Roadmap 7
Increased field performance and reliability
Rationale for support
From multi-MW utility scale down to small systems on resi- dential roofs, electricity generated by photovoltaic systems is changing the energy landscape as we know it. GWs of capacity are added worldwide year after year where the cumulative 1 TW goal could be achieved already in 2022. By the end of the next decade the TW annual market could become reality. PV already represents a share of more than 8 % of the electricity generation in some countries (Italy, Germany, Greece, to name a few) and with these val- ues in mind the penetration levels will quickly reach the double-digit all-over Europe. It is within this scenario that the PV sector must ensure that the installed power capac- ity in GW can also reliably generate TWh of electricity for an extended lifetime.
The introduction of novel technologies and novel PV sys- tem design makes the need of increased field performance and reliability a continuous industry demand. Solutions and services which are already available in the market or close to the market will need to be continuously updated and redefined to capture innovation trends. Moreover, new technologies can introduce new degradation modes once in the field.
Status
Field PV diagnostics, mainly in the form of infrared (IR) and electroluminescence (EL) imaging – and recently the emerging ultraviolet fluorescence (UVF) imaging – are PV O&M tools typically auxiliary to string/inverter level PV monitoring. Often combined with analysis of electrical signatures, these inspection methods can identify, with high spatial resolution, the (potential) presence or evo- lution of different failure modes of PV modules and their exact physical location in a PV plant. At earlier research/ pilot level, several research groups have implemented and demonstrated experimental setups for aerial-IR and day- light EL imaging inspections and fault diagnosis in MW- scale PV plants, employing drones or unmanned aerial ve- hicles (UAVs). Typical IR inspection rates reach up to 4 MW (of PV system size) per hour, corresponding to net flight time, though complete studies (i.e. inspection, manual data treatment, diagnostic analysis, reporting) require 6-8 times longer time for the same system size. More recently, broad adoption of EL- and (mostly) IR-based PV diagnostics has been accelerated through technical standardization, technology collaboration platforms and recent technologi- cal advances in drone-based imaging and digital mapping. As such, today, aerial-IR imaging attracts high attention, emerging among the best practices for PV O&M and as the cornerstone for advanced, large-scale failure diag- nostics for PV plants. Turnkey aerial-IR inspection servic- es are offered, including artificial intelligence (AI)-based data analytics, fault diagnostics and reporting as well as consulting, i.e. recommendations for corrective mainte- nance actions to PV asset owners and O&M engineers
The detection of failures in the field and the subsequent action are triggered by:
- Periodic field inspection which are contractual obli- gations for O&M operators
- Alarms generated by monitoring systems
State-of-the-art commercial solutions for PV monitoring, allow for monitoring the operational state of PV systems and pinpointing performance issues in real-time and high temporal granularity, from system up to string/array or inverter level. In principle, as defined in the IEC 61724, such solutions involve: i) monitoring hardware for on-site logging of acquired electrical outputs (inverters, strings, meters) and weather data (e.g. irradiance, ambient tem- perature), coupled with ii) management software for re- mote performance management, data visualisation, KPI calculations, reporting, alarming and ticketing. The most advanced PV yield monitoring and fault diagnostic tools offer software-driven quantification and classification of string/inverter-level failures, as well as data analytics for soiling rates and performance degradation. On the other hand, other platforms offer supervisory control and data acquisition (SCADA) features, tailored for utility-scale PV plants. Yet, particularly for utility-scale PV systems moni- toring and diagnostic needs are significantly complex and demanding. As of today, detection and assessment of un- derperformance in PV plants are typically executed in a semi-manual top-down approach, analysing low perform- ing components (e.g. PV modules) by drilling down from substations, inverters to strings and junction boxes.
On this basis, monitoring-based fault analysis and diag- nosis are time-consuming, expert dependent and often of insufficient spatial granularity. As a result, several under- performance issues and failure modules – especially on PV module level – may either remain undetected, trigger “false alarms” or their root-cause stays unidentified.
Targets, Type of Activity and TRL
The introduction of novel technologies and novel PV system design makes the need of increased field per- formance and reliability a continuous industry demand. Solutions and services which are already available in the market or close to the market will need to be continuous- ly updated and redefined to capture innovation trends. Moreover, new technologies can introduce new degrada- tion modes once in the field.
Development of algorithm for predictive mainte- nance to avoid component failures
Embedded sensors and use of on-site autonomous UAV to enable continuous and cost-effective field diagnostics for optimal O&M strategy and analysis of failure evolution
The conceptualisation, innovation and deployment of EPC and O&M friendly PV components and sys- tem designs
Hybrid or integrated monitoring-diagnostic imagery solutions for maximum spatiotemporal granularity and diagnostic resolution. Multispectral imagery in- spections linked with electrical signature; synchro- nisation of field techniques with monitoring
Diagnostic and field inspection enabled by novel features in PV components (fully automated diag- nostic techniques)
Early alert detection system for Potential Induced Degradation (PID)
Effective and large scale use of metrics (e.g. CPN) to optimise O&M strategies
“Complete” diagnostics, e.g. by IR inspections → fault detection, identification and loss analysis up to module/submodule level → minimize or even eliminate any dependence of IV tracing (time-con- suming, costly, yet still required and applied today)
Interoperability, standardization and auto-config- uration of sensors, data acquisition, inverters and communication systems within PV plants and be- tween PV plants and central monitoring systems (Industry 4.0/internet of Things)
Smart control/tracking systems (e.g. coupled with real-time monitoring data, e-yield forecasting, EMS, meteo, etc) for performance optimisation in specif- ic PV applications (e.g. optimised self-consumption in micro-grids; optimised energy/crop production in agri-PV; “self-protection” under extreme events in harsh environments, e.g. dust/snow storms).
Development of data-driven and/or physical mod- els / Reliability models of PV modules, inverters and other BOS components to predict the lifetime based on field data including climate dependent stress factors
Creation of a large-scale database of PV plant per- formance to increase the knowledge in terms of performance ratio, performance loss rates, climate and other stress factors dependency to be used for the development of algorithms, models, etc +
Interoperability of databases at EU member state level for incentivised PV systems (Mandate through RED directive to share performance of incentivised PV systems as open data in compliance with GDPR)
The role of digitalisation
Digitalisation will enable the creation of BIM/Digital Twin concepts which will allow an asset to be properly followed along the whole value chain down to component level. From the manufacturing phase, through EPC, O&M and end of life
KPIs
| KPI | Target Value (2030) |
|---|---|
| Inspected PV plants using (semi)-automatic EL/PL | 420 MW/day |
| Inspected and analysed PV plants using aerial IR (referring to low-altitude IEC compliant detailed IR inspection) | 6 MW/h |
| Failures or underperformance issues identified (root-cause analysed) and recovered or isolated; | >90 % |
| Cost Priority Number of PV system (total cost of O&M, insurance, warranty, etc.) | <10 Euro/kWp/year |
| Diagnostic accuracy for automated aerial IR imagery: false negatives/positives | <10 % |
| Diagnostic accuracy: modelled / calculated power loss for automated IR imagery | >95 % |
| On PV plant level, common annual performance ratio (PR) including periods of unavailability and after correction for expected degradation in the field. | 85 % for residential and small commercial plants and 90 % for other plants |
| Proven system energy output per year; (verified by extrapolating performance loss rate analysis and defining contribution at single component level,) | at least 80 % of initial level for 40 years by 2030 PV module degradation 0.4 %/y |
| Cost reduction on today’s per-schedule preventive or corrective O&M as a result of reducing failures and limiting unnecessary O&M tasks and predictive maintenance | by 10-15 % |
| Size of large-scale PV performance database | 50 GW included in the database with at least 3 years of average operational time by 2025 and 100 GW with at least 7 years of average operational time by 2030 |