Photovoltaic Power Plant Performance Analysis and Evaluation: Most PV power plants have operation and maintenance loss loopholes
Mar 02, 2018
We rely on the camera to quickly extract and lock the suspects from the crowd of film and television shows are no stranger to the behind the "face recognition" technology not only reduces labor input costs and improve efficiency, with the recognition technology is gradually maturing, its The scope of application is also getting wider and wider.
For PV power station owners, if the power station can also perform performance identification, such as quickly locating faulty components or inverters from large power plants and intelligently analyzing the operation of power plants through big data analysis to find out the greatest impact on power generation Factors, and give specific recommendations for the implementation of improvements, which will undoubtedly greatly enhance the efficiency and value of power plant operation and maintenance to get more power station benefits.
On January 19, 2018, Prosperous Energy Apollo Photovoltaic Co., Ltd. announced a photovoltaic power plant performance evaluation and evaluation product - Apollo Ensight. It is understood that this is the first large-scale power plant performance evaluation product in China, which drives the intelligent operation and maintenance of power plants through data analysis , And integrates years of experience with the US PV Innovation Center.
They are one of the few teams in the industry to have a fully automated power plant loss high accuracy algorithm. Some of the experts have worked at the National Renewable Energy Laboratory in the United States, the United States SunPower, First Solar and others.
Have a monitoring system, but will not use the data?
The cumulative installed capacity of China's PV power plants has been increasing year by year, and has now exceeded 130 GW, accounting for nearly 10% of the total power installed capacity. Behind the rapid development of photovoltaic power plant operation and maintenance services has become increasingly prominent issue, which is a trillion-level market, how to operate hundreds of power plants at the same time?
In order to effectively operate a large number of scattered power plants, the mainstream photovoltaic power plant investors, the owners are almost all equipped with a set of monitoring software. However, the monitoring system to transmit the power station data reliable? Whether it can truly reflect the operation of the equipment? What is the real situation of power station operation?
Monitoring software real-time access to data, mass data transmission, management and extraction? Such a huge amount of data is difficult to pass through the human line analysis, to accurately find the point of loss of power generation and increase power generation. On the other hand, due to the quality of communications and equipment, there are serious problems with the data quality of PV power plants.
The unfiltered data can seriously affect the effectiveness of data analysis, giving wrong decisions and wasting human and material resources. After several failed verifications, the big data brought by the monitoring system became dispensable management embellishment.
How to use big data? In some people's view, intelligent monitoring can make good use of big data. However, at this stage, the accuracy and practicality of intelligent operation and maintenance of domestic PV power plants are often criticized. Some relatively advanced intelligent monitoring systems can only solve the extraction of huge amounts of data. The data screening, analysis, data identification of power plant performance, operating status and a series of problems still can not be solved.
In fact, there is a very strong threshold in the field for effective analysis of plant data. In developed countries, the loss decomposition of PV plants is generally the subject of continuous research by scientists in major laboratories. In China, because of this shortage of talents, there is also a lack of specialized agencies to provide first-line operation and maintenance personnel with efficient power plant loss decomposition services.
Data analysis drives accurate operation and maintenance
"The traditional off-line manpower operation and maintenance methods can not keep up with the development of PV power plants and help us manage power plants more efficiently and scientifically through high-tech big data analysis technologies." Pascal Wehrli, Algorithm Leader at Vision America Innovation Center, said , Our main idea of developing Ensight solar is to hopefully evaluate the performance of power plants through data analysis and get suggestions that can be improved to help owners or operation and maintenance teams improve their power generation.
Traditional monitoring products only provide KPI at the station level, which can not identify the root cause of the problem, which leads to many problems difficult to be found and can not assess whether the operation and maintenance team is doing well, because of these, some PV plants have been operating at low performance for many years .
"We start with the raw data, through the data cleaning algorithm, including loss analysis algorithm, generating capacity algorithm to find potential problems of power plants, various loss analysis, and then automatically summarize, evaluate the results and give improved operation and maintenance It is recommended to prioritize high-loss issues in terms of revenue loss, to help owners improve their power generation, where the raw data, all of which is directly available, explains Pascal Wehrli.
Practical operation: When cleaning the most cost-effective power plant?
The purpose of big data analysis is to evaluate the operation and performance of power plants to identify problems, provide suggestions for improvement and guide operation and maintenance. Among the many operation and maintenance problems, power station maintenance is a common problem.
In China, dust has a great impact on the power station, with a serious power loss of up to 30%. If frequent cleaning, the cost is very high, if reduce the cleaning frequency loss caused by relatively large, how to weigh? When do you need to clean the power plant, how to clean the most cost-effective?
First of all, dust pollution and the physical environment are closely related to the regional level of dust pollution analysis, different regions need to be treated differently.
Pascal said: "Ensight will accurately quantify the loss of dust, cleaning only those who need cleaning, automatic detection of the regional level of dust accumulation rate, taking into account the loss of dust and cleaning costs, the region by the optimal cleaning strategy to maximize return on investment. "
Taking a 6MW roof distributed power station as an example, there are two kinds of installation inclination, one low inclination of 2 degrees, one high inclination of 13 degrees with a chimney in the middle, and the power station is in a more complicated dust pollution environment.
After obtaining data on this plant's operation, Ensight cleaned and automatically detected the data, and clustered the loss of dust from each inverter. The loss of dust in the less polluted area was estimated to be about 0.3% / day. Heavy polluted areas And low-angle zones have tripled so Ensight can automatically detect dust accumulation rates for different areas.
The next step into the dust cleaning optimization strategy, based on cost-benefit maximization, the use of cost-benefit analysis, the system gives suggestions for improvement: for low inclination area, monthly cleaning, annual cleaning 12 times, for high inclination area The number of annual cleaning 6 times, do not need regular cleaning, so as to reduce part of the cleaning costs, on the other hand this data is based on TMI data acquisition, real power plant data, will continue to use real data iteration, iteratively iterate this algorithm , Get the best time to clean next time. Pascal Demonstration Road.
There are inverter downtime problem.
As an important component of photovoltaic power generation system equipment, inverter shutdown of the power generation capacity is very large, then why the inverter downtime? What is the reason for downtime, how long downtime? Stop loss will cause much damage?
Apollo Ensight automatically logs and identifies information on each equipment downtime, including downtime, lost time, and more, giving you a variety of graphical presentations of its evolution, analyzing why it is down and allowing one-click export of reports. Specific analysis of specific issues, let the data speak, thereby enhancing the power generation capacity, the formation of a proactive operation and maintenance mechanism.
It is reported that at present, Apollo Ensight has been widely promoted and used in the European and American markets, this is the first time in the domestic public, has been the China Guangdong Nuclear Power, Guodiao, Trina Solar, Sky Blue Sky, Sumeida, Zhongkang Electric, Yingli investment, Art Sri Lanka, crystal Australia, Zhenhua Heavy Industries and many other PV power plant owners concern and interest, the next step will start the pilot project.
