In the past, great efforts have been made to optimize pumps in the sense of a Product Approach: The energy efficiency at the pump's point of optimal operation was maximized by improving its design. While this has enabled considerable improvements in the optimal efficiency, in practical applications, where one finds oversizing or varying loads, 90% of the pumps are operated at partial load and thus not at their best operating point [1].

This insight has led to the Extended Product Approach [2], which considers pump and motor as part of a system with various operating points. A further step towards practically relevant energy assessment is the System Approach: here, the interactions of several components in the surrounding system are taken into account. This is of high relevance, since 85% of the energy consumption associated with pumps are actually dissipated in the system [1]. To address this, the systemic optimization of fluid systems was investigated in a joint project of TU Darmstadt, MLU Halle and KSB SE & Co. KGaA [3].

Employing the methodology Technical Operations Research (TOR), algorithms from discrete optimization were used to design optimal systems. In a current project of VDMA pumps+systems TOR is applied to a booster station for water supply in skyscrapers - a typical example of a fluid system. In this paper, we present the application of all steps of the TOR-methodology for a downscaled booster station. This includes data collection for modeling, global optimization as well as validation through simulations and experiments. First, the function of the system is described using load profiles that have to be fulfilled. Then the aim of the optimization - minimization of life cycle costs - is defined. For modeling the set of possible components, we use manufacturer's data. Based on this, a mathematical optimization model is developed. Only simplified models, e.g. without transient start-up procedures, can be considered within the optimization.

For this reason, a simulation with Modelica is carried out in the next step. Afterwards, the optimal configuration is set up on a test rig and the feasibility of the configuration is checked. The fluid-system test rig is 6-meter-high and has five outlets to ambient pressure on different levels, which represents the downscaled water supply in a skyscraper. In total, 13 speed-controlled pumps are available, of which up to six can be operated and measured simultaneously as a booster station. A modular piping system allows the simple set-up of different system configurations. If shortcomings emerge either in simulation or experiment, the optimization program can be adapted.