A friend from the solar sector recently get in touch. Their customer who invested in SPP thinks that the production levels are less than expected, whether I can communicate with them or not. Replied of course I can and got the contact information.
Even though there was a lot of low performance in solar energy investments, the situation of this customer was slightly different. He had used three different panels and three different inverters with a capacity of 6MW. It’s kind of a research topic.
First, we connected the central infrastructure of the monitoring system and received data files. Unlike many investors in the field; radiation sensors, ambient and panel temperature sensors were installed. Since we do not see a data logger, a pyranometer / reference cell in many fields, we’re happy to see it now 🙂
First we downloaded the data and converted it into data that the software could understand and process, and then we put it into the PR(performance ratio) analysis.
The raw data of the sensor data [/ caption]
After receiving and compiling the sensor data, we compiled SPP conjugate production data.
An example of the data we followed on excel before transferring it to the software. [/ caption]
If there is a problem, you can catch some of the deviations between the inverters, but there is no major deviation between the inverters. We completed the data compilation process in a short time and passed the analysis.
In our first review, the productions were below expected. A situation that should be a nuisance in the entire field, but I also came across a very problematic field.
When we look at the production breakdown in the analysis, it was possible to see some loss of temperature, but it was also possible to produce less than the loss of heat in the analysis? Something’s not right. It was not possible to go into more detail with the data. We decided to move to the field and make more detailed analysis. At the beginning of these analyzes were the measurements we call the IV-Curve curve.
When we got to the field, they welcomed us with tea. After a pleasant short chat, the team started to fulfill our duties in the field. We gave priority to inverters with relatively low production status. My priority was that skeptic panels had a manufacturing defect, but the fact that the problem had spread throughout the site was in contradiction with this suspicion.
While the IV Curve measurements were taking place, the data was flowing to our analysis station on the back side. At first, a few IV-Curve failures made us think that the question was still in the panels.
As a result of our investigation, we have come to an explanatory reason why the inverter is lower than the others; The lightning arrester in the immediate vicinity of the sequences made a shadow and distorted the IV characteristics of the arrays. Reminded me that why I don’t like lightning rods to friends who have worked with me before, why I insist on catch bars?
This explains why the inverter produces less than the others, but it does not explain the low PR area wide. When we continued the IV measurements, the graphics were very smooth.
My suspicion had disappeared, the panels were intact, where was the problem?
Our promise to the companies that received service from us never been sending a test report only, we were there to solve this problem; we had no intention of returning.
In the meantime, the first thermal images are receieved in the data station. The drone thermal had only a few ghosting and dust-induced hot-spot and was consistent with IV-Curve measurements. In the case of hand-held measurements, we could only detect that several junction-boxes were overheated and that several cables were exposed to the sun.
EPC has fulfilled all obligations. The cables were drawn very well, the construction was not forced, the clamps were not applied much force. Cable heads, connection components, switchgear materials, panel designs were very smooth. If I wanted to write a deficiency that would be forced to write for that reason.
The main tests used in the examination of SPP facilities are Thermal (Infrared), Current-Voltage curves (IV-Curve) and EL (Electroluminescence) tests. The hand-held test and IV-Curve test are related to each other. This is reflected in the IV-Curve test if there are problems in the EL test. After dark, we performed EL measurement by energizing the arrays. It is not possible to measure the EL test in the entire field, so we have examined certain points in the morning in the IV-Curve tests. There was still no serious problem with the panels. We found a few micro-cracking panels in the EL test, but they were too small and did not affect production.
EL Test [/ caption]
And why? There was only this question in mind when we were going to the hotel in the evening. There were three different panels on the field, but all were smooth.
The next morning we started the day by measuring the performance of the inverters. We watched an inverter for about 4 hours regularly with our special devices. I would like to point out a wrong point about the inverters. The efficiency of the inverters is not one-dimensional but two-dimensional.
The efficiency of the inverters is shown by the change of the instantaneous output power, but this is not enough.
The efficiency of the inverters also varies according to the voltage of the array. Taking into account this fact, we carried out the measurements per second and examined the behavior of the inverters. There was no problem in the inverters; they were acting just as they should.
I realized that we were looking for the error in the wrong place. The problem was in another place that escaped our eyes.
I shifted my focus to the SCADA system reporting low PR at the plant, maybe there was something wrong there? I first directed the team to check the calibrations of the sensors in the field. There was no problem in the calibrations of the devices and they were compatible with the y = ax + b curves. I had to enter the SCADA system in more detail when the sensors were intact.
That’s right there; Hundreds of pages of SCADA looked at me, smiling, error “Gradient”.
Sensors were analog devices if not specifically ordered with RS485 connection
They provide a resistance, current or voltage across the measured value. We convert these analog values to digital values with devices. There are two parameters during this conversion
Y = Gradient * X + Offset
The gradient and offset values are determined by the sensor manufacturers. An error in the offset is easily detected in the analysis phase (the sensor gives the values that should not be given, for example, night light), but it is very difficult to detect an error in the Gradient by looking at the data. 0 is 0 where it should be, but a radiation sensor with 800 w / m2 during the day can easily show 850 w / m2. When the error is detected, it is easy to correct the retrospective data.
It was obvious what we had to do at that moment. The value to enter 100 was entered into 110, then the sensor values were incorrectly reported to the SCADA system and the PR calculations were made incorrectly.
At this stage, both the system’s gradient was corrected and we rearranged the data to be 100. The result was more than 85% PRs. The only obstacle in this well operating power plant was the faulty introduction of the sensor to the system.
Executive note: EPC companies can lower the gradient value of the power plant and increase the PR value of the power plant. As an investor, it is very important to check the accuracy of this variable. 🙂
So far, we had the opportunity to perform control over 130MW in Turkey. As we found many errors in these plants due to lack of maintenance, EPC error and material error. We also come across excellent plants which have no errors to report. Do not be afraid of investing Turkish solar power because of some rumours like “All the plants were poorly constructed, panel dumpster”; there are domestic suppliers, EPC companies that produce very nice products and produce very good materials in this country. The important thing is to get the right people to do your right and convenient choices. In the case of supplying 25-year performance-guaranteed panels, this investment should not be considered short-term.
I would like to extend my sincere thanks to you if you are a semi-technical investor who has read this article so far. Very few people can come to this stage. 🙂 If you think that your Solar facility produces less than it needs or you have any problems, please contact us. Let’s find the problem in your facility together.
You can fill in the form below to contact us.