Estimation device, estimation method, and power storage facility

Abstract

An estimation device 55 estimates an internal state of a cell 61 on the basis of an AC impedance characteristic of a diffusion region of the cell 61, the AC impedance characteristic being obtained from a response made while an AC signal is applied to the cell 61 via a power line L for DC charging / discharging purposes.

Description

Estimation device, estimation method, and power storage facility

[0001] The present invention relates to a technique for estimating the internal state of an energy storage element.

[0002] In recent years, the introduction of renewable energy sources such as solar and wind power has expanded, and fluctuations in power generation output due to weather have become an issue. In order to stabilize the power grid against output fluctuations due to weather, it is important to ensure adjustment capacity, and one means for achieving this is energy storage facilities (e.g., energy storage systems (ESS)). Energy storage facilities have a large number of storage batteries.

[0003] Since the capacity of a storage battery (hereinafter also referred to as a cell) decreases due to its own deterioration, it is desirable to perform optimal control after grasping the battery capacity.

[0004] There are two ways to determine the battery capacity of a cell: (1) Actual measurement by DC charging and discharging (2) Prediction using a mathematical model

[0005] In method (1), cells are charged and discharged under predetermined conditions using direct current while the energy storage equipment is in operation, and the battery capacity is calculated from the amount of electricity [Ah] charged and discharged and the voltage difference [V] between the cells before and after charging and discharging. This method (1) requires charging and discharging an amount of electricity equivalent to at least several tens of percent of the SOC (State of Charge), which takes several hours. Furthermore, the battery capacity can be obtained only after charging and discharging an amount of electricity equivalent to several tens of percent of the SOC, but not before that.

[0006] Method (2) involves carrying out cell degradation tests under various conditions (voltage, current, temperature, etc.) and expressing the data obtained using a mathematical model (such as the root law or Arrhenius law), and predicting the current battery capacity from the mathematical model using the cell voltage, current, temperature, etc. during operation of the energy storage equipment as input (see, for example, Patent Document 1 below). Method (2) differs from actual measurements using direct current charging and discharging in that it can predict battery capacity before charging and discharging. However, while the energy storage equipment is in operation, it is necessary to store all of the cell voltage, current, and temperature as historical data from the start of operation, and the amount of historical data to be stored increases as the operating period becomes longer.

[0007] Patent No. 6428957

[0008] In recent years, in addition to the above methods (1) and (2), the AC impedance method has been considered promising. The AC impedance method is a technique for measuring the response when an AC signal is applied to a cell, and can evaluate the internal state of a cell indirectly, non-destructively, and in a short time. One aspect of the present invention uses the AC impedance method to estimate the internal state of a cell.

[0009] An estimation device according to one aspect of the present invention estimates the internal state of a cell based on the AC impedance characteristics of the diffusion region of the cell, obtained from the response when an AC signal is applied to the cell via a power line for DC charging and discharging. The "diffusion region" of the cell will now be described. When an AC signal is applied to the cell using the AC impedance method, the impedance is measured while sweeping the frequency. The resistance component that contributes to the impedance varies depending on the frequency band, with ohmic resistance contributing primarily at high frequencies, charge transfer resistance at intermediate frequencies, and diffusion resistance at low frequencies. The diffusion region is the region in the 10 to 100 mHz frequency band where diffusion resistance makes a dominant contribution to the impedance.

[0010] As mentioned above, the method of measuring battery capacity by DC charging and discharging takes several hours. However, the inventors discovered that applying an AC signal to a cell through a power line can provide the AC impedance characteristics of the cell's diffusion region in a relatively short time, leading to the above-described configuration. This configuration allows the internal state of a cell to be estimated indirectly, non-destructively, and in a short time, without removing the cell from the system or connecting a dedicated AC impedance measuring device to the cell.

[0011] Block diagram of the energy storage system Block diagram of the CMU (Cell Management Unit) Block diagram of the BMU (Battery Management Unit) Diagram showing the measurement principle of the AC impedance method Cole-Cole plot of the cell Graph showing the relationship between R and -X of the cell Graph showing the relationship between the cell's battery capacity and -X / R Equivalent circuit of the cell Cole-Cole plot of the cell Graph showing the relationship between SOC and OCV (Open Circuit Voltage) of the cell Flowchart of the estimation process Current waveform (waveform of the AC signal) Voltage waveform (response waveform) Lookup table Block diagram of the energy storage system

[0012] An outline of an embodiment is described below. (1) The estimation device estimates the internal state of the cell based on the AC impedance characteristics of the diffusion region of the cell, which are obtained from the response when an AC signal is applied to the cell through a DC charging / discharging power line. This configuration allows the internal state of the cell to be estimated indirectly, non-destructively, and in a short time using an AC impedance method. This configuration is advantageous in that it can measure the internal state of the cell in a short time compared to actual measurements using DC charging / discharging, and in that it does not require the cell's operation history data to be stored, thereby reducing the capacity of the storage unit, compared to predictions using a mathematical model.

[0013] (2) The estimation device described in (1) above may acquire the AC impedance of the diffusion region of the cell based on a response of the AC signal applied to the cell from the second cycle onward. This configuration allows the AC impedance to be acquired while the cell is in a stable state, thereby improving the accuracy of estimating the internal state.

[0014] (3) In the estimation device described in (1) or (2) above, the internal state of the cell may be the capacity retention rate or battery capacity of the cell. With this configuration, the amount of power that the cell can discharge over a predetermined period of time can be predicted from the estimated capacity retention rate. Furthermore, if a significant decrease in the cell's capacity retention rate is observed, the cell or the energy storage module including the cell can be replaced, thereby solving the problem of limited charging and discharging due to cell deterioration. Similar effects can be obtained when the battery capacity of the cell is estimated as the internal state.

[0015] (4) The estimation device according to any one of (1) to (3) above may estimate the internal state of the cell based on the ratio of the resistance to the reactance in the diffusion region of the cell. With this configuration, the internal state of the cell can be estimated with a relatively simple calculation.

[0016] (5) The estimation device according to any one of (1) to (4) above may be provided in a power storage facility and estimate the internal state of the plurality of cells connected in series to the power line. This configuration makes it possible to monitor the states of all of the plurality of cells constituting an operating power storage module or a power storage bank (described later). The techniques according to (1) to (5) above can be applied to a power storage facility or a method for estimating the internal state of a cell.

[0017] <First embodiment> 1. Structure of energy storage system 10A Fig. 1 is a block diagram of an energy storage system 10 A. The energy storage system 10 A may be connected to a power grid 1 to adjust the supply and demand of power.

[0018] The power system 1 may be provided by an electric power company, or may be an electric power system independent from the power system of the electric power company.

[0019] The energy storage system 10A includes a power conditioning system (PCS) 20 and a storage battery system 30. The PCS 20 is a device that charges and discharges the storage battery system 30, and is provided on a power line L that is a main circuit. The PCS 20 may be housed in a container of the storage battery system 30.

[0020] The PCS 20 includes a bidirectional inverter 21 and a power control unit 23. The bidirectional inverter 21 is a power conversion device capable of both forward conversion (AC to DC) and reverse conversion (DC to AC). The power control unit 23 controls the bidirectional inverter 21 based on the state of the storage battery system 30, and adjusts the supply and demand of power.

[0021] The battery system 30 is composed of one or more power storage banks 50. In this example, the battery system 30 is composed of two power storage banks 50A and 50B connected in parallel. The power storage banks 50A and 50B are connected to the PCS 20 via a power line L.

[0022] The power storage banks 50A and 50B are each made up of a current sensor 51 that measures the current of the power storage bank 50, a BMU 55, and a plurality of power storage modules 60 connected in series to the power line of the power storage bank 50.

[0023] The power storage module 60 includes a plurality of cells 61 connected in series and a CMU 65. The plurality of cells 61 are fixed to a frame to form a unit. The cells 61 are chargeable and dischargeable power storage elements. Various types of cells can be used, such as lithium-ion secondary batteries (an example of a non-aqueous electrolyte secondary battery).

[0024] As shown in FIG. 2A , the CMU 65 includes a CPU 66 and a storage unit 67 , and monitors the temperature of the power storage module 60 and the voltage of each cell 61 .

[0025] A BMU 55 is provided for each of the power storage banks 50A and 50B. As shown in Fig. 2B, the BMU 55 includes a CPU 56 and a storage unit 57, and monitors the states of the power storage banks 50A and 50B.

[0026] Specifically, the BMU 55 monitors the current of the power storage banks 50A and 50B based on the measurement results of the current sensor 51. The BMU 55 is also connected to each CMU 65 via a communication line, and acquires information on the temperature of the power storage module 60 and the voltage of each cell 61 from the CMU 65 to monitor the state of each power storage module 60.

[0027] 2. Estimation of the Internal State of the Cells When utilizing the storage battery system 30 as a power supply and demand adjustment capability, the following issues arise: (1) Because the capacity of the cells 61 decreases due to degradation, it is desirable to constantly grasp the internal state of the cells 61 and perform optimal control. (2) Ideally, it would be possible to individually grasp the internal state of the multiple cells 61 connected in series and parallel and monitor all of the cells.

[0028] In this embodiment, in order to contribute to solving the above problems (1) and (2), the internal state of each cell 61 of the storage battery system 30 is estimated using an AC impedance method.

[0029] 3 is a diagram showing the measurement principle of the AC impedance method. The AC impedance method is a technique for measuring the response when an AC signal (e.g., an AC current signal or voltage signal) is applied to the cell 61, and can evaluate the internal state of the cell 61 indirectly, non-destructively, and in a short time.

[0030] When an AC signal is applied, the impedance is measured while sweeping the frequency ω, thereby making it possible to obtain a Cole-Cole plot (complex impedance locus) of the cell 61 as shown in FIG.

[0031] The Cole-Cole plot of cell 61 shows that the resistance components that contribute to the impedance differ depending on the frequency band, with ohmic resistance contributing primarily at high frequencies, charge transfer resistance at intermediate frequencies, and diffusion resistance at low frequencies. In other words, by analyzing the Cole-Cole plot, it is possible to indirectly evaluate the phenomena occurring in each part of cell 61.

[0032] In a Cole-Cole plot, the resistance component R and reactance component −X of the AC impedance Z in the diffusion region have a linear relationship, as shown in Figure 4, and the slope of the line Y is expressed as −dX / dR. The diffusion region is a frequency band of 10 to 100 mHz where the diffusion resistance becomes dominant among the resistance components contributing to the impedance.

[0033] A plurality of cells (e.g., lithium ion batteries) 61 with different battery capacities were prepared, and the AC impedance was measured when an AC signal with a frequency of 10 to 100 mHz, which corresponds to the diffusion region, was applied. The battery capacity is the capacity [Ah] that can be extracted from the fully charged cell 61, and is also called the full charge capacity.

[0034] 5A and 5B are graphs showing the measurement results of AC impedance in the diffusion region, where Fig. 5A shows the relationship between AC impedance R and −X, and Fig. 5B shows the relationship between battery capacity and −dX / dR, where R is the resistance component (real part) and X is the reactance component (imaginary part).

[0035] 5B, a correlation was observed between the battery capacity and −dX / dR. In other words, if the relationship between the battery capacity and −dX / dR is known through a preliminary test for cell 61 of energy system 10A in operation, it is possible to estimate the battery capacity of cell 61 by measuring the AC impedance of the diffusion region and calculating −dX / dR.

[0036] The reason why the battery capacity can be estimated from the −dX / dR of the diffusion region is explained below. As shown in Figure 6A, the characteristics of the AC impedance Z of a cell 61 can be expressed by an equivalent circuit consisting of an ohmic resistance R0, a charge transfer resistance R1, an electric double layer C1, and a diffusion resistance Rw.

[0037] Of these, the diffused resistance Rw has a characteristic in which the slope of the reactance component X with respect to the resistance component R is 45 degrees in the diffused region on the Cole-Cole plot (straight line Y in FIG. 4). However, in an actual cell 61, there are cases where this angle is greater than 45 degrees. In such cases, analysis becomes possible by adding a differential capacitance Cdiff as a capacitance component, as shown in FIG. 6B.

[0038] The differential capacity Cdiff means the differential capacity dSOC / dOCV in the SOC-OCV characteristics of the cell 61. In other words, because the AC impedance measurement itself involves charging and discharging the cell 61, the SOC and OCV are likely to increase or decrease during measurement, particularly when a low-frequency AC signal is being applied (during charging and discharging over a long period of time), and as shown in FIG. 7, this effect appears as the impedance 1 / jωCdiff of the differential capacity in the diffusion region of the AC impedance characteristics.

[0039] The differential capacity Cdiff is a parameter based on the SOC-OCV characteristics, and differs between new and deteriorated products, as shown in Figure 8. When the battery capacity decreases due to deterioration, the curve representing the SOC-OCV characteristics shrinks, and the differential capacity Cdiff decreases.

[0040] As a result, the impedance 1 / jωCdiff of the differential capacitance Cdiff in the diffusion region of the AC impedance characteristic increases, and the magnitude (absolute value) of −dX / dR also increases.

[0041] As described above, since there is a correlation between the battery capacity and −dX / dR of the diffusion area, it is possible to estimate the battery capacity of the cell 61 of the energy storage system 10A in operation based on −dX / dR of the diffusion area.

[0042] FIG. 9 is a flowchart of the process for estimating the internal state of the cell 61. The estimation process is composed of five steps, S10 to S50. The estimation process in FIG. 9 can be executed at any timing while the energy storage system 10A is in operation. It is desirable to execute the estimation process when the storage battery system 30 is not charging or discharging DC current through the power line L. Alternatively, the estimation process may be executed while the storage battery system 30 is charging or discharging DC current.

[0043] <S10 (Application of AC signal)> When the control unit of the battery system 30 (in this example, the BMU 55 of the power storage bank 50A) detects the start of the estimation process, it instructs the PCS 20 to output an AC signal and further notifies the BMU 55 of each power storage bank 50A, 50B of the start of the estimation process.

[0044] Upon receiving the instruction, the PCS 20 uses the bidirectional inverter 21 to simultaneously apply AC signals to each of the power storage banks 50A, 50B of the battery system 30 via the power line L for DC charging and discharging. The frequency of the AC signals is not limited, but is preferably in a frequency band within the diffusion region. AC signals in the frequency band within the diffusion region can be generated relatively easily by the bidirectional inverter 21 of the PCS 20. The AC signals may have either a sine wave or a square wave waveform. If the AC signal is a square wave, it is not necessary to control the AC signal at a fine level, making it easier to control the bidirectional inverter 21.

[0045] In this embodiment, the bidirectional inverter 21 applies, as an AC signal, a square-wave current signal included in the frequency band of the diffusion region to the storage battery system 30, as shown in FIG. 10A.

[0046] The number of cycles of the AC signal to be applied may be one cycle, since it takes time for the state of the cell 61 to stabilize immediately after application, but it is preferable that the number of cycles is several (two or more).

[0047] Furthermore, the larger the amplitude of the applied AC signal itself, the larger the amplitude of the response, improving measurement accuracy. On the other hand, if the amplitude is too large, the influence of heat generated by current flow becomes more pronounced. The C rate of the current signal can be any value, but it is desirable that it be in the range of 0.05 to 0.3 [CA]. The range of 0.05 to 0.3 includes 0.05 and 0.3. The C rate of the current that can be output by the bidirectional inverter is, for example, ±1 CA.

[0048] <S20 (Measurement of Response)> While the PCS 20 is applying the AC signal, the following measurements are performed for each of the power storage banks 50A and 50B. The BMU 55 measures the AC signal (in this example, a current signal) applied to the power storage banks 50A and 50B using the current sensor 51, and stores the measurement result in the storage unit 57. Furthermore, as shown in FIG. 10B , the CMU 65 measures the voltage of each cell 61 as a response of each cell 61 to the AC signal, and stores the measurement result in the storage unit 67.

[0049] When an AC signal is applied from the PCS 20 for several cycles, it is desirable to measure and use data from the second cycle onwards, rather than using data from the first cycle.

[0050] <S30 (Calculation of AC Impedance)> Subsequently, the time series data of the current and voltage measured by the BMU 55 and CMU 65 are subjected to a discrete Fourier transform.

[0051] Specifically, the Fourier coefficients Va(f) and Vb(f) of the voltage and the Fourier coefficients Ia(f) and Ib(f) of the current are calculated using the following equations.

[0052]

[0053] where f is the frequency to be extracted (specifically, the frequency band of the diffusion region), n is the number of voltage and current measurement points, t is the measurement time of the voltage and current, V(t) is the time series data of the voltage, and I(t) is the time series data of the current.

[0054] Thereafter, each CMU 65 acquires necessary data from the BMU 55 and calculates the AC impedance Z(f) of each cell 61 using the following formula: This gives the AC impedance Z(f) of the frequency band included in the diffusion region.

[0055]

[0056] R(f) is the real part of Z(f), and X(f) is the imaginary part of Z(f). Basically, AC impedance Z(f) is calculated from time series data when an AC signal of frequency f is applied.

[0057] <S40 (Calculation of -dX / dR)> The CMU 65 calculates -dX / dR for each cell 61 from the AC impedance Z(f) of the diffusion region calculated in S30.

[0058] <S50 (Estimation of Internal State)> In this embodiment, the relationship between −dX / dR in the diffusion region and the capacity maintenance rate is determined by a preliminary test, and the data is stored as a lookup table in the memory unit 67 of the CMU 65 (FIG. 11).

[0059] Therefore, the CMU 65 can estimate the capacity maintenance rate [%] of each cell 61 by referring to the lookup table shown in FIG. 11 for the −dX / dR of the diffusion region calculated in S40.

[0060] The capacity maintenance rate is the ratio of the initial battery capacity C0 to the current battery capacity Ct, and is expressed as follows: The capacity maintenance rate is also referred to as SOH (State of Health).

[0061] Capacity maintenance rate = (Ct / C0) x 100

[0062] If the characteristics of the diffusion region of AC impedance change depending on the temperature or SOC, correlation data (lookup table) between the diffusion region's -dX / dR and the capacity maintenance rate for each temperature and SOC can be acquired (e.g., stored in the storage unit 67) and the correlation data can be selected and used according to the temperature and SOC. Improved accuracy can be expected by performing this estimation in an SOC region where the dV / dQ characteristics of the cell 61 have a slope. V represents voltage [V], and Q represents quantity of electricity [Ah]. This is because the region where the SOC-OCV characteristics of the cell 61 have a slope is characterized by a large increase in dV / dQ (change in V per unit quantity of electricity Q) associated with the degradation of the cell 61 (see FIG. 8 ).

[0063] 3. Explanation of Effects In the storage battery system 30, the PCS 20 and the BMU 55 / CMU 65 work together to measure the AC impedance Z of the cell 61, and it has become possible to estimate the capacity maintenance rate of the cell 61 using the AC impedance characteristics of the diffusion region (specifically, -dX / dR), without the need for a dedicated measuring device.

[0064] In addition, with this configuration, for each storage module 60 of each storage bank 50A, 50B, the capacity maintenance rate of each cell 61 connected in series can be estimated individually and simultaneously, making it possible to monitor the status of all cells.

[0065] The estimated result of the capacity maintenance rate of the cell 61 can be used for the operation and control of the energy storage system 10A. For example, it is possible to predict the amount of power that can be discharged over a predetermined period of time.

[0066] Furthermore, if there is a tendency for the capacity maintenance rate of some of the cells 61 to decrease significantly, the storage module 60 including the deteriorated cells 61 can be replaced, thereby solving the problem of the deteriorated cells 61 limiting the charging and discharging of the storage bank 50.

[0067] <Other Embodiments> The present invention is not limited to the embodiments described above and illustrated in the drawings, and the following embodiments, for example, are also included within the technical scope of the present invention.

[0068] (1) In the above embodiment, the estimation technique is applied to an energy storage system. However, the estimation technique may be applied to other power storage equipment, such as a UPS, a DC power supply, a PCS with a storage battery, etc. Furthermore, the estimation technique may be applied to cells for mobile objects, such as automobiles and motorcycles.

[0069] (2) In the above embodiment, the capacity retention rate and battery capacity of the cell 61 are estimated based on the AC impedance characteristics in the diffusion region of the cell 61. Instead of the capacity retention rate and battery capacity, the internal resistance of the cell 61 may be estimated. The capacity retention rate, battery capacity, and internal resistance are examples of the internal state of the cell 61.

[0070] (3) In the above embodiment, the internal state of the cell 61 is estimated based on the ratio of the resistance component to the inductance component in the diffusion region. However, the internal state of the cell 61 may be estimated by another method based on the AC impedance characteristics in the diffusion region. For example, the internal state of the cell 61 may be estimated based on a change in the inductance component.

[0071] (4) In the above embodiment, the capacity maintenance rate of the cell 61 is estimated by the CMU 65. However, the capacity maintenance rate may be estimated by the BMU 55. The CMU 65 may be eliminated, and the functions of acquiring AC impedance and estimating the capacity maintenance rate may be integrated into the BMU 55. The BMU 55 and / or the CMU 65 are examples of the estimation device of the present invention. The estimation device may also be realized as an integrated circuit. The estimation device is preferably provided in a container of the energy storage system 10A or the battery system 30 or near the system (on the edge side), and estimates the capacity maintenance rate and battery capacity of the cell 61 in almost real time while the system is operating.

[0072] (5) The estimation device may have any configuration as long as it estimates the internal state of the cell based on the AC impedance characteristics of the diffusion region of the cell. For example, in the above embodiment, the AC impedance of the cell 61 is calculated using a discrete Fourier transform, but any method for calculating and obtaining the AC impedance may be used.

[0073] (6) In the above embodiment, in S10 of the estimation process, the PCS 20, which is an example of a power conversion device, is used to apply an AC signal to each cell 61. When a DC / DC converter 70, which is another power conversion device, is provided in each of the power storage banks 50A, 50B as in the energy storage system 10B shown in FIG. 12 , the DC / DC converter 70 may be used to apply an AC signal to each cell 61 of the power storage banks 50A, 50B.

[0074] (7) In the above embodiment, an AC current signal is applied to the multiple cells 61 connected in series, and the voltage of each cell 61 is measured in response. However, the present invention can also be applied to a single cell. In this case, an AC voltage signal may be applied to the single cell, and the current may be measured in response.

[0075] 10A, 10B Energy storage system 20 PCS 30 Battery storage system 50A, 50B Storage bank 51 Current sensor 55 BMU (estimation unit) 60 Storage module 61 Cell 65 CMU (estimation unit) 70 DC / DC converter

Claims

1. An estimation device that estimates an internal state of a cell based on AC impedance characteristics of a diffusion region of the cell obtained from a response when an AC signal is applied to the cell through a power line for DC charging and discharging.

2. The estimation device according to claim 1, wherein the AC impedance of the diffusion region of the cell is obtained based on the response of the AC signal applied to the cell from the second cycle onwards.

3. The estimation device according to claim 1 or 2, wherein the internal state of the cell is a capacity maintenance rate or a battery capacity of the cell.

4. An estimation device according to any one of claims 1 to 3, which estimates the internal state of the cell based on the ratio of the resistance to the reactance in the diffusion region of the cell.

5. An estimation device according to any one of claims 1 to 4, which is provided in a power storage facility and estimates the internal state of a plurality of said cells connected in series to said power line.

6. An energy storage facility comprising: a plurality of cells connected in series to a power line; and an estimation device that estimates the internal state of the cells based on the AC impedance characteristics of the diffusion regions of the plurality of cells obtained from responses when an AC signal is applied to the plurality of cells through the power line.

7. An estimation method, comprising: estimating an internal state of a cell based on AC impedance characteristics of a diffusion region of the cell obtained from a response when an AC signal is applied to the cell through a power line for DC charging and discharging.

Images