Having had some experience with pumps operating in parallel working in similarly poor regions of their curves, I suggest that you not be surprised by maintenance problems being much worse than expected. Your best hope may well be for them to hold up long enough to get from one scheduled shut-down to the next and then needing to do a very thorough repair during each scheduled shut-down. These recurring needs may well include repair of surprisingly substantial casing erosion damage.
If the actual head vs. flow curves prove to be flatter than expected, the actual slip characteristics of the individual driving motors can become a part of the operating problems with sharing the load between the pumps. Normally, the published pump curves are based on operating at synchronous speed throughout the entire range while during actual operating conditions the each pump operates at the speed where the required driving torque matches the available torque its induction motor's slip speed that produces that torque. In most cases, this is a matter of trivial concern, but when the real pump curves are too nearly flat or the pumps are too imperfectly matched, it can significantly influence the sharing of the flow between the pumps due to the affinity characteristic of the head varying with the square of the shaft speed. Also, where pump matching is already problematic, details of any differences of the piping associated with the individual pumps should be given all due consideration.
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.
