Retrofitting a fixed-speed pump to a variable-speed drive (VFD) is typically analysed under IPMVP Option B: measure kW and flow at several operating points before and after, fit a regression model kW = f(flow), and compare at identical operating points.

In the baseline case the fixed-speed pump is throttled by a control valve, so kW stays high while flow is modulated wastefully. After the retrofit the VFD modulates pump speed, and by the affinity laws shaft power scales roughly with the cube of flow, which is precisely what Option B (retrofit isolation with all parameters measured) is designed to verify. A non-linear regression kW = a*Q^3 + b*Q^2 + c*Q + d is fitted across the measured operating range, then evaluated at matched flow points to compute savings. Typical savings on this kind of ECM range from 30 to 60 percent versus a fixed-speed pump with a regulating valve.

Always span the full operating envelope when sampling: a regression built only around the design point will badly mis-predict savings at the low flows where most of the VFD benefit actually materialises.