New equipment is showing up to address the limitations of standard single-inverter systems. Conventional batteryless grid-tied systems generally use a single maximum power point tracker (MPPT) that is housed in the inverter to condition the energy coming from multiple PV modules. The process tracks one maximum power point from array input, varying the ratio between the voltage and current delivered to get the most power it can.
But each PV module has differences that will cause it to have a different MPP than its neighbors. This can be caused by manufacturing tolerances; dust and dirt; different module temperatures; partial array shading; and so on. The result is that the array’s MPP becomes a compromise between all the intermodule variances, and additional power that may be available from some modules (i.e., the unshaded, cleaner, cooler ones) is lost.
One solution is distributed MPPT (or “distributed power harvesting”) equipment—electronic boxes, wired to individual PV modules, to independently track each module’s MPP. This hardware can manipulate the maximum available power to deliver either a higher current or voltage, depending on the configuration of the system (series versus parallel wiring). Included in the category of “distributed MPPT” equipment are specialized DC-to-DC converters, microinverters, and AC modules (i.e., a microinverter packaged with a PV module and UL listed as a single product).
To understand how these devices tap into power that might otherwise be lost, we need to understand a few rules about combining PV modules.
Series configuration. When modules are wired in series, the voltage of each module is summed to obtain the total voltage of the string. The string’s current will remain that of a single module. For example, let’s say we have ten 175-watt modules in series—each at 35 V and 5 A.
Theoretically, the array should have 350 V and 5 A, or 1,750 W. However, what if each module’s amperage isn’t identical to the others, as would result if a module was partially shaded or in a different plane of orientation? When we wire dissimilar modules together in series, voltage will be additive. However, the current will be just above the lowest current in the string.
Let’s say there were nine modules at 35 V and 5 A, but partial shading on the last module reduced its output to 35 V and 3 A. The resulting array’s power would be about 350 V and 3.2 A, or 1,120 W. Notice the loss in the entire string (1,750 ;W – 1,120 W = 630 W) is much greater than the loss from the underperforming module (175 W – 105 W = 70 W).
Although this example is simplified, and bypass diodes on the array would keep the losses lower (see “Bypassing the Bypass” for more detail), the goal in a series configuration is to keep all the currents as equal as possible (discussed later)—or consider parallel wiring.
Parallel configuration. When modules are wired in parallel, the current of each module is summed, with the array’s voltage remaining that of a single module. When dissimilar modules are wired in parallel, the amps add up, and the voltages will average. Wiring the same 10 modules in parallel results in an array that will have 35 V and 50 A, or 1,750 watts. But now, with the last module producing 35 V and 3 A, the resulting array’s output would be about 35 V and 48 A, or 1,680 watts—a significant improvement over the series scenario.
Microinverters and some DC-to-DC converters (such as eIQ Parallux vBoost or Tigo Energy’s Module Maximizer EP series) accommodate parallel configuration. But instead of leaving the module and array voltage low, they first boost the voltage to several hundred volts, proportionally reducing the current. Then, the high voltage (which has also been converted into AC in the case of the microinverter) and low current is wired in parallel. The result is that the shaded module is still able to contribute its full power, without affecting the rest of the array.
Per-module MPPT & series configuration. A distributed MPPT device wired to each module facilitates tracking the highest power point (voltage and current) for each module, but it can also manipulate that extracted maximum voltage and current to reduce potential string losses.
Although each maximization technology is a bit different for DC-to-DC converters, they all use DC-to-DC conversion circuits (“DC boost” raises voltage and lowers current, while “DC buck” drops voltage to raise current). Some solutions lower the voltage output of the weak modules, raising their current to match that of the stronger ones. Others raise the voltage of the stronger modules, lowering the voltage of the weaker ones (buck/boost) to create a fixed string voltage and current.
In the example system, each module produced 35 V and 5 A (175 W), but the reality is that every module has a slightly different peak voltage and peak amperage. The MPPT adjusts the load for each module to find that per-module peak wattage. Then the maximum available power for the entire array (i.e., the sum of all the individual module’s peak watts) can be harvested either in a parallel configuration, or in a series configuration by transforming the voltage to align the string current.
Although other factors, such as module production tolerance or temperature variances, can change a module’s peak power output, let’s continue with the simplified shading example (9 modules with 35 V, 5 A output, and one with 35 V, 3 A). The theoretical total watts available would be 1,680 W [(9 x 175 W) + 105 W]. Now, we can set the MPPT equipment output voltage to whatever we want and transform the voltage in the DC-to-DC device to set the current as well.
In the case of DC-to-DC converters using a series configuration, the output voltage can be set to one that is efficient for the centralized inverter. For example, the SolarEdge PowerBoxes are typically configured so that the total output voltage of the array (i.e. the inverter input voltage) is 400 V to work with their SolarEdge inverter. The output of each PowerBox is set to have a consistent amperage, and the summation of the voltages will equal 400 V.
Following our example: 1,680 watts ÷ 400 V = 4.2 A—this sets the current (amps) goal of each module/MPPT pair (for consistent amperage in series). Each PowerBox will alter output voltage to achieve 4.2 A. With the nine similar modules, voltage will be boosted to about 41.7 V (175 W ÷ 4.2 A = 41.7 V). The voltage of the PowerBox attached to the shaded module will be dropped to 25 V (105 W ÷ 4.2 A = 25 V).
A series configuration would yield 400 VDC at 4.2 A for 1,680 W [(9 x 41.7 V) + (1 x 25 V)], not including conversion losses, which are discussed later. Compare this to the 1,120 W that comes from this series configuration not using distributed MPPT equipment.
The values used in these examples are theoretical and do not account for the impact of bypass diodes (see “Bypass” sidebar). Explanations also are simplified. In reality, there are several ways to achieve this end result: For example, Tigo’s equipment uses a technique called “impedance matching” via transistors, rather than directly alternating voltages.
Several of these products have added benefits, such as module-level monitoring (see the “Equipment” table). MPPT hardware attached to each individual module also can capture module performance data for monitoring each module in the array. Typically, this data is uploaded to the manufacturer’s Web site via a separate systems communications box that is linked to an Internet gateway.
While module-level data monitoring is common for microinverter systems, when included with DC-to-DC converters, this feature also gives centralized inverter systems (for new systems or as a retrofit to older systems) the same capability. Monitoring each module’s performance can help users quickly spot problems within the array. For example, a module with a bad diode (which can keep an entire section of a PV module from producing power) will be easy to spot. Without module-level monitoring, that problem would likely remain undetected.
Another feature available on some units is the ability to limit PV power at the array, remotely. Normally, PV arrays cannot be shut down when the sun is shining on them—even if you disconnect the balance-of-system equipment, high-voltage DC will still exist at the PV array, unless the modules are covered. With some distributed MPPT systems, commands can be issued to each box to disconnect each module from the next.
In systems utilizing DC-to-DC conversion, the voltage of an array normally operating at up to 600 VDC can be reduced to the voltage of a single module (from 17 to 70 VDC for crystalline PV modules), or less with some distributed MPPT equipment. The means for accomplishing this varies by product. For example, Tigo uses a “safety” button (via the on-site communications box or activated online, but the communications box requires AC power to do so). SolarEdge units are programmed to switch into “safety mode” whenever the inverter is shut off. In microinverter-based systems, shutting down the AC disconnect (or PV backfed breaker) automatically shuts down each microinverter and thus each module’s AC output.
Forthcoming safety requirements could significantly impact the distributed MPPT market. A potential National Electrical Code requirement (slated for 2011) calling for “DC Arc Fault Circuit Protection” devices may be implemented. This protection device could be built into distributed MPPT equipment to limit the potential for arc faults in PV arrays. Placing them at the module level increases the protection for the rest of the PV circuit, as opposed to placing that protection further downstream (for example, at the inverter). In anticipation of this new requirement, a few distributed MPPT equipment manufacturers are already including this feature in their equipment (see table).
So, what are the drawbacks? The most obvious is increased up-front cost compared to the traditional centralized inverter approach. But does the increased energy production and access to additional features (monitoring and safety) pencil out with the additional equipment purchase and installation time? It depends on the PV system’s specifics: the amount of array shading (the more shading, the more likely increased energy can be harvested with these devices); various orientations/tilt angles; or different models of PV modules within the same system. Chances are, if your system does not fit into these categories, the additional cost of the equipment will not balance out with the increased energy yield. However, the desire to add module-level system monitoring or the safety features included may be enough to tip the scales for some system owners. Several manufacturers are taking steps to integrate their MPPT devices into the modules to decrease overall system cost and installation time. National Semiconductor’s SM3320 unit is one example.
Another consideration is that these devices use some energy to perform their functions. The question is, then, does the increased energy harvest exceed the energy lost in the device? If so, by how much? Again, the answer depends on the PV system’s specifics, such as the amount of shading and any module mismatch. DC-to-DC converter efficiency ranges from about 97.0% to 99.6%. Parallel configurations are on the low side of that range, due to the significant voltage boosting that they must perform. To fairly compare these systems against microinverter-based ones, you’ll also need to account for the centralized inverter’s efficiency.
These products are new to the industry—there is not yet a track record. All claim to have performed some field-testing and “accelerated life testing” on their products, but it is difficult to predict if the thermal cycling these units endure will take a toll on their electronics—yet most carry 15- to 25-year warranties.
Currently available, UL-listed DC-to-DC converters and microinverters are compared in the table, including the following:
Sage solar advice is to install PV arrays on shade-free, south-facing rooftops. While good counsel, what about sites that have a little shading, multiple module types, or multiple orientations? The boom in residential grid-tied systems is pushing manufacturers to devise solutions, since shading, module variations, and using various roof planes are a reality for many installations. Distributed MPPT products are one solution, yet viability for your installation depends on the circumstances.
Another indicator that this technology is on the rise is the many more manufacturers—in addition to those listed here—that are in various stages of developing their own distributed MPPT products, including Enecsys, Exeltech, Greenray, SMA/OKE, Direct Grid, and SolarBridge. In the years to come, it will be interesting to see which products will ultimately be standing out in the sun—or rather, under the shade of the modules.
Justine Sanchez is a NABCEP-certified PV installer, Home Power Technical Editor, and Solar Energy International instructor.