac{m^3}{h}$ , $H_{max} = 60m$ , Power= 1hp at 1450 rpm

What I need to know what happens if I operated both pumps simultaneously, knowing that both pumps will face the same resistance in the system?

• Why would you expect anything to happen other than a lot of water circulating? – Carl Witthoft Mar 14 '16 at 11:38
• How are they connected? In series? – Carlton Mar 14 '16 at 12:35
• @CarlWitthoft: I'm afraid of back flow from the piston pump to the centrifugal pump – Mohamed AbdelAlim Mar 14 '16 at 12:37
• @Carlton they are connected in parallel – Mohamed AbdelAlim Mar 14 '16 at 12:37
• I suspect nothing much, but please provide a P&ID sketch with the elevations marked. – mart Mar 14 '16 at 13:36

Assuming the pumps are connected as in the following diagram, and that the inlet and outlet pipes are sized appropriately, the total flow rate will be the sum of each individual pump's flow rate. Unless the pipes are severely undersized, you probably don't need to worry about backflow from one pump to the other. This link has some additional information that may be useful.

• Even if both pumps are running at different speeds, and the size of the pipes is different: – Mohamed AbdelAlim Mar 14 '16 at 13:14
• D (piston suction)= 1.25" D (piston discharge)= 1" D (centrifugal suction)= 2" D (centrifugal discharge)= 2" – Mohamed AbdelAlim Mar 14 '16 at 13:15
• The flow rates can be different, the net flow rate will still be the sum of the two pumps. If one pump stops, then you could have back flow though. – Carlton Mar 14 '16 at 14:32
• I think the 60m head pump will overpower and cause some amount of back flow to the centrifugal pump (w/ only 9.5 m of head) – Mark Mar 31 '16 at 5:17

You need to provide pump curves, especially for the centrifugal pump, and a system curve for the pipework. Or better: don't provide them, do the work yourself. Here's how:

You don't want to push water into the pressure side of your centrifugal pump. This happens when the head (static + dynamic) at the outlet is higher then the max. head the pump can deliver. Find the static head and the dynamic head for different flowrates (=system curve).

I assume pipes from both pumps will join shortly after the pumps and the hydraulic losses are neglible. If you think the piping between pumps and T-piece is significant, please provide a P&ID sketch of your system.

Anyway you don't even need to find your exact working point, just look a your system curve at the combined flow rate of both pumps and see where the static + dynamic head are. If lower than the max head for your centrifugal pump, you are clear. The actual working point of the system will be a bit lower than this, but as I read your question you don't need to know this exactly.