Power Electronics Application Solutions in Off-Grid Solar Photovoltaic System Design

The rapidly declining cost of photovoltaic solar installations is making this technology a practical solution for off-grid applications. The National Renewable Energy Laboratory (NREL) calculated the installed cost of a solar system at just over $7/W in 2010, while SolarBuzz says system pricing is heading towards $15.50 per kilowatt-hour. This is pricing for the entire system, including the solar cells, energy storage devices, and charger and inverter power electronics .

From portable highway construction signs and flashing lights to remote pumping stations and communications networks, the opportunities for off-grid applications seem endless. The main barrier to implementing such projects is high cost, but falling prices are gradually breaking down this barrier and making the implementation of projects feasible. This article will focus on power electronics systems and some key opportunities and trade-offs to keep in mind when designing off-grid solar systems. When solving most system-level problems, it is usually easiest to start with the end application requirements and then work back to identify, define, and size the entire system.

load

A load can be almost anything, but off-grid applications are off-grid for a reason. An application may need to be portable, such as a construction information sign or a simple hazard warning flasher. Given its portability, it is obviously not feasible to connect each load to the grid. Or, perhaps the application is located in a remote location, such as a cellular communication tower or a remote pumping station.

We should first identify some of the key factors that need to be considered when developing an off-grid solar solution. A high-level system block diagram is shown in Figure 1. When using the energy balance approach, it is critical to understand the load, including its type and characteristics over time.

Figure 1: High-level system block diagram

First, determine the type of load and any special requirements. Is it a constant or variable load Is it used only during the day or only at night Is it used intermittently or continuously Knowing the type of load and its characteristics will help determine how the system will be implemented. For example, construction hazard flashers are often constant pulsating loads that may only be needed at night. Therefore, we should charge and size the batteries for these loads during the day and then run these loads at night. Information signs may be pulsating loads, but they need to be running both during the day and at night. In this case, the system needs to be determined to support constant load operation during the day while charging the battery bank to support continuous operation throughout the night. Similarly, pump loads need to operate day and night and are not necessarily constant loads. Here, the system needs to be able to cope with the worst case scenario, or a backup system is needed to address the worst case scenario. For example, a pump station used to remove rainwater may not be a good fit for an off-grid solar application because there is not much sunlight when it rains to charge the batteries. Obviously, the types of loads that can be used are almost endless, and Figure 2 only lists some of the load types discussed above.

Figure 2: Common load types

The key is to know the average daily load, given the amplitude and frequency and operating characteristics. When the load and operating conditions are well understood, specifying the energy storage requirements becomes straightforward. Sizing Energy Storage

Since the sun rises and sets every day, we can calculate the basic energy storage requirements using a simple 24-hour energy balance. Note Figure 3, where the operating scenarios are identified. Clearly, the energy storage must be sized to support the scenario on the left side of the diagram.

The load may be of a well-defined magnitude, such as a number of construction hazard warning flashers, or it may vary greatly, such as a pump application. When dealing with varying loads, it is best to consider the following two scenarios. The first scenario is the "normal" scenario, which covers 95 percent of the operating conditions. The energy storage components should be charged during the day and their capacity should be set to be sufficient to drive the load for the rest of the night. The second scenario is the worst-case normal operating scenario, which covers more than 95 percent of the conditions (i.e., the pump starts operating at dusk and runs at full capacity all night).

The following formula captures the worst-case energy storage requirement by incorporating the hourly maximum load power during the period when no charging is taking place.

There is also a worse case scenario, where the battery bank is not fully charged when the load needs to work and there is no sunlight. The main consideration is the cost of failure (i.e. no flash or no suction). If the pump does not pump, the cost can be considerable. The most obvious solution is to increase the capacity of the energy storage component. However, there is always a worse case scenario. In situations where failure cannot be tolerated or maximum power is rarely needed, it makes the most sense to install a backup system (such as a diesel generator) and size the rest of the system for normal conditions only.

Determining the size of your solar array

With a better understanding of the load requirements, we can determine the size of the solar array. Based on Figure 3, the solar array must be sized to provide the required charge to meet the energy storage requirements within the specified charging time while supporting the average load output during that time period. The following formula identifies this high-order relationship.

Solar array output power = energy storage capacity/charging time + average load power

We can estimate the size of energy storage and solar array components using a simplified energy balance approach. However, there are a number of internal and external factors that need to be understood to accurately adjust these estimates. From an external factor perspective, one of the most important factors is the location of the off-grid application, especially the latitude. This alone can be used to predict the peak amount of solar radiation and how it varies throughout the year. For example, the location of the application site relative to the sun alone can predict that solar radiation will be lowest in the winter and highest in the summer. There are other external factors, including cloud cover and ambient temperature, that can also have an adverse effect on the amount of sunlight the system is expected to receive and the efficiency of energy conversion. It should be understood that these external factors will vary depending on the application and location.

Additionally, internal factors such as the system architecture (especially how the devices are connected) will affect the size of the components. Unfortunately, it is impossible to achieve 100% conversion efficiency, so losses must be taken into account. In the above discussion , the size of the energy storage and solar array determines the energy and power that can be provided. To calculate the power that needs to be generated, the power electronics need to be considered. Power Electronics Topology

Although the system block diagram in Figure 1 helps understand the energy balance, more details are needed to consider the internal factors that affect component size. Figure 4 describes the system implementation in more detail. It also raises issues that affect the power electronics strategy.

A microcontroller-based power strategy provides great flexibility. It allows standard reference designs to be used in a variety of applications while still meeting application-specific needs and implementing advanced features. It not only supports basic power conversion, but also provides flexibility in the selection of core components, supports variations in various operating conditions and enables optimization. In addition, advanced features such as communication and diagnostics can be easily implemented. This is not possible with dedicated discrete power converters.

In this implementation, the biggest question comes from the load; the key question is what is the nature of the load and what does the "load control" need to look like? Does it need voltage or current? How accurate does the voltage or current set point need to be? The load control may be as simple as a relay or as complex as a 3-phase inverter. In any case, it requires a charger function (i.e., power electronics) to charge the energy storage device using solar energy . And, if the system allows, it can also provide maximum peak power tracking (MPPT) function.

Perhaps one of the first decisions to make is whether to use a common rail architecture or a distributed rail architecture. Figure 5 shows the difference between the two, and the load characteristics will likely determine the choice. If the load requires a constant voltage, the common rail shown in Figure 5a may be the best choice. In this case, the load controller is a simple relay or solid-state switch. The solar DC/DC converter maintains the common rail at the voltage setpoint, and the battery charger draws power from the bus to charge the energy storage device. The advantages and disadvantages of this approach are both in the power conversion steps. Considering the average power conversion efficiency of 85%, this means that there is a 15% loss in each conversion. If the solar DC/DC converter can support the load, then only one power conversion step is required. However, to charge the battery, two power conversion steps are required ( solar DC/DC converter to common rail and common rail to bidirectional DC/DC converter), plus an additional conversion (bidirectional DC/DC converter to common rail) to support the load.

The common rail can also be used if the load is only used when the solar DC/DC converter is not operating (i.e. at night). In this case, the solar DC/DC converter can be removed, and the bidirectional DC/DC converter on the energy storage device can be used to charge the battery from the solar array; alternative solutions can also be used to power the load. In this case, the energy only needs to go through two power conversion steps (solar DC/DC converter to bidirectional DC/DC converter, and bidirectional DC/DC converter to load).

The distributed architecture in Figure 6 is more flexible and can support changing load requirements. In this case, the solar DC/DC converter can be used to support the energy storage rail (i.e. charging), and the DC/DC converter can support the load requirements. The disadvantage of this approach is that there are always two power conversions. But in general, this is the optimal solution if the solar array and load are expected to operate simultaneously.

Simple Example

After looking at the power structure from a high-level perspective, we will now look at a simple low-power example. Consider a "construction zone hazard warning flasher" that can often be seen on top of a construction barrel or concrete barrier. From a high-level perspective, the hazard flashers only operate at night, and the battery will be charged at all other times. This characteristic allows us to use a common bus architecture because the hazard flashers are either charging or flashing, not both at the same time. We can further simplify the topology by combining the solar DC/DC converter, bidirectional DC/DC converter, and load control into a single bidirectional converter. Figure 6 shows the recommended circuit design.

Figure 6: Suggested circuit diagram

The proposed circuit design uses Microchip's PIC16F690 microcontroller and two MCP1630 analog PWM controllers to drive a bidirectional flyback converter. During the day, this configuration uses solar energy as input and charges the battery. At night, due to negligible low energy detected on the solar array, the converter begins to power the LED lights in a programmed "blinking" mode. Table 1 lists these assumptions and calculation results.

Table 1: Application assumptions and results

in conclusion

Distributed applications will continue to take advantage of the decreasing cost of solar installations. End-application requirements will dictate system topologies and highlight key performance trade-offs. Microcontroller-based power conversion architectures offer great flexibility in supporting a wide range of end applications and in supporting the continued evolution of photovoltaic solar technology. This flexibility means that current designs will remain viable in the future.