login

Search documents

Browse topics

Document details

 
Title:
 
Performance Optimization of PV Systems through Advanced Data Analytics
 
Author(s):
 
G. Mütter, B. Eizinger
 
Topic:
 
PV Systems and Storage – Modelling, Design, Operation and Performance
Subtopic: Operation, Performance and Maintenance of PV Systems
Event: 36th European Photovoltaic Solar Energy Conference and Exhibition
Session: 5DP.2.3
ISBN: 3-936338-60-4
 
Price:
 
 
0,00 EUR
 
Document(s): presentation
 

Abstract/Summary:


This presentation is a report out of practical life with a huge theoretical background and the experience of operating and monitoring of more than 2 GWp large scale PV plants, located in Europe and India, analyzed and optimized by ALTESO in strong cooperation with the asset owner and the local O&M teams. The presentation will show examples including quantitative results for improvement all over the live time of a PV plant. Case studies and some impressive pictures of real faults of PV plants in Europe and India will show the audience real results of using big data and AI for performance optimization. The presentation will explain how the integration of advanced analytics into PV plant supervision allows a shift away from full-scope and scheduled maintenance towards prescriptive maintenance to achieve highest performance at the lowest cost and ultimately PV asset value maximization. Limited by the presentation time the focus had to be reduced to following examples: 1. Initial phase Common issues: Start-up losses and construction mismatch case study: Gaining 11.000€/year on rewiring only 30 string in a PV plant in the mountains 2. Daily operation phase Common issues: small local failures, O&M team performance, animal impact, soiling and vegetation case studies: a) hidden issues on block-wise underperformance and partial data losses, b) guiding cleaning team to dirtiest areas resulting reduced OPEX and higher performance 3. Wear out phase Common issues: Inverter wear out, module degradation, other equipment wear out case study: back sheet degradation after 7 year, within 6 month 90% failed strings, savings with support of advanced data analytics >300.000€/MW/year (no new modules!)