Is VFD always the best solution?

Choose any centrifugal pump with a full speed duty point flow significantly to the left of the Best Efficiency Point flow. Then, operate that pump in a system with a comparatively high set point pressure (e.g. a single stage centrifugal with a 20m full speed duty point and an 18m static head or minimum set pressure - you could also use 60m/55m and get a similar effect). Then, check the specific energy for that duty point (i.e. the kWhr/kL which, for a fixed total volume pumped, is a firm indication of how much power you will use). Then, simulate the low speed duty and identify the specific energy for that duty point and you will more than likely see a significantly increased kWhr/kL compared to the full speed duty point.

Note that the actual kW value drawn by the motor will probably be lower with the reduced speed. What the increased kWhr/kL value tells you is that the pump will have to run much longer to pump the same volume and this will end up costing your more in energy consumption (i.e. power bill) for the volume that you have to pump. e.g. if your pump uses half the power but runs for 3 times as long to deliver the same volume then you have used 50% more energy and not saved 50%.

I am confident in saying this as I have over 20 years experience in this area. I've been responsible for making contractual performance guarantees on behalf of Gozuk Co Ltd (China) for nearly all of their largest projects for over 10 years. In some cases, the guarantees have included performance at reduced speed duty points at the VFD supply terminals. Furthermore, I've presented on this topic to water authorities, councils, consultants and large mining clients. Whilst there may not be a high static head in a water supply system, there is typically a minimum pressure setpoint which, if the pump is operated at relatively low speeds at those pressures, will induce a similar effect on pump performance.