The infrastructure supporting the contemporary society must operate with very high reliability. Internet server farms and communications exchanges rely on a very mature technology—lead-acid batteries—to ensure nearly 100% “time to failure†or system availability. Data storage centers use high-tech. Typically, these critical nodes and many other important departments are equipped with backup power. The first layer of backup power is generally an inverter, and the inverter is assembled with a valve-regulated lead-acid (VRLA) battery or a similar sealed colloidal battery. The group provides power conversion.
There are many reasons why this traditional technology is widely used, especially lead-acid batteries are economical and have outstanding reliability. But although outstanding is not perfect. VRLA batteries have a limited life span (typically 12 years in design), and critical systems typically use this type of battery as a backup power source, but periodically replace them. Faults can and do occur from time to time. In a typical backup power system, the batteries function just as they do—they always remain fully charged waiting for the main power supply to fail. The fully charged state is maintained by a continuous, low-current "floating" charge. If the float current is below a certain set limit, the gas generated by the internal electrolysis of the battery will recombine. In this case, even if the float voltage is slightly higher than the standard value of 2.27 V for a single battery, the battery may be damaged. A small overvoltage will cause the electrolyte to precipitate more gas than the recombination process. These untreated gases will escape through the safety valve. If the battery temperature is too high, even if the charge voltage is appropriate, it will cause loss of electrolyte.
Other failure modes include early sulphation, poor pole and grid connections, poor plate and grid connections, electrolyte stratification, and accelerated grid corrosion. There is also a mode of failure, which is a catastrophic failure mode, though it is rare. It is a failure mode unique to VRLA and colloidal batteries that can cause explosions and fires. The only way to prevent thermal runaway is to monitor the internal temperature of the battery.
Monitoring the battery voltage only has a very limited effect on the detection of lead-acid battery capacity drop, which has been recognized in the industry. When the battery performance is declining, the nominal voltage is usually present until it can be revealed when a large current is discharged, and at this time its capacity has been seriously reduced and the terminal voltage has dropped prematurely. The battery condition is determined by measuring the exact specific gravity of the electrolyte. This method is not suitable for sealed VRLA or colloidal batteries. Conventionally, the only way to check the battery capacity is to discharge the entire battery pack to a controlled state, but this method requires The battery stops using. In addition, deep discharge can also reduce the life of lead-acid batteries; this test scheme is used to determine battery life on systems that regularly perform battery discharge tests on their backup batteries and their main power supplies are highly reliable.
Recently, non-intrusive electronic methods that can perform continuous monitoring can detect the near-failure status of a single battery. This method can not only save costs but also maintain the availability of the entire system. Predecessors of such systems usually measure the voltage of a battery or battery pack (battery industry term for multiple batteries housed in the same housing)—although its limitations are well known—plus charge/discharge current and ambient temperature. Some systems attempt to measure or estimate the internal resistance of batteries, and their effectiveness varies.
LEM's Sentinel system is a leading product based on analog measurements that rely on simple basic parameters. It has now evolved to the third generation, Sentinel III. It integrates analog and digital technologies on monolithically custom-designed SoC (system-on-chip) integrated circuits. The device is configured in a module that measures terminal voltage, internal battery temperature, and internal impedance. It is a design-by-design system for systems that can provide accurate measurement results at a cost that is affordable for most backup system configurations. Key elements.
The battery temperature and/or exponentially increasing internal impedance values ​​(Figure 1) are indicative of near-failure. The data logging system monitors the trend of the data over time and identifies potential near-failures. All Sentinel III modules are equipped with an external temperature measurement probe or patch that can be attached directly to the individual battery or battery pack housing to keep the battery temperature as accurate as possible.
figure 1
When the battery is being used or is being charged, a mature technology can be used to assess the internal impedance. Usually, a weak AC voltage is superimposed on the floating DC voltage, and the AC voltage and current at that time are measured. Then, the internal impedance is calculated based on the measurement results. The specific implementation methods are different. However, this method has some limitations, it can only handle the shape of the exponential curve. A single battery that is about to fail will show a good state before the data logger recognizes its failure trend during the failure process. Conversely, when the failure problem occurs, the battery may completely fail in a short period of time.
LEM has developed a more sophisticated algorithm that can detect the performance of a single battery that is attenuating as early as possible. The result is a very reliable test method that can completely penetrate the energy layer of a single battery, ensuring maximum reliability. Based on the popular Randles equivalent circuit, it presents an electrochemical cell as a network of electrical components. Each electrical component is related to a physical factor that constitutes a single cell. (See drawing board)
Figure 3 shows the progression of various parameters over a single battery life. The same characteristics have also been confirmed during discharge or capacity reduction. All impedance factors of the equivalent circuit follow an approximate curve; there is no significant change in the early failure or capacity reduction phase. If the impedance is used as the primary indicator of the operating status of a single battery, it will not give any meaningful indication unless the capacity drops by more than 25-30%. Because the industry standard is a battery whose replacement performance falls below 80% of the specified performance, it is obvious that possible failures must be identified as soon as possible.
image 3
However, in the Randles equivalent circuit, there is a parameter that will change early in a single cell failure (except for pure metal corrosion, this failure mode will appear through the increase of Rm parameters), which is Cdl, double layer capacitance . The lowermost curve of FIG. 3 shows its characteristics; in addition, the shape of the Cdl curve is similar for a normal battery that is in the normal discharge phase, and a dead battery that is assumed to be fully charged.
Monitoring technology
This article does not describe this monitoring technique in detail. It is briefly described below.
The test signals are fed into the individual cells one by one, without injecting large currents in the entire battery pack and without interference with the external system's DC connection. The bipolar test signal was used to improve the original algorithm, but the results proved that the unipolar signal is more reliable. However, DC drift occurs when testing with unipolar signals. Simply eliminating this drift does not preserve the characteristics of the dataset and its characteristics are necessary to accurately determine the parameters. Rearranging signal pulses of different frequencies (including test signals) in a frequency sweep manner allows the battery voltage response to agree with the predetermined curve.
Once the potential drift curve becomes regular, a firmware algorithm can be set up to model this drift and eliminate it, resulting in an average zero voltage data set that fits directly into the Sentinel algorithm. This method reduces the drift error to less than 0.1% without causing significant distortion in the data set. Therefore, this algorithm can also be used in waveform measurement, so that the accuracy of the equivalent circuit parameters is higher.
Many measurement functions and algorithm processing are integrated into a single integrated circuit. The Sentinel module measures both individual cells and the entire 12V cell). Up to 250 measurement points are measured in modules. The measurement results can be submitted to the battery data logger, S-Box, via the dedicated data bus. In a large-scale battery pack system, several such data streams can be synthesized so that the local or remote upstream management system can use these data streams via a standard bus or Internet connection using an integrated web server in the S-Box.
By using measurement SoCs to determine the true state of each battery, it is not only possible to provide the functionality that is needed to detect a mature monitoring architecture that is near failure; other functions and services can also be set.
For example, the internal resistance of a single battery within a battery pack is usually different. Over time, this state creates problems. The SoC intelligent control system can quickly detect these individual batteries. The terminal voltage optimization system can transfer floating current around a single battery that cannot continue to charge...
Real-time charge management extends battery life: With the same terminal voltage, the float current in the VRLA battery is higher than in the flooded battery. This may accelerate the corrosion of the anode plate and reduce the useful life of the battery up to 30%. Eliminating float charge for a certain percentage of life can reduce this undesirable effect. However, this side effect of cycle life has an advantage, which is to reduce the incidence of thermal runaway.
A battery mounting module can also provide end voltage and temperature records throughout the life cycle for use by manufacturers and users.
Over-discharge protection: This type of device is common in charger/UPS systems, especially battery monitors, which terminate the discharge according to the average individual cell voltage to protect the battery. However, the terminal voltage of a battery with poor performance may be much lower than the average voltage of the battery, and it discharges well until it reaches the termination voltage. Therefore, a high-precision 'Time To Run' algorithm was developed to give a warning when any individual battery is about to run out.
Backup battery parameter monitoring must be as detailed as possible in order to generate results that best reflect the status of the battery. This is not just a technical issue but it is also an economic issue. It is indispensable to avoid battery failure, but premature replacement of a battery that is not yet near end of life is extremely uneconomical. In addition to measuring the voltage, impedance, and discharge performance of each battery, LEM also monitors the battery's internal temperature for standard functions; this is a world leader. Currently, LEM is developing a floating-fill sensor using fluxgate technology with a resolution higher than 10mA and no or almost no temperature drift. There is almost no remanence after high-current discharge, and measurement repeatability is higher. Integrating these advanced features, the battery monitor is no longer an expensive additional system, but an extremely cost-effective overall life management system.
illustrate:
figure 1
The battery internal impedance is not a valid indication of near failure. The exponential curve means that early failures are hard to notice, but later performance deteriorates very quickly. .
figure 2
Randles equivalent circuit for electrochemical cells.
image 3
The Randles parameter progresses with battery life or discharge. Different resistance parameters exhibit the same curve shape, and early changes in double-layer capacitance can be detected.
Drawing board
Randles equivalent circuit
figure 2
Each element of Randles (Figure 2) represents a physical process and/or failure mode of an electrochemical cell.
Rm is a metal resistor that represents the resistance of the metal and the junction of the components.
Re is the electrolyte resistance: electrolyte loss may be the main reason for premature failure.
Cdl is an electric double layer capacitor that represents the effective plate area and dielectric strength of the electrolyte.
Rct is the charge transfer (inductive current) resistance due to the limited rate of chemical reaction at the plate/electrolyte interface.
Wi, Warburg impedance, represents the diffusion material transport process. It is a low-frequency electrical component that does not exist during discharge.
After dividing these equivalent electrical components (each component represents a certain performance constraint), the energy layers of a single battery behave as simple electrical components that can be eliminated during the test.