Online measurement technology solutions for hybrid power supplies

Integrating fuel cells with batteries, supercapacitors or other energy storage devices to form a hybrid power source can solve many dynamic power supply and heat generation problems. However, this solution also has its own power management problems.

Hybrid Power

In the power architecture discussed in this article, we call the combination of fuel cells and batteries a hybrid system. This architecture is widely used in a variety of fuel cells and batteries, and replaces power storage devices such as ultracapacitors or supercapacitors. However, each hybrid power implementation is specially designed to meet the unique requirements of the selected fuel cell and battery.

The main components of the hybrid power system include fuel cells, fuel cartridges, batteries, system loads, DC input power supplies and power controllers (see Figure 1). The combination of fuel cells and batteries is called a hybrid power supply (HPS).

The above system can be used as three energy sources and two loads at different stages of use. When the system is not plugged into a DC power source, the combination of fuel cells and/or batteries can power the system load. In addition, when the DC power source does not exist, the fuel cell can also charge the battery to maximize the performance of the end-of-power-shutdown or to achieve better system dynamic power response characteristics. When the DC power source is available, it charges the battery and powers the system load.

With this complex structure, we must precisely control the power path management of the system to ensure that the operation of the system load can always meet the end-user's usage requirements. The key control time is when the available power drops to a certain level, at which time the power supply can no longer power the system load, resulting in a restricted usage configuration or even a controlled shutdown operation.

To achieve this precise control, the power controller must be able to detect multiple factors to generate key data such as effective power and total effective power peak. These key data are defined as follows: Effective power peak is defined as the power that the hybrid power supply can provide in a certain short period of time, such as the power required for disc operation when a DVD player is started or shut down. The peak period depends on the load distribution characteristics of the terminal equipment. The total effective power is defined as the total power that the hybrid power supply can provide, which is independent of the discharge ratio.

Figure 1 Power supply system using hybrid power supply

System monitoring

The battery can be monitored using standard fuel gauges currently available on the market, such as using the bq20z75 to monitor two, three or four series-connected lithium-ion batteries, or using the bq27210 to monitor a single series-connected lithium battery. These monitoring solutions can provide the power controller with the required voltage, current, temperature, charge state and other data.

The battery monitoring system is connected to the power controller via a data bus such as I2C, SMBus or HDQ. Through this interface, the power controller can obtain a very accurate battery state of charge (SOC) to ensure that the battery can be used safely during the charging and discharging process.

Monitoring fuel cells and fuel cartridges is more challenging. The type and amount of available fuel in the fuel cartridge, as well as the current and average efficiency of the fuel cell, are factors that need to be considered when monitoring the available power of the fuel cell.

In many cases, the fuel cartridge is a system-specific device, so the fuel type data can be stored in the power controller. In some other battery monitoring system implementations, we need to provide the fuel data stored in the fuel cartridge and pass it to the power controller through a similar interface bus.

In the implementation of a fuel cartridge with data storage function, the best method is to write the measured remaining fuel data back to the fuel cartridge through the power controller or the fuel filling system. However, this method may only be applicable to power systems where the fuel cartridge can be removed and reinserted.

In addition to the fuel data of the fuel cartridge, other parameters need to be monitored for the fuel cell, including temperature, fuel injection rate, output voltage and output current. These parameters are used to calculate the current efficiency of the fuel cell. For example, the temperature parameter can be used to determine whether the fuel cell is currently in the best working state.

In addition, we also need to measure data such as DC power supply and system load power. With this data and data from the monitoring subsystem, we can calculate the total effective power and peak effective power values. The effective operation time of the terminal equipment depends on these four factors.

When analyzing the characteristics at the end of a power outage, the responsiveness of the fuel cell power output and the size of the battery will also bring new issues, which require further understanding.

Predicting HPS Runtime

The battery and fuel cell monitoring subsystem can provide the main system with total and peak power data, allowing the main system to determine various required user data. In this example structure, we use a power controller, which has many advantages. One of the main advantages is the ability to manage data and subsystems so that the hybrid power source is like a standard battery power source during use.

The power controller is responsible for receiving monitoring data and managing the battery usage process to achieve the highest performance during the expected life of the HPS. This is particularly beneficial in two aspects.

Charging the battery via the fuel cell ensures that the peak available charge is at an optimal level even when no DC power is available. The state of charge (SOC) of the battery is managed to maximize the availability of this architecture. Managing the SOC characteristics is a departure from the way batteries are used in most current portable applications. Typically, the battery is the only wireless power source, so it must provide all the power to the main system. Therefore, the battery should safely store as much power as possible to achieve the longest system operating time. Similarly, the battery charging time is critical, and the shorter the charging time, the better. There is a trade-off between battery charging time and life, but this is not common in current consumer products. For HPS, these two usage dynamics do not work, so a power controller can achieve a better balance between the best conditions of batteries and fuel cells. Ideally, the batteries in the HPS will continue to operate for the entire life of the HPS without needing to be replaced. To achieve this goal, the power controller can provide battery charge management functions such as charging at a lower voltage, charging at a slower rate, and temperature compensation for charging voltage/rate. The power controller regulates the battery charging current to ensure that there is sufficient DC power when the system load is connected.

The recently launched Smart Battery Data Set (SBDS) supplement adds fuel cell data to the existing battery-supported data set, enabling the host to access and control the use of fuel cells and batteries. With a power controller capable of handling complex HPS functions, the SBDS fuel cell supplement can help the host system use the HPS more efficiently.

Increasing the total effective power of the fuel cell and battery enables the main system to realize basic functions such as effective operating time indication, remaining time alarm (RTA), or remaining capacity/power alarm (RCA).

The formula for predicting the running time is as follows:

AtRateTimeToEmpty (ARTTE) = total available power/AtRate()

According to this formula, the main system can determine the effective operation time according to the user's operation intention, such as playing a DVD or starting system diagnosis. It would be better if the main system can further understand the energy consumption in different modes and different programs.

Figure 2 Comparison of power usage

Controlled power outage and maximum HPS uptime

Since the future is always unpredictable, predicting runtime for the user is a “crystal ball” exercise. However, with data from the power system, we can implement a controlled system shutdown when the battery is low. The more precise this control function is, the longer the system runtime will be.

Controlled shutdown must account for measurement errors to ensure that it is achieved under all conditions. Improved accuracy in measuring the total available power will directly increase the amount of power available to the user. Therefore, full exploitation of the energy potential will result in longer operating time (see Figure 2).