Use of optimizers with string inverters in photovoltaic projects

Optimizers are increasingly present in photovoltaic projects. Some factors may be decisive for this, mainly the evolution of market knowledge regarding this technology.

It is common for many photovoltaic designers, when they know the MLPE (Module-Level Power Electronics) solution, and how the optimizer works, to already remember several projects carried out in which the optimizer would solve numerous problems and increase performance.

If you still want to know more about the MLPE solution and projects with optimizers, it is recommended to read the articles MLPE and power optimizers for photovoltaic modules, Understand optimizers for photovoltaic systems It is Minimizing mismatch losses with power optimizers.

The optimizers are connected individually to the modules, meaning that each module has its own MPPT (Maximum Power Point Tracker).

Some advantages that can be highlighted with the use of optimizers:

• Individual monitoring with data from each module instead of data from the entire string or array;
• Safety technology, reducing the output voltage of the modules in the event of an electrical arc or accident at the installation site;
• Optimization of modules, making them generate their maximum, even when shaded, and without impacting other modules connected in the string.

Some optimizers require the use of their own inverters, while other manufacturers provide optimizers that can be used with any string inverter, regardless of the manufacturer, and present the possibility of operating even in strings in which not all modules receive optimizers.

Optimizers with string inverters

As is already known, string inverters are made to receive several modules connected in series, forming one or more strings.

The figure below shows an example of a photovoltaic system with a string inverter. In this case, 20 modules are divided into two strings, with 10 modules in each string, and each string is controlled by an individual MPPT.

This photovoltaic architecture is very traditional and the investment pays for itself within a period (payback time) acceptable to the end customer. However, the string system has disadvantages in some situations. Some of them:

• The monitored data comes from several modules in series, making it difficult to detect a defective module or a module that degrades faster than normal;
• If an arc, fire or accident occurs, the only protection that can be activated is anti-islanding, but the system will still present a high continuous voltage in the string, with values generally above 400 V, reaching up to 1500 V. ;
• Modules connected in series will always be limited by the weakest module, so if shading occurs in just one module, the entire string will have its generation reduced.

These three disadvantages are resolved precisely by the three advantages of using optimizers mentioned at the beginning of the article.

However, depending on the project, adding optimizers to all modules can make it financially unviable. With this in mind, some optimizer manufacturers create a modular solution, where there is no need to optimize all modules, and without the need for a specific brand of inverter, which can be applied with any string inverter.

If in the previous system of 20 modules there was a daily shadow in seven modules (4 in string 1 and 3 in string 2, for example), placing optimizers in all modules could solve the problem, however, it would be important to analyze the viability of the investment.

On the other hand, if the designer places optimizers only on modules where shadowing should occur, as shown in Figure 2, he solves the problem of daily shadowing without the need to buy 20 optimizers and without the need to use a specific inverter brand.

This solution can even be applied to systems that are already installed, where the shading problem is detected after the system starts operating, without the need to change the inverter.

To better understand the solution, it is important to understand how this type of optimizer for string inverters works.

Operation of optimizers for string inverters

To understand how optimizers for string inverters work, it is interesting to understand how a IV curve of a partially shaded string.

In our case, we will only analyze string 1 in our example. We have ten modules, seven of them receiving irradiance without obstacles and three of them receiving only a portion of the irradiance. Although the string has four optimizers, let's say in this example that only three modules are shaded.

The IV curve of this string is represented in Figure 3 and, in most inverters, the system would operate at the lowest power peak, so all modules would be limited by the weakest modules (with the lowest current).

Figure 3: IV and PV curves of string 1 with three shaded modules, highlighting the power values of the peaks

Some more modern inverters feature an MPPT solution for shading. Thus, the inverter has the ability to choose which peak to work on. This solution reduces losses due to shading, but does not completely solve the problem, because either the shaded modules will be “bypassed” (have their bypass diodes activated), or the non-shaded modules will generate less according to the shaded current. .

Only the solution with optimizers makes each module really generate its maximum. The objective of optimizers is to make the current in the string the same as that of the strongest module (with the highest current).

Thus, acting as a buck-type DC-DC converter (voltage step-down), the optimizer increases the current leaving the shaded modules, meaning that the current in the shaded modules is not limited to that of the other modules.

In string 1, the shadowless modules want to work at their maximum power point (MPP), where Impp = 9A and Vmpp = 45V. The shaded modules, however, want to work at their MPP, Impp = 2A and Vmpp = 45V.

Then, the optimizers operate by increasing the current coming from the shaded modules, so that it is at the same level as the non-shaded modules. Through the shaded module, the same level of current (2A) will continue to come out, but through the optimizer output, the same level of current will come out from the non-shaded module.

However, the optimizer does not “create power” – the shaded module can only generate its maximum power at that moment. So, if the optimizer increases the current, it must decrease the voltage, to keep the input power equal to the output power.

Considering the lossless optimizer, we have that the input power must be equal to the output power. Remembering that power is the product of current and voltage, we can calculate how much the optimizer's output voltage will be.

P(input)=P(output)

V(input).I(input)=V(output).I(output)

45.2=V(output).9

V(output)=10V

Next, we show an example where three modules of string 1 are shaded. It is interesting to note that the optimizers work to ensure that this shading does not interfere with non-shadowed modules.

Figure 4: String 1 with 3 shaded modules with optimizers acting for each module to generate its maximum power

Another interesting detail is to note that when the module is not shaded, the optimizer just lets its current pass without correction.

If the traditional inverter (without optimizers) had a good MPPT system, capable of detecting shading situations, it could decide which of the peaks of the PV curve (figure 3) to work on, thus the system would generate 987 W or 2848 W. However, With the system with optimizers connected to the string inverter, all modules generate their real maximum and the photovoltaic system can reach 3105 W.

Thus, with this solution it is possible to make the system generate its maximum without the need to install optimizers in all modules. They are only necessary in modules that the designer considers will suffer shading.

Solutions in application in Brazil

The North American company Tigo has optimizers that operate with any string inverter and can only be applied to modules that need optimization. Furthermore, something interesting proposed by the company is to have different optimizers with different functions, and to be able to use them in the same string.

The TS4 family has three models (TS4-AM, TS4-AS, TS4-AO). The first only presents monitoring functions; the second has monitoring and security functions; and the last one is the complete optimizer with monitoring, security and optimization.

According to Tigo, the designer can design strings with the TS4-AO where shadowing may occur, and:

• install the TS4-AM on other modules to continue having the individual monitoring function on all modules, or
• install the TS4-AS in other modules for individual monitoring and security across the entire string.

Figure 5: Tigo presents three possible solutions in systems that do not require complete optimization: (a) optimization only where shadows occur; (b) optimization where shadows occur and individual monitoring of all modules; (c) optimization where shadows occur and all individual monitoring and security functions.