The HECTOR project, funded under the German Ministry of Education and Research (FKZ 03EN2006A‑F) and part of the ENPRO 2.0 initiative, ran from 2018 to mid‑2023. Its aim was to develop an efficient modular chemical plant concept by optimising module selection, detecting inefficient operating states, analysing module interactions and handling planning uncertainty. The research was carried out by a consortium of the Technical University of Darmstadt (TUDa‑Opt and TUDa‑Fluid), the industrial partners Klaus Union, Lewa, NETZSCH, the consulting firm IAV, and the chemical company Evonik. TUDa‑Opt led the work packages on the optimal module kit (AP 1) and system‑optimised plant design (AP 3), while TUDa‑Fluid coordinated the wear‑testing work package (AP 2). IAV supplied measurement data and modelling expertise, and Evonik contributed real‑world operating data and helped shape the practical relevance of the module kit. The consortium met monthly via video conferences and held semi‑annual status meetings with the project sponsor, Dechema, to keep the project on track.

The technical core of the report concerns wear detection on two pump types. For the centrifugal pump, vibration sensors were mounted at several positions on the impeller housing. Analysis of the sensor signals revealed that a sensor placed vertically on the housing (channel 2) provided the clearest indication of wear when the impeller gap, throttle gap and axial bearing were altered. At rotational speeds of 1 450 rpm and 2 900 rpm, frequencies above 1 000 Hz showed a pronounced shift in the vibration spectrum for both artificial and abrasive wear. Orbit analysis, which requires two orthogonal sensors, demonstrated that wear changes the kinetic trajectory of the vibration signal, producing larger orbit displacements. However, the displacement was not consistent across the full flow range and did not always provide a distinct diagnostic marker. The study concluded that vibration alone is insufficient for reliable field diagnostics and that additional sensors such as pressure or flow should be integrated to increase robustness.

For the piston‑membrane pump, Lewa performed extensive wear tests on the fluid valve (FV), hydraulic valve (HV) and piston‑bushing pair (KBP) across normal, light, medium and severe wear levels. A broad sensor suite—rotational angle, hydraulic pressure, discharge pressure, speed, torque, flow and acoustic emission—was used in a laboratory setting to identify measurable differences. Normalised hydraulic pressure plotted against rotational angle revealed that FV wear caused an earlier pressure rise at points A1 and A2, whereas HV and KBP wear produced a delayed rise. Torque measurements across operating points (21, 88, 220 rpm and 5, 15, 30 bar) showed distinct patterns for each wear type, confirming that torque is a useful indicator. The analysis highlighted that a minimal set of sensors, particularly hydraulic pressure and torque, could discriminate between wear locations, but that field deployments would benefit from a combined sensor approach.

Overall, the HECTOR study demonstrated the potential of sensor‑based wear detection but also its limitations. The consortium did not develop a final diagnostic concept; instead, the focus remained on detailed wear measurement and analysis to inform future soft‑sensor development. The findings provide a foundation for integrating wear diagnostics into modular plant design, thereby supporting the project’s overarching goal of improving energy efficiency and reducing downtime in chemical processes.