Projektdetaljer
Beskrivelse
Abstract:
One key aspect of the global CO2 emission reduction targets is the usage of renewable electricity to produce hydrogen, liquid fuels and chemicals. One crucial element of this advancing field of Power-to-X (P2X) technologies are power electronic converters. They convert electric energy directly from the renewable source or from the grid to the requirements of the P2X process. Components of the converter are subject to aging induced by environmental conditions or the applications mission profile. Due to variations in these as well as manufacturing tolerances it is a challenge to estimate the converters Remaining-useful-Lifetime (RUL). Usually, during converter design several population wide assumptions are made to account for the lack of knowledge of individual conditions and tolerances. This leaves the potential for error towards a RUL-Estimation (RULE). However, a precise RULE is desired for several reasons:
- Levelized Cost of Energy (LCoE): To minimize investment cost it might be desirable to maximize the usage of the installed materials or reduce design margins.
- Safety: especially in close proximity to plants with potentially leaking reactive substances it is of great importance to warn about potentially explosive EoL events or other component failures
- Predictive maintenance: especially in remote P2X applications reducing down-time and planning for optimal timing of maintenance is a significant economical factor.
- Well-informed economical decisions: in the last phase of the converters operation the RULE is needed.
- 2nd life / recycling of converter or components: after decommissioning of the P2X plant a good RULE may enable future 2nd life related business models
In this project physics-informed machine learning is applied to enrich the known analytical / empirical aging models of high-power semiconductor modules and DC-Link capacitors with their individual characteristics. With regards of desired low LCoE of P2X applications the focus of this work is on sensor less condition monitoring possibilities such as temperature sensitive electric parameters.
Funding: The PhD project is financially supported by the CAPeX ‐ Pionercenter for Accelerating P2X Materials Discovery project running from 2023 to 2036.
One key aspect of the global CO2 emission reduction targets is the usage of renewable electricity to produce hydrogen, liquid fuels and chemicals. One crucial element of this advancing field of Power-to-X (P2X) technologies are power electronic converters. They convert electric energy directly from the renewable source or from the grid to the requirements of the P2X process. Components of the converter are subject to aging induced by environmental conditions or the applications mission profile. Due to variations in these as well as manufacturing tolerances it is a challenge to estimate the converters Remaining-useful-Lifetime (RUL). Usually, during converter design several population wide assumptions are made to account for the lack of knowledge of individual conditions and tolerances. This leaves the potential for error towards a RUL-Estimation (RULE). However, a precise RULE is desired for several reasons:
- Levelized Cost of Energy (LCoE): To minimize investment cost it might be desirable to maximize the usage of the installed materials or reduce design margins.
- Safety: especially in close proximity to plants with potentially leaking reactive substances it is of great importance to warn about potentially explosive EoL events or other component failures
- Predictive maintenance: especially in remote P2X applications reducing down-time and planning for optimal timing of maintenance is a significant economical factor.
- Well-informed economical decisions: in the last phase of the converters operation the RULE is needed.
- 2nd life / recycling of converter or components: after decommissioning of the P2X plant a good RULE may enable future 2nd life related business models
In this project physics-informed machine learning is applied to enrich the known analytical / empirical aging models of high-power semiconductor modules and DC-Link capacitors with their individual characteristics. With regards of desired low LCoE of P2X applications the focus of this work is on sensor less condition monitoring possibilities such as temperature sensitive electric parameters.
Funding: The PhD project is financially supported by the CAPeX ‐ Pionercenter for Accelerating P2X Materials Discovery project running from 2023 to 2036.
| Status | Igangværende |
|---|---|
| Effektiv start/slut dato | 01/09/2024 → 31/08/2027 |
Samarbejdspartnere
- University of Copenhagen
- Technical University of Denmark
- Aarhus University
- University of Southern Denmark
- University of Toronto
- Stanford University
- University of Twente
Fingerprint
Udforsk forskningsemnerne, som dette projekt berører. Disse etiketter er oprettet på grundlag af de underliggende bevillinger/legater. Sammen danner de et unikt fingerprint.