Due to chiller manufacturer constraints or just simple engineering habits (i.e proven-to-work, simple controls), back in the 20th century, and in particular towards the end of the 90s, it was industry standard to design a chilled water plant in a primary-secondary variable flow arrangement. In such an arrangement, the primary chilled water pumps are used to maintain a constant flow through the evaporator section of the chiller while the secondary chilled water pumps are sized to deliver chilled water to the building loads.
In some instances, these secondary pumps were originally designed and installed without VFDs; the control method employed in this scenario was to simply turn on the pump via a motor control center, and then let the pump “ride its curve” based on the variation of the building loads. In other instances, the secondary chilled water pumps were designed with VFDs or retrofitted (i.e change its motor) such that the pump speed can be controlled by a VFD. For a hospital, emergency backup for a chilled water plant is a need not a want; it’s not uncommon, even for plants that are getting close to the end of their useful lives, to see emergency chilled water piping connections extended to the outside wall of the mechanical room.