Pumping Configurations

Pumping Configurations

When it comes to chiller plant design, there's a surprising amount of research and experimentation that went into it. Decades of trial and error took the entire industry, worldwide, in quite a few different directions.

The oldest chilled water systems had a pump for every chiller. The purpose of the pump was to overcome pressure loss in the piping and supply the required flow rate to the cooling coils.

The general rule of thumb was that chillers are best kept at a constant water flow rate. This simplified control and design and protected chiller evaporators.

To allow for load variation, a bypass line would be placed at each fan coil or air handling unit, allowing for flow control at the load level. However, this presented a problem, which is that as the heat load decreased (at night for example), keeping all chillers operational would cause a severely low delta T. Solution: turn a chiller or two off.

But having one pump per chiller meant that chiller sequencing (turning chillers on/off depending on the load) will also turn its respective pump on/off. This drops the flow rate of water in all cooling coils, and hence their cooling capacity.

To separate the chilled water production side from the distribution side, a secondary pumping arrangement was added, with a decoupler line connecting it to the primary pumping side. This arrangement allowed for constant TOTAL flow rate through the distribution piping.

Upon the implementation of VSD (variable speed drive) on chilled water pumps, it became possible to change the flow rate of a single pump depending on the load demand. In addition, modern chillers can now operate at a variable water flow rate, given that the minimum flow rate is met.

So, we can divide these into three possible arrangements: (from oldest to newest developed)

1- Constant flow primary pumping only

2- Constant primary with constant/variable flow secondary pumping

3- Variable flow primary pumping only