Taiwan’s thingnario delivers PV production insights from 3,200 PV sites


Thingnario’s report, titled “Taiwan PV Performance Report,” presents its insights courtesy of over 100,000 IoT devices including inverters, power meters, pyranometers, weather stations, and more, and covering 45% of the new installation market in Taiwan. The total weight of the data totals more than 273 terabytes.

While the report focuses on avoiding losses, other interesting data nuggets emerge: although Taiwan covers just 394km from north to south, southern Taiwan has most of the island’s solar plants to avoid more persistent clouds and rain in the north. 

Avoidable losses

The report focuses more on recoverable energy losses, which averages 2.16% of all energy generated per plant. It shows at least one-third of the losses can be fixed through monitoring and swift, proper O&M, says thingnario.

The report also looks at avoidable, recoverable losses and quantifies what those are. Identification of the avoidable losses comes from separating recoverable and unrecoverable losses, the latter of which is not considered in the report. These losses are considered irreversible, and relate to module losses and orientation, ohmic losses in wiring, inverter conversion losses, and so on.

Recoverable losses are further addressed. These losses are then split between availability, referring to complete losses due to entire equipment failures or outages, and performance losses. Performance losses are down to equipment effectiveness and inefficiencies, where maximum efficiencies are maintained through better maintenance and control. An example is if a fan on an inverter fails, or the inverter overheats. This is recorded as a performance loss, given these losses are able to be rectified through correct management namely quality monitoring and proper maintenance.

Thingnario’s data points to better outcomes for solar parks with proper O&M. The report suggests that performance losses in particular, can be minimized via monitoring systems that combine with data analysis, providing both accurate and instant notifications, followed by rapid action from qualified maintenance teams.

The company suggests that using its own AI monitoring system combined with a well-trained and responsive O&M team, helps save US$4,200 per 1 MW. That figure comes from data suggesting that the average downtime for equipment repair sits at 123.6 hours. However, a better O&M team can reduce that downtime to 52.1 hours. In turn, on average, that helps to reduce recoverable losses to only 2.16% of all generated energy.

More from our partners

Per megawatt, the losses on average are 29,3768 kWh, but a better O&M performance can reduce this average to 8,475 kWh, a 70% saving.

AI magic

General PV plant availability can be analyzed through inverter, MPPT, and string downtime monitoring. However exact performance losses remain complicated to evaluate, with production data unable to tell the full story. Thingnario’s approach is to use AI to understand the amount of electricity that is being lost and where improvements can be made, by learning patterns and comparing the real-time performance and real-time environmental changes.

Thingnario developed deep-learning algorithms that discover different kinds of on-site, and sometimes hidden abnormalities. The system can calculate the possible loss, and immediately inform the O&M team of the issues. The process is improved further by the learning system collecting swift feedback from the O&M team, creating an iterative system that keeps the algorithm upgrading and evolving. In turn, this provides an accurate, comprehensive analysis of the performance and abnormal issues.

FIT causes flux

Taiwan’s high feed-in tariff rate (FIT) has previously insulated domestic PV sites from significant losses from poorer O&M practices. FITs started as high as $0.35 per kWh for small scale roof-top solar, and $0.25 for ground-mounted utility scale solar installations. However, with lower FITs now taking over, priced at around $0.14 per kWh, PV site monitoring is now much more important, and monitoring is the first step towards actionable management, via correct and transparent information.

William Kao, COO of thingnario, noted the transition catching out some PV investors. “The energy revolution is happening now,” he said. “It is vital to pay our attention to know how to utilize each equipment we invest in for renewable energy at this critical turning point year.”

The moves in Taiwan to reduce FITs follow similar situations in Australia, UK, and Germany, which reduced large-scale PV tariffs in 2011 following drastically decreasing costs for PV installations and caused shifts in operational approaches.

This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.