EnergieReductie door conditieGestuurd Onderhoud

Publieke samenvatting / Public summary

Semiotic Labs' SAM4 provides a novel and unique approach to condition monitoring for critical AC motors and rotating equipment such as pumps, compressors, fans, and conveyors. Unlike traditional solutions that install vibration sensors on the asset in the field, SAM4 detects upcoming failures by analyzing electrical waveforms from inside the Motor Control Cabinet. But in addition to condition monitoring we now want to use the same data to reduce energy consumption. ERGO will focus on research for developing features within SAM4 that provide system-level monitoring as a basis for detecting energy-inefficiencies, and providing data-driven insights for rightsizing motors. ERGO aims to deliver tools for 15-30% energy reductions, +85% reduction in unplanned downtime incidents, and validation of the results of ERGO through specialists as a basis for commercialization. Because SAM4 has the potential to be applied across (virtually) 100% of AC motors in the Netherlands, and these motors represent 53% of total electric power consumption, the potential in terms of energy reduction are considerable.

The research of ERGO aims to develop energy waste reduction features, based on system-level modelling and asset load monitoring over time. ERGO also validates the new condition monitoring features of SAM4 at a site-level. These features will be made available in the commercial version of SAM4 after this project. Goals • Develop models that detect inefficiencies in processes with the aim to reduce energy waste. Semiotic Labs' data across 375 assets suggest that this could result in 2-5% energy reduction across the monitored assets. • Develop dashboards that calculate the energy waste reduction potential of rightsizing and controlling motors, based on the data that SAM4 provides, offering long-term energy consumption reductions by an additional 10-35% - but there is a caveat, explained in chapter 4 - Sustainability and societal relevance. • Deliver a proof of performance study for these energy reduction claims. • Generate new insights of the Proof of Concepts in a large-scale field test, deploying SAM4 on a site-level at several corporate partners. • Provide a technical and economical overview with essential process design and Integration boundary conditions.

Korte omschrijving
The current model is designed for one parameter, in this project it will be developed to be able to predict more parameters, such as rightsizing of equipment and to give insight in other yet unknown parameters, e.g. advanced process control. To perform the research SAM4 will be installed by Semiotic Labs on the premises of several corporate partners. Once installed, the data science team works with corporate partner's domain specialists to turn data into information about the energy savings potential for AC motors. This will lead to new models of industrial processes as a basis for detection of process-inefficiencies. TPA is responsible for reporting and developing best practices with the industrial partners. The reporting focusses on developing proper assumptions with regards to savings potential, validation of the results, and dissemination of the results. Best practices are to be developed that provide a roadmap for capitalizing on the energy savings opportunity.

As part of ERGO, we will: • Install more than 1000 units for data collection on 2 – 5 end users sites. • Investigate, implement, and validate energy efficiency models with the new generated data. • Develop, implement, and validate energy reduction potential dashboards (rightsizing) also with the new generated data. • Investigation report whether energy consumption of monitored assets reaches the projected reduction of 10-20%. • Achieve a substantial improvement of the unplanned downtime incidents prediction by at least 85%, by using the developed models. • Validate the results in a report through a specialist. • Provide integration guidelines. • Provide an innovation system analysis. • Identify roadblocks to technology diffusion.