Parallel pump system operations can get very weird very easily. Identical pumps with identical motors connected by functionally identical piping configurations are the non-existent ideal. The messy end of the spectrum involves dissimilar pumps connected at different points within a complex piping system with numerous dissimilar loads. Always remember that the pumps and piping system including all pumps and loads are always operating as one complete interactive system. Each change in or to any element will influence everything in the system. In general, pumps with relatively steep head vs. flow curves work better in parallel pumping situations, but that is no assurance that everything will always work nicely. Your question suggests that you may have little experience with parallel pumping, so plan to do plenty of very careful, very open-minded studying of everything. If you are hoping for some simple, easy solutions to your problem, you will surely be disappointed.
Whether you are dealing with an existing system that has problems or must be modified or an entirely new system, there will be no substitute for developing a very thorough understanding of all of the characteristics of the system and its elements.
One caution, if the pumps (or some of the pumps) have relatively flat head vs. flow curves, very small changes in shaft speed can have dramatic effects. As an example, I have seen pumps shift loads effectively instantaneously where one pump momentarily "hogs" the load, and the flow rate at another pump "stalls." Since that pump no longer has much (or any) flow, its shaft speed increases because its motor operates with less "slip" due to the lower torque required. Since head varies with the square of shaft speed, the pump now can produce greater head, and it takes over the load forcing another pump into the "stall" situation. Meanwhile, the first pump that "hogged" the load has its shaft speed reduced because of the greater torque load on its motor. Since its head is reduced because of the reduced shaft speed, its ability to deliver to the system is reduced sometimes to the point of going into its "stall" in a continuing sequence of similar events.
Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.
