“Data really powers everything that we do!
Harness the data to harness the power from the sun.”
The aforementioned quotations incites curiosity and forms the motivation for this blog. The world as a whole is at the tipping point of transition from extensive use of fossil fuels to using renewable energy to generate the energy that we need to lead our lives. Over the years, this shift has been constantly increasing because of the targets set by several countries across the globe to generate electricity from renewable resources, with an ultimate goal of cutting-down the greenhouse gas emissions. In this aspect the introduction of sustainable policies and regulatory frameworks has resulted in phenomenal growth of PV industry. For instance, a record high 114 GW of PV installations is said to be achieved by end of 2019, which is almost 17.5 % increase in growth compared to 2018. This is evident from Figure 1, where the total PV installations is shown based on the data from IEA. Eventually, these sustainable practices has culminated an increase in the Compounded Annual Growth Rate of PV Installations to 42% between 2000 and 2015 and it is set in the range of 100 GW/yr installations in past few years. The above mentioned figures is a clear-indicator of increased industrial growth, credits to rise of new PV markets and increase in global demand.
Figure ‑1 Global PV Installations
Challenges faced by PV industry and Evolution of Big Data
As mentioned earlier, the PV industry has expanded rapidly in the last few years. These large scale expansions has led to more possibilities as well as complications. One of the challenges is most of the large scale PV plants produce several terabytes of data which needs to be analysed to understand the real-time PV performance. These large scale solar plants record meteorological and electrical data which are logged every minute. In addendum, the presence of numerous sensors and high frequency of data collection yields a large dataset with several million points per day which has to be analysed. Though most of the plants have data acquisition systems and communication devices, the quality of data that is retrieved is not known. The recorded data should be summarised in relevant metrics and visual representation of data enables quicker comprehension. In a nutshell, “every data recorded is important and what that data tells us is even more important”. However, the PV plant users do not have time to manipulate and interpret these extensive datasets. As a consequence, most of the PV plant owners end up in bad business decisions due to poorly presented data analysis. The analysis of outdoor PV data solves a variety of questions such as
- Is the PV plant operating normally?
- What are the differences observed in AC and DC energy yield (predicted and measured)?
- Is the loss in PV Performance due to thermal effects, soiling, inverter faults or flaws in
electrical measurement equipments? - How is the data quality and reporting accuracy of measurement equipments?
Significance of Big Data
Big Data acts a powerful decision making instrument and it has greater significance for optimization of PV plant performance. But, the challenge lies in transforming big data into valuable information. One must be aware that the data recorded should not suffer from any inaccuracies, however in reality the raw dataset is bound to have certain errors. Hence, after data retrieval the quality of dataset has to be checked as erroneous data is hard to detect. In this regard, from the acquired dataset analytics helps in visualizing data in the form of simple charts as well as complex functions, thereby enabling to determine inaccuracies if any. Future lies in investing in machine learning tools which can very well solve performance related issues and removing bad data. Also with the advent of Artificial intelligence and machine learning tools paves way for predictive analytics, from which PV performance can be predicted based on current patterns and historical dataset, of course this is beyond the task of human analysts.
In this aspect, more attention has to be paid to the outdoor data analysis, because it can help us in understanding outdoor PR and enables a realistic and accurate comparison of the rated and obtained values, however this still remains a tough nut to crack. Hence, to solve these issues the PV industry definitely needs reliable data tools to analyse outdoor data; in turn the findings from outdoor data can help in reducing the cost of solar energy which will stimulate newer markets.
Final Notes
With the emergence of Big Data, PV industries are investing in data tools, which ensures reliable tracking, monitoring and data evaluation. For instance, market giants like IBM(Watt-sun), DNVGL(Veracity) are building self-learning weather models and data platforms which can forecast failures, commercialization of such technologies will benefit the PV industry. In a nutshell, the PV industry is not searching for one specific data tool we need a combination of many ways such that data can be ideally used to reduce the cost of solar which eventually proves crucial to transform PV to a mainstream energy source. Conclusions drawn from outdoor PV data analysis can definitely help in increasing the PV performance. Focusing on data acquisition, data organisation, data governance and data validation, coupled with the mastery of Data Analytics is vital to achieve higher levels of PV performance as each percent of additional performance contributes to profitability.
Our projects are driven by great ideas. We will be glad to hear yours too. Strike up a conversation with us!
