Case study: fire in photovoltaic solar inverter

See in this article the importance of monitoring PV systems to avoid accidents
10 minute(s) of reading
Estudo de caso: incêndio em inversor solar fotovoltaico
Result of a fire starting at the input terminals of the photovoltaic inverter

There is no doubt that photovoltaic solar energy is one of the most promising forms of energy generation in Brazil. All photovoltaic solar systems are becoming more accessible and the market is showing high growth rates.

Consumers are increasingly interested in benefits. Manufacturers increasingly invest in new technologies, developing products with greater efficiency and lower cost. Equipment distributors increase their inventories and offer more products and options to the market.

Finally, there is a significant increase in professionals and companies operating in the sector. All these factors contribute to making the market increasingly competitive.

However, this accelerated growth – combined with the lack of knowledge of a large part of consumers – allows the entry of unprepared companies and professionals, which puts the quality of projects and installations at risk.

All electrical installations – especially power generation projects – require monitoring by a competent and qualified technician. Failure to comply with this rule can put material assets and the lives of people who decide to install a photovoltaic system at risk. An example of a problem that can occur is seen in the image below, which shows the result of a fire starting at the input terminals of the photovoltaic inverter.

Estudo de caso incêndio em inversor solar fotovoltaico

Occurrence description

It is a solar system composed of a 40kW inverter and 125 photovoltaic modules. After going into operation, the system needed to undergo maintenance due to the low efficiency reported by the customer.

Upon arriving at the installation site, the maintenance team turned off the circuit breaker that connects the inverter to the AC (alternating current) network and also the DC (direct current) disconnector switch integrated into the inverter. Theoretically, by turning off these two switches, the safety conditions necessary for system maintenance would have been met.

The next step would be to disconnect the MC4 connectors (photovoltaic module inputs) from the bottom of the inverter. The first disconnection attempt generated a electric arc and the same was enough to almost instantly start a fire in the place. What could have happened, given that the system was off on both the AC and DC sides?

Technical analysis of facts

The inverter used in the system has 8 inputs strings and 2 MPPT (maximum power point tracking) inputs, as well as fuses and an integrated DC disconnector switch. In Figure 1 below we show the electrical diagram of one of the inverter's MPPT inputs, where we verify the existence of 4 inputs for connecting strings photovoltaics.

Figure 1: -One of the inverter's MPPT inputs

The disconnector switch is positioned after the busbars where the strings are connected in parallel, as we can see in Figures 1 and 2. In this way, disconnecting the DC switch keeps the input buses energized and consequently the strings remain electrically connected to each other, as seen in the figure below. This shouldn't be a problem, however.

Figure 2 – Strings with different numbers of modules connected to the positive and negative buses of one of the inverter’s MPPT inputs

After extinguishing the fire, a technical inspection was carried out on the system and the following configuration was identified in the project, namely: strings connected to the MPPT2 input:

MPPT2

  • ST1 – 17 modules;
  • ST2 – 16 modules;
  • ST3 – 15 modules;
  • ST4 – 15 modules.

It can be clearly seen that the arrangement was composed of strings containing a different number of photovoltaic modules and connected in parallel, as shown in Figure 2. This situation must be avoided and is prohibited by the design standards for photovoltaic electrical installations. Any well-prepared designer or installer knows that they should not be placed in parallel strings different, either with different quantities or types of modules.

Reverse current

The parallel connection of strings different causes a potential difference between the strings with more modules (higher voltages) and the strings with fewer modules (lower voltages). With the potential difference and the existence of a path, there is reverse current circulation in the modules of the smaller strings.

This situation is illustrated in Figure 3, where we see an example (different from the configuration of the case studied) in which there are three strings with 16 modules and 1 string with 14 modules connected in parallel. As the figure indicates, the strings of 16 modules will supply electrical current in its normal direction, while the string of 14 modules will receive current in reverse direction.

Figure 3: Parallelism of strings with different numbers of modules can cause reverse current circulation in the smaller string

Reverse current in solar modules, although unwanted, can occur in photovoltaic systems when there is any inequality between strings connected in parallel. Even with strings identical, the reverse current may occur if there is a difference in the incidence of solar radiation on the strings.

Thus, the strings with less light (shaded or dirty, for example) will have reduced voltage and will receive reverse current imposed by the strings more illuminated. To avoid this problem, it is necessary to use fuses in series with the strings in any photovoltaic project where there are three or more strings parallel.

What was the cause of the accident?

By disconnecting one of the MC4 connectors at the inverter input, a direct current electrical circuit under load was disconnected, a condition very conducive to the generation of the electric arc, which was what caused the fire.

There was current circulating between the strings, even with the inverter turned off, with the strings smaller under reverse current, as illustrated in Figure 4. Turning off the AC circuit breaker and the DC disconnector switch could not prevent the existence of electric current in the modules in this case.

Figure 4: Reverse currents dividing between strings with fewer modules.

As previously discussed, the fire was caused by the presence of current in the circuit during sectioning by removing the connector. In a well-designed system, with strings of equal size, there should be no current circulating through the photovoltaic modules.

Strictly speaking, the cause of the accident was a design error – a very basic error, the explanation of which can only be based on the lack of preparation of those who designed and also those who installed the system. The occurrence in question is not an accident, but a technical error.

Why didn't the fuses work in this case?

Fuses in series with the strings present inside the inverter could have acted, preventing the existence of current in the system. Photovoltaic modules are specified to withstand a reverse current that is typically twice the rated short-circuit current, i.e., around 18A in modules currently available on the market.

The interrupting current of the fuses must be positioned below this value (18A) and obviously above the nominal short-circuit current value (9A). Typically, fuses with a rated interrupting current of 12A would be used in this design, sufficient to protect the modules if reverse currents above 12A occur.

In a correctly designed photovoltaic system, with strings of equal characteristics, reverse current circulation is possible, but unlikely. Unless the solar irradiation conditions are very different between strings, the reverse current will not flow and it is actually very rare for this to happen. So what happened in this project?

Analyzing the circuit in Figure 4, we see that there are currents circulating through all modules, with the strings smaller ones continually offering a path for the passage of the electric current supplied by the larger ones. This way, even with identical solar irradiation conditions, there will always be some current circulating through the modules.

The fuses could have tripped if we had a greater number of strings, generating conditions for the creation of high intensity reverse currents. However, in this case the reverse currents were divided between the strings smaller, causing currents that were certainly below 12A, the interrupting value of the fuses.

Even so, in the event of an electrical arc, why didn't the fuses operate? The answer is simple: it is not the function of fuses to protect the circuit against electric arcs. If the intensity of the electric arc is below the fuse's interrupting current, the fuse will not even notice the existence of the arc.

It is a common mistake to think that the fuses present in photovoltaic circuits (on the DC side) prevent accidents like this. The function of fuses is to prevent overcurrents in the circuits (to protect the modules) and has no relation to extinguishing electric arcs.

Conclusion

There are obvious design and execution errors in this case studied. The root of the problem was the parallelism of strings with different numbers of modules, a major conceptual error that any well-prepared designer or installer should not make.

It is essential that the electrical characteristics of the strings connected in parallel are the same (the maximum allowed deviation is 5% in any variable – power, voltage, current). Also the inclination and azimuth of installation of the modules strings parallel lines must be the same (also with maximum allowable deviation of 5%).

In this case, we can see a succession of errors made, from design, execution and entry into operation. The system was poorly designed, poorly installed, and was not commissioned (inspected) before commissioning. Unfortunately, the consumer who purchases a photovoltaic system does not know that they are being harmed by bad professionals, who are quite unprepared, who damage the image of the photovoltaic solar energy market.

Even with all these design and installation flaws, a qualified commissioning team hired to inspect the project would certainly have detected and identified the system's flaws and would have prevented the fire.


This text was published in the author's Linkedin articles area, and is reproduced here with his consent. Its content was edited, complemented and reviewed by the Canal Solar editorial team.

Contributed to this article Bruno Kikumoto It is Marcelo Gradella Vilallva  

 

Picture of Jobson Carvalho
Jobson Carvalho
Specialist in operation and maintenance of photovoltaic systems and Installation Quality Leader at EBES - Brazilian Solar Energy Company.

4 Responses

  1. Good morning, I'm interested in learning everything about distributed photovoltaic energy systems, I've already taken some courses but they only teach the basics, I'm not interested in these types of courses, I really want to learn how to design and install with greater safety and efficiency.

Leave a Reply

Your email address will not be published. Required fields are marked *

Receive the latest news

Subscribe to our weekly newsletter